a critical cattell's model of personality - brunel university research
TRANSCRIPT
A CRITICAL
0f
ANALYSIS
CATTELL'S MODEL OF PERSONALITY
P SAVILI, E PhD Thesis Brunel University
March 1978.
Education Department.
CONTENTS -------- Page
ACKNOWLEDGEMENTS 9
ABSTRACT 10
SECTION 1: GENERAL INTRODUCTION 11
1. INTRODUCTION 11
TH -- PERSONALITY Q ESTIO: 1 AIR. 12
16 = TIE : ACTOR ANA YT_C APP _t. P-L7RSO`. {AL IT'1i i' OAýri 4 CATTELL' S MODEL OF PET, 1SONALITY 19
DESCRIPTION OF CATTELL'S PRIMARY SOURCE TRAITS 25 CATTELL' S MEASUREMENT OF PERSONALITY BY QUESTIONNAIRE 50 CRITICISMS OF CATTELL' S MODEL CF PERSONALITY 52 CATTELL'S METHODS OF SCALE CONSTRUCTION 57
SECTION 2: HYPOTHESES, METHODO-L, OGY, PROCEDURE, THE SAMPLE 64
2. AIMS AND HYPOTHESES 64
INTRODUCTION 65 GENERAL HYPOTHESES 65 HYPOTHESES ON GROUP DIFFERENCES IN THE PRIMARY SOURCE TRAITS 66 HYPOTHESES ON THE RELIABILITY OF THE PRIMARY SOURCE TRAITS AW
70 HYPOTHESES ON THE INTERNAL CONSISTENCY OF THE PRIMARY SOURCE TRAITS 71 ALTERNATIVE MODELS IN THE PERSONALITY DOMAIN 72 AIMS
3. METHODOLOGY 75
INTRODUCTION 76 SAMPLING 80
4. PROCEDURE 83
THE PILOT STUDIES 84 TRAINING THE MAIN FIELDFORCE 88 INTERVIEWERS' INSTRUCTIONS 89 THE BACKGROUND QUESTIONNAIRE 102 DATA PROCESSING 108
5. DESCRIPTION OF TF SAMPLE 109
DESCRIPTION OF THE SAMPLE CHARACTERISTICS 112 THE PROBLEM OF NOP, i_RESPONSE 113
I
SECTION 3: GROUP DIFFERENCES AND THE PRIMARY SOURCE TRAITS
6. CROSS-NATIONAL AND SEX DIrFE. R . NCES
INTROD , -C T1ON CROSS-^. y' TONAL, DIFFERENCES: RESULTS SEX DIF E. RENCES: RESULTS DISCUSSION OF THE CROSS_1N. ATZONAL ANL SEX DIFFERENCES
7. SCCLkL CLASS DIFFERENCES
,. ]T: - -cr: c r
SCCI LASS DIr'_ ERELCr $: ES LTS SOCIAL CLASS DIFFERENCES: DISCUSSIO°.
8. AGE DIFFERENCES
INTRODUCTION RESULTS FOR PRODUCT MOMENT AND ETA COEFFICIENTS CATTELL' S QUADRATIC AGE CORR. c'. 'TIONS RESULTS ON CATTELL'S QUADRATIC AGE CCR. RsECTIONS CATTELL'S AGE CORRECTIONS: RESULTS AGE DIFFERENCES : DISCUSSION GROUP DIFFERENCES: CONCLUSIONS
SECTION 4: RELIABILITY AND THE PRIMARY SOURCE TRAITS
9. ALTERNATE FORM RELIABILITY OF THE PRlf; lARY SOURCE TRAITS
INTRCDUCTION TO RELIABILITY THEORY INTERNAL CONSISTENCY RELIABILITY TEST-RETEST RELIABILITY ALTERNATE FORM RELIABILITY THE STANDARD ERROR OF MEASUREMENT ALTERNATE FORM RELIABILITY COEFFICIENTS: RESULTS
10. AGE AND SOCIAL CLASS AS MODERATORS ON RELIABILITY COEFFICIENTS
r
IT. TROD` TION AGE AS A MODERATOR ON RELL, %BILITY COEFFICIENTS : RESULTS SOCIAL CLASS AS A MODERATOR ON RELIABILITY COEFFICIENTS: RESULTS AGE F.,. D SOCIAL CLASS AS MODERATORS CN RELIABILITY COEFFICIENTS:
DISCUSSION
11. VARLL. 1 . CNS IN THE STANDARD ERROR OF :, IEAS, REMENT
INTRODUCTION AGE AND SOCIAL CLASS AS MODERATORS C: 7 THE SEMS: RESULTS AGE L, D SOCIAL CLASS AS MODSPJ. TORS C THE SEMS: DISCUSSION RELIABILITY AND THE PRIMARY SOURCE TRAITS: CONCLUSIONS
Page
115
115
116 118 119 119
129
132 136
139
140 142 166 167 168 184 185
187
187
188 190 191 192
193 195
202
203 204 208
210
214
21.5 223 223 232
2
Pace
SECTION 5: INTERRZAL CONSISTENCY AND TTI-E PRIMARY SOURCE TRAITS 235
12. INTRODUCTION 236
THE PROBLEM OF REPLICATING CATTELL'S FACTOR SOLUTION 238 THE
-PROBLEMS GF RESPONSE FORMAT AND CORRELATIONAL METHODS 241 THE HSPQ STUDY 249
13. ESTI: ": ATTIýNG THE DIRECT VALIDITIES OF TIHE PRIMARY SOURCE TRAITS 258
I_ ODuc_Ia. ' 259 I `, AL1y ES OF TEE SOURCE .t AITS: PROCEDURE 251
RESULTS AI: ý SCÜSSICPý 262 SUi"iYARY AND CCiNCLUSIONS 269
14. FACTOR ANALYSIS AND THE PRIMARY SOURCE TRAITS 271
INTt CDLCTION 272 FACTOR ANALYSIS: RESULTS AND DISCUSSION 278 CONCLUSIONS 280
SECTION 6: SUiO1aP`_' AND CONCLUSIONS 281
15. SUi1NL. RY AND CC CLUSIONS 281
SU"21-. RY OF FI (DINGS ON THE PRIMARY SOURCE TRAITS 282 CONCLUSIONS ON GROUP DIFFERENCES AP 322 CONCLUSIONS REGARDING THE RELIABILITY OF THE PRIMARY SOURCE TRAITS 326 CONCLUSIONS ON THE INTERNAL CONSISTENCY OF THE MODEL 327 CONCLUDING RE. -, a. RKS 329 AREAS CF FURT}ER RESEARCH 330
16. REFERENCES
SMTTICN 7: APPENDICES
332
1-142
3
LIST0FTABLES
Pa e
SECTION 1: GENERAL INTRODUCTION
1. INTRODUCTION
1.1. Example of a Hypothetical 5x5 Correlation Matrix 18
. 2. First O- er (2- of 22
1.3. Second Order Factors of the 16PF 23
SECTION 2: HYPOTHESES, i. ': ETHODOLOGY, PROCEDURE, THE SAMPLE
2. AIMS AND HYPOTHESES
2.1. Sex and Social Group Differences and Age Trends in the Primary Source Traits Hypothesised from Cattell's Personality Model 68
5. DESCRIPTION OF THE SAMPLE
5.1. The 1971 British Standardisation Sample 110
SECTION 3: GROUP DIFFERENCES AND THE PRIMARY SOURCE TRAITS
6. CROSS-NATIONAL AND SEX DIFFERENCES
6.1. Significant Personality Differences Between the American and British Adult Groups Adjusted to Age 30 Years: Females 125
6.2. Significant Personality Differences Between the American and British Adult Groups Adjusted to Age 30 Years: Males 126
6.3. Significant Personality Differences Between British Men and Women 127
6.4. Summary of Cross-National and Sex Differences on Cattell's Primary Source Traits 128
7. SOCIAL CLASS DIFFER iNCES
7.1. The Sixteen Personality Questionnaire: Correlations Between Social Class and First Crder Factors 131
4
Page
7.2. Eta Values (n) Regressing the Primary Source Traits on Socio-Economic Group: Females: Forms A+B 133
7.3. Eta Values (n) Regressing the Primary Source Traits on Socio-Economic Group: Males: Forms A+B 135
7.4. Correlations Between rge, Educational Age and Social Class in the 16PF: Forms A+B Adult Sample: Both Sexes 137 (N = 2012)
8. "GE DIFFERENCES
6.1. -Ste Values o-)n the rimary Source Traits For Six Age Groups Derived from Cattell's Age Curves 143
8.2. Conversion of N-Sten Age Curve Values to Z Scores Based on Total Group Statistics 144
8.3. The Sixteen Personality Questionnaire: Correlations Between Age and First Order Factors 162
8.4. Eta Values (n) Regressing The Primary Source Traits on Age: Females: Forms A+B 163
8.5. Eta Values (n) Regressing the Primary Source Traits
on Age: Males: Forms A+B 164
8.6. 16PF Age Trends in Personality Form A+ B Males Age Corrected Scores 169
8.7.16PF Age Trends in Personality Form A+B Females Age Corrected Scores ýº 170
8.8. Eta Values (n) For Age Corrected Scores Form A+B Females 171
8.9. Eta Values (n) For Age Corrected Scores Form A+B Males 172
SECTION 4: RELIABILITY AND THE PRIMARY SOURCE TRAITS
9. ALTERNATE FORM RELIABILITY OF THE PRIMARY SOURCE TRAITS
9.1. Group Alternate Form Reliability Coefficients (Forms A&B) For Each Primary Source Trait 196
9.2. Standard. Errors of Measurement: Primary Source Traits (Stens) - Forms A/B 197
9.3. Conversion of the Alternate Form Reliability Coefficients to Fisher Z Statistics: Adult Males and Females 199
9.4. Comparison of Adult Reliability Coefficients urith Those of Undergraduates: Form A with B 200
5
Pa e
10. AGE AND SOCIAL CLASS AS MODEFATORS ON RELIABILITY COED FICIENTS
10.1. Rank Frequencies of the Alternate Form Reliability Coefficients Classified by Age Group 204
10.2. Summary Table: Reliability Classified by Age 205
10.3. Conversion of Primary Source Trait Reliability Coefficients to Fisher Z Statistics: Both Sexes by Age 206
10. _.
Rar t: Fe '1cies .. _ e Alt? rra: e ý'o m tRc2llab_ý"i f_y
Ccetficients Classified by Socio-Econor^ic Group 207
10.5. Social Class as a Moderator on the Reliability of the Primary Source Traits: Summary Table 208
10.6. Conversion of Primary Source Trait Reliability Coefficients to Fisher Z Statistics: Both Sexes by Social Class 209
11. VARIF. TIONS IN THE STANDARD ERROR OF MEASUREMENT
11.1.16PF Form A SEMs According to Age and Social Class: Both Sexes 225
11.2. 16PF Form B SEMs According to Age and Social Class: Both Sexes 226
11.3. SEM Rank Frequencies for Social Class: Both Sexes: Form A 227
11.4. SEM Rank Frequencies for Social Class: Both Sexes: Form B 227
11.5. Summary Tables: SEMs and Social Class as a Moderator Variable: Forms A and B 228
11.6. SEM Rank Frequencies For Age: Both Sexes: Forms A and B 229
11.7. Summary Tables: SEMS and Age as a Moderator Variable: Forms A and B 230
SECTION 5: INTERNAL CONSISTENCY AND THE PRIMARY SOURCE TRAITS
12. INTIRODIXTION
12.1. Correlation Between the Point-Biserial and Biserial Approaches to HSPQ Item Scale Correlations 252
12.2. High School Personality Questionnaire: Form A, Girls NL mber of Higher Extra than Intra Scale Correlations for Each Factor (Biserials) 254
12.3. High School Personality Questionnaire: Form A, Girls Number of Higher Extra than Intra Scale Correlations = for Each Factor (Product Moment) 255
6
Pace
12.4. High School Personality Questionnaire: Form A, Girls Correlation Between Product Moment and Biserial Approaches (All Items) 257
12.5. High School Personality Questionnaire: Form A, Girls Correlation Between Product Moment and Biserial Approaches (Factors) 257
13. ESTIMATING T}E DIRECT VALIDITIES OF THE PRIMARY SOURCE TRAITS
13.1. Composite Item-Scale Correlations (Direct Validities) For Cattell's Primary Source Traits 263
13.2. Correlations Between the Source Traits of the 16PF as Scale Scores: Forms A and B 266
13.3. Comparison of the Item-Composite Correlations with Scale Intercorrelations 267
14. FACTOR ANALYSIS AND THE. PRIMARY SOURCE TRAITS
14.1.16PF Factor Analysis: Males: Eigen Values and Percentage of Trace 274
14.2.16PF Factor Analysis: Females: Eigen Values and Percentage of Trace 274
14.3. Promax Factor Analysis: Males: Table of Factor Loadings 275
14.4. Promax Factor Analysis: Females: Table of Factor Loadings 276
14.5. Correlations of the Oblique Primary Factors: Males 40 277
14.6. Correlations of the Oblique Primary Factors: Females 277
Slr-TION 6: SUMMARY AND CONCLUSIONS
15. SUMMARY AND CONCLUSIONS
15.1. Consistency of Sex, Social Class and Age Differences on Cattell's Source Traits fgr the Seven Factor Solution 323
7
LIST0FFIGURES
SECTION 1: GENERAL INTF%OD'1CTION
1. INTRODUCTION
Pacge
1.1. Cattell's Factor Theory of Personality 24
1.2. A Graphical Representation of Cattell's Concept of 'Balance Seecifics' in Scale Construction 63
SECTION 2: HYPOTHESES, ý"IETHODOLOGY, PROCEDURE, THE SAMPLE
2. AIMS AND HYPOTHESES
2.1. Abstracted Personality - Age Relationships for each of Cattell's Primary Source Traits 69
2.2. A Study of Reliability and Structure in Cattell's Model of Adult Personality 74
ý. PROCEDURE
4.1. Contact Sheet 85
4.2. Background Questionnaire 86
4.3. The Sampling Units and Standard Regions of the 16PF: British Standardisation Sample (N = 2227) 107
SECTION 3: GROUP DIFFERENCES AND THE PRIMARY SOURCE TRAITS
8. AGE DIFFERENCES
8.1. Analysis of Personality Age Curves 141
8.2. - 8.17. Age Trends in Personality - Factors A- Q4 146 - 161
SECTION 5: INTERNAL CONSISTENCY AND THE PRIMARY SOURCE TRAITS
12. INTRODUCTION
12.1. Item Analysis Procedure 251
I
8
ACKNOWLEDGEMENTS
I would very much like to thank Janice Field, Clive Downs and
Iris Parry of the Test Department of the National Foundation for
Educational Research for all their help in this project. I would
also like to thank B-111 .: bey. Joan MacFarlane Smith and
Sue Rodger, previously of the British Market Research Bureau
for their contribution to the fieldwork organisation.
Thanks must go to Jill Tarryer, previously of the NFER, for
her statistical advice and to her staff for dealing with the
large quantity of data generated by this study. I would also
like to acknowledge discussions with Steve Blinkhorn, who shares
a common interest in personality measurement.
Very sincere thanks are given to Irene Stevens for all her
excellent work in typing this thesis and for remaining patient
through the various drafts and amendments which this thesis
has undergone.
Finally, I would like to state my thanks to Professor Furneaux
and Dr Ormerod of the Education Department at Brunel University
for all their help and advice.
AP
it
9
ABSTRACT
The personality theory expounded by R. B. Cattell and his
collaborators is one of the most influential models of
personality in modern-day psychology. The central and most
controversial elements of this theory are the sixteen primary
source traits, which make up the model.
This thesis describes research into the primary source traits
as represented by Cattell's Sixteen Personality Factor
Questionaire (16FF). From the responses of a large sample of
general population British adults (N: - 2,000) evidence is
presented to suggest that: -
1. inconsistencies exist between the relationships of Cattell's
personality scales with variables such as age and social class
and the relationships postulated from his model.
2. the precision of measurement afforded by certain scales is
extremely low, especially in lower social class subjects.
3. the model as represented in the scales of the 16PF is
untenable in terms of its internal consistency.
4. Cattell's personality scales are not in fact the most
parsimonious description of the personality domain.
SECTION 1:
GENERALINTRODUCTION -------------------
0
1. INTRODUCTION
1ll
THE PERSONALITY QUESTIONNAIRE
It is generally believed that Francis Galton devised the first
questionnaires in the 1880s for his studies of mental imagery at his
Kensington laboratory. His statement in the 'Fortnightly Review' of
1884 that 'the character which shapes our conduct is a definite and
durable "something" and therefore .... it is reasonable to attempt to
measure it' is not unlike many contemporary views of personality.
Despite the problems associated with the questionnaire approach to
personality measurement - one is reminded of the comment of Sir
Frederick Bartlett reported by Buzzard (1971), 'I don't know ... they
may be all right ... they always seem to me to overestimate the
self-knowledge of the subject and to underestimate his sense of
humour' - the technique flourished after World War I, until, in the
1970s there are literally hundreds of published instruments.
Most self-report personality questionnaires consist of a series of
statements or questions about the individual, his interests and attitudes.
These the subject endorses according to his strength of agreement. To
make scoring easier and more dependable, the subject is usually
required to select the most appropriate answer from choices provided
in the questionnaire form. Because item formats will concern us at a
later stage we shall briefly consider how items are typically presented.
First, there is the true/false technique of, for example, the Eysenck
Personality Inventory, where the subject has to choose between two
clear-cut alternatives.
12
I enjoy going to parties.
a. b.
Yes No
This approach has the advantage of being statistically manageable (cf
the pass/fail dichotomy cn an ability test item) and helps to overcome
certain types of bias. Ho. ", ever, it can become frustrating to the
; 'ho may find -t both di ficu_: l t and unnatural to answer
every item on a yes/no basis. There is the danger that the item
format will cause the subject to challenge the sense of questions
being asked him, and so become inconsistent in his answers. Indeed,
it is not unknown for respondents to invent a middle response category.
A second approach is the use of the forced choice technique, where
one or more of a number of alternatives are chosen by the subject as
most or least characteristic of himself. For example: AP
Mark the statement which shows which you prefer doing:
a. Going to parties.
b. Reading a serious book.
Inventories based on this system are the Gordon Personal Profile, the
Gordon Personal Inventory and the Edwards Personal Preference Schedule
(EPPS). One attraction of this approach is that by matching the items,
social desirability response sets (the tendency for subjects to put
up a good front and show socially acceptable characteristics) can to
some extent be controlled. In practice it has proved difficult to match
itens exactly in terms of social desirability and it appears that
instruments like the EPPS can in fact be deliberately faked to create
13
more favourable impressions (Dickens, 1959). It is probable also that
items matched in social desirability for one occupation or set of
circumstances will not be matched for other occupations or other sets
of circumstances. In addition, because certain personality types are
more prone to presenting answers in a favourable light, it has been
argued that attempting to eradicate the effects of social desirability
is like throwing the baby out with the bath water.
Another problem concerns scaling properties. The strength of each
variable is gauged relative to the strengths of all other variables
and as a result, it is impossible to gain high or low scores across
all scales. Cattell (1944) has termed this 'ipsative' measurement
and strictly we should not compare scores of different individuals,
although this is often done and extensive norm tables produced.
The third type of item format is an extension from the true/false
item to three or more possible choices. For example:
I like going to parties.
a. I,. C.
True False ?
or I like going to parties
a. b. Co
Frequently Occasionally Never
14
or I like going to parties.
a. b. c. d. e.
Always Frequently Occasionally Rarely Never
The first example is similar to the method of item presentation in
the Minnesota rultiphasic Personality Inventory (? 4PI) and the
secor. _. the personality questionnaires developed by RB Cattell.
This approach is more acceptable to the subject, although the
dimensionality of the middle response is often a problem, especially
if it becomes a waste-bin category of 'don't knows', 'in betweens',
'occasionally' and 'I don't understand the question'. In these
multiple choice formats the effects of faking and social desirability
are not so much controlled but detected by lie, motivational distortion
and other such keys. In the case where there are three choices the
responses become statistically less manageable, whilst five or more may
cause a practical problem in the hand marking of the questionnaires.
Obviously with most questionnaire items the prior assumption has
already been made that the underlying trait represents a continuum.
It is not usually acceptable for a subject to endorse two answers even
though he may feel that both are equally appropriate and it is this lack
of flexibility which many criticise in the questionnaire approach.
Goldberg (1970), corrsnenting upon the current proliferation of
personality inventories ('inventory' and 'questionnaire' will be used
interchangeably throughout this thesis), distinguished between those
inventories developed in response to pressures from society to deal with
specific problems and those stemming from theoretical concepts about the
15
nature of personality. In the first group are questionnaires like the
California Psychological Inventory (CPI), designed to measure practical
outcomes of social behaviour, and the Minnesota Multiphasic Personality
Inventory (MMPI), which measures aspects of personal adjustment, and
those which are concerned with vocational choice, such as the Strong
Vocational Interest Blank. Instruments of this kind tend to be
constructed on an empirical basis where items are selected which
discriminate between different groups of people.
The second category of questionnaire includes the inventory measures
of introversion-extraversion, those based on Murray's (1938) system
of needs like the Edwards Personal Preference Schedule, those which
are psychoanalytic in conception like the Dynamic Personality Inventory
and the various gqestionaires which have been based on factor analytic
methods of-construction.
THE FACTOR ANALYTIC APPROACH TO PERSONALITY MEASUREMENT
Many questionnaires have been produced, each, as Cronbach (1970) has
pointed out, using items from its predecessors, adding new ones and
scoring them in new combinations of traits and scales. Cattell (1973)
has written "... let us be clear that nothing stops anyone from making
up questionnaires for any trait he cares to imagine and since there are
an indefinitely large number of conceivable traits and roughly 50,000
psychologists, questionnaires have in fact been the prolific rabbits of
the psychomobic world! ".
16
Trait names become a source of much confusion. What was 'Introversion'
to one author was not introversion to another; 'Dominance' in system X
bore a striking resemblance to the trait of aggression in system Y.
It was the construction of what appeared to be arbitrary personality
scales that the factor analysts attacked. The only way to resolve
this Tower of Babel, the factor analytic school argued, was to
'discover' clearly defined personality variables, building blocks,
which would serve the psychologist as the table of elements serves
the chemist. Rather than have a multitude of scales to predict all
sorts of social, educational and clinical outcomes (the criterion -
related or empirical approach), is it not possible to find a basic
set of constructs by which we can understand human personality?
The factor analysts believe it is.
Although there is considerable variation within the school, the basic
principles are simple. Working with a set of questionnaire scales
which has been administered to a large group of subjects, it is possible
to measure the degree of association between the variables by correlating
every scale with every other scale. The correlations between each pair
of variables are represented in a correlation matrix as in the example,
Table I. I.
'r
Looking closely at Table 1.1, we can ask whether any of
these hypothetical variables are related to any degree with one another.
For example, do subjects consistently gain a high score on one scale
and not on another? The latter is clearly the case with scales R and S;
17
TABLE: 1.1. Example of a hypothetical 5x5 correlation matrix
VARIABLE
R S T X Y
R -. 92 . 13 -. 11 . 82
pq S -. 92 . 01 . 27 -. 86 T . 13 . 01 . 89 . 08 X -. 11 . 27 . 89 . 21 Y . 82 -. 86 . 08 . 21
the negative correlation of -0.92 means that there is a strong tendency
for those getting high scores on Scale R to get low scores on Scale S
and vice versa. The direction of the relationship is of no major
importance because it merely reflects the scoring conventions of the
scales, but what is important is the presence of a strong association
between the two variables. Indeed, the relationship is so high that
we might wonder whether these scales were in any way distinct and
whether R and S are simply different names for the same underlying
variable.
Visual inspection of this correlation matrix suggests the presence of
two underlying dimensions or 'factors' - the first common to variables
R, S and Y and the second to T and X. This kind of visual examination
of correlation matrices of course illustrates only the logic of factor
analysis, However it is in fact true to say that one of the most
sophisticated factor theories of personality, that developed by R. B.
Cattell, was to some extent formulated by such methods (Cattell 1943).
18
CATTELL'S MODEL OF PERSONALITY
There can be little doubt of the importance of RB Cattell and his
collaborators to current personality theory. Cattell's research into
personality has spanned more than thirty years of large scale study
and experiment and his writings have introduced not only new ways of
investigating the domain but also new ideas on the very structure and
functioning of personality. Howarth (1976) writes 'Cattell's long
continued and widely documented research into personality factors
represents the single most monumental effort to operationalize those
sources of individual differences often referred to as 'temperament'
or 'personality' (and is comparable to that of Guilford on cognitive
factors)". According to Berg (1972), "Cattell's major contribution
will undoubtedly be seen as the discovery of the major factorial
dimensions in rating, questionnaire and objective tests".
Cattell (1950) has provided us with a deliberately general definition
of personality as "that which permits a prediction of what a person
will do in a given situation. The goal of psychological research in
personality is thus to establish laws about what different people
will do in all kinds of social and general environmental situations ...
personality is ... concerned with all the behaviour of the individual,
both overt and under the skin".
19
Thus, Cattell's theory of personality stresses both the empirical and
descrip: -ive and is concerned with establishing a set of personality
eiemen s shared by all individuals. Rather than have a multitude
of scales to predict all sorts of social, educational and clinical
outcomes, Cattell set himself the task of discovering a basic set
of constructs by which human personalty could be fully understood.
It was in fact by way of behaviour ratings of one group of people
by another ('L' or 'Life data') that Cattell's major dimensions of
personality were derived (Cattell, 1943). Using the adjectives and
adverbs descriptive of behaviour listed in the English dictionary,
Cattell first defined. the clusters upon which his traits were to be
used by eliminating synonyms and visually inspecting correlation
matrices. Cattell then extended this work to self-ratings (questionnaires
or 'Q' data) and to objective tests (laboratory experiment or 'T' data). s
At the first level of Cattell's model of personality there are the
many surface traits of personality. Surface traits are directly
observable aspects of human behaviour, the labels which are frequently
used in our everyday experience. At the next level are the first order
or primary source traits of personality which are combined to produce
the complex and many observable surface traits. It is the primary
source traits which are the universal basic variables in the total
personality structure. They are defined both by letters of the
alphabet, technical names and popularly descriptive labels.
The order of the primaries (which are labelled from A through to Q4)
are based on their descending contribution to behavioural variance
20
as estimated from Cattell's studies. Whilst Cattell refers to the
first order factors as functionally unitary, they are oblique
factors, ie. they show some correlation between themselves.
Because the primary source traits are in fact correlated, Cattell
has factor analysed his variables to some eight second crder factors.
ii list o; these and the pr m r`y' source tja "`_s is contained in Tables 1.2.
and 1.3.. In addition, Figure 1.1. shows the relationship between the
primary traits and the major second order factors.
To Cattell the problems of studying personality are extricably wrapped
up with the problems of measurement in psychology. In early studies
he used correlational techniques on personality ratings ('L' data)
but as computer technology became more developed, his ideas and
theories rested more and more on the technique of factor analysis AP
(leading at times to bitter argument with other members of the factor
analytic school, Cattell, 1972 and Eysenck, 1972). Still later work
was with self-ratings into what has become the more heavily researched
area of questionnaire measurement and it is the personality questionnaire
which has become the most developed vanguard of his model. Certainly,
Cattell's personality scales have had very wide applicability in the
western world, only rivalled by the '? IPI and the measures developed
by Eysenck.
t
I
21
TABLE: 1.2.
FIRST ORDER (PRI -ýSRY; FACTORS OF THE 16PF
LOW SCORE DESCRIP'ION Dr^ SC T HIGH SCORE
I, -z DESCRIPTION
RESERVED, DEIACrtED, CRITICAL, OUTGOING, WARMHEARTED, EASY- ALOOF A GOING, PARTICIPATING
(Sizothymio) (Affectothymia, formerly cyclothymic)
LESS INTELLIGENT, CONCRETE. MORE INTELLIGENT, ABSTRACT- THINKING
' B THINKING, BRIGHT (Lower scho csric mental capacity) (Higher scholastic mental capacity)
AFFECTED BY FEELINGS, EMOTIONAL- EMOTIONALLY STABLE, FACES LY LESS STABLE, EASILY UPSET C REALITY, CALM, MATURE
(Lower ego strength) (Higher ego strength)
HUMBLE, MILD, ACCOMMODATING, ASSERTIVE, AGGRESSIVE, STUBBORN, CONFORMING E COMPETITIVE
(Submissiveness) (Dominance)
SOBER, PRUDENT, SERIOUS TACITURN HAPPY-GO-LUCKY, IMPULSIVELY , (Desurgency) F LIVELY, GAY, ENTHUSIASTIC
(Surgency)
EXPEDIENT, DISREGARDS RULES CONSCIENTIOUS, PERSEVERING. , FEELS FEW OBLIGATIONS
' " STAID., MORALISTIC
(Weo <er superego strength) (Stronger superego strength)
SHY, RESTRAINED, TIMID, VENTURESOME, SOCIALLY BOLD, 4P- THREAT-SENSITIVE H UNINHIBITED, SPONTANEOUS
(Threctio) (Pormio)
TOUGH-MINDED, SELF-RELIANT, TENDER-MINDED, CLINGING, REALISTIC, NO-NONSENSE I OVER-PROTECTED, SENSITIVE
(Harris) (Premsic)
TRUSTING, ADAPTABLE, FREE OF SUSPICIOUS, SELF-OPINIONATED, JEALOUSY, EASY TO GET ALONG L HARD TO FOOL WITH (Ataxia) (Protension)
PRACTICAL, CAREFUL, CONVENTION- IMAGINATIVE, WRAPPED UP IN INNER AL, REGULATED BY EXTERNAL URGENCIES, CARELESS OF PRACTICAL REALITIES, PROPER (Proxernio) (Autio) MATTERS, BOHEMIAN
FORTHRIGHT, NATURAL, ARTLESS, SHREWD, CALCULATING, WORLDLY, UNPRETENTIOUS N PENETRATING
(Artlessness) (Shrewdness)
SELF-ASSURED, CONFIDENT, APPREHENSIVE, SELF-REPROACHING, SERENE Q WORRYING, TROUBLED
; J-troubled adequacy) (Guilt proneness) CONSERVATIVE, RESPECTING ESTAB- EXPERIMENTING, LIBERAL LISHED IDEAS, TOLERANT OF TRADI- T
, ANALYTICAL. FREE-THINKING IONAL DIFFICULTIES (Conservatism) (Radicalism)
GROUP-DEPENDENT, A "JOINER" AND SELF-SUFFICIENT, PREFERS OWN SOUND FOLLOWER Q2 DECISIONS, RESOURCEFUL'
(Group adherence) (Self-sufficiency)
UNDISCIPLINED SELF-CONFLICT, FOL- CONTROLLED, SOCIALLY PRECISE, LOWS OWN URGES, CARELESS OF 43 FOLLOWING SELF-IMAGE PROTOCOL (Low integration) (High self-concept control)
RELAXED, TRANQUIL, TENSE, FRUSTRATED, DRIVEN, UNFRUSTRATED Q4 OVERWROUGHT
(Low ergic tension) (High ergic tension)
22
TABLE: 1.3.
SECOND ORDER FACTORS OF THE 16PF
SECOND CHIEF FIRST LOW SCORE ORDER HIGH SCORE ORDER FACTORS
DESCRIPTION FACTORS DESCRIPTION INVOLVED
INTROVERSION EXTRAVERSION A+, E+, F+, Invia QI Exvia H+, Q2-
LOW ANXIETY HIGH ANXIETY C-, H-, L+, O+, Adjustment QII Anxiety Q3-, Q4+
SENSITIVITY TOUGH POISE A_, I-, M- Pathelria QIII Cortertia
, E+, L+.
DEPENDENCE INDEPENDENCE E+, L+, M+, Subduedness QIV Independence Q1+, Q2+
NATURALNESS QV DISCREETNESS N+, A+, M.., O-
COOL PRODIGAL REALIST QVI SUBJECTIVITY I+, M+, L-.
LOW HIGH INTELLIGENCE QVII INTELLIGENCE B+
LOW SUPEREGO HIGH SUPEREGO STRENGTH QVIII STRENGTH G+, Q3+, F-
23
aw
HW as
N Self-
Cj sufficient
H
ö
En
ii
w c7 CH++
Ü
c.., Happy-go--lucky
41 Assertive
d Outgoing
H Tender-minded
Experimenting
0 Conscientious
z Shrewd
z Imaginative
ai Intelligence
x Venturesome
Emotional Stability
Tense
o Apprehensive
A suspicious
`"ý Controlled
24
L! J1
ýa w ýý Ha
4J H
N
aj
N N
8
I
DESCRIPTION OF CATTELLtS PRIMARY SOURCE TRAITS
As is to be expected with a trait psychologist, Cattell's theory is
very much rooted in the constructs he uses in describing individual
differences in personality.
We shall see that Cattell borrows heavily from a long list of other
psychologists to describe the traits derived from his factor analytic
research. He makes reference for example to the work of James, Freud,
Jung, McDougall, Thurstone, Maslow, Eysenck and Guilford. Indeed Cattell
frequently uses nautical expressions and analogies, seeing himself and
his researchers as modern-day explorers discovering and re-discovering
concepts and ideas observed by previous less well equipped voyagers.
This is not to say that Cattell denies the possibility of traits
unique to the individual but he believes that the primary goal of
psychology is to discover and measure the core of primary
source traits common to all individuals. Once this is achieved, it
will then be possible to observe the inter-relationships of such
traits which make each individual unique and which will allow actual
predictions of their likely behaviour under different environmental
conditions.
FACTOR A- RESERVED VS. OUTGOING
(a) Low Score Description
Reserved, Detached, Critical, Cool (Sizothymia, previously Schizothymia)
The person who scores low (step of 1 to 3) on Factor A tends to be
stiff, cool, skeptical and aloof. He likes things rather than people,
25
to be precise and 'rigid' in his way of doing things and in personal
standards, and in many occupations these are desirable traits. He
may tend, at times, to be critical, obstructive, or hard.
(b) High Score Description
Outgoing, Warmhearted, Easy-going, Participating (Affnctothymia, previously Cyclothy ia)
The person who scores high (sten of 8 to 10) on Factor A tends
to be goodnatured, easy-going, emotionally expressive (hence
naturally Affectothymia), ready to cooperate, attentive to
people, softhearted, kindly, adaptable. He likes occupations
dealing with people and socially-impressive situations. He
readily forms active groups. He is generous in personal
relations, less afraid of criticism, better able to remember
names of people. Ap
(c) Origins in Psychological Theory
Cattell traces this concept back to the cyclothymia-versus-
schizothymia dichotomy of Kretschmer (1921) which in turn was
adapted from Bleuler's (1911) cyclic-versus-schizophrenic label
of the functional psychoses. In rating studies and questionnaire
studies this is the largest variance factor.
(d) Life Associations
Cattell claims that affection is associated with a warm home
background. Higher scorers on affectia get married earlier,
usually stay married longer and have an endomorphic bodybuild.
Women score significantly higher than men. A fairly substantial }
genetic component is hypothesised for this trait.
26
(e) Age Trend
Cattell states that this trait decreases until about 10 years of
age; then there is a steady and relatively rapid increase to 30
, rohere it levels off.
FACTO ? :3- LESS Ts'TELL:: GES i VS. MORE INTELLIGENT
(a) Low Score Description
Less Intelligent, Concrete-Thinking (Lower scholastic-mental capacity)
The person scoring low on Factor B tends to be slow to learn and
grasp, dull, given to concrete and literal interpretation. His
dullness may be simply a reflection of low intelligence, or it
may represent poor functioning due to psychopathology.
r
(b) High Score Description
More Intelligent, Abstract-thinking, Bright (Higher scholastic mental capacity)
The person who scores high on Factor B tends to be quick to grasp
ideas, a fast learner, intelligent. There is some correlation
with level of culture, and some with alertness. High scores
contraindicate deterioration of mental functions in pathological
conditions.
(c) Origins in Psychological Theory
Cattell's view of intelligence remains very similar to the
hierarchical theories of British psychologists like those of
Spearman and Vernon (Cattell 1965). He has certainly not rejected
27
the meaning of an overall processing factor ("g") although he
has spent much time in trying to isolate two main influences
within such a construct. These he has named fluid (processing
ability unaffected by cultural influences) and crystallized
(intelligence gained as a result of learning experiences)
general ability.
(d) Life Associations
Cattell (1965) believes that some intelligence is important to
predicting behaviour in a great many situations such as future
academic and occupational performance.
(e) Age Trend
Following increases up to adolescence, fluid ability then starts
to decline whilst crystallized ability can improve until about
40 years of age when it declines.
FACTOR C- AFFECTED BY FEELINGS VS. EMOTIONALLY STABLE
(a) Low Score Description
Affected by Feelings, Emotionally Less Stable, Easily Upset (Lower ego strength)
AP
The person who scores low on Factor C tends to be low in frustration,
tolerance for unsatisfactory conditions, changeable and plastic,
evading necessary reality demands, neurotically fatigued, fretful,
easily emotional and annoyed, active in dissatisfaction, having
neurotic symptoms (phobias, sleep disturbances, psychosomatic
complaints, etc. ). Low Factor C score is common to almost all
28
I
forms of neurotic and some psychotic disorders.
(b) High Score Description
Emotionally Stable, Faces Reality, Calm, Mature (Higher ego strength)
The person who scores high on Factor C tends to be emotionally
ma-tu-ne, stable, realistic about life, unruffled, possessing ego
strength, better able to maintain solid group morale. Sometimes
he may be a person making a resigned adjustment to unsolved
emotional problems.
(c) Origins in Psychological Theory
Cattell claims that this pattern in questionnaire items is clearly
the reality behind Freud's concept of ego structure strength.
v (d) Life Associations
Low ego strength we are informed by Cattell is associated with
unemployment, truency and delinquency; high scores go with group
participating and skilled occupations. It is low in those from
family backgrounds of matriarchal dominance and high among those
who are disciplined by reasoning as opposed to punishment by the
father. Low ego strength is also shown in some chronic physical
disabilities and in marital instability.
(e) Aqe Trend
Increasing strongly with age and then levels off at about 40 years.
29
FACTOR E- HTJ N. B: ýE VS. ASSERTIVE
(a) Low Score Descriotion
Ht,; ble, Mild, Accomumodating, Conforming (Submissiveness)
The person who scores low on Factor E tends to give way to others,
to b= docile, and to conform. He is often dependent, confessing,
anxious for obsessional corr`ctnoss. This passivity is part of
many neurotic syndromes.
(b) High Score Description
Assertive, Independent, Aggressive, Stubborn (Dominance)
The person who scores high on Factor E is assertive, self-assured,
and independent-minded. He tends to be austere, a law to himself,
hostile or extrapunitive, authoritarian (managing others), and W
disregards authority.
(c) Origins in Psychological Theory
This trait has a long history (Maslowl, 1940,1954). It has a
second tributary in animal ethology from McDougall (1932).
(d) Life Associations
Those high on this factor are seen as unconventional and rule
breaking in group behaviour. There are significant relations we
are told to an early authoritarian family atmosphere and to
stricter application of discipline. Among 18-22 year olds there
is a tendency to reject sexual restraints (Maslow, 1940), to have
30
a great number of social contacts, to drink immoderately and to
have less than average sleep needs. This factor is high in
psychopaths and low in neurotics, alcoholics, drug addicts, all
psychotics except manics, in stably married wives and unstably
married husbands (Cattell and Nesselroade, 1968). Males score
significantly higher than females on this factor.
(e) Aqe Trend
Increasing rapidly in early life, then levels off and declines
in the middle years.
FACTOR F- SOBER VS. HAPPY-GO-LUCKY
(a) Low Score Description
Sober, Prudent, Serious, Taciturn (Desurgency)
The person who scores low on Factor F tends to be restrained,
reticent, introspective. He is sometimes dour, pessimistic,
unduly deliberate, and considered smug and primly correct by
observers. He tends to be a sober, dependable person.
(b) High Score Description
Happy-go--Lucky, Impulsively Lively, Gay, Enthusiastic (Surgency)
The person who scores high on this trait tends to be cheerful,
active, talkative, frank, expressive, effervescent, carefree.
s
He is frequently chosen as an elected leader. He may be impulsive
and mercurial.
31
(c) Origins in Psychological Theory
Cattell points out that in early years this was confused with
his second order factor of exvia (extraversion). In Cattell's
model, surgency is a distinct component, similar to one of the
four components in extraversion conjectured by Jung.
(d) Life AGscx-iations
Cattell (1973) hypothesises that this trait is related to high
economic status of the family but fewer work responsibilities.
There is low restraint in sexual activities, there are many
social contacts and in small groups a high scoring person is often
voted an effective speaker and leader. There will be appreciable
cultural differences on this trait we are informed: for example,
it is high in Australians and low in the Chinese and Japanese.
0 (e) Age Trend
Rising in adolescence but decreasing in adulthood.
FACTOR G- EXPEDIENT VS. CONSCIENTIOUS
(a) Low Score Description
Expedient, Evades Rules, Feels Few Obligations (Weaker superego strength)
The person who scores low on Factor G tends to be unsteady in
purpose. He is often casual and lacking in effort for group
undertakings and cultural demands. His freedom from group influence
may lead to anti-social acts, but at times makes him more effective,
32
while his refusal to be bound by rules causes him to have less
somatic upset from stress.
(b) High Score Description
Conscientious, Persevering, Staid, Rulebound (Stronger superego strength)
The person who scores high on Factor G tends to be exacting in
character, dominated by sense of duty, persevering, responsible,
planful, 'fills the unforgiving minute'. He is usually
conscientious and moralistic and he prefers hard-working people
to witty companions. The inner 'categorical imperative' of this
essential superego (in the psychoanalytic sense) should be
distinguished from the superficially similar 'social ideal self'
of Q3+.
" (c) Origins in Psychological Theory
The concept of conscience, Cattell points out, is dateless; it is
discussed by Hebrew prophets and Greek dramatists. Of course this
was an important construct in Freud's personality system but
Cattell claims that the factor-analytic evidence of a distinct
and measurable unitary structure in questionnaire data is only
about twenty years old.
(d) Life Associations
Higher superego strength is related to such life associations
as: election to leadership in small groups, higher sexual morals,
having suffered a bereavement, working long hours and being
dedicated to some form of work. In the home background high G
33
is related to more years of education of parents and greater
warmth of the parental relationship. Women tend to score
significantly higher than men on this factor.
(e) Aqe Trend
Moderate but decelerating increase with age.
FACTOR H- SHY VS VENTURESOME
(a) Low Score Description
Shy, Restrained, Diffident, Timid (Threctia)
The person who scores low on this trait tends to be shy, with-
drawing, catious, retiring, a 'wallflower'. He usually has
inferiority feelings. He tends to be slow and impeded in speeh
and in expressing himself, dislikes occupations with personal
contacts, prefers one or two close friends to large groups, and
is not given to keeping in contact with all that is going on
around him.
(b) High Score Description
Venturesome, Socially-bold, Uninhibited, Spontaneous (Parmia)
The person who scores high on Factor H is sociable, bold, ready
to try new things, spontaneous, and abundant in emotional response.
His 'thick-skinnedness' enables him to face wear and tear in dealing
with people and gruelling emotional situations, without fatigue.
34
However, he can be careless of detail, ignore danger signals,
and consume much time talking. He tends to be 'pushy' and actively
interested in the opposite sex.
(c) Origins in Psychological Theory
Like surgency this factor has been mistaken for extraversion but
Cattell claims that it has been isolated in questionnaire data
as a sub-factor of exvia since the late 1940s. Parmia is an
acronym to express the theory that it represents parasympathetic
predominance at the Hi- pole and sympathetic system predominance
at the H- pole. The latter amounts to high susceptibility to
threat - hence threctia.
(d) Life Associations
This factor is associated positively with the demonstration of W
affection by parents and negatively with strict discipline
(Barton, Cattell, 1972). High H is significantly related to
numerous social contacts, a rating as cooperative and friendly
as a child but lazy and unorganised, an indication of association
with endomorphic or mesomorphic, pyknic build. Parmia tends to be
high in psychopaths and moderately young delinquents but it is
low in alcoholics, neurotics and attempted suicides. Men on
average tend to score higher than women.
(e) Age Trend
The trait increases strongly through middle age for men but only
slightly for women.
35
FACTOR I- TOUGH-MINDED VS. TENDER-MINDED
(a) Low Score Description
Tough-minded, Self-reliant, Realistic, No-nonsense (Harries)
The person who scores low on Factor I tends to be practical,
realistic, masculine, independent, responsible, but skeptical of
subjective, cultural elaborations. He is sometimes unmoved, hard,
cynical, smug. He tends to keep a group operating on a practical
and realistic 'no-nonsense' basis.
(b) High Score Descriptioli
Tender-minded, Dependent, Overprotected, Sensitive (Premsia)
The person who scores high on Factor I tends to be tender-minded,
day-dreaming, artistic, fastidious, feminine. He is sometimes v
demanding of attention and help, impatient, dependent, impractical.
He dislikes crude people and rough occupations. He tends to slow
up group performance, and to upset group morale by unrealistic
fussiness.
(c) Origins in Psychological Theory
Cattell points out that the concept of a dimension of this kind is
even older than extraversion and goes back to the tender-versus-tough
minded trait of William James and earlier sources traced by Roback
(1927). As a demonstrably distinct questionnaire data factor it
was replicated in the mid-1940s by Cattell and soon after as a
pattern in attitudes by Eysenck (1952). This factor, it is claimed,
is particularly important in terms of the behavioural variance it
36
accounts for in children. In some studies the premsia-versus-harria
pattern has been called masculinity-versus-femininity but Cattell
claims that this is a gross oversimplification in that only a
fraction of the male-female difference resides in I.
(d) Life Associations
Pransia has significant positive relations to lax discipline
in the home background, warmth in permitting free aggression
and more control by reasoning than by punishment. It is also
associated with poor health, chronic illness, hypochondria and
low athletic interest and performance. This factor also distinguishes
those who go to higher education, take. up social responsibilities
and who enter the professions. Women score higher than men in
premsia.
(e) Age Trend
Steep decrease into adolescence followed by slight continuous
rise throughout life.
FACTOR L- TRUSTING VS. SUSPICIOUS
(a) Low Score Description
Trusting, Adaptable, Free of Jealousy, Easy to Get on With (Alaxia)
410.
The person who scores low on Factor L tends to be free of jealous
tendencies, adaptable, cheerful, un-competitive, concerned about
other people, a good team worker.
i
37
(b) High Score Description
Suspicious, Self-opinionated, Hard to Fool (Protension)
The person who scores high on Factor L tends to be mistrusting and
doubtful. He is often involved in his on ego, is self-opinionated
and interested in internal, mental life. He is usually deliberate
in his actions, unconcerned about other people, a poor team member.
(c) Origins in Psychological Theory
In his survey of questionnaire factors Cattell (1946. ) found several
suggestions of a dimension resembling paranoid introversion. In
the questionnaire realm he defines this factor as a general use
of projection as a defence mechanism. The low end of the pole,
alaxia, implies the quality of resigned relaxation and acceptance
of a socially low expression of individuality. AW
(d) Life Associations
As a child the protensic person had a high confidence in himself,
found school unpleasant but felt superior. In adulthood he is often
a strong drinker, does not marry early. and tends to be socially
unpopular. This factor is high in criminals, drug addicts,
neurotics and attempted suicides. Males tend to score higher
than females.
(e) Age Trend
Declines entering the middle years but increases appreciably
after forty.
38
FACTOR M- PRACTICAL VS. IMAGINATIVE
(a) Low Score Description
Practical, Careful, Conventional, Regulated by External Realities, Proper (Praxernia)
The person who scores low on Factor M tends to be anxious to do
the riggt things, attentive to orac tical matters, and subject to
the dictation of what is obviously possible. He is concerned
over detail, able to keep his head in emergencies, but sometimes
unimaginative.
(b) High Score Description
Imaginative, Wrapped up in Inner Urgencies, Careless of Practical Matters, Bohemian (Autia)
The person who scores high on Factor M tends to be unconvention, a,
unconcerned over everyday matters, Bohemian, self-motivated,
imaginatively-creative, concerned with 'essentials', and oblivious
of particular people and physical realities. His inner-directed
interests sometimes lead to unrealistic situations accompanied
by expressive outbursts. His individuality tends to cause him
to be rejected in group activities.
(c) Origins in Psychological Theory
Cattell claims that this source trait had no early recognition
clinically but only became evident through his factor analytic
work in the 1940s. Those high in autia have been hypothesised
to produce a greater intensity of images and ideas relative to
39
sensory stimuli. In essence their behaviour is one of accepting
an intense inner subjective life relative to the reality of the
outer world and its demands.
(d) Life Associations
Autious people are likely to choose further and higher education
rather t-hen a job, take part in social organisations but are
dissenters and are rejected by groups although they are called
significant members. There is some history of delicate health.
Their parents disapproved of their sex life and companions;
they lack regard for order and are untidy in dress. This factor
appears in scientists and writers. They are accident-prone and
often change jobs. Culturally, the latin countries are more
autious than, for example, Britain and Australia. In American
groups females tend to score higher than males on this factor. AW
(e) Aqe Trend
Increases in early life and then declines.
FACTOR N- FORTHRIGHT VS. SHREWD
(a) Low Score Description
Forthright, Natural, Artless, Sentimental (Artlessness)
The person who scores low on Factor N tends to be unsophisticated.,
sentimental, and simple. He is sometimes crude and awkward, but
easily pleased and content with what comes, and is natural and
spontaneous.
40
(b) High Score Description
Shrewd, Calculating, Worldly, Penetrating (Shrewdness)
The person who scores high on Factor N tends to be polished,
experienced, worldly, shrewd. He is often hardheaded and
analytical. He has an intellectual, unsentimental approach to
situations, an approach akin to cynicism.
(c) Origins in Psychological Theory
Again, this factor was not recognised in the clinical literature
but was identified hesitatingly Cattell informs us in the 1940s
and has been confirmed since then: it does not account for a
great deal of questionnaire variance but apparently adds to
criterion variance predictions.
e
(d) Life Associations all
Positively related to social status and freedom from pathology.
In small groups such people are positive influences in problem
solving and are rated high in keeping to the point. This trait
relates to less social integration in marriage (Barton et al 1972. )
and, we are informed, manic depressives and schizo-affectives. Males
tend to be higher on this factor.
(e) Age Trend
Tends to show a steady slight increase throughout life.
i
41
FACTOR 0- PLACID VS. APPREHENSIVE
(a) Low Score Description
Placid, Self-assured, Confident, Serene (Untroubled adequacy)
The person who scores low on Factor 0 tends to be placid, with
unshakable nerve. He has a mature, unanxious confidence in
hi^: saiý and 's-. is caoacity to deal with things. He is resilient
and secure, but to the point of being insensitive of when a
group is not going along with him, so that he may evoke antipathies
and distrust.
(b) High Score Description
Apprehensive, Worrying, Depressive, Troubled (Guilt proneness)
The person who scores high on Factor 0 tends to be depressed, moody,
a worrier, full of foreboding, and brooding. He has a childlike
tendency to anxiety in difficulties. He does not feel accepted
in groups or free to participate. High Factor 0 score is very
common in clinical groups of all types.
(c) Origins in Psychology Theory
This trait has a long history in clinical psychology - under such
titles as melancholy agitation, depressive tendency, moodiness and
cycloid tendency. It is not a liability to particular pangs of
guilt so much as a global sense of inadequacy, loneliness and tears.
It might also be considered (Cattell 1973) as a heightened demand of
the gregarious erg which leaves an insatiable loneliness when it is
not met.
42
(d) Life Associations
Those of high guilt proneness are loyal in friendship, take social
failure badly and more observed to be moody. In home background
guilt proneness is associated with discipline by physical punishment
and less overall behavioural guidance. Females tend to score
significantly higher on this factor. High guilt proneness is
associated with less than average school participation and
occupationally is high in those of need of rehabilitation.
High 0 is almost universal in clinical cases.
(e) Age Trend
Falls after early adolescence,, tending to rise slightly after
middle age.
FACTOR Q1 - CONSERVATIVE VS. EXPERIMENTING
(a) Low Score Description
Conservative, Respecting Established Ideas, Tolerant of Traditional Difficulties (Conservatism)
V
The person who scores low on Factor Qlis confident in what he has
been taught to believe, and accepts the 'tried and true', despite
inconsistencies, when something else might be better. He is
cautious and compromising in regard to new ideas. Thus, he tends
to oppose and postpone change, is inclined to go along with
tradition, is more conservative in religion and politics, and tends.
not to be interested in analytical 'intellectual' thought.
43
(b) High Score Description
Experimenting, Critical, Liberal, Analytical, Free-Thinking (Radicalism)
The person who scores high on Factor Ql tends to be interested in
intellectual matters and has doubts on fundamental issues. He is
skeptical and inquiring regarding ideas, either old or new. He
terc's to be more well informed, less inclined to moralise, more
inclined to experiment in life generally, and more tolerant of
inconvenience and change.
(c) Origins in Psychological Theory
This factor was discovered as a unitary trait of large variance
by Thurstone and Chave (1929) and by Guilford before Cattell's
early factor analytic work. It has consistently appeared in
attitude studies conducted by Eysenck (1944) but apparently shows
instability of pattern across subcultures such that it is difficult
to use the same scale across groups varying in terms of age and sex.
Cattell regards this factor not as a narrow trait of political
attitude but as a more general predisposition to try the new.
(d) Life Associations
There is a positive relationship with affiliation with progressive
political parties and the conservative pole is related to strong
family ties and more than average male dominance in the family.
The trait is moderately associated with good school achievement
and entering higher education. Males tend to score significantly
higher than females on this trait.
44
(e) Aqe Trend
Cattell claims no clear age trend for this factor.
FACTOR Q2 - GROUP-DEPENDENT VS. SELF-SUFFICIENT
(a) Low Score Description
Group-dependent, A 'Joiner' and Sound Follower (Group adherence)
The person who scores low on Factor Q2 prefers to work and make
decisions with other people, likes and depends on social approval
and admiration. He tends to go along with the group and may be
lacking in individual resolution. He is not necessarily
gregarious by choice; rather he needs group support.
(b) High Score Description
Self-sufficient, Prefers Own Decisions, Resourceful (Self-sufficiency)
The person who scores high on Factor Q2 is temperamentally independent,
accustomed to going his own way, making decisions and taking action
on his own. He discounts public opinion, but is not necessarily
dominant in his relations with others (See Factor E). He does not
dislike people but simply does not need their agreement or support.
(c) Origins in Psychological Theory
This is one of Cattell's major introversion traits which he is.
claimed to have broken down into sub-factors. It seems that those
high on self-sufficiency have a lack of gregariousness and a
steadiness of emotion that favours dependence on the self and a
45
lack of rewarding response by others.
(d) Life Associations
Positively related to those whose parents were maritally harmonious;
those with this trait were seclusive as children and stubborn in
school. They developed early, were considered mature and in general
showed good scholastic ach-levement. No clear pattern of differences
between males and females has developed on this primary source trait.
(e) Age Trend
No age change in males; slight increase in females.
FACTOR Q3 - UNDISCIPLINED SELF-CONFLICT VS. CONTROLLED
(a) Low Score Description all
Undisciplined Self-conflict, Careless of Protocol, Follows Own Urges (Low integration)
The person who scores low on Factor Q3 will not be bothered with
will control and regard for social demands. He is not overly
considerate, careful, or painstaking. He may feel maladjusted,
and many maladjustments (especially the affective, but not the
paranoid) show Q3-.
(b) High Score Description
Controlled, Socially-precise, Following Self-image (High self-concept control)
The person who scores high on Factor Q3 tends to have strong
control of his emotions and general behaviour, is inclined to be
46
socially aware and careful, and evidences what is commonly termed
'self-respect' and regard for social reputation. He sometimes
tends, however, to be obstinate. Effective leaders, and some
paranoids, are high on Q3.
(c) Origins in Psychological Theory
Catteil (1965) follows the notion of a dynamic structure different
from either the ego or superego strength from McDougall (1932).
Cattell also acknowledges the work of Thurstone who discovered a
reflective factor in questionnaire. This factor is supposed to
represent the clarity with which the individual has developed
a self-concept and the strength with which he adheres to it.
(d) Life Associations
As a child the person of high sentiment was usually healthy and the
family harmonious. He controls his temper and especially likes
science and mathematics at school. The loading on scholastic
success is usually positive and this trait is usually higher in
those following professional occupations. Males tend to score
higher than females on this factor.
(e) Aqe Trend
After adolescence it increases positively levelling off in middle
adult years.
47
FACTOR Q4 - RELAXED VS. TENSE
(a) Low Score Description
Relaxed, Tranquil, Torpid, Unfrustrated (Low ergic tension)
The person who scores low on Factor Q4 tends to be sedate, relaxed,
composed, and satisfied (not frustrated). in some situations, his
oversatisfaction can lead to laziness and low performance, in the
sense that low motivation produces little trial and error.
Conversely, high tension level may disrupt school and work.
performance.
(b) High Score Description .
Tense, Frustrated, Driven, Overwrought (High ergic tension)
The person who scores high on Factor Q4 tends to be tense, excitable,
restless, fretful, impatient. He is often fatigued, but unable to
retrain inactive. In groups he takes a poor view of the degree of
unity, orderliness and leadership. His frustration represents an
excess of stimulated, but undischarged, drive.
(c) Origins in Psychological Theory
This factor can be initially traced back to Freud's 'id', McDougall's
'total instinctual energy' and even Aristotle's 'conation'.
(d) Life Associations
In home upbringing, ergic tension is associated with use of
punishment and lack of demonstration of affection. This trait is
associated with high blood pressure and poor ability to perform under
stress, loading negatively on school achievement. No clinical groups
48
are low on ergic tension. This factor is high in alcoholics,
homosexuals, exhibitionists and all neurotic syndromes.
Frustration and depression increase this factor; low rank in
school and job, unhappy love affairs, bereavements. Females
tend to score higher on this scale.
(e) Aqe Trend
Increases in early life, then declines throughout adulthood
until about 40 years of age.
40
49 1
CATTELL'S MEASUREMENT OF PERSONALITY BY QUESTIONNAIRE
As we have already noted, the Sixteen Personality Factor Questionnaire
(16PF), develop--d by Cattell and his associates in the United States,
has found wide acceptance in Britain, Ireland and the rest of the
Western World. Buros (1972) shows a yearly average of 80 references
i : ieen -19r. 5
.d 11-6`ß for ý.: e 16E': and the ýýf nd 1s uowards.
The instrument has been put to a multitude of uses, impossible to
review here but the list includes studies on academic achievement
(Warburton et al., 1963; Crotis, 1968), teacher training (Start, 1966,
1968), computer programmers (Morris and Martin, 1972), birth order
(Allman and White, 1968), racing drivers (Johnsgard and Ogilvie, 1968),
sensory deprivation (Arnhoff and Lean, 1963), angina pectoris
(Bakker and Levenson, 1967) and hypnotic susceptibility (Hartman, 1966).
Cattell has never held that the sole or even the best method of
investigating human personality is the questionnaire. Whilst it is
true, possibly as a result of the practical advantages and economies
of the self-report technique, that his work has in recent years
centred on the questionnaire, Cattel2 has strongly emphasised the
importance of ratings and laboratory tests in personality measurement.
Cattell regards the life record (IL--data), self-rating (4-data) and
objective test (T-data) as different ways of measuring the fundamental
traits which are descriptive of human personality. Much time has
been spent, not always with complete success (Becker, 1960), in
attempting to match the factors across these three approaches to
50
personality measurement.
It is perhaps a little surprising that Cattell should have so early
adopted this view. After all it is not unreasonable to suspect that
the constructs used by an individual in rating himself may vary from
those used to rate others. Such difficulties have been acknowledged
in practice by the presence, for example, of four factors in the
16PF which have not been discovered in L or T data.
The first editions of the 16PF appeared in the early 1950s with the
issue of revisions over the next 15 years or so. The latest and
most refined versions were published in 1968 and 1969 - parallel
(alternate) Forms A and B, each of 187 items and the somewhat easier
language Forms C and D each of 105 items. Forms A and B take
approximately 50 minutes each to complete and C and D approximately
35 minutes. In addition to the main primaries and secondaries,
Cattell has provided weights to apply to the first order scales
for the calculation of four 'derived criteria'. These are amorphous
measures related to external criteria.
We should also mention the questionnaires which Cattell has developed
for measuring children's personality (High School Personality
Questionnaire, Children's Personality Questionnaire and Early School
Personality Questionnaire), motivation - (Motivation Analysis Test
and School Motivation Analysis Test) and pathological dimensions
(Clinical Analysis Questionnaire).
51
CRITICISMS OF CATTELL'S MODEL OF PERSONALITY
Despite Cattell's prodigious research output, his theories have not
gone without question. It is true to say that most criticism has
centred on the primary source traits which make up his model. Here
three main areas of debate can be distinguished. First there is the
matter of the consistency of group differences postulated by Cattell
from his model. Second is the reliability of measurment afforded by
Cattell's first order personality scales and third, the internal
consistency and very structure of the model itself.
1. Group Differences
In a study of undergraduate personality using the 16PF, Saville and
Blinkhorn (1975) reported that the primary source traits showed an
inconsistent pattern of differences between arts and science and
student and non-student groups depending upon the particular form
of the questionnaires in use.
Similarly these authors were able to demonstrate that the relationship
between Cattell's source traits and academic attainment as measured
by degree class showed considerable variations between questionnaire
forms. They conclude: -
"There is very considerable inter-form variation in the direction
and size of inter-group differences, such that to specify differences
in terms of 'Cattell's Factor N', for instance, rather than '16PF
Form A, Factor N', may be to invite immediate contradiction by those
52
using another form of the test. "
and they continue: -
"This inter-form variation may be thought to account for the
vanishingly small increment in multiple correlations with degree
CiS... .. =Lned by sun: nj scale scores on a1 ternate forms. It
would seem that at least for the purposes of multiple regression,
16PF 'alternate' forms should be treated as independent. "
We have also seen that Cattell places a great deal of emphasis
upon the relationship of the primary source with variables such as
age, sex and social class. How the source traits change over age
for example, is of very great importance to the understanding of
his model. Indeed, Cattell is one of the few authors who provides 40,
graphical and mathematical expressions of age trends in personality.
Despite this fact, there has been no independent research on this
aspect of the theory nor for that matter on other important group
differences which Cattell has postulated from his model of personality.
2. The Reliability of Measurement Afforded by the Model
One of the principal areas of criticism of Cattell's primary source
traits has been the severe lack of reliability which certain scales
demonstrate. Saville and Blinkhorn (1975) in their study of
approximately 2,500 British university students found that as
judged by the alternate form method on Forms A and B of the 16PF,
sever, factors had reliability coefficients between values of . 44 to
53
. 60 (C, E, G, I, Ql, Q2 and Q3) whilst four scales (B, L, M and N)
had an estimated reliability as low as . 43 or less.
In an interesting paper Eysenck (1972) considered the question of
the unreliability of Cattell's source traits and their inter-
relationship with other scales. Taking the published table of
inter_"or el ations be' : vee the 169= SC31C scores of 423 male and
535 female American college students (Cattell et al, 1970),
Eysenck corrected the coefficients for attenuation (unreliability),
paying particular attention to the five 'anxiety' scales (C-, L, O,
Q3- and Q4) and five 'exvia' scales (A, E, F, Q2- and H).
On the five scales which contribute to Cattell's anxiety second
order factor, twelve of the twenty intercorrelations exceeded unity.
On this evidence Eysenck states that "as far as one can see, "
practically all the information contained unreliably in the primaries
is contained reliably in the second order factor; very little
information indeed is left over for contribution by the primaries".
In the case of the exvia-invia factors, the position was less clear
cut in that whilst eight of the twenty coefficients exceeded . 9,
twelve correlations were below this level. Eysenck points out
this in part is due to certain of Cattell's exvia primaries
overlapping with the anxiety and other second orders. He concludes:
"The figures given are not incompatible with a general view which
would regard the primaries advocated by Cattell as random groupings
of items either measuring extraversion or neuroticism or occasionally
both. Such a view would also be compatible with the fact that several
54
writers have found it impossible to replicate Cattell's factors
in independent analyses, using both his items and his methods of
analysis and rotation".
3. The Internal Consistency of the Model
li' ldc itiv`rl o the reliability of Cattell's source
traits as scale scores, considerable debate has centred on the
internal consistency of the model as represented by the items
in Cattell's personality questionnaires.
In an early study Levonian (1961) using Form A of the 16PF on
252 American psychology students, intercorrelated item responses
and factor scores and found low internal consistency for many of
Cattell's factors. He concluded, "while item homogeneity is not
the sine qua non of test construction, the extreme item
heterogeneity demonstrated in the study would seem to indicate
a critical appraisal of the 16PFQ".
Becker (1961) who had earlier demonstrated low correspondence
between Cattell's Questionnaire ('Q' data) and rating ('L' data)
studies, showed in relating the 16PF with some of Guilford's scales,
that with psychology students, certain of Cattell's factors were of
low reliability and that at best only eight factors could be
distinguished in the 16PF.
In two well-known papers Howarth and Browne (1971) and Howarth,
Browne and Marceau (1972) claim that despite its prominence Cattell's
55
16PF has been developed along unsatisfactory lines and as a consequence
does not measure univocal primary factors as is the "often reiterated
claim of its originator and proponent". They point out that Cattell's
preferred method of analysing clusters of inter-related items is
probably based not upon any technical superiority but rather that
total item factoring was not available in terms of computer
ceehniiüq ý hen the c3ie$ were originally deeveioned. They point
out that Comrey and Duffy . (1968) tried to group Cattell's items
into 'factored homogeneous item dimensions (FHIDS)' leading them
to conclude that Cattell's factors were complex rather than unitary.
Arguing that items chosen to represent or 'mark' a factor would
intercorrelate highly among themselves and not outside the given
factor, Howarth et al (1972) took the responses of 567 university
students and calculated the some 17,000 inter-item correlations 4P
on Form A of the 16PF. They, in fact, found that only about 30%
of the intra-factor item correlations were significant at the
. 01 level; the most homogeneous item groupings being found in
Factors F (Surgency), H (Social Boldness) and Q4 (High Tension).
In addition to the packaging of items into FHIDS, they criticise
Cattell's method of targetting his factor solution to the twelve
factors, which he observed in two early rating studies (Cattell,
1945,1947); for when the item intercorrelation matrix was factor
analysed by the method of Householder-Wilkinson only ten factors
were extracted.
The latest appraisal of Cattell's system is in a recent paper by
56
Howarth (1976). Whilst the author remains in sympathy with Cattell's
strategy regarding multivariate definition and comprehensive choice
of variables, he believes there were questionable aspects to Cattell's
tactics. Rather than a system defence, Howarth argues, what is
now needed is a movement of investigators towards "a consensus
buttressed by co-ordinating studies". As one important factor
in such work, he continues, populations should be tested from
various social levels.
CATTELL'S METHODS OF SCALE CONSTRUCTION
All the studies summarised above have in some respect or the other
revolved around the reliability of Cattell's scales. Indeed, the
issue of test reliability and more specifically item homogeneity, a
is central to the dispute on the validity of Cattell's model of
personality. The papers by Howarth et al (1971 and 1972), Eysenck
(1972), Levonian (1961) and Becker (1961) have all taken issue with
the low homogeneity of item content of Cattell's first order
primary personality traits as represented in the 16PF.
Perhaps it is not altogether surprising that the central issue
in personality measurement by questionnaire has been reliability
theory. Because we can never be absolutely sure of what we are
measuring in psychology, because our constructs are always to some
extent arbitrary in nature and because they have no real existence
in themselves, it is imperative with general. purpose instruments
that they cover the domain to be measured as completely, succinctly
57
and accurately as possible. What is more the use of factor analysis
in terms of identifying or 'discovering' the major dimensions of
personality means that any model rests to a great extent on the
homogeneity of the items concerned.
It is certainly true that a classical test and factor theory have
emphasised the importance of measuring clear cut, non over-lapping
variables and this is generally achieved by producing scales made
up of items of high intra-factor consistency and low extra-factor
correlation. Moreover, it is perhaps significant that the aim of
apparently less orthodox approaches to measurement, such as
Rasch Item Analysis and Scaling (1960), is to produce scales
of high homogeneity.
In answer to the criticisms on his first order factors, Cattell
has been clear and prolific in reply. "The primary factors give
one most information and we would advocate higher strata contributors
only as supplementary concepts ... it is a mistake, generally, to
work at the secondary level only, for one certainly loses a lot of
valuable information present initially at the primary level"
(Cattell 1970. Cattell, 1974 ).
On the connected and important matter of item structure, Cattell
has taken a radical departure from orthodox thinking and argued the
case that in many respects. the aim of ever increasing homogeneity
is unrealistic. In simple terms he reasons that if we accept a
definition of homogeneity as the average correlation between all
the items in a scale, then perfect homogeneity suggests a scale
58
which provides no more information than a single item. In personality
measurement he claims that the desire for scales of high internal
consistency has led authors to repeat almost identical items; a
practice which may be suitable for measuring narrow, specific
factors but not appropriate to the measurement of general influences
in behaviour.
Cattell (1972) has argued: "The charge that item intercorrelations
are low .... are therefore correct; but one would have thought that
well informed psychometrists today would have recognised that ....
this might actually be a virtue. Indeed, it is actually a deliberate
feature of the design of the 16PF and other tests by the present
writer which aimed to measure a broad. personality factor over so
widely diversified a set of items that no specific variance due
to situations will be spuriously included to raise the correlations". 40
Cattell continues"..... the ideal homogeneity is not the greatest
attainable but an optimum low value based on various combined
psychometric considerations".
Cattell in this paper entitled 'The Importance of Factor Trueness
and Validity Versus Homogeneity and Orthogonality in Test Scales'
(Cattell 1964) goes on to point out that classical test theory has
almost exclusively concerned itself with what he has termed 'item
metrics' (Kuder and Richardson 1937, Cronbach 1951). He points
out that it is not sufficient to examine the homogeneity of the
test as a whole but "good, adapted construction often requires
that the test, like a watch, shall have parts with different functions,
properly balanced in certain ways". The essence of Cattell's
59
argument is that the best combination of items to measure a factor
are those which have high loadings with the pure factor but in fact
have low correlation amongst themselves. He develops the concept of
factor trueness as the extent to which the scale as a whole measures
only the required factor. Contrary to orthodox reliability theory,
he makes the case that the addition of suppressor items, that is
72 1oadi r: (j on ,- ___ ý a- which
would usually be considered unsatisfactory by conventional reliability
methods, can have a definite advantage in purifying the measurement
of the underlying trait afforded by a given scale of items.
Thus, his'concept of direct validity' is concerned with the multiple
correlation of a set of items with the pure factor. Clearly, as with
any multiple regression problem, maximisation of the relationship
of the sum of item responses which represents the scale score tend
to be higher with low rather than high homogeneity. The primacy of
high homoege; ty is rejected in that it tends to bring lower total
validity for the multiple correlation of items with the pure factor.
Cattell (1964) has provided a simple graphical representation which
is reproduced in Figure 1.2. Here items j and k are loaded to the
extent of . 71 on the wanted factor w but in opposite directions on
the unwanted factor u. They are said to suppress each other on the
unwanted and re-enforce each other on the wanted factor. It is then
possible to get full validity in that j+k gives a factor true
scale score with complete heterogeneity in that the two items
correlate zero.
60
Thus, it has become clear that the validity of Cattell's theory is
very much rooted in the psychometric characteristics of the scales which
he has developed to measure the variables in this personality model.
More specifically the arguments have centred upon the internal
consistency of the 16PF and Cattell's aim to produce 'buffered'
scales of low homogeneity. Cattell (1974) also indicates that much
criticism has been based upon obsolete editions of his questionnaires
collected on university samples.
Whilst scale homogeneity has been the central issue in the controversy,
there have been other and more general matters related to reliability
theory; questions focused on reliability of measurement as a person
as much as a test characteristic. First, there is the point mentioned
above that very little is known about the reliability of questionnaire
measurement, let alone the 16PF, on groups other than university
students. This is concerned with the effect of what Ghiselli (1967)
has termed 'moderator variables' on reliability. In test theory
is is commonly assumed that error variance is uncorrelated with
true variance. In moderator theory however, error scores may be
related to other variables. Probably the more important moderators
(Cattell 1974) are likely to be the sex, social group and age of
respondent but whereas it is fair to hypothesise that the accuracy of
measurement might vary from one specific subject cluster to another,
there have been no reported studies of its effect on personality
scale measurement.
In summary then, as have been pointed out by Howarth (1972) a critical
evaluation of the item structure of the 16PF, how it is affected by
61
certain moderator variables such as age, social class and score level
and an exploration of Cattell's concepts of factor trueness and
direct validity as they relate to orthodox reliability theory,
are long overdue.
Certainly, if Cattell is correct in his assertion that item
homogeneity has aS an L: nn, -ce S3iv and n-,: ý-alistl, c goal -in
psychometrics, it will have far ranging implications for conventional
measurement theory as it is practised today. It is the aim of this
research to investigate these areas as well as the topics of
suppressor action in scale measurement, and the role of potential
moderators in personality measurement in the questionnaire realm.
AP
62
FIGURE 1.2.
A GZAPHICAL REPRESENTATION OF CATTELL'S CONCEPT OF BALANCE SPECIFICS' IN SCALE CONSTRUCTION
W
Wanted General Personality Factor
Unwar*ed General Factor
63
SE tTION 2:
HYPOTHESE S, METH0D0L0GY,
PR 0C EDURE. THE SAMPLE
#º
2. AIMS AND HYPOTHESES
64
INTRODUCTION
In view of the criticisms outlined in the previous chapter, it would
seem that Cattell's theory of personality is much in need of independent
research. Like most complex theories, there is unlikely to be a
definitive once and for all test of Cattell's model. What is
required is careful programmatic research conducted on various
aspects of the model with the aim of observing how well the theories
fit well collected data bases. Accordingly, a general hypothesis
has been developed from the literature review and from this three
groups of subordinate hypotheses have been distinguished.
GENERAL HYPOTHESES
The general hypothesis has been developed that: -
Cattell's theory of personality rests on a system of primary traits
which are untenable in terms of: -
(a) the verifiability of important group differences in personality
postulated from Cattell's theory.
(b) the reliability of measurment afforded by the personality scales
which represent the model.
(c) the internal consistency and structure of the items defining the
primary source traits
and that despite the existence of some independent, reliable and
descriptively useful factors, Cattell's model of personality is in part
65
made up of redundant traits. That is, Cattell's primary source traits
cannot be described as the simplest, most reliable definition of the
personality domain.
The first of the three groups of subordinate hypotheses concerns the
strength and direction of differences between certain groups of subjects
which have beer. mrpoth`sised from Cattell's model. Of course, there are
an infinitely large number of such groups but the most important would
seem to be the effects upon the primary source traits of cross-national
variation, sex, social class and age. Second, there is the question
of the reliabilities of the primary source traits and their fluctuation
as a result of sex, social class and age acting as
moderator variables. Finally, there is the related issue of the
internal consistency, homogeneity factor trueness and construct
validity of the model.
mo
HYPOTHESES ON GROUP DIFFERENCES IN THE PRIMARY SOURCE TRAITS
We have seen (Chapter 1) that Cattell's theory of personality, at least
as regards the primary source traits, is relatively rich in potential
hypotheses. It is difficult at times to. be absolutely sure which of
his statements remain theoretical and which are backed up by scientific
evidence. Nevertheless, it is possible to take these factor descriptions
and generate meaningful hypotheses on expected cross-national sex and
social class differences and age trends in the primaries. These
hypotheses are listed in Table 2.1.
The table contains a total of sixty-six hypotheses, twenty of which are
the null hypotheses of no difference between means. It is thought
66
unnecessary to review these here, especially as they originate in the
descriptions of the primary source traits as postulated by Cattell which
were summarised above.
As for the nature of these hypotheses, it is evident that the sex
difference cells for each of the primary traits in the table
simply contain either the null hypothesis of no difference between
the sexes, or the hypothesis that males or females will be found to be
the higher scoring sex. The social group variable is hypothesised
either to have no relationship with each of the primary source traits
or to show a moderate or slight and positive or negative linear
association; this gives a total of five types of hypothesised relationships.
In addition, we shall also seek to test the hypothesis of the non-linearity
of the regression of personality on social class.
The age variable is a little more complicated. Whilst certain personality
factors have a hypothesised positive or negative straightline relationship
with age, the majority, Cattell claims, depart from the simple linear
model. Accordingly, the hypothesised age trends in personality have
been abstracted in Figure 2.1., which shows more clearly the six trend
lines which Cattell envisages for his source traits.
Cattell (1972) has recently provided age curves for the primary source
traits but it is necessary to replicate these on groups of subjects
selected by better sampling methods than those originally employed by
Cattell. As with social class, it will also be possible to test for
the non-linearity of regression of the primary source traits on age.
67
TAFLE: 2.1.
SEX AND SOCIAL CROUP DIFFERENCES AND AGE TRENDS IN THE PRIMARY SOURCE TRAITS HYPOTHESISED FROM CATTELL'S PERSONALITY MODEL
PRIMARY SEX CROSS-NATIONAL SOCIAL SOURCE DIFFERENCE DIFFERENCE AGE GROUP TRAIT (ON AVERAGE) (ON AVERAGE) TREND DIFFERENCE
A Females American Increases 17-30 yrs, No Outgoing higher higher then levels off. Differences
B No No Negative Moderate positive Intelligence difference difference relationship relationship
C Increases up to 40 E: No ycea r s, n? ev els Slight posit ive
Wince () f. re 1 tons 1p
E Males American Level up to 40, Assertive higher higher then slight No
decline. Differences
No American Rises in adolescence Slight negative I{appy_Go-Lucky difference higher then decreases relationship
G Females No Moderate positive Conscientiousness higher difference but decelerating Moderate positive
increase with age relationship
H Males American Strong positive Venturesome higher higher males, Weak No
positive females. Differences
= Females No Slight positive Tender-minded higher difference Slight positive relationship
L Males No Declines up to 40, No Suspicious higher difference then increases. Differences
Females American Declines after Slight positive imagination higher higher adolescence relationship
N Males No Slight positive Slight positive Shrewd higher difference increase relationship
O Females No Falls after adoles- Apprehension higher difference cence, rises Slight negative
slightly after 40. relationship
01 Males No Moderate negative Experimenting higher difference No clear trend relationship
C22 No No No trend in males, Self-sufficient difference difference slight increase Slight positive
in females. relationship
Q3 Males British Strong & positive, Controlled higher higher levels off at Slight positive }
40 years. relationship
(24 Females No Increases until 40, Slight negative Tense higher difference then levels off. relationship
68
FIGURE: 2.1.
P e r sT or na ai
t -i.
Y
ABSTRACTED PERSONALITY - AGE RELATIONSHIPS FOR EACH OF CATTELL'S PRIMARY SOURCE TRAITS
A: NEGATIVE: POSITIVE/ZERO/NEGATIVE RELATIONSHIPS
L (Suspicious) 70 (Apprehensive)
Zero
B (Intelligence) M (Imagination)
Age
P
e r sT or na ai 1t i t y
B: ZERO: POSITIVE/ZERO/NEGATIVE RELATIONSHIPS
-Se 5X->
4o
Zero ', Zero Ql (Experimenting) Q2 (males) (Self-sufficient)
E (Assertive)
P e r sT or na ai 1t i t V
Age
C: POSITIVE: POSITIVE/ZERO/NEGATIVE RELATIONSHIPS
G (Conscientious) H (Venturesome)
ye I (Tenderminded) N (Shrewd)
Qov Q2 (females)(Self-sufficient)
Zero A (Outgoing) - /T _t_ _. __I rr. . .-..
Age
'. ' zmouonat ýtab1llty) F (Happy-go-Lucky)
Q3 (Controlled)
Q4 (Tense) }
69
HYPOTHESES ON THE RELIABILITY OF THE PRIMARY SOURCE TRAITS
We have seen that a great deal of controversy has revolved around the
question of the reliability of Catteil's primary source traits.
Eysenck (1972), for example, does not so much deny the existence of
lower order factors, but claims that at present they are too fickle
and of such low reliability as to be of little practical value within
a no; nothe is personality system. Moreover, if such scales are of
very low measured reliability, the fractionalization of more general
order variables like extraversion and neuroticisminto sub-factors
is at best arbitrary and at worst propagating the existence of traits
which have no functional independence or substance.
An associated issue which is likely to extenuate the reliability
problem is the fact that estimates have more usually been made for
relatively sophisticated student samples, which are closely cornparab*e
with groups used in the original model development. This is the quest-ion
of the transferability of the reliabilities of the primary source traits
to other, less sophisticated and older groups of subjects. Should there
be a demonstratable loss of reliability across various sub-groups of
adults, there are likely to be serious consequences for the generalisability
of the factor structure.
Ghiselli (1967) refers to this as the role of moderator effects upon
reliability; a test which is reliable or for that matter valid in one
group of the population may not be so for another. The number of
empirical studies, however, on moderator effects have been remarkably
few and the literature appears to contain no large scale research on
moderator effects in personality data. I
7G
The issue of moderator effects is of course crucial to the general
hypothesis stated above that Cattell's model of personality has been
derived from inadequate data bases for a generalised theory. Moreover,
it is likely that the more divorced the particular sub-group under
investigation is from the student population the less transferable
the theory becomes. Thus, the hypotheses have been developed that
the reliabilities of the primary source traits will be lower in .
aiult than in student groups and in the lower socio-economic
and higher age groups of the adult population. It is hypothesised,
on the other hand, that the sex of the subject will have no appreciable
effect upon the reliabilities of the primary source traits.
HYPOTHESES ON THE INTERNAL CONSISTENCY OF THE PRIMARY SOURCE TRAITS
We have seen that the debate over the item structure of Cattell's
primary source traits has been a long and complex one. Cattell (1972) claims
that his items are deliberately heterogeneous in order to maximise
the multiple correlation of the items with the pure factor which the
scale is designed to measure. Whilst many critics (Howarth and Browne, 1971,
Levonian, l961, Eysenck, 1972) have analysed the 16PF by conventional factor
and item analysis methods, none has investigated whether Cattell's specific
claims hold up under closer empirical scrutiny.
Accordingly, the first hypothesis that we shall test is that the primary
source traits will show low intra factor homogeneity of item content.
Assuming that Cattell's factor solution can in some way be re-derived,
we shall also test the hypothesis that items will show high multiple
correlations with their own scale. That is, they will have high factor
71
validity. These investigations and whether estimates of factor purity
can be made through conventional reliability theory, will form the third
major section of this thesis.
ALTERNATIVE MODELS IN TH PERSONALITY DOMAIN
Should it be necessary to reject the hypotheses on factor validity,
`L; - '- -1 ._i L_Lr. ^c J nd p-- , aý nl ).
_ rs; oL. S , planes i ity of 'i r} r- rcni CST
of the personality domain will be investigated. Perhaps the most obvious
alternative model to that of Cattell is Eysenck's (1972) three factor
description of personality. From the information collected in the
various sections of this thesis, it will be possible to test, for example,
hypotheses on Cattell's main anxiety primary source traits to see whether
they have separate identities demonstratable in terms of group differences,
reliabilities and the internal and factor structure.
40
AIMS
The aims of this research can be summarised as follows: -
1. To administer parallel forms of Cattell's personality inventories
to a large sample (N '72000) Qf general population adults
representative in terms of age, region and socio-economic class.
2. To test specific hypothesis on Cattell's model of personality
under three main headings: -
2.1. the verifiability of important group differences in
personality hypothesised from Cattell's theory.
t
72
2.2. the reliability of Cattell's primary source traits.
2.3. the internal consistency and structure of the model.
3. To investigate alternative theoretical models in the personality
domain.
SUMMARY
In summary we are to research in Section 3 of this thesis, the
relationship between the parameters of Cattell's model of personality
and such external factors as sex, age and social class. In section
4 we shall next investigate the precision of measurement afforded
by the model as expressed by Cattell's first order 16PF personality
scales. Finally in Section 5 of this thesis we shall undertake a
number of researches into the question of the internal consistency
of the questionaire items which are defined by and are central to,
Cattell's model of personality at the first order level.
73
N
CV
u H
H
0
H
a 0
U)
H U
z H
w a H U
E/I
Cl z
H
H
w
4
0
a
U
:j -4 4J ý 10
m a) M 4J 111
43 r4 Ö Ö U)
4-3 O C -4 0 d) ß, " "d -P S- to
0 -0 v >
4-4 O
04 S:
`° a 0 u "d
, cö .ý -P P0 - C CO . ri ý },
ll -P a) ' v r-1 N ý4 dO - ý v
Rf U1 E0
d rýö
) C) '-t r-I r-1 ri E
_ý u! b
4-) ri (a u IQ a) ul r 0 ` 'dd
Q) 0 N ri Q) d r 4 -A . "r q ýr p O (n E I 0 -H z v
.ý >
4. to
E O G)
U) p. - la
tue 0
-N S-i U) co 0 14 Ä HR `( 0
O 0ý . c 0 .N U -P "
Q1 43 U -1-) U] ý., 4-) a)
.C I Ö H C ý
H O fl
rd tU +j .c r
P H H (a A . HE-
N
4 v .Q UU] . ºJ W 0 0 0)
-r a) V)
41 rf `° (0ö .0 w s fo - 4
(a ° b z
O-P-P i 0) + 3
S: ý O EO a) (a ý r-A U En u E C t r-4 U t ö E -H -q N C2 4
O to 14 Q) , r r- I to O 4-4 C O aJ
{ QJ U) UN 0) U) (0 0
. Q) 5 Q) -P 4-) U O0 ul ul fO - ?'
O 4 O r SCI
N U) 4i -1 .a4 S4 "ri
0S (tl S -I NO
U) N 4P 0 4- -1 (0 a) u t7 S! (a r. Ul (U ý4
v O. H W4 tttddd v m ", I LI Ui r-i H $4
. f. b -P ,
p. "
to ul 4J
4 S j
04 NÖ (a
.d 'A
(11 -1 C (7i H a -1 rl "9 dO "r1 H IL uO u1 "ýI S-1
to U
N "
N b
C. O O H "i A .ÄL
C
9 0) $4 N rl r. 'D N U)
4-J vO ci. -i X -P
u C U 4- r- IN
-P C ý4
0 U2 3
.CVO S1
4-1 - ,5 0 u ro - ö av .
3 4 , 0 b V to N to Q
-P $4 Ll 4i Cu .d . tJ "rf O r--A ra 0( N o -1 + -P , U c o P -4 ö . 14 . 14 (Uý 0 "
)- PP P r4 -ri ° v -i 5 (L)
a S-l ) ro 0
a F! > i7ß. U$ 44
Q
I
74
r]
3. METHODOLOGY
75
INTRODUCTION
The first methodological problem was that of selecting the sample.
An important aim of the research was to investigate Cattell's
model on adult rather than student groups. It was also clear that
in order to undertake studies on the effect of moderator variables
and of social class and age trends in personality, a relatively
large sample size would be required.
The problems of gaining large representative samples of adults,
however, should not be underestimated. Any sampling procedure of
adults contains some bias. If there are dangers in taking results
from students as if they represented firm conclusions for the general
population, then it is equally problemmatical to infer that military
personnel, subjects known to the experimenter, those employed as
clerks, apprentices, managers or indeed those employed in any Alp
occupational field will represent the archtypical personality type.
It is true that most personality inventories have been developed
on little more than ad hoc occupational, clinical, educational
and student groups (Eysenck, 1964, C6ttelJ, 1968). This may be of
less importance for standardisation purposes in that an author
can go some way to justify his procedures by using the random
error philosophy. That is, whilst all of the constituent groups
will have their own specific bias, these will in some convenient
fashion cancel out when placed in the statistical melting pot.
This is a rather dangerous philosophy if any form of multivariate
analysis is to be used however. Child (1966) and Adcock (1973) i
76
specifically warn us for example against what they have termed
'oddballs' in the sample if factor analytic techniques are to be
employed in data analysis. Should cluster analyses be used,
moreover, it is not difficult to see that a certain structure
might evolve in one sample which would be difficult to replicate
in another.
One possible method was that of requesting the return of completed 16PF
answer sheets from British test users. After a predetermined duration,
samples could then have been stratified and weighted by current
census data. However, although this procedure has proved practicable
in the States and was the principal method employed in the American
16PF standardisation, a considerably lower answer sheet usage
rendered this approach unsuitable in Britain. Added to this was the
unknown nature and degree of bias introduced into questionnaire
data by the test-taking attitudinal and motivational state of the
subject. One serious source of this distortion is direct sabotage.
Another is where the subject assumes, whether consciously or not,
a role congruent with the particular testing circumstances, as,
for example, where an inventory such as the 16PF is used in the
first instance for purposes of vocational guidance and then for
industrial job selection on the same person or population.
Clearly the best recourse would be some form of random sampling
such as that used by the Scottish Council for Educational Research (1965
when standardising the Wechsler Intelligence Scale for Children.
Such a procedure would in fact prove very difficult in practice in
i
77
that even lists such as the electoral register are inevitably
out of date, whilst the costs of tracking down each respondent
would be prohibitive.
An alternative strategy would be to draw up some random or
stratified sample of names and undertake a postal survey.
J-iow. qever, postal surveys ha'. 'e the problem that response rates
(that is the number of returns as a percentage of those contacted)
for completion of relatively large questionnaires can be as low
as 25%. In addition there are the added problems of the quality
of the data: how much can the subject be relied upon to complete
age and social class data or indeed the questionnaire itself
accurately, and can we really be sure that the individual contacted
actually was the one completing the questionnaire? Another
drawback to the method is that using well known personality
inventories in a postal survey might have contravened the
professional and ethical controls on test standards of the
British Psychological Society.
Alternatively, it may have been practicable to test groups of
a
previously selected individuals in centres (schools, church halls,
etc) throughout the British Isles. The major objection to this
method is that difficulty is often encountered in gaining a
representative sample in that females and the lower socio-
economic classes tend to predominate in an obtained response
which is frequently below 10%.
The only solution then was to use some form of quota sampling of 11
78
people in their own homes. Ideally there would be a good representation
of regions, age and social class in the sample. Two forms of Cattell's
Questionnaire 16PF Forms A and B were to be administered under-
controlled conditions and every effort would have to be made to
keep the response rate as high as possible. If any investigation
of item structure and of particular sub-groups within the sample
was to be carried cult (Lo ca L_'gories for the sex of the subject
and five for social class would give ten possible data cells for
example), then it was decided that the sample size would need to
be approximately 2000 cases.
Cattell's Sixteen Personality Factor Questionnaire Forms A and B
each consist of 187 items which on average take 45 minutes each
to complete. Collection of socio-economic demographic and other
information would add about 30 minutes to each interview giving
approximately two hours on average for each subject. A simple
calculation gives an estimate of 4000 interviewing hours for
the survey. With the addition of travelling and preparation time,
it is evident that it would take one experimenter three years
of full-time work to complete the fieldwork.
The alternative was to employ more fieldworkers; hopefully people
who had experience in administering questionnaires and interviewing
techniques. With training, it was thought that such individuals
should be capable of administering the 16PF on an individual basis,
recording the background information and completed forms for data
processing and analysis.
4 F
79
Suitable bodies which could provide such fieldworkers were considered
to be specialist teams of certain research councils, the Office of
Population Census and Surveys and certain market research organisations.
Contacts were first made with the Office of Population Statistics
and Surveys but it soon became clear that a project of this kind
c"; o_tld he outside their reit. Similarly, the SSRC Social Survey
Unit was not principally involved in conducting its own research
surveys. For these reasons, it was decided to employ the British
Market Research Bureau (BMRB) for the fieldwork. This organisation
had the advantage of having a large fieldforce which was experienced
in social surveys of this kind.
The next problem was one of funding the research. The National
Foundation for Educational Research agreed to finance part of the Ar
survey, providing materials, card punching and initial computer
resources. In addition BMRB itself agreed to fund a portion of
the fieldwork costs on the condition that a self-completion
questionnaire on consumer habits could be left with the respondents
after the main survey. The remaining costs were covered by a
number of small donations from commercial organisations which were
willing to make a contribution to work of this kind.
SAMPLING
The sampling frame finally adopted was based on a method developed
by the British Market Research Bureau over a nw-nber of years and t
80
known as Random Location Sampling.
This technique involves the drawing up of a representative sample
of 200 Parliamentary constituencies in England and Wales and those
parts of Scotland south of the Caledonian Canal, selected with
probability proportional to their population size. These
constituencies were selected at rasdon using the currula ive
sum and fixed sampling interval technique, after stratification
by the distribution of votes cast at the 1970 General Election
(as an indicator of economic status) within conurbation or urban
or rural area, within standard region. For each of the 200
constituencies, the polling districts in it were placed in
random order within ward, the wards in random order within
administrative areas and the administrative areas in random order
within constituency. Groups of streets were then selected with
probability proportionate to the electorate in a way which was 40
consistent with polling district. Interviews were then spread
uniformly over these streets.
In random location sampling one is adopting probability techniques
down to and inclusive of the selection of street. This has an
important advantage over traditional quota sampling in that the
freedom of the interviewer to choose her subject is minimised.
In quota sampling, certain controls (sex, age, social class etc)
are to be satisfied, allowing the interviewer freedom to select
subjects in accordance with these constraints. Clearly this was to
be avoided and was one of the principal considerations when
adopting the described sampling technique. I
81
In summary, the methodology was to ensure the collection of data
on approximately 2000 adults in the general population. Ideally
the respondents would be tested in their own homes on the Cattell
inventories, Forms A and B, and the sample would be representative
in terms of geographic region, marital and educational status,
age and socio-economic class.
AP
rr
82
0
4. PROCEDURE
83
THE PILOT STUDIES
Two pilot investigations were undertaken in the lower socio-economic
areas of Ealing and Acton in London, for it was from such areas that
most difficulties were anticipated. Seven subjects were chosen in
each of those districts and the author observed as a female fieldworker
': n 't -hrou:, h the inter`. i ew, ring procedure. ( It should be pointed out
at this stage that female fieldworkers were to be preferred in work
of this kind because many housewives are understandably reluctant
to admit a lone male into their homes. )
The materials were prepared as follows: -
1. A contact sheet containing the initial introduction to the purpose
of the study. (Figure 4.1. )
2. A number of ball point pens.
3. One 16PF Form A and Form B questionnaire printed so that their order
was reversed in alternate booklets.
4. A number of 16PF answer sheets.
5. One serial-numbered background questionnaire (Figure 4.2. ).
A number of problems came to light as a result of these pilot surveys.
The most serious was one of using answer sheets with older subjects or
those with poor eyesight. Accordingly, it was decided to employ the
84
FIGURE : 4.1. CONTACT SHEET
I am from the British Market Research Bureau and we are carrying out a National Research Study of people's interests and attitudes for a Research
Body. We would appreciate it if you would help us in our study.
The research is on people's interests and views and includes a couple of questionnaires. Amongst other things this study will be used to find out the views of various groups of people. Over 2000 men and women all over the country are helping us in this work.
P1 }l0'.: _research
t"( th this questionnaire has been done. 'ten of ner parts of th
.... r . Is the tine jh Eis U en carried Oi in i; 1C1. an
. 'JCO'-lvnd/
Wales (use appropriate country). it is hoped amongst other things that the
results will be of great value in building up a picture of the attitudes of
people in various countries.
ESTIMATED Result (put down CONTACT SOCIAL Questionnaire number Full reasons for
NO. AGE ý SEX LADE if contact successful) any refusal I
r
s
V.
t
85
FIGURE. 4.2. BACKGROUND QUESTIONNAIRE
n ONNA; RE No. (Cols. 2/3/4/5)
BLOCK CAPITALS. PEEASE: -
N. m',, e Coss ti tuency
tPlr. icc t. ite Mr.. Mrs. or Mis. and yive initials)
`ulI t. , taI . ddre, s
Iýterviewer Code no.
Supervisor Time of interview:
OJ-e of inrrrview _1970
Frofl
FAMILY UNIT (related by blood, rarria9e or adoption)
Jab No.
A AEA CODE Tv
To RESPONDENT IS:
- --- ---- -------
Man I I. I:
NORMAL occupational status Marital Has paid job No paid job Status
SEX Aue last
F"II Part-time Ret. from Wid/
Line Re lat ion sF. i 0 birth- t ime
Under F/T Div/
no. to re+;: o. j"ir., t m F ddvv 31 + 8-29 8 job Others S M sep.
1. RESPONDENT
2.
3"
4
i" 8.
11 OTHER MEMBERS OF HOUSEHOLD (not related to respondent)
9.
10.
11.
12.
13.
14.
15.
16.
-------------------- - - OFFICE USE
S. G. A 14.12 g 11
cl 0 c2 I D2 E3
CHILDREN Under I year (14. ) 4
BY AGE I-4 years 5 5-9 years 6 10 - 15 years 7
None 8
RESPONDENT Yes (14. ) 9
fiEAD 0P No A
H-)'JSEHOLD
HOUSEHOLD One 15.12 SIZE Two 11
Three 0 Four I Five or more 2
WEEK, OF One (15. ) 3 INTERVIEW Two 4
Three 5 Four 6 Five 7
VI At what age did you finish
your full time education?
14 or under 16.12 15 11 16 0
17 1 18 2 19 3
20 4 21 - 23 5 24 or more 6
Still studying 7 Refused 9
86
Working 30+ hrs 12.12 Working ß-29 hrs 11 Not working 0
IV RESPONDENT'S AGE:
15 - !9 (12. ) 1 20 - 24 2 75 - 34 3 35 - 44 4 45 - 54 5 55 - 64 6 65 or over 7
V RESPONDENT'S MARITAL STATUS:
Single 13.1 2 Married 1 1 Widowed/divorced!
separated 0
IF 'Sin le' Are you engaged or planning to get married'
Yes No A
IF 'flurried' How long is it
since you were first married? Less than 1 year (13. ) 2 1-4 years 3 5-9 years 4 10 - 19 years 5 20 years or more 6 Refused 7
Ii11 (I3. ) ----------- ---- ------------------------------
VII Do you (or some other member of your household whom you have mentioned) own this house/flat, or rent it, or do you live here rent-free?
Owned 17,12 Rented 1) Rent free 0
IF 'Owned' Is it owned outright or is it being bought with a mortgage or loan?
Owned outright 1 Mortgage/loan 2 Don't know A
IF 'Rented' Is it rented from the council or from someone else?
Council 3 Someone else 4 Don't know A
Viii How long have you (or your family) been living in this house/flat?
Less than I year (17. ) 5 1-4 years 6 5-9 years 7 10 - 19 years $ 20 years or more 9
FIGURE: 4.2. (Cont'd)
jx" uhlch mcn'her of your hot,. ehold Iiviny here: Actually owns it? STATE LINE NUMBER
(FROM HOUSEHOLD
Is responsible for the rent? Q
COMPOSITION BOXES) (i. e. has his/her name on the rent book)
OF THE PERSON Is responsible for this CONCERNED household having it rent free?
x
IF LINE NUM3ER ENTERED IN iX a IS ANY NUMBER FROM 9-16 ASK STATE LINE NUMBER
IX b Which member of your family (i. e. respondent and relatives) living here FROMI is actually responsible for you having this accommodation?
If member of (o. uily unit
TICK 5.! 41": 11 yr detJt. -
iYa 1Xb APPLE
he/she is now occupied (even if temporarily out of his/her present or usual work) occupation
Man or he/she is not now occupied but has means other than single only state pension. etc. his/her former occupation woman he/she is not now occupied and has no means other occupation of senior
than state Dessins. etc* breadwinner` her husband is now occupied (even if temporarily his present or usual occu-
out of work) pation even if he is not a
Married her husband is not now occupied but he has means his former occupation woman other than only state pension, etc.
her husband is not now occupied but he has only occupation of senior state pension, etc. breadwinner'
she has no means other than state pension etc. occupation of senior
, breadwinner' Widowed,
she has means other than (or in addition to) state her husband's present divorced,
pension. etc., and part is obtained from her former occupation if he is alive separated, husband (if he is alive) or as a pension from her or his previous occupation etc., former husband's revious employers (if he is dead) if he is dead woman she has means other than (or in addition to) state her i
pension, etc. but no part of this is obtained present or prev ous
occupation from former husband
The "senior breadwinner" is the oldest full-time occupied male member of the family unit or, if there is no such man, the oldest full-time occupied female member of the family unit. If there are no full-time workers, take the oldest part-time occupied male member of the family unit, or if there is
no such ran, the oldest female part-time worker.
IF "NONE", WRITE "Dependent on state pension" BELOW
xi
Enter in this column occupational details
OCCUPATIONAL DETAILS REQUIRED: of the person indicated above in X
What type-of firm or organisation does/did (........... ) work for?
(STATE TYPE OF FIRM, WHAT FIRM MAKES/DOES, etc. )
What job does/did (...........
actually do?
IF IN CIVIL SERVICE, FORCES, POLICE, etc.
What is his/her rank or grade?
IF "OTHER" Does/did he/she hold any particular position in the organisation? (e. g. foreman, typing supervisor,
office manager, company secretary, etc. )
IF PROPRIETOR CF BUSINESS CR A 25 o. more MA NIAG EA
10 - 24 Roughly how many people work at the place where (......... "") Less than 10
works? (INCLUDE THIS PERSON) (STATE NUMBER-
ASK FOR ALL Has ...........
) any qualifications? (Such as apprenticeships, professional qualificat+ons, university degrees, diplomas, etc. ) (STATE WHAT QUALIFICATIONS HELD)
87
Xli NOW ASSESS S. C.:
AB CI C2 D E
Xiii Have You a TV set In your home?
Yes 18. 12
ýº'ý' No A
IF 'Yes'
Can your set (any of your
sets) receive: Yes No
(i) BBC I A A
(ii) BBC 2 (18. ) 11 A
(iii) ITV 0 A IF 'ITV' On average how 19. 1 or less many hours do 2 you watch 3 independent q TV progrannes 5 each night? 6 or rare
Does your set (any of your sets) receive colour?
Yes (19. ) 9
No A
:. <'w ý.: y` '.? k'ý,
i
16PF booklets only in the main study, subjects recording their preferred
answers by ringing the appropriate choice to each item directly in the
booklet. Although this had the disadvantage of increasing the material
costs considerably,. the important benefits resulted of shortening
response time by about thirty minutes and removing one clerical stage
from the operation in which transcription errors might occur.
As further outcomes of these preliminary inquiries, it was also
established that the following controls would operate in the main
standardisation sample: (1) All testees would be in the 16 to 70
year age range. (2) All testees were to be literate and capable of
reading the booklet for themselves. If required, help was to be
given with occasional words and sentences. In the case of the subject
being unable to read either through illiteracy or severe physical
or mental infirmity, the interview was to be diplomatically terminated.
(3) All testees were to hold a six-month residence qualification in
the British Isles prior to November 1971. This was imposed in order
to exclude overseas visitors and temporary immigrants.
THE TRAINING OF THE MAIN FIELDFORCE
Armed with the information from the pilot studies, certain questions
were made more clear and materials printed and collated for the main
fieldwork.
From editors' ratings of the total B. M. R. B. fieldforce, 200 experienced
fieldworkers were chosen and briefing sessions set up during October
and November 1971 in five main centres: I
88
London: South East England
Bristol: South West and Wales
Birmingham: Midlands and East Anglia
Manchester: Northern England
Glasgow: Scotland
The author took the necessary materials which were arranged in packs
to each of the interviewer briefing sessions by road. The written
ins __tctions prepared or ±ach interviewer are given below. At
each centre the interviewers were carefully taken through these
instructions step by step.
INTERVIEWERS' INSTRtLTIONS PREPARED FOR THE BRIEFINGS
Written instructions were prepared at the briefing sessions and a
copy provided for each of the fieldworkers. These are reproduced
on the following pages and can be distinguished by a vertical line
drawn down the side of the page. º
INTERVIEWERS' INSTRUCTIONS
Together with these instructions you should receive:
Three contact sheets Ten pens One red biro One questionnaire booklet containing Form A and Form B
(serial numbered) Twelve answer sheets Form A and twelve Form B Twelve questionnaires (serial, numbered)
Purpose of the Study
Basically, the aim of this study is to explore the interests and
attitudes of various types of people. You will probably find that,
unlike most surveys, this study may well arouse people's curiosity
so it will be as well if you yourself have as full an understanding as
possible into what this is all about. You will find a long Introduction
89
for you to use when speaking to the respondents at the top of your
Contact Sheets. This explains that:
1. This is a National Research Study and not a Market Research
Survey.
2. The Research Body is interested in people's opinions,
attitudes and interests. The questionnaire is a standard
inventory used in the social sciences and the main purpose
of the project is to get British data on it.
3. Similar studies have already been done in other parts of
the world (e. g. America, Europe) but have never been done
here before.
4. The results of this Study will help the Research Body
amongst other things to build up a picture of what people AP
are like in different countries.
By explaining this to your respondents, you will probably find that
they are interested and willing to help. It is important that the
response rate is kept as high as possible.
Contacting Respondents
You are to contact ten adults above the age of 15 within the streets
with which you have been issued, namely:
5 men
5 women of whom 3 should be housewives, 1 should be a working houswife and 1 an other woman
90
A six months' residence qualification in the UK is to apply to the
sample.
We would like to know about every contact you make who is eligible:
so please use your contact sheet every time. For those contacts who
refuse to take part put down their estimated aqe and social class,
? tnei. _asors for rcý, usal. For yo_. -- successful contacts wri `c°
down their questionnaire number in the space provided.
When introducing and discussing this interview do remember always
to use the word study and never survey - this is very important.
Each interview will take about an hour and a half, so in order to
be able to complete the interview, it is important that you do not
mislead respondents as to how long it will take. Remind him/her
that all the information collected is strictly confidential. 40
Before starting the interview, do check that the booklet and the
verbal questionnaire have the same serial number on them. If this is
not the case then always alter the numbers to the number on the verbal
questionnaire.
The Form of the Interview
The interviewing procedure is really quite straightforward:
1. The interview starts with you giving the respondent the booklet,
instructing him/her to read the first page, then turn over and
start answering the questions using the pen to ring the answers.
It may be as well to remind the respondent that all their answers
91
are kept absolutely confidential.
2. When the respondent reaches the half-way point in the booklet, to
give him/her a short break, you will ask him/her the questions in
section I (only) on the verbal questionnaire.
3. Thurn, as:. the respondent to return to the booklet, first reading
the instructions and then answering the questions as before.
4. Administer section 2 and 3 of the verbal questionnaire, remembering
to use the Introduction at the top of section 2. Then collect
classification details.
N. B. 5. Please write the serial number that is on the verbal questionnaire
onto the front cover of the self completion booklet before handing 40
it to the respondent.
The Questionnaire
The questionnaire is very straightforward and only a few points need
to be made: -
1. Remember to check whether the respondent is married before asking
Q. 1,2,3.
2. Q. 2. We want to know about all children under 21. If the respondent
has no children under 21, then do not ask Q. 2 (b) or (c).
3
92
3. Only ask Q. 3. of married women under about 40. If you find the
wording of this question rather abrupt in some cases, then you
may alter it to something like 'Are you likely to be having a
baby in the near future? '.
4. Q. 6. If the respondent wears contact lenses and spectacles, then
only ring the code 32.11 for contact lenses, because this is
mainly what we are interested in.
5. Q. 7. Here we are interested in birth order i. e. whether the
respondent is the first, second child etc. If the respondent's
brothers or sisters are deceased, we are still interested in the
answer to the question.
6. Remember to check respondent's age if he is male before asking Q. 9. 0
7. Section I ends after Q. 9. At this point instruct respondent to
return to the booklet and at the end of the second half of the
booklet, continue with Q. 10.
The Booklets and Answer Sheets
1. Some of you will have the booklets stapled in the order A, B, and
the others in the order B, A. This order must not be changed in
any way.
2. As mentioned earlier, the respondents will be asked to answer these
questions in the booklet itself by ringing the most appropriate
letter, 'a', 'b' or 'c'.
93
3. Whilst the respondent is answering the questions:
(a) Give help with instructions if necessary, but do not look over
the respondent's shoulder as they answer - guess:
(b) Stress that they are to work quickly, giving the first natural
answer as it comes to them, and tell them not to spend time
pondering over questions.
(c) If the respondent wishes to change an answer, get him or her
to clearly cross out their incorrect answer with the red biro.
(d) If the respondent asks for advice, merely say 'it is up to you,
or make some other equally noncommittal remark.
4. After the respondent has answered the first questionnaire, and y . lst
he or she is completing the second one, check that the respondent
has given ONE and ONLY ONE answer to each question. On completion
of the second questionnaire check this one also. If there are any
instances where more than one answer has been given, refer the
question back to the respondent and ask them which answer they
finally wish to decide on, and remember to cross out the incorrect
replies with red biro.
5. Recoding of respondent's completed questionnaires
(a) For every Form A and Form B questionnaire there is a corresponding
answer sheet. You are to recode the information from each
questionnaire onto the PINK answer sheet for Form A and the
BLUE answer sheet for Form B.
94
(b) Follow this procedure
First of all write the serial number that is on the front of
the booklet onto both answer sheets, A and B.
Then at the bottom of Form B write whether the respondent
is male or female. The appropriate word should go at the
bottom right-hand corner beneath the large column of empty
boxes.
Now recode page by page. One page of the questionnaire
corresponds to one complete column on the answer sheet.
Transfer the answers from the questionnaire to the answer
sheet carefully, ensuring that you are transferring the
respondent's answer into the correct little answer box 'a',
'b' or 'c' and that at all times the number of the question gº
corresponds to the number on the answer sheet.
N. B. A constant check on this is that whenever you turn a page
in the questionnaire the answer to that first question should
always go in the first set of answer boxes at the top of a new
column.
When all answers have been transferred and you are quite
satisfied that there are no errors due to putting an answer
in the wrong set of boxes for that question, please GO BACK
AGAIN AND RECHECK THE ENTIRE RECODING.
When completed, questionnaires and answer sheets will be
inspected for accuracy before analysis.
95
(c) It is suggested that the recoding of answers be carried out,
whilst the cuestionnaires are being completed by the next
respondent at the following interview.
(d) If you make a mistake cross out in red biro - if the answer
sheet is a total mess (eg missing out one question, leaving all
Cl _e.:. ions out , ý7r up icate.
6. Gen egal Points
(a) If the respondent is unable to read, diplomatically terminate
the interview. Give assistance on odd words. Tolerate slow
readers.
(b) In the questionnaire, the first question refers-to 'test'.
If this is questioned, please explain that there are no rigIft
or wrong answers and that the questionnaire is not a test in
any sense. It is unfortunate wording.
(c) If the respondent completes the first questionnaire, but
refuses the second, please explain the importance of this study.
If necessary, ask if you could come back at a more convenient
time. (Under no circumstances should you agreed to leave the
booklet and return for it when completed later - you must always
be present when the questionnaires are being answered. ) If the
respondent still refuses to complete the second booklet, the
first booklet and answer sheet should still be submitted, with a
bold note across the unused answer sheet stating 'Refused to
Complete'. But collect classification information if you can.
96
(d) There is a chance although a very unlikely one, that the
respondent may recognise the questions. If he is obviously
familiar with the materials, then merely say that you are of
course involved with the standardisation of the questionnaire
in the United Kingdom.
this stage the interviewers were actually put through the 16PF
questionnaire themselves. This was to ensure that they would be
familiar with the item content of the inventories.
INSTRUCTIONS ON COMPLETING THE QUESTIONNAIRES
In order that you can become fully familiar with the materials you
are going to use, I am going to ask you to complete the questionnaires
yourselves.
0
HAND OUT THE. QUESTIONNAIRES
Please don't get the materials confused.
Now, some of you will have a white page uppermost arid others a
blue or pink page uppermost. If you have a white page uppermost,
will you please sit on the right hand side of the room; if you
have a blue or pink page uppermost, please sit on the left hand side.
ALL MOVE
Thankyou, now I will deal with this side of the room and Miss Bodger
with the other side.
97
You all have a white page uppermost, will you please answer the
questions in the first booklet by ringing the best answer for you
in the booklet itself.
O. K.
? d1 ii `tOU ri. ): L , ý5e r=: a `_? hit lY Lr,, ct1. ii1J dill, :. 'I read 1
aloud.
READ
What to do:
Inside this booklet are some questions to see what attitudes and interests J
you have. There are no 'right' and 'wrong' answers because everyone has
the right to his own views.
Vr
There are three possible answers to each question. Please ring the letter
'a' or 'b' or 'c' to indicate your answer choice. Remember that you should
have one answer for each question.
When you answer please keep these four points in mind:
1. You are asked not to spend time pondering. Give the first,
natural answer as it comes to you. Of course, the questions
are too short to give you all the particulars you would sometimes
like to have. For instance, one of the questions asks you about
'team games' and you might prefer football to cricket. But you
are to reply 'for the average game', or to strike an average in
98
situations of the kind stated. Give the best answer you can
at a rate not slower than five or six a minute.
2. Try not to fall back on the middle, 'uncertain' answers
except when the answer at either end is really impossible
for you - perhaps once every four or five questions.
3. Be sure not to miss anything out but answer every question,
somehow. Some may not apply to you very well, but give your
best guess. Some may seem personal; but remember that the
booklets are kept confidential.
4. Answer as honestly as possible what is true of you.
Do not merely mark what seems 'the right thing to say'.
Please ask the interviewer - in this case that's me - if Ap
anything is not clear.
O. K. ? Any questions?
Will you please write at the bottom of the white page the time
you are starting this first questionnaire.
PAUSE
When you have finished you should also put the time on the white
instruction page.
I
99
It you have any questions, please ask theam now. Once you have
started, please don't confer with your neighbours.
PAUSE
Fine - Would you like to begin now, remember to work quickly
g; vn Your :E _oL natura answer.
Check that all is going well. After twelve minutes say'Work'as
quickly as possible. When the interviewers have finished, remind
them to put the time they finished if they have not already done
so and check.
Fine - Now we will do the second form. This one however, you
will be answering on the answer sheet provided. Would you like i
to now read the instructions on the next page and I'll read them
aloud.
READ
(From outside cover of 16PF booklet)
Well, thank you all for your help. Now I am going to ask you
to recode the answers which you put in the booklet, which you
completed first, onto an answer sheet. We will now hand out the
relevant answer sheets.
PAUSE WHILST ANSWER SHEETS HANDED OUT
Follow the recommendations in your briefing instructions - this
will stop errors creeping in.
loo
Recode page by page. One page of the questionnaire corresponds
to one complete column on the answer sheet. Transfer the
answers from the questionnaire to the answer sheet carefully,
ensuring that you are transferring the respondent's answer
into the correct little answer box 'a', 'b' or 'c' and that
at all times the number of the question corresponds to the
on the ? nswer sheet.
A constant check on this is that whenever you turn a page
in the questionnaire the answer to that first question should
always go in the first set of answer boxes at the top of a
new column.
When all answers have been transferred and you are quite
satisfied that there are no errors due to putting an answer
in the wrong set of boxes for that question, please GO BACK
AGAIN AND RECHFCK THE ENTIRE RECODING.
TIME
Go round check on the Interviewers' method of recoding.
TIME WHEN RECODING COMPLETED
40
Fine, now you have each done one Form of the questionnaire by
marking your answers in the booklet and one Form by marking your
answers on an answer sheet. In the pilot study which we carried
out we found that too many respondents were finding difficulty in
101
using the answer sheet. For this reason all respondents will be
answering by using the booklets for ringing their answers 'a', 'b'
or 'c'. Now you will recode these answers onto an answer sheet for
each respondent. We want you to do this so it will be done
accurately, the respondents might make mistakes if using an
answer sheet.
Remember to recode as laid down in the instructions. I will go
over them.
READ
CONCLUDING REMARKS
Are there any outstanding points? As was stated in the introduction,
this is a very important study and this the first time that it has
been carried out in Britain. The usefulness of the results carer ?t
be emphasised enough and, of course, good results now depend on you
and your colleagues.
THE BACKGROUND QUESTIONNAIRE
In addition to the personality questionnaire completed by the subject,
the interviewer was required to complete a background questionnaire
for each respondent (Figure 4.2. ). Information on the major questions
in this part of the survey are as follows: -
(a) Standard Region
Apart from the East Midlands and East Anglia which were treated
as one unit, the ten geographic regions employed in the study were I
102
those of the Census General. The total adult group
was made up of testings conducted in the following forty-eight
British counties: -
Northern England: County Durham, the North Riding of Yorkshire
and Northumberland.
Yorkshire and Humberside: The East and West Ridings of Yorkshire and Lincolnshire.
East Midlands and East Anglia: Derbyshire, Leicestershire, Northamptonshire, Nottinghamshire and Cambridgeshire, Norfolk and Suffolk.
West Midlands: Shropshire, Staffordshire and Warwickshire. --
Greater London: Testings were conducted in a total of 30 Parliamentary constituencies within the
administrative boundaries of the Greater London Council.
South East England: Bedfordshire, Berkshire, Buckinghamshire, Essex, Hampshire, Hertfordshire, Kent, Oxfordshire, Surrey and Sussex.
North West England: Cheshire and Lancashire.
South West England: Cornwall, Devon, Dorset, Gloucestershire, Somerset and Wiltshire.
Scotland: Aberdeenshire, Angus, Dumfriesshire, Dunbartonshire, East Lothian, Fifeshire, Lanarkshire, Midlothian, Perthshire and Renfrewshire.
Wales: Brecknockshire, Denbighshire, Glamorgan, Monmouthshire and Radnorshire.
103
(b) Social Grade
Rather than rely on information such as economic standing or
educational level, which undoubtedly affect social class in
varying degrees but which are often difficult to assess objectively,
social grading has been based upon the occupation of the head of
tee family unit as the single factor most responsible for the
social status of the family. For this reason detailed information
was collected on the job, position, type and size of organisation,
and any professional qualifications of the senior breadwinner.
Social class was then carefully assessed by way of a detailed
guide of which the following serves as an illustration.
Social Grade A: The Upper Middle Class
Professional and Semi-Professional: Doctor, Solicitor, Senior 40
Civil Servant (Principal and above), University Professor,
Chartered Accountant, Airline Pilot.
Commerce and Industry: Proprietor, Senior Executive, Company
Director and Manager of large business (25 or more in establishment),
Senior Industrial Scientist.
Police and Armed Forces: Superintendent, Chief Constable, Lieutenant
Colonel, Wing Commander, Commander (R. N. ) and above.
Social Grade B: The Middle Class
Professional and Semi-Professional: School Teacher, University
Lecturer, Executive Grade Civil Servant, Journalist.
Commerce and Industry: Owner of medium size business or shop, r
Managerof small branch office of bank or insurance company
104
(10-24 in establishment). Manager of a department in a large
factory or business.
Police and Armed Forces: Inspector and Commissioned Officers in
the Armed Forces.
Social Grade Cl: The Lower Middle Class
Professional and Semi-Professional: Junior Civil Servant, Bank or
Local Government Clerk, Pharmacist, all Students other than
Apprentices.
Commerce and Industry: Lower Managerial and Supervisory Grades,
Clerk, Typist, Laboratory Technician.
Social Grade C2: The Skilled Working Class
All skilled manual workers: Foreman (manual), Deputy (mining), sip
Chargehand, Overlooker, Overseer (manual), Superintendent (manual),
Supervisor (manual).
Social Grade D: The Semi-Skilled and Unskilled Working Class
Most semi-skilled and unskilled workers. Most adult female manual
workers. Unskilled assistants, labourers and mates of the
occupations in grade C2, all apprentices to skilled trades.
Social Grade E: Those at the Lowest Level of Subsistence
Social grade E is assigned only to those households which are
dependent on state pensions etc. and where no member of the
household is working. i
105
(c) Marital Status
Marital status was classified by one of four categories:
(a) Married, (b) Widowed, divorced or separated, (c) Single and
engaged, (d) Single but not engaged.
^1: Working Hours
Information was gained on whether the respondent was working
full-time, part-time or not working at the time of the study.
(e) Terminal Education Age
The age at which the subject completed full-time education
was established. If a respondent had worked for a period
between uninterrupted education, the most recent age was coded.
w
(f) Head of Household
The head of household was defined as the member who owns or is
responsible for the accommodation occupied by members of the
household. Suitable criteria were laid down to ensure correct
evaluation in the event of unusual household composition.
(g) A2
Subjects were required to give their age in years at their last
birthday; if this was refused, as was the case in approximately
20 instances, age was estimated and then coded by one of seven
intervals. i jr
106
FIGURE: 4.3.
THE SAMPLING UNITS AND STANDARD REGIONS OF THE 16PF
BRITISH STANDARDISATION SAMPLE (Nm 2227)
i,
107
Once all the forms had been received, edited, scored and card punched,
the data viere initially set up on the NFER IBM 1130 computer system
which was to undertake the first stage analyses. The more sophisticated
analyses at the second stage were to be undertaken by a much larger
computer installation at the Edinburgh Regional Computer Centre on
an IBM computer.
DATA PROCESSING
As subjects were tested the completed questionnaires were returned
for editing and coding of the social class data. Checks were also
made on the coding of responses from booklets to answer sheets and
those with complete and correct data were passed for handscoring,
cardpunching and further editing by computer.
When training the interviewers, the importance of control over the
quality of data had been stressed and in fact the vast majority of
scripts received for editing were complete and accurately coded.
In twenty-one cases either only Form A or Form B of the 16PF had
been answered by the respondent but after editing, results for the
complete sections were sent for card punching.
As is always the case in fieldwork of this size, odd questions had
been omitted by subjects but rather than reject the entire record,
information was filled in where possible or simply left blank.
(Computer programs had been adapted to deal with this and the associated
problem of variable sample sizes in the data analysis. ) i
108
0
5. DESCRIPTION OF THE SAMPLE
109
w
FI
Ni
z U H
Cq H Q C 4 Q L
c-ý (J.
G; t.
f"N
W
E-+
E V C'' N0 IO
iC C' itD 'N Ic")
13 co co C) !M tom ; G? J]
'
I
-: J C , t
14 ä Cl! MN C7 lN O (O 11ý
"I`- UO t0 jq IG716 ; tD ' T. IN Ito j CO "o
-. 1 u- CL
I( I!
I
f
"m tD co CO L7
icA
M tý t7 tD Iq
ici IW iV
ISO IM CO 'l,
C
CL t° 0 CC) cri cD sm C7) !v 1- t, i
U nE
bM C) 7 III N
IM
to m6 aö :. n co ri Its ei U'ý
C
Q (j C
OO 10 to
i a, co in CO
I r,
117
tD Qi M G) "7 Ir NI CO I OO ; I. -
CI I r ýl cI co rx 11 ý ý; JvL
LTJ WU Cy C
r9Cý W lo ý-pü aNi I c '- .
f-,;
O I-Hi.. U N p'r, ý
, Ö: 6 öÖ p 'ý
1 ; I C) W , ýý
z} ,ý (7 V) Z (f) Cn
aýs _ QI 73 r
- n -0 0 C) C CO
''J V -O
n n
0; N NOC . p) C. _
U
ýCýWZfncn
1 i0
M ;ýN IN
(>J Ir L) "6
r- iN 'C'7 iN
c lci IM Nay
NC lw
r- NMN
LO In 7 -7 N C) N'CD
Ir IV IN
O) AGO'C: / IN
c N G1
ö
Q. 3
W
0 n< 4J10
a o
Ux
F- -
cD ICD iO c0
ca i. -
ij
N Co CO
Ln is cD
co jr": 1 v,
r IM N
N
Cl) , _-o ra
gý
N ; iC) ci N
ca a. Q
Vl
'D
°o c iV±
a
c9 C) a) x
ýý C
X cr.
r
I t
f
Cn !- iC,
.- .r Iq
}ý iN
4
i` N IfJ
IN
V7 iý C'
1
öio
Ö cý ýi ýN C- OO'IC
L7) C CfC
.1 .Y OCOp
CL S, o ,,
CD C) C) x cý m
W.
z 0 H H lzýc L H
'7 Q d!
r
", i V) 41 O &)
UH
H
f
Fwý ý,
C) ö. E
r- <o - o
C) ö. C to
C OO
C1 ýi
C) O
? fin.
O Q E R
_C 7
C) a 00 EL
0-i i C, 4 LO (n
MTM
C)ý11I ;5a -mai W0 N 1121 t: ýi
Ü u7 '7 `ýio'j
CL
CD c, ý Y, ýJ
ýJ ý`
ýC
ýýý M
f e-
iý iN
i
ýýl ýi
i
M
cam, l0Z
in
NN (/)
co N O IM
ý
0ý Lo
ico T
T cn
I re '0
I Cl?
N, - N
N I(14 O 04 C14
Nib m_
Ico
N CO c6
c 6
O M
CO N N
ý; 6? OCl) O O O NN (V
N CO L7 .- GO V ýL)
uý
{
N
p
y
ö Cu (1,
}}} C
} }
o N Cl) { co N
? 1
In L7 Ln L7 zr NM Ln a_
0L LN N
Q C
J - O
'o ä mm_Q l
, ýý ý u' ýJ U U co
l--__ ý
crvü, zc j .2 n<n
O
Ü CO
CO O a-O
O C N G E D
öi v c
0) U rn
i
n
i
3ý
7ý
r
111
DESCRIPTION OF THE SAMPLE CHARACTERISTICS
A total of 2,227 subjects cooperated in the study. From an analysis
of the contact sheets it was found that the response rate was
approximately 78% and it was evident from interviewers' estimates
that those refusing tended to be female, older and of lower socio-
C_C than `> it ioipcants in the research.
The first computer run involved counts and percentages on the main
demographic and socio-economic variables contained in the background
questionnaire. These are represented in Table 5. l.. The population
estimates in the table come either from the Registrar General's
Annual Estimates and the National Readership Survey or from the
Target Group Index -a national survey of some 25,000 adults.
40 Before proceeding to consider the results in more detail, it is
important to remember that this sample specifically excluded the
illiterate, more seriously infirm and older subjects (70 years
and more). This is not the case with the population estimates
which can only be regarded as the best available statistics and not
the most relevant. The sample and population estimates do not share
identical parameters and direct comparison on certain variables
could be misleading. Whilst bias due to non-response is also
involved here, the greater proportion in the sample of those of
higher social grade for example, reflects to some extent the
controls used in the procedure.
it can be seen from Table 5.1. that in terms of region the sample I
112
seems relatively representative nationally with the maximum discrepancy
between sample and population estimates of 1.1%. With regard to social
group, the sample has a somewhat higher proportion of social class A,
B and C1 subjects but this is to be expected where older subjects
were excluded from the sampling frame.
mI! -re is also a relatively good representation of marital status although
the Widowed/Divorced/Separated category is a little low. Similarly,
by reference to the table, it is apparent that working hours, terminal
education age and age statistics show effects which are probably
due to the controls adopted in sampling mentioned above.
THE PROBLEM OF NON-RESPONSE
Alk
A dilemma which besets all researches of this kind and which must surely
make the exact measurement of the modal national adult personality an
unattainable goal, is the unknown degree and nature of bias introduced
into results by those who refuse to cooperate. From the very outset the
important practical consideration was recognised that adults have to be
willing to take part in research of this nature and that without full
cooperation it is impossible to apply true random sampling. It was
for this reason that great care was exercised not only in the phrasing
of the introductory remarks to the study, but also in the method of
recording responses, the provision of suitable incentives in the form of
a premium bond prize, the stressing of the confidentiality of results,
and freedom for the subject to be interviewed by appointment. Moreover,
the experience and known proficiency of the selected interviewers for =
113
gaining cooperation and their patience and persuasiveness in ensuring
that a session which was anything between 30 minutes and 3i- hours long
was not prematurely terminated, were also of crucial importance.
Although the refusal rate of 28% is somewhat higher than might have
been expected with particular subsets of the population, which are
,. s--:? 1', v rKo. _c e ;: dL av rec en 'r ýý "a1 3b I E? d? 1ý Lv? "c U_"r =i "ý? O'"1SE'ti: ^ýCý^r s (: r.
school 'leavers, for example), it is highly probable that this figure
is to some small extent spuriously high as a result of contacts,
ineligible for the reasons of illiteracy or age, declining assistance
before their ineligibility could be established by the interviewer.
As occurred in 21 of 2,227 cases in the present study, even on having
entered the home of the subject there remains the further danger that
the session will be terminated before completion. This issue of non- 40
response is of course enduring in studies of this type. Short of
providing fieldworkers With revolvers, an action which would secure
entirety but not necessarily accuracy of response, the researcher's
recourse can only be to efficient incentives whether they be social,
monetary or idealistic.
In all it is thought that despite certain limitations on response rate,
the sample will serve as a good data base to test the hypotheses
outlined in a previous section. Indeed knowing the problems of
collecting data from adult groups with an ab initio research design,
Cattell (1972) wrote " ... not even in America itself, where research
support is apt to be more lavish, has any one of these standardisations
quite reached the closeness to the ideal model that has now been }
achieved for Britain in this undertaking".
114
SECTION 3:
GR0UPDIFFERENCESAND
THEPRIMARYS0URCETRAITS
40
6. CROSS. -NATIONAL AND SEX DIFFERENCES
115
INTRODUCTION
Once data have been punched and verified, computer analysis proceeded
ui ti-I the calculation of descriptive statistics for the primary and
secondary source traits on each 16PF form combination (A, B, A+B)
according to the particular classifying variable under consideration.
with sixteen first order, eight second order factors and four derived
criteria measures, three questionnaire form combinations, three groups
for sex (males, females and both sexes combined), six age groups
and seven socio-economic classes, it is clear that an enormous volume.
of data was generated even at this early stage of the study.
The first step was to produce counts, frequency distributions,
means and standard deviations for each of the primary source traits
(16) by form combination (3) and by sex (3) giving a total of 144 tables.
These data were then converted by hand to the standard score system'O
used by Cattell, ie. the n-sten. This system relates raw scores to
a mean of 5.5 and a standard deviation of 2. Hence, standard scores
range from 1 to 10, from which arises the name 'sten' (standard ten).
Specifically, the n-sten is determined by using a normalised area
transformation: that is to say, by imposing a normal distribution
with the required mean and standard deviation in the reference
population. This procedure is discussed by Tatsuoka (1969) in
'Selected Topics in Advanced Statistics'.
Following the hand calculation of n-stens (these data are held
separately from the main thesis), a computer program was written to
automate the calculations and further tables were computed for separate
age and social class groups. These tables are contained in Appendix 6.1
116
In the case of cross-national and sex differences, 't' tests were then
calculated for each factor and these data will be incorporated into the
main body of the text. It is important to note that these results deal
with statistical significance as hypothesised in Cattell's model; they
do not of course necessarily imply practical significance for questionnaire
users.
There is the further problem that Cattell's primary source traits
are oblique factors, that is to say, there are non-zero correlations
between the scales. For instance, we may find that the difference
between the sexes on Factor C in our sample is of about the same order
as the difference on Factor Q4. Yet it may be that the difference found
is simply reflected in both sets of scores because of their inter-
correlation. It is important to note, therefore, that when scales are to
some extent inter-dependent, the use of multiple 't' test tends to
exaggerate underlying differences. But our main concern is as much
with the characteristics of individual scales as with the characteristics
of the sample. Whilst it would be misleading to suggest that multiple
't' test results give an accurate picture of the extent of the.
differences between various groups, it is certainly the case that they
give a good account of the characteristics of the individual scale.
The 't' tests which follow in the various sections are two-tailed and
adopt the conventional . 05 probability level or beyond as the standard
of statistical significance. It should also be pointed out that about
5% of the differences between means can be expected to be statistically
significant purely by chance, although this is unlikely in any way to
seriously distort our findings.
117
CROSS-NATIONAL DIFFERENCES ON THE PRIMARY SOURCE TRAITS
Because the British groups are of somewhat older mean age than the
American data provided by Cattell (1971)(mean 30 years), it was not
possible to test for cross-national differences without first having
taken account of age discrepancies. Accordingly, to bring all mean
factor scores to a common age, the British group means were corrected
on the quedratic model provided by Cattell (ly? 1)(see beloýý).
(a) Female Cross-National Differences: Results
After the application of the age correction weights, it is evident
from Table 6. l. that the British females were significantly higher
than American females on Factors L (p -:: L-. 01) ,N
(p c. 05) ,
Q2 (p< . 001) and Q4 (p G . 01) and significantly lower on Factors
G (p0.001), H (p<. 001), I (p C. 05) and M (p<. 001).
Differences were not significant on the remaining eight factors.
Then the British female group are described. as more suspicious (L+),
shrewd (N+), self-sufficient (Q2+) and tense (Q4+) and less
conscientious (G-), venturesome (H), tender-minded (I-) and
imaginative (N-) than the American female group.
(b) Male Cross-National Differences: Results
In Table 6.2. the respective American and British male groups are
compared. Eleven of the sixteen source traits are in fact
significantly different at the 5% level or above: British males
being higher on Factors B (p ` . 001), E (p C . 05), L (p-< . 001),
Q1 (pC . 001), Q2 (p<----. 00l) and Q3 (p . 01) and lower on
118
A (p C . 001), G (p x. 001), Z (p x. 001), M (P", _ . 001) and
O (p. < . 001).
These results give a picture of group differences of the British
male being more intelligent (B-), assertive. (E+), suspicious (L+),
experimenting (Al+), self-sufficient (Q2+) and controlled (Q3+)
than the average for. American males. The British male group had
significantly lower scores in terms of outgoingness (A-),
conscientiousness (G-), tough-mindedness (I-), imagination (M-)
and apprehension (0-).
SEX DIFFERENCES ON THE PRIMARY SOURCE TRAITS
The means and standard deviations for the male and female samples
on Forms A+B of the 16PF are to be found in Table 6.3.. Fourteen of
these differences are significant at the . 01 level or beyond, only
Factors F and G do not show statistically significant differences
between the sexes. Males had significantly higher means on Factors
B, C, E, H, L, M, Q1, Q2 and Q3 and females were significantly higher on
A, I, N, O and Q4.
These personality differences give a picture of the female group
being more tender-minded (1+), emotionally labile (C-, 0+ and Q4+),
conservative (Q1-) and less dominant (E. -) than the male group.
DISCUSSION OF THE CROSS-NATIONAL AND SEX DIFFERENCES IN THE PRIMARY SOURCE TRAITS
Results for the cross-national comparisons on the primary source traits
support the hypothesis of no difference between British and American
119
groups on Factors B (intelligence), C (emotional stability),
0 (apprehension) and Q1 (experimenting) in females and Factors C
(emotional stability), N (shr&v. dness) and Q4 (tense) in males.
In addition British females scoring lower than the comparable
American group on Factor Ni (imagination) and British males being
lower on Factors A (outgoing), NI (imagination) and higher on Q3
sc-1-conc? )4- cor trot ) ere all differences in . P. ? hyno hesi sed
direction.
Departures from hypothesis usually involved statistically
significant differences in results where none had been envisaged;
only Factor E (dominance) in males with the British rather than the
American group scoring more highly showed the opposite tendency
to that expected.
Thus the hypothesis that the British group would show a more introverted
personality pattern is only partially supported by the data. Whilst
differences on Factors A (British less outgoing) and Q3 (British more.
controlled) were in the predicted direction, no significant differences
were found on Factors F (happy-go-lucky) and H (social boldness). As
we have noted above, one major extroversion component in males,
Factor E (assertiveness), even showed the opposite trend, with the
British group scoring higher.
Factors which show consistent differences between the American and
British groups are G (Americans higher on superego strength), I
(Americans more tender-minded), M (Americans more imaginative) and
Q2 (British more self-sufficient). These source traits do not group
120
together in any major second order factor such as extraversion or
anxiety nor are they factors like E, C, O and Q4 which are particularly
susceptible to the social desirability response sets (Cattell 1973)
and so it would seem less likely that these personality differences
merely reflect dissimilar research methodologies.
.. ýý.. cu, h and other primary source traits reach statistical
significance,, Ae must not lose sight of the fact that the groups in
terms of mean differences are reasonably close. The largest difference
in Z-score units is . 37 for males on Factor Q2 which still represents
considerable overlap between the two curves. Certainly most of these
differences would have little effect on the standard score tables
(n-stens) used in profiling results.
What is probably more important to the reliability analysis in the "
next section is the apparent greater dispersion of factor scores in
the Arerican female group. Fourteen of the sixteen primary factors
show higher variance in the latter group (sign test, p-<-'. 05).
The mean scores being relatively close suggests a higher incidence
of more extreme scores in the American female sample. This may well
have its origin in Cattell's method of collecting data by pooling
results from distinct and less typical sub-groups of the population.
It could also be true that variations in the motivational set of his
samples is at work here. For example, the large differences between
variances on the emotional stability factor (C) may have its origin
in a response set for those tested under client centred conditions
(eg. therapy) to fake low and those involved in organisational
121
decision making (eg. job applicants) to fake high. Combining such
groups would tend to polarise the data, perhaps having little effect
upon the central point but increasing dispersion about the mean.
As a concluding remark, it is important to remember that because of
the nature of the American data, too much confidence should not be
? ', ä Gi SV Ci oss-r Flt.:. cr _ý ?: _ ýl "7cß'1:! j3 (19-C.. )
1S
pointed out if it is not really possible to specify the exact modal
personality of a nation, then it is not possible to talk definitively
of personality differences between nations. Inferences regarding
cross-national differences can only make sense within the context
of sample characteristics, the research design and the conditions
of testing. We may conclude, however, though differences are often
statistically significant, their actual size is usually small and
for our purposes, we do not appear to have particularly unusual
data with respect to Cattell's. n
Turing to the sex differences on the primary source traits, ten of
the factors showed differences in line with those hypothesised.
These were namely Factors A (females more outgoing), E (males
more assertive), H (males more venturesome), I (females more tender-
minded), L (males more suspicious), *O (females more apprehensive),
Q1 (males more experimenting), Q3 (males more socially controlled),
Q4 (females more tense) and F (no statistically significant difference
between the sexes). The hypothesis of no difference between the sexes,
however, must be rejected in the case of Factors B (intelligence),
C (emotional stability) and Q2 (self-sufficiency), where males
achieved a significantly higher mean score than females. Moreover,
122
although the mean of Factor G (conscientious) was higher for the
fer; aie group, the difference failed to reach statistical significance.
Finally, on Factor M (imagination), with British males higher and
on Factor N, with British females higher, reversals of the expected
sex differences in personality are evident.
r_ st: nele , r: ýup scoring significantly ; nigher on average than
the female sample on Factor B (intelligence) is perhaps a little
surprising, especially in view of the fact that the females were
of some hat lower mean age than the male respondents in the study.
Reference to Cattell's (1973) norm supplements shows that in just
about every case, the American male and female groups have identical
means and standard deviations on this factor. This suggests that
Cattell uses this same trait as a classifying variable when
compiling his tables and, if this is so, these data call into
question the validity of the method. 0
The sex differences on the primary source traits were in fact consistent
across both Forms A and B, except for Factors G (males higher on Form A,
females higher on Form B), N (females higher on Form A, no significant
difference on Form A, males higher on Form B). Why these factors
should fluctuate in such a manner is not immediately apparent. It
may be significant, however, that as measured by coefficients of
equivalence between Forms A and B, Factors G, N and Q2 are three
of the least reliable source traits (Cattell et al, 1970).
It should be remembered that these sex differences are based upon
groups of mean age approximately 38 years. The nature and extent
123
of these differences can only be generalised to other ages with
great caution. There is evidence that the 16PF personality factor
age changes in the American culture at least, show dissimilar
patterns for the sexes, (Cattell, Eber and Tatsuoka, 1970).
For example, crystallized intelligence demonstrated a slight
fall in women on Forms A+B in the 1961-62 edition of the 16PF
^h er^. i . ̀-din for r males. al The . "'yamestL. ion o `Jerson iy is not d_sc :
... '
differences between the sexes at various points in the age
distribution will be taken up later in this chapter.
As concluding remarks, we can say that the British sex differences
on the primary source traits are roughly in line with Cattell's
model and results reported in the united States. The differences
between males and females in personality were more predictable
than the cross-national trends, although this could have been due
in part to comparing data from studies which used very different
AP
methodologies. It is certainly true to say that from these results,
the sex of subjects matters far more in understanding personality
than their nationality.
4
124
4I
H
I..
0 M
W U
0 EA
A W
:n
O Nm N m N V
N
O
NO N
OI N
Q Q
4 f0
m R
p Q °
\G
ö R can v cv
N ° ° r�
Z
O MR
N
( N
M I°
I
Q
p
-
V N
1
S
R^
' NV
N Jn
ý
Q _' N
Q
M Q
{
Qf R
O CJ
m O
Q
6
O
J
V L7
Q
O
C
M
ýA ;
ö
O
V ý
N O
u'f Q N
;n O 1
N ý�
m C7 f")
MO N
Oý tD
L! i N
OJ V
O
H ^
M
00
V
LL
W
R m 1_
N
O^ mO
t0
m C7 U
ýJ
lLi
Ct
(
C"ý
R
N N
R Q
M
N
°
N
R
v
M
O
O
r
p
O
ýI
N
V
V
G
öR
v OV
!JIQ
-I
i,
r
ÖI
r. ý
C]
"-
"
M M °
N
0
v
°
°
o
ci
[h _
Z
9qO Q mi N
I 6
fn 0..
G_>
CS C q C2 �
C G1
Q
v
I ! [1 G 1 CI '. ' V.
a
12ý
(V
W
W 4
C7 ß.
0
0
H
H E-4 H
z u
H
w Uz W a w w
Q
H ý-7 4 z 0 vý w a h Q U H HIS
H ýl
j-ý
UNý
14-4
O M
0
H U7
ýI
V O
OU V
p
t7
_ N _
cý c Of
T
OU N
C1 p
1
d
^ ps
v co NSW ýJ
7 ýQ ý- p
or .r `i
ý ^m N .. < °' 'Q ö
C 10
C7 N(ý
dU
NU
O
O 1
i
IS u CD _
` OO : ") r
_
MO
0
O
O
hfl 7
QN 4
L7
U N
O
I
I
O
ý
pO
ýn
O Ca
U
R-
v ', (O
:np
m
CJ
O 1
I
"
t7
'f N, m
Oa Lfl
T lý1
"mal ý.
O1 (7
O
O
9uG Cý h
O ICN
.nZ
©f G
yu
01
v
WO I'ý O
V
O
O O
jv
cJ
N
O O
V
n! O OO cc v
cý V
O Uý
VZ Ü
N Lý
O_
NV
(. l
U9 O nO
C., `i
6
UO O
QV
C
V
O
Ci
?I
ýJ
i
126
rý: a C1 H
z gW
ti
z
H
ro
r-1 H
V)
C) c O
C: r ,-
_ O
I
O
Ö to p `i
.n Ci : tt!
R O
c? N 'ý
d m QQ N r, ý ýr/
n V 'r
0 - `
r; n t
u ö N j
V
ý ý! r! c UO
N N V
n o2 ö LS N 7
tO q fý O
b M
ý
r
N
dO
(ý
r M
O m
C. 7 ý
Q N
ýr
LLý tý N
OH O
LL N NrV tL- . y V OG O
W O M! C ', C7 hO
V O
J ý rv 'n M ýO
o dNv N a ýJ
n c c
- c-i ^n ý M! 1/
`
m' NN , rý O
Q N O UG
N i v
L O y CO
L to
C `c
v=
C1 o Y c+
VN
jo
127
N rn x x
O vý v7 ý
ý:: ( ý H 1
m H
C. )
WI
H'I
W
Qli
ö
z CA V)
Hi
HI U2 f Ga
z
ri Q
CLI Q Ul U) V] (/l
z
z z z z
b v) vý U) Z z z
Q z ¢ w w
U U1 ", -I Vý W vý "d o)
E-1 a E, rý c a c a . U) ä äw ä C). ! )l
c
s4 (u 0 v "r-, a
-rA c ä 41 ö ä) ö 0 ý
W,
4 t)
.C 0 tr s4
.0 r1 0 to
0 "HI U
-P u 4
. .H 4 -1 i
V) C "rl U 4-i 0 rf "ri >1 "d CP 4-i
r-1 (' "d rl I ( a
u a a "d ä 0 0 Q ) 0. bý
- C 0 0) co
+3 U ul 0) N u1 "'. 4 "H eH
$4 -P v rs 9
vi d cn i ýý z
128
Aw
7. SOCIAL CLASS DIFFERENCES
129
INTRODUCTION
In an earlier section hypotheses were developed from Cattell's model
on personality differences between socio-economic classes. In this
study subjects were assigned to social groups classified on a similar
basis to that used by the Census General; these categories have been
described above.
Linear relationships have been hypothesised for the majority of the
primary source traits and such trends will be investigated by way of
conventional product moment coefficients. However, to test the
hypothesis that the regression of personality on social class is
non-linear, eta coefficients will be calculated on the data.
The correlation ratio or eta (Y) is a general index of association
particularly useful when analysing data in which there is a curved
regression. Eta allows us to predict on the basis of array means
and can be defined as the proportion of variance in the first variable
explained by and associated with the second. Eta assumes no particular
type of functional relationship between two variables and it can be
regarded as the maximum size of correlation between them. Like Linear
correlation, two regression lines are found for eta. However, these
may differ in shape and slope and so two eta coefficients can be
calculated:
dl y d
y
Eta for the regression y on x
130
N
N c4 H
W a H
a ö
0
H U)
a
a
a
0
0
z
z
a
H
Ö
Nx w
U) 0 ww
U
H Hq
U
O O 0) 1-4 %D N N I -i 'cr M ko ;: r N C) N i- CO 00 O O O. r-1 . -1 O O O
N M m l0 00 r-r O. N O CO 01 I- O 1.0
O "
r-1 O 9 O H O "
O O.
I 1 1 I 1 I 1 1 t
Cl v. co N CO lo off" S H CO M, l0 Lfl . N- N H r-1 O O O O O O O O O O
IT f'7 M: r CO t- IT N CO 110 % ö ON ý C ö e-1 r . -1
. C)
. C)
. H
p
" 1 1 1 t "
i
H H H O ri CV O r
Ö r i
CV N lJ Ol kD ('') ri fl O lfl tf) O ON L) f"1 CO Ln N O O O O O O- O O O
1 I
l0 N 01 Ln CV) ßl O C) LV) to C) 01 H" 01 tlý O i- N N N N r-1 N N, N N
1 1 ! 1 1 1 1 1 t
C) ýo ko (A r- C14 co 4w in C%j C) (n IV I MY en r- N (D O H r-I r-4 Cl O C) O j
O ý ko ON En L V r'r O . 0'. co
O O O. r "
tý . -r O O O
1 I i I 1 I 1 I" I
H CO %. 0 H O i -1 - O rI N r-{ r 1 - r-1
1 1" I 1 I" t "I 1ý
O in C) CO M C14 Ln c
In . 00 N C) Ö
O O O O O O. C) C)
1 1 1 1 I 1 t
' %D M lfl CO N co O r-r '-1 O O O O
" O O O O C)
i 1 1 t 1 N C
dý Ol H O co H CO "Cr r- A 43) - tD
" . Q, O O O O C) C) 1 1
H r Ln CV) CA C) 00 CO r 1 ß -I C) N ß in ý
O ri H N° 1 1" S 1 I I 1 t I"
00 Ö CO - 1` CO (W) %D H C*4 N f"1 l'M M
N CV)
M
1 I " 1 I I 1 I" U)
H M C . 1 - r-1 rl LO N
Ö CV)
Ö
O O O O O O O Cl
N O
In ýy^
co gOct
cr, h: ýv d aý ý: }
131
d1 Eta for the regression of x on y
Q x
where d1= standard deviation of the yl values predicted from x
d1= standard deviation of the x1 values x rr¬dIcted from y
d and d= standard deviations of the total distributions yx
We shall be concerned with the first of these, testing the hypothesis
that the regression of y (personality) on social class (x) is linear.
If we wish to test for the linearity of regression, Guilford (1973) has
proposed an F test based on an analysis of variance approach where only
a knowledge of eta and the product moment correlation coefficient (r) t
is required.
r2) (N - K)
(1- 2) (K - 2)
where K= number of columns or rows
SOCIAL CLASS DIFFERENCES: RESULTS
The product moment correlation coefficients between Cattell's primary
source traits and social class are contained in Table 7.1.. Because of
computer restrictions, these data are based on slightly reduced group
sizes in that only subjects with complete data records were included
in the analysis. The social class variable was coded by assigning the
i
132
S N
E-4
oa + d
O
y C1 b
Gý+
d' M pp rý t0 U1
O " z
co 0 I ri
" I
O "
Ol CO N N O O .N tt
a. z
rn cM 0 0 ýn ul
z O
co co C- O Q M cp
z z
N N r. Q
O N
Ln O ý-1 . -t rt ßl1 O
N CO O
O
rn rn ý
x o to
. LO Ln CD , -i o 0 0
o LO
` ö rn ö , -ý w O N
Ö d p ul
N N y
U " ö z t
t` c+1 M ul
z O N t 1
M iý
z
u ý v ýry! W
4-) u :: s u 41 -r4
4J -q v 44
44 ri
1J 4Ou 04 0.6 : E: U
133
character 1 to the highest social group (A) and the character 6 to the
lowest group (E). The number scale bears an inverse relationship,
therefore, to the social class variable. Those coefficients significant
at the 5M level or beyond are marked in the table. Appendix 7 also
contains the means and standard deviations on the primary source traits
for each social class, 16PF form combination and sex.
Turning to the product moment correlations, which of course are appropriate
only when investigating linear trends in data, it is evident that there
are sizeable correlations between social class and the primary source
traits of B (intelligence) and M (imagination). Two other factors
which show consistent associations. across form and sex are C (emotional
stability) and 0 (worrying, troubled). In addition, there are relationships
in the female group with Factors H (social boldness), I (tender-minded),
L (suspicious) and Q4 (tense). These findings are relatively consistent 40
across the two questionnaire forms although Ql (experimenting) seems to
be related on Form B but perhaps in the opposite direction in Form A.
As to the direction of these relationships, higher social class in general
is related positively with Factors B, C, H, M, I and Q3 and negatively
with Factors L, Q4 and 0. This gives the picture of members of the
'higher' socio-economic groupsas more intelligent, emotionally stable,
imaginative and controlled and less apprehensive. Additionally, females
in the higher socio-economic groups appear more socially bold, tender-
minded and less tense and suspicious.
Eta coefficients, representing the regression of the primary source
traits (Forms A+ B) on socio-economic groups are contained in Table 7.2.
134
0 Mý
N co H
a 0 6 H
H
8 En
Ö
En H H
En U)
aW Cry
0 Lv
N
N U') 0 rl H 0 0 O
O M
co Lr) -4 Cf) r1 O M 9 a I M
ul N 0
N O O O V1
a r-I ýD d'
H O r-I V]
ff i N
z
N N tp
O z N
0 1 d Ö
N
O 0 0 M M 0
z i z N m
O 0 01 a z
p ö ö -i 0
d'
O ö x " z
a ö H i N
e CC) 4 r1 H O lp ° " N
In fV W V " " " z
N
M
" ý
Ö 0 O ° `n Q . z N
U U C ý.
C N u
ý U (a 4-1 U
4 j ` V v - w to ä 8 ý' '-
4-4 (a 0 t o. ääßü U)
135
With regard to the data for females, the F values suggest that six
of the 16 factors show significant departures from linearity. These
are the primary source traits of F,, G, I, L, M and Q3. The corresponding
analysis on-the male data (Table 7.3. ) shows significant F values for
: 'actors ß, 1;,, F, C, I, N, Q3 and Q4.
SOCIAL CLASS DIFFEhENN; CES: DISCUSSION
The largest coefficient is between social class and Factor B (intelligence).
so as the cu t-egori_ attion of social group is based upon the
occupation of silo hoadý of the family unit, this is perhaps not unexpec'ýed.
As is evident from the correlation matrix reproduced in Table 7.4.,
social class is also positively related to the amount of full time
education received by the subject, which clearly is likely to have ý.
some effect on a verbally orientated ability scale like Cattell's Factor B.
Another primary trait with a significant relationship with social class,
is Cattell's Factor M (practical vs. imaginative). This variable shows
a significant departure from linearity and reference to the group means
and standard deviations (Appendix 7) suggests that the regression line
tails off slightly at the extremes with the maximum acceleration taking
place between social groups Cl (the lower middle class) and C2 (the
skilled working class). It is interesting that it is precisely this
>I I; ýti':. _ch _ I'csec' _ divisic be t. en the professional or 'ideas'
jots or 'thing' orientated occupations. These
results are in line, therefore, with the initial hypothesis that those
in the more practical occupations would score lower on this factor.
136
TABLE : 7.4.
CORRELATIONS BETWEEN AGE, EDUCATIONAL AGE AND SOCIAL CLASS
IN THE 16PF FORMS A+B ADULT SAMPLE: BOTH SEXES
(N = 2012)
AGE EDUCATIONAL AGE SOCIAL CLASS
AGE 1 -. 41** -. 01
EDUCATIONAL AGE 1 -. 37**
SOCIAL CLASS 1
40 correlation significant at . 01 level or above
40
I 137
Three other trends in the data which support initial hypotheses all
concern the major anxiety variables in the 16PF with the higher social groups
being more emotionally stable (C+), less apprehensive (0-) and, in females,
less tense (Q4-). Only Q4 for the male adults showed evidence of a non-linear
relationship . In addition tender-mindedness (I) in females showed a
significant correlation with social class, which again supported the
ot'r. e. is of a slight. positive relationship.
In all, results supported or tentatively supported hypotheses on some
ten of Cattell's primaries: -
FACTOR
A Outgoing
B Intelligence
C Emotional Stability
E Assertive
H Social Boldness
I Tender-minded
L Suspicious
M Imagination
O Apprehension
Q4 Tense
SOCIAL CLASS DIFFERENCE
No differences
Positive relationship
Positive relationship
No differences
Positive relationship - (females) No differences (males)
No differences (males) Positive relationship (females)
Negative relationship (females)
, No differences (males)
Positive relationship
Negative relationship
No differences (males) Negative relationship (females)
Of the six unsubstantiated hypotheses, probably the most serious
concerned the Factor N (shrewdness) where no clear relationship with
social class was found despite Cattell's contention that this trait
is closely associated with social status.
I
138
a
$. AGE DIFFERENCES
139
IN'T'RODUCTION
Cattel? (1972) points out that the curves of age trends in personality
need to be determined in order to make proper comparisons among
individuals of different ages. Beyond this consideration, reliable
information about the development of personality traits is also
fundamental for personality for as Ang]eitner (1971) has shown, the
age dependency of personality scores shows in every culture yet examined.
Cattell analyses age curves as the sum of endogeneous and exogeneous
components. The former describes the effects of internally determined
a maturation and involution and is virtually synonymous with a pure
genetic component as it manifests itself over time. The latter is
made up of all the various averaged environmental influences on the
individual and can be divided again into an ecogenic, representing
what happens to all people as an average across all historical human
ecologies and an epogenic component expressing the effect peculiar to
an epoch and a given culture.
Methodologies for separating the ecogenic components have been discussed
by Baltes (1968), Cattell (1970) and Schaie and Strother (1968) in terms
of comparing longitudinal (the follow-up of age cohorts with re-testing)
with cross-sectional data (the study of different age groups at one
point in time). However, Cattell's conclusions on age trends in
140
FIGURE-' 8 . 1.
ANALYSIS OF PERSONALITY AGE CURVES
ENDOGENOUS
genetic ccuponent)
EXOGENOUS
(nurture)
DCOGENIC
generalised effects across all human ecologies
Innate maturational and involutional influences on personality
Environmental influences on personality
EPOGENIC
41
effects peculiar to an epoch and culture
a
I 141
personality remain based on simple cross-sectional data where averages
on the primary source traits are available for a particular decade and
for groups of people at different age levels. These curves thus
confound the ecogenic with the epogenic components but are,
nevertheless, important as a first step in understanding age
differences in the primary source traits. Moreover, in that both
Iris Judy and Cattell's conclusions are based on a cross-sectional
design, it becomes possible to test for the similarity of age trends
in the primary source traits.
AGE DIFFERENCES: RESULTS FOR PRODUCT MOMENT AND ETA COEFFICIENTS
An indication of the hypothesised age trends in personality was
given in an earlier section describing the primary source traits.
Cattell (1972) has more recently provided a picture of the curves 0
for the primaries across an age range from about 5 to 60 years.
These curves are represented in n-sten units about the total group
mean and standard deviation. Two associated problems arise in the
comparison of Cattell's trend lines with those collected here.
First, Cattell's curves are for a different age base. In order
to ensure comparability, it will be necessary, therefore, to
restrict our considerations to changes over the adult years,
16-60 years. Secondly, this change in age range will need a
corresponding adjustment to the group means and standard deviations
upon which Cattell's standard score system rests.
}
142
* Eý
6 5 C4 w a w
a U) a
4 X H N
Ö w Ey H
9
a
w
4' I) N
0 10
0 10
0 M
O to
O CC)
O OD
O . -f
O CO
O tO
O In
In N
m (V
o O
C'J " t0
" I! 1 U1 V; If) If) ltd tý l0 l0 l0 N
" l4
" N
to In 0 0 0 0 In o 0 to 1n O 0 0 cn fV (V U) 0 O\ CO N In O CV N O N O Cf .
d' "
4l .
to .
t0 .
to If) .
d ..
trl r
lf) ul LI) LI) N
N In N
LO N
O to
11 N
Ln N
u1 N
O 0
10 IN
0 0
O 1-1 EN
o 0 lf)
cV C5
O O
i! ) to N It) to ul Lý I) 10 lD 11 V In N
{ o to
0 In
O In
0 In
0 Un
0 Le)
0 1n
0 In
0 V)
0 LI)
o II)
O In
0 to
0 O
Q . " . . . . . . . " " . "
N In I! ) I[1 In IA LA L1) In LA ý Ifl N
O !O
l0
O O
to
O U)
It)
I; N
If)
0 to
Ul
0 Lf)
CO
0
.n N
u1 Ca
N
0 O
N
i'') 5
l0
O Gý
10
u) CV
N
N d'
lD
O U
N
ýJ N
to
N
to
in
N
to
U)
0
lß
N
CO
0
tO 10
11
l0
t11
c0
O
N
0
E
N
l0
O
N
If) N
"
0 Oý
"
O u)
" to
0 ul
" In
0 0
" to
0 0
" d'
It) N
"
O In
"
O O
. ul
O O
" U)
to N
"
0
r 0
N N
"
p
" 0
N
ºa " %D
" 10
tin "
it)
d' "
to
I
" u)
" tO We
" Ott
0
. In
00) e
ýi'
O "
If)
0
" t[1
W "
0
O r
(%4
H O O
in
O r-i
In
0 in
in
tn N
in
0 O
l0
0 rt
l0
0 10
N
0 O
co
'0 H
a)
LI) C\t
O
0 0
co
O 0)
N
a0 N
to
0 p
N
x O d'
IT
O fý
It
0 tf
tn
in N
II)
In N
d'
0 to
cM
O 0
d'
O H
d'
0 c
d'
U) N
dý
0 0
. t'
0 c0
M
t[1 M
dý
N
N
%r
0 Ln d'
tin
in W0
to In
to
N
to
In
d' S
d' 0
In
N
to
0
to
0
to
0
u)
o
N
0
t0
N
u1
(D 1
in
in
0 0
to
d'
d'
u0'ß
c0
0
t0
)
to
N 4
N O)
M
0 O pip
.4
0
N
W O p
" In
O d'
" 10
O in
" U)
0 H
" U)
to N
"
0 N
"
O G0
" M
O 0
" 44
O CO
"
O It)
" M
O , -.!
" CO
LO N
" N
d' N
"
O O
" tV
0 p
d'
O U1
d'
O it)
ui
0 U)
(n
O N
tn
0 to
d'
0 CO
M
0 01
en
0 d" '
d'
0 mal'
d'
0 0
00
c*>
co
"
o
" N
O 0
" tIt
0 tO
" U)
0 o "
to
0 0
" In
O in
" [
0 "
d'
0 0
" ul
0 10
" to
O LO
r In
O 0
V"
0 LI d" '
0 0
ý3" '
N 0%
<! " '
0 0 N
AC Lý d'
O to
tý
to
LI)
to in
tO 0 U)
O I'll) U1
O 0
l0
O I lQ
001
U'1 If)
M
IA
p
N
C7 4
N r-i
N N
0 en
0 d'
0 u1
0 to r-I N CO d' U) cß t11
L S3'TVW
1 S3'IVw33 1
143
4 N
Wa
cn
Eý N a
6 A W N4
W',
N O H
W
a U W 4
13
ör M
W
O U
O d' O
" ('M '
" M I
"
N Ids
"
S IM "
N IN "
11 m N
"
r d'
"
o', N
"
O H
"
dý "
-i In "
M N "
O O
" tß N
m a
I 0 H d' m OM I d' (ý 'rA O O
0 0 0
" U1 N
Ný N O
"
N O
"
N 0
"
N 0
"
N 0
"
N O
"
lD I) "
N O
"
d' . -I
"
41 ,
"
CV O
" r-1 rI
"
tV N
"
0
" L7 N
Q E; O 0 O 0 0 O O O 0 0 0 0
0
tr! 0 Oi
U) (V
0 O +N O H i l)
cri ct
CV O
N LO
01 M
N N
d r
N cV
G M
N ý"
p
l0 N
z Ian I0
"
(M R "
IcM+1 .
I. -or) 1
"
Ö
" IrI
" O
co
" . -i
" ýT
"
(f) "
N
"
0
"
H lD N
d N
O O
O' M
M M
d' '!
t0 (M d'
IN r1 I CN I
d! , -{
4V O
l0 IM N N
0 O
[P CV
a r-1 dý r-i N
0 I
d' I . -ý Ol (O w
M M v O
ON N
O ýM d' M IO ýCf 0
ý! '3 N
m GO
d' (d0
d' I lfl
N I to
01 IM
Iq IM
rl d' rl lD
W lfl
d' N
r-I $p
W to
N O
D N
M 0
CC) . -ý
CD Ln d'
0 N
M I d'
ap I . -1 M ' . -1
M O
to IO 'M M O
N
O a $cn
" ýN
"
N
"
N C.
N
" H
" IN
" " '0
" O ce)
. O I0
"
04 O
" " Cl
Lf N
ý d1-1 '
M O IN Iw0 to to
M 10 t- IM ON $ '1 'Itt dl co OD 0 O
d' N
W M
" " 1O
" " N
" IO
" tV
" IH
"
(c'
" IcV
" Iul
" IN"
"N 0
0 N
V N Ö N N M
O ' IN IO !O N $M d' O
d' N
O "
M
"
N
" O 1w
"
r-4 IN "
ýD IV'
"
lq O "
M
" N
.
d' 0
"
H N f" tD
d' "
N M
"
O 0
" dý N
M ý Ö Ö (rl I'l Iý co Ö
cy) M (ý
m N ýp O
M O
u1 N
C7 4
tý r-1
N N
0 M
O d'
O In
O 0
N . -i
N N
0 M
0 d'
0 t! 1
0 t0
N C)
A "
s-nvw 11 5iVM3 I 144
Accordingly, a grid was prepared to facilitate the taking of readings
in n-seen units from Cattell's (1972) figures at the age points of 17,
22,30,40,50 and 60 years. These data were recorded for males and
females separately, treated as raw scores and then converted to total
group standard score units with a mean of .0
and a standard deviation
of 1 (z scores). It was these values (Tables8. sand g. 21 which were
plotted on the graphs contained in riqLtres 8.2. to 8.17. (blue lines
for males and red for females).
In a similar fashion the subjects of this study were grouped into one
of seven age cohorts with mean ages of 17,22,30,40,50,60 and 68
years. Raw score means and standard deviations were then computed
for each factor and age group by sex by 16PF form (A, B and A+B).
The 16PF Form A+B data for each sex were then converted to z score
units based on the total group (male + female) means and standard
deviation for each primary source trait (Appendix 8). Figures 8.2. to 8.17. 40
contain the results of plotting out these statistics in graphical form.
This scaling procedure allows comparisons to be made between those
personality differences predicted by Cattell and the results obtained
for the samples of male (N= : ='=l000) and female (N=1000) subjects in
this study. In addition, comparisons are now possible across the two
questionnaire forms as well as the different primary source traits.
The product moment correlation coefficients representing the linear
association between the primary source traits and the chronological
age of the respondents in the study are represented in Table 8.3.
Statistically significant linear negative trends are discernible on
Factors B, E, F, L, Ql and Q4, whilst Factors G, N., Q2 and Q3 show I
145
9
.8
.7
.6
.5
---- .4
.3
_ .2
H
0 o 0
FIGURE 8.2.
AGE TRENDS IN PERSONALITY:
FACTOR A
Ö \\
L
- rý
5I__.
-. 6
-. 7
-. 8
146
.8 t G: ý TRýi jS :N_ : iýs'JidiLI'_
-
14 FACTOR B
.6
ý- '-5I.
.3
1ýXý n
20 30 40 50 60 68
-,. _.. ý (YEARS)
Male Form, A --
----- 147- - --_"- .. __,
.8
. .7
- .6
.5
i1GU:,,.. 8.4.
AGE TRZIII)s
FLC oR C
/
0 O -- -----.... .
ý- -- --_ -'ý .4 --
- -. 7_
20 30 40 5O 60 68
AGE OL-EARS) F0 -r 7,1 ~- -- -_. _
Kiiiie Form.. A Form, B
148 -- --.
ýý... _ _--- ";
--_ _. _.. _.
ý __ _
__:.:.. _ _
__ `^
-, . ý.
.9
_8
.7
.6
.5
x ý.
x
FIGU 8.5.
FEGE TRENDS IN PL Gi. ý : ITY:
FACTOR E
ýý "`-yam _ ...
ý` `.
IS,
Ü
-- - -- N--. 1 - -. - -. ,x
4-
_ -- - 20 30 40 SO 60 "- - 6G
AGE NEARS) Temcl Farre F. ------ - -- -- psm F
.9 FIGUR"- 8.6.
$ AGB TRT-' DS IN PERSCNALIT' :
FACI'CR F
.7
.6 "\
14
-. 5 .
-. 6
I- ---
ý; - -. 7___.
\
\- ---ý
20 30 40 50 60 62
. _.. . . _. _... tiGE (17EARS) ... - ...
Fe. °-Gle Form 'h Form E
-1'ä1e Form Ä---------- " ... _. _ý. _... -- -- --- -ý--------
ý_. _ýý------ ---- __ --- - Form--B ..... _. ----
-...
.9 FIGURE 8.7.
-7 { FACTOR G .
.6
-. 5- ýc
.4
- ýý
-I-/ /1'
2aý
/
-__. _.. -.. xoe
ai . _..
-- ------- k/7 JbKc
ý. -. 7
20 30 40 50 60 68
AGE (: '£ARS) Y _.. _
-- _. _. __. _-_. _. _.. _ "---- c1E FC--7r, A
J_ 'ra Form. P-,
15-1
.9
.ß,
. 7_
.6
-. 5
FIGURE 8.8.
FACTOR H
a ýKý. _. ý_... _? c
.. _. X ý_ ýý
20 30 - 40 50 60 68
ýý -- __. --_- kC-E (YE; -. R£) Fend er crin A
Orr,
'' '--_ -- ---. -- --- --- ___ -- - .. Form- $
G
.B r
.%
.6
i
.Sý..
0 U Cl)
---- ; -. 4
-- --p
/
/
2 30 40 50 60 66
AGE (YLF, RS) F
- --- - --Made Form. A-_-
Jý. _...... _ _ý ._ -- F orni B --
FIGURE 8.9.
AGE TREK IS IN PERSONALI_- :
.9
.8i
.7
.6
.S
rx
/I, "-
.8 20 30 40 50 60 Ea
; G£ (YEARS).. pemcle Forrn i, Yon. h
-- -- _ ____- -- leF orii r------- Fern- B _.. _
.. _. __. 2.. 4
FICURE 8.10.
AGL IN PERSGNA_, IT'_ .
FP C^CR L
.ö
ý_
.7
.6
.5.
C. 8.11.
. =GL ; READS IN jTY.
FACTOR M
-4 - - TT It I_
_ '\ _
/ ýI
" - ý\- - '-- ------ --2-- . -
ý f. _ ------ -- -
- - i T t ý Y
----- 8----. 2-- -.
- .. _:. z. 3. -- - -- ---- - -- - ..... -_--.... -. ___..
__
- -. 6 -
20 30 40 50 60 68
" -- AG= (YEARS) F'e111c1 EF p271 A "Form
-_ _` -- --`- -- - --- - ---
ýý. _
^ý. _. --- -^ _Mae.
Form
.9 FIGURE 8.12.
_8
. .7
.6
AGE TREI DS IN P. ERSONALrY:
FACTOR N
5-
8__.
20 30 40 50 60 65
"-_. - ---- .-- jiGE (ice; RE) gi-ec1E Form i.
. --_. ''- -- ---___ Form F-
--- - -- Mal e . Form A - -"-_ -- "-- -- --. _. ---. ý.. _------------ __.. _. ----- ______
---- - ---- --- -,.. _ ý. -F crni B"--
.7
.6
.%
.6
.5
. -'1G; -F rö. 13 .
. GL TflrS IN ptisSOicALIT:
7z
N
20 30 40 50 60 68
AGE (Y-kRs) 1 elrýýle }'orr l:
_- -= --- ---... Fcrr,
\-ý -- - . _.. __- __-- ----
-- -- --. -- - Male Form--A--- -- ---- ý1
- ... -ý
---- _- Form B-"-., -
_ Ecu: X3.14 .
.8 ý\\ TAGE TRENDS I PERSONALITY:
FACTOR Q1
.7 \
-- --- . -- - - -
20 30 40 50 60 68
7. GE (YE; --RS) F'err cý E f'orr,, t, -
Form-B
"9 f FIGURE 8.15.
-8
+I t Gr i iu:,; ýJL5 I1: FýrcSONALi X.
FACTOR Q2
.7
.6
ý-- _---. 2 -- -. -...
ý- -6
20 40 30 50 60 68
- ýOY'Tt i AGE (YMRS) }orn. h -ý
fir` -----_ ___---------- _- -- --- -- --- ----------- -- ___. _.. _ _. ____. _ ; _.
------" - -- - --- ý Form .. B - -- --------- - --- ---
.9 rIC E 8.16.
.8 AG TRENDS IN FERSOXA. -I I'.
FACTOR Q3
.7
i
.6
_ -r
o
q, -
20 30 40 50 CIO 68
_ J, GE (YEARS) Fe ae Torn. A ---- _. _. -,
Form Ä
Q
.8
.%
.6
x __ .s
ýý g. 17.
AGE. TRENDS 1IN
FAC^"'C, R Q4
T"4 x"\ -T::. _. _3 x
__. _--
E, -- __ --_- -- -
o r 0 V i
20 30 40 50 60 6E
ý .. - -- AGE (YEARS) Female Form i;
`- - -- - _ } orri ao
w... - :- -- ---- - ---- Form
r Ct
c-+
N 0
0
0
z
X
H
r, '
Ow 0
AI L
z
w
CY)
HHV
I co 4 c, to o V)
I. D rl C3
t f co f{ C,
}1 -a o.
1i
ch tie -, crn .t I r-4
cý1 1 --4 LO N '+. " c7
ell I
04 co C*4 (7 N M N N N `
i 1 1 :' t
in %D O
ýIý
d' c'7 O "i t
to in O
1'
C) N O
i
"cT O
f.
N ýp O
t.
N Oý O
%D O
!t
CO Q
i" .
-+ tJD Ln Cß M N
N in Get
N CO) N CV N N N N
r4 V-4 ai r- n co . co
I ("
O O r-4 M - ö Q O n O O O O O ,
ý 1_ i" ý" 1 1ý
to "N a% M O N Co N l1w m OD co C3 r- (IN O Ö
r-1 . - O O rl O C) .4 " I 1ý - 1- 1. 1" 1 1" t 1
to ' Gl p C) "C' to to C%j M l1ý ! "ý "ý ch . N O ý4 ý- O r+ O O O
ý O'
i ! [- C) rn in w M 03
° in to ý ö ä ö ö ö o .
40 f- CV) Cý 04 410 cy) Ln N Lti LO LO C N C -1 N M N
M C '-)
co Gl O co /`') O N C"I O Xn (V . rl E? 1
in C') c" M ý
ýi`
1 1ý !f 1f f 1
"I Co . -# to t- M O N l0 -ý . to to n . 'W " t- at in N " -t N C-1 O . -ý r" 1 C)
" . -ý
t It 1 i i t F
fi C '' a in C) O r-f a C14 4 O O O O O O IO
" aJ O. O O
1 1 I 1 t ! 1
-co M l7 (N (n Co M N- CO
r4 N r -f "-! EEE i t" i i i i i ý I
N f) (D tf) M (n Ln l'ý E'er . -4 Ll) C' N lfl MC) .s CJ. , N O O Q O O O. O O Q
tý p
"-i t t! 1
Iv C 7 4rß
ti
162
W t7 4 Ö
W
O
R
U) Eel
w a4 U) C :4
W
ä Q) 0
ýýw
Q1 h
o
M M
N CO N
N : O lp z
C
Co M O
O N
ri O N C! ]
C) rn o co N M N vi
z z
0 o rn vý 0
r-1 r-I" CO O O M
H O d' to H " " " z
r-i
ý N M O N
x z
M ý'ý1 lp V!
r-A cm C%
w z r-,
rn C(I
w .. z d1 O
M O
M M ý
r1 r N N pý Ln O
1 N
0 Ln 0 L
ý, ar
v (J
tu p4 ý, ;2
0 4-1
tu >
ääzu 0 14
ý. ccn
163
0
En El H
9 H
ä
cA a+
.. N
W0
c'ý wý
`" " 0 z 1
M N cý O
" " " CY)
00 N "4 M O
M
M M lO VI ý5 " " " z,
a O -I
N l0 t0 O O N Z
0
M rl Y) M In bj
z z '--ý
O N N
ý+ o O z o . -+
to r1 i L v)
z 0
N N
ýý " " O Z
01) LO N 0 A
" " "
V
O
d' Ci O CD
r- (D co U-1 d' d' O
w O N
N N In N O
" " " " O N
N c CO O O 01 y
U " O" " z
C\l CV
OD 1 M
ý
CU " " 9
O M
N d' O O U)
4 " ö z
4-J 0)
(a z 4J Nu +' :3 U w 4 a s
w -1 a 04 - 0 4)
ää Ü cU)
164
significant positive relationships. The remaining six primary source
traits (A, C, H, I, M and 0) show no clear linear pattern.
In order to test the hypothesis that the regression of personality
on age is linear, eta coefficients were calculated on the primary
source trait means and standard deviations, classified by age and
con, ained in the appendix.
The eta coefficients representing the regression of personality on
age for Forms A+B of the 16PF are contained in Table8.4. -for females
and Table8.5. for males. F tests of the significance of departures
from linearity were then undertaken between the etas and product
moment coefficients and the results of this analysis are also
presented in these tables.
i In the female group Factors B (intelligence), L (suspicious) and
Ql (radicalism) show significant departures from linearity. In
males, six traits have significant F values, these being B (intelligence),
E (dominance), F (happy-go-lucky), G (superego strength), Q2 (self-
sufficiency) and Q3 (social precision).
i 165
CATTELL'S QUADRATIC AGE CORRECTIONS
It was noted in an earlier section that in addition to graphical
representations, Cattell (1972) has reported precise mathematical
expressions of the relationship between age and the primary source
traits. These are based on cross sectional data and are provided
for the benefit of the researcher who wishes to compare the
personality of individuals or groups which differ in age. These age
'corrections' which reflect both the endogeneous and exogeneous
components are usually quadratic in form and of the type: -
Yadj =Y- (Y' - Y'
=Y- bl(X - XS) - b2(X2 - XS)
where X= age Y= observed score
v
If the primary source traits represent the skeleton, then these age
expressions are part of the flesh of Cattell's theory. In his earlier
work, 1940-1960, Cattell was primarily concerned with the identification
or 'discovery' of the source traits. Although considerable effort has
been spent in defending and extending the system, since those dates
Cattell has been concerned to provide a meaningful contextual
explanation of his traits. Indeed, it is only reasonable to expect
that once a theorist has made claim to certain constructs, he will
seek to define their nature vis a vis external realities.
166
But in Cattell's case this need was more acute because he was
particularly vulnerable to the charge that his work lacked clinical
impact or, in more extreme terms, that his theories were nothing more
than matrix algebra in search of meaning.
Any mathematical expression representing the relationship between
`ü 'v: l contain scre degree of error. YV ýi a;. per: - oý litt' iL. h- lst of
acknowledge the existence of such error, these expressions are
constructed with the aim of reducing the common variance between a
personality variable and age. One expects that such formulae
produce results which at least show a lower correlation between
age and personality than was the case before they were applied.
RESULTS ON CATTELL'S QUADRATIC AGE CORRECTIONS sr
The descriptive statistics (Forms A+ B) for the male and female adult
groups and for each age cohort were corrected by the formulae provided
by Cattell (1972). These data are contained in Tables 8.6. and 8.7.
respectively. Following these calculations, eta coefficients were
computed (Tables 8.8. and 8.9. ).
In the male group the two factors of F and N show significantly lower
eta coefficients after application of Cattell's age correction weights.
Factors B. E, G, I, Ql and Q2 show little difference between the two
sets of coefficients. The remaining eight factors show etas which are
in fact higher for the age 'corrected' scores; these being Factors A, C,
H, L, M, 0, Q3 and 04. Most of these discrepancies are not only
167
statistically significant but very large; the most extreme example
being Factor 0 (age uncorrected eta = . 07, age corrected eta by Cattell's
formula = . 78).
Turning to the female analysis, Factors F and N again showed significantly
lower eta coefficients after application of Cattell's weights.
Differences in et-as were small for Factors H, E, G, H, I, L, Q1 and Q3.
The primary source traits of A, C, M, 0, Q2 and Q4 all showed
significantly higher eta coefficients on the age corrected data.
CATTELL'S AGE CORRErTIONS: DISCUSSION
The results of the comparison of eta coefficients between corrected
and uncorrected primary source traits and age, show a fair degree of
consistency between males and females. In both groups Factors F
w (happy-go-lucky) and N (shrewdness) showed a significant decrease in
correlation after application of Cattell's weights, whilst B (intelligence),
E (dominance), G (superego strength), I (tender-minded) and Qi
(radicalism) demonstrated no appreciable differences.
Those source traits with the most marked increase of correlation with
age were Factors A (outgoing), C (emotional stability), M (imagination),
0 (guilt-proneness) and Q4 (tension). In fact eta coefficients on the
corrected data were so high in certain cases, especially with males,
as to make one wonder at the very origin of Cattell's age expressions.
In the case of C, 0 and Q4 for males, the 'corrected' data seem to
bear little or no relationship to Cattell's hypothesised trend lines
reproduced in figures 1 to 16 and so we must conclude that errors
of mathematics are at work here.
168
9 TIi A
a 0
a w H
Ö
U W
4
rn
4
w Hy
N a w a rl
,pR C6
a
IN to lfl rl Ol M M O N OD 14, N
a N d' M H Co Co N CV N N /-1 c1 d' Oý Lý m Co rn al ýt' a, a' 1ý () O ri C Y) ul to a)
N N N N N CV Co
lf) In C\I Co lf) N 00 If, a1 d' O N N
cr CO O O (7) N N r-I N N r-1 <V N
N Co C 1S ^ 0 ! fl co
. CJ H 711
Cr r4 co Iz CV H H H r-1 r-I
d' Co O) Co Co Co N E- Co N O " . " W 0) 0) w
N N N ý1 C \I, IV . 4t 00 ON f) NT Co N OD M tf )
Ln co N Co Co 0 CV CV CV N CV c*)
C4 0 cr% (A N H
rl 1 H
Q\ if) 0, N a) H H . " " " . "
H H H H iH H.
to N N Cý H 0 Co d' d' 0) N 0 x M N H
CV CV M
[` N H 0 t0 Ol M N O
Co CO cm CV N N
M
IV 01 [N U) q; r
Co M (V N N
N a% N N l0 Op O l0 M C\l N Q) W " " CO Co Co [,
CV N N N N R
N H O N O u) N N
M O CNJ c C Y) cn M
N N O c'1 0 H " co " " " " Ln in in d' C') S M
H H H H H '-4
N C\j Co N M N
r co N H -4 r N
N
qz?
Q i o e uý t t If) c p ý-I (V M LO
169
A W H
Ö
U
W
ca
4
O Gv
z H
A
ýi
ä
E'ý ýý
LO N o "
ON "
co "
(V "
N "
Cl) "
co N N N
N r-I
M lc: t to
O O1
N . -I
C) U)
"1 l0
to M
r4 N 4 l() (- N N N N N N
N O
N N 1' a% to r- N
co
O N
ON 1-1
ON ri
a1 r-1
O CV
v CV
(Y) 0)
In rn
U) o
CV M
co p
O . . . co (A
CY) N N LO N N
0 ý
CO 0) '1) N Oo
N N N N cm N
Lr) CV
M (V
N
r-I CV
rl
CV CV
O
M N.
U) N
Ol
Lp CV
CV dI
(5)
Un
O
to H
i
In cM
N
"
H " N " N
C.
N
C.
N
0 IIZP
N
N M
N G" r-f
(5) N
r-I CO
(+) CV
d' rl
d' N
N CV
C') C\ I
pl
N M N
M
(n N
U1
N
9
14V N
-
r-1
N
d'
N
LL0 "
N
N N N N N N
W " N
r4
N
" co
N .
p
" 0
r. 4
N
d0
N
u
ul
0 M
co N
CD N
IC) CV
N
N M
co
N N.
N "
M
-i M
"
M
OD co LX)
" -
to hfl
"
rl
to "
-4
N
, -4 nj
5-4
O N 5 -4
tto N
d'
C\l N
C
N
n
-4 N
t0
C\l
O
N
4 ui r--I
ui I
CV
d'
Lo M ýn
IV LA Ui
in
170
4 Ö W
v
ac ? I-fl 1-4
H
U)
p w ua
4
kD rn N M ri H
"
M M O M " " " a
N N O G" H M
CM " . a
(y CV p
M 0 M M r-1 N
0 "
H H
rn N
Co
z
In co N O
N C` In N S O
01 H N O H O
H " " .
0
N M co
N
W " ,
N r-I 0 W
IWO G CM en C U :
fV N
'4 " "
O
l 0 M ri N
4 "
0) 0 4 Ö te 4 -1 4U 0) 'a ' Q5 Qý q
171
Aº
w ö
cý
u w u 4 Ö W y
ä
oý w °°
C1 ý WW
Lr)
CO l0 N d' M
cyl
a ri N r-I N N O
N " " " O I
M p rI O
N
O " "
ri d' M
. -+ M N z "
N O N ri M
d' N N " " O
H
cm Co O M
O O O t7 "
N O
W
N
N tp
N O
N N 0 O . /ý W " "
Co N O
78 ü
.. i i i
e0 4-3 0) u (0 4-4
Qaý q
AP
172
In summary, Cattell's age corrections had little appreciable effect
on about half of his primary source traits. For the remaining factors
they introduced biases which were far more extreme than those in the
untreated data. It would seem that these corrections have indeed
been poorly developed and should not be used until such time as
new formulations can be researched.
In the interim, two possibilities are open to the researcher who
wishes to take account of age differences between one or more groups
of subjects. First, he may use the separate standard score norm
tables prepared for each sex and age cohort contained in Appendix 6.
Alternatively, it is possible to use the product moment coefficients
in a conventional regression approach, which despite certain but
usually small departures from linearity, is likely to be a more
reliable method than Cattell's age correction formulae. In any
event, there can be no doubt that the quadratic age expressions have
produced anomalous results, even against Cattell's (1972) own
graphical trend lines.
SUMMARY OF RESULTS
FACTOR A (OUTGOING)
Consistent very small negative correlations with age, although
the majority are not statistically significant.
Eta was not significantly higher than the product moment coefficient -
that is no evidence to suggest that a curvilinear relationship is
a better fit to the data.
i
173
Only female Form B curve approximated the hypothesised
increase to 30 with levelling off thereafter. Males on Form B
showed a slight linear decline from 18 years as did females on
Form A. Males Form A showed a decline up to 30 years and then
slight but non-significant increase.
Gene-rail sad s rate-: e. over form and sex seems to be a flat
or perhaps very slight linear negative relationship.
Obviously sex differences are more important than age variation
here - at every age significant differences remain with females
seen as more outgoing than males.
FACTOR B (INTELLIGENCE)
There was a significant negative correlation between this primary
source trait and age for both males and females. Tests of the
departure of linearity of regression were also significant, the males
showing a rise up to about 25 years of age, a levelling off until
about 30 years and more rapid decliKie after 40 years. Females on
the other hand showed a relatively shallow gradient up to about
50 years of age, when the decrease accelerates considerably.
In all the data are in rough agreement with Cattell's trend line,
although the point of deflection appears somewhat later. at 40-50
years and not at 30 years as postulated by Cattell.
i
174
FAC' RC (EMOTIONAL STABILITY)
This factor showed no significant linear trend in either males or
females. There was some tendency for adults of both sexes of about
50 years of age to have lower scores on this factor, although the
overall line demonstrated no significant curvilinear results. The
data do not in fact shop: any real correspondence with Cattell's
hypothesised differences in age; his curves being a marked
exaggeration of the empirical results found here. From Figure 8.4.
it can be observed that sex differences on this factor appear
more important than age differences.
FACTOR E (ASSERTIVE)
This factor showed significant negative linear correlations with
age in both groups. For the male sample there was a slight but
statistically significant departure from linearity (p----. 05).
Although Cattell's quadratic age correction formulae in fact had
little effect in reducing the co-variation of age with this primary
source trait, the visual impression from Figure 8.5. suggests that
Cattell's graphical' representation of age differences are relatively
close to those discovered here. In males, there is an increase in
means up to about 30 years of age, when there is a decline in
assertiveness. Females show a similar pattern only the decline is
somewhat earlier at 22 years.
p
275
FACTOR F (HAPPY_GO--LUCKY)
The trend in this trait is for a very pronounced negative correlation
with age (r = . 4) for both sexes. We have already noted that this
was one of only two factors where Cattell's age correction formulae
significantly reduced the co-variance with age. The graphical plot
of age cohort means shows a relatively close similarity with the
trends postulated from Cattell's model and these data support the
hypothesised direction of differences.
FACTOR G (SUPEREGO STRENGTH)
The strong positive correlation with age is clearly marked. in the
graphical plot of the 16PP means. This trend is strong in both
males and females. _ "
One explanation of this trend is of course a tendency for superego
strength to increase as an individual ages. We more usually think of
the process of socialisation and the development of superego strength
as occupying the years from birth to adolescence. But here is evidence
that perhaps the process continues considerably past these early years
in fact throughout life up to some 70 years of age.
This fits in well with popular prejudice, ie. reckless expedient youth
versus the conscientious moralism of the later years. In fact, this
is probably one of the major discriminators of the younger versus
the older personality. An alternative explanation of this phenomenon
is that the younger individuals in this study have experienced a
176
particular epoch where superego strength has in some way been eroded.
They will certainly have lived their adolescence' through the nineteen
sixties' where there was a revolution in the attitudes and morals of
society. And such changes in attitudes no doubt had more effect on
the younger 15-45 than the older 50+ age groups.
The furý. er possibility exists trat 'he sýmpiing method employed
in this study was responsible for this trend, for example, that the
young people in the sample were unusually low and the older people
of particularly high superego strength. One sector of the community
which was not tested for example, was prisoners who are no doubt
low on this primary source trait. This explanation does not seem
very likely however, there being-no clear reason why such an
unusual sample bias should have been introduced.
f With regard to hypothesis, the evidence does support a pattern of a
moderate but decelerating increase with age.
FACTOR H (SOCIALLY BOLD)
The results for this primary source trait, one of Cattell's major
extraversion components, show a relatively flat pattern of differences
across the six age cohorts. When Cattell's age differences are
superimposed on these data, results for females are relatively close.
However, the male curves are very dissimilar in shape, there being
no evidence of the pronounced inverted U shape proposed by Cattell
This discrepancy was also noted in the section on quadratic corrections
in that Cattell's formulae had little effect upon the female age
177
differences whilst eta on the male group was considerably inflated
after application of his weights. We must reject the hypothesis
then of a positive relationship between the source trait and age.
FACTOR I (TENDER-MINDED)
The general trend of differences on this factor is slight, positive
over age group, particularly in males. Cattell's postulated curves
are in fact relatively closely related to our findings but he may
tend to exaggerate sex differences somewhat. Nevertheless., the
variance associated with the sex of the subject is very large.
We have found some support., therefoize,. for the hypothesis of a.
low positive relationship between this primary source trait and age...
FACTOR L (SUSPICIOUS) i
Cattell has postulated aU shape relationship between this factor
and age in a study of cross-sectional design. More explicitly he
suggests (1972) that L (suspicious) declines up to 40 years of age
and then increases.
The data contained in Figure 9 show a slight decrease in scores for
males over age with no significant departure from linearity. Females.
differences on the other hand did depart from the simple linear
association, suggesting that scores decline until about 50 years of
age where they begin to increase again. This was about the strongest
non-linear trend in the personality data. The picture which emerges
on this source trait is best described as a form of rotated "D".
178
where differences between the sexes are relatively small at the
age of 20 years, slowly decline in males and more rapidly fall in
females thereafter, until approximately 60-68 years where the
differences between the sexes are again relatively small.
The exact reasons for these findings are difficult to establish
i. a study like this. Perhaps marriage in females has amore marked
effect on this personality variable than it does for males. It is
possible, for example, that young single females tend to be more
suspicious and hard to fool, that the level of this trait declines
through and as a result of marriage and as the years of widowhood
approach, begins to increase again as a natural defence. mechanism.
Obviously explanations such as this remain speculative; this study
was not principally designed to investigate such possibilities. However,
we can say that Cattell's suggested U shaped relationship between this
personality trait and age was supported in the case of females but
not with the male group. Indeed, the overlap between age and the
quadratic corrected scores in the case of males was very much higher
than eta based on uncorrected scores.
FACTOR M (IMAGINATION)
Results on this source trait indicate relatively small differences
across both age and sex groups. Product moment coefficients were
generally non-significant as were statistical tests for the
significance of departures from linearity.
179
Cattell's claim for an inverted U shape relationship was not supported
by these data and his corrections merely increased the correlation
between this variable and age.
As an additional observation, it is interesting that despite this
trait's relatively high correlation with Factor B (intelligence),
it dos not da-ionstrate the characteristic sharp decline in
performance at about 40 years, shown by many ability tests. It is
not improbable, therefore, that this scale represents the overlap
between the ability and personality domains; what some researchers
have termed'cognitive style'(Witkin 1970).
FACTOR Ni (SHREWDNESS)
This factor shows a moderately strong relationship with age with no
significant departure from linearity. In all, Cattell's mathematical
corrections were reasonably successful in reducing the common variance
between this trait and age. We have evidence. to support the hypothesis,
therefore, of a positive association with age.
FACTOR 0 (APPREHENSION)
The overall trend is for no significant relationship between this
primary source trait and age. Figure gi3 . is in fact similar to the
data contained for Factor C above, ie. a relatively flat profile of
age cohort mean scores with perhaps a slight rise at about 50 years
in the direction of increased anxiety of which this trait is an
important component. Again the effect of 'correcting' by the use of
180
Cattell's weights was to increase the correlation between age and this
particular personality trait. In the case of males the weights for
Factor 0 were so inappropriate as to produce a corrected eta value
of . 78; so high a result as to suggest some serious mathematical
error in the equations.
Ln view of these results, we must reject our initial hypothesis of
a fall in the personality trait of apprehension after adolescence
followed by a slight increase after 40 years. It would seem that
there is no clear linear or curvilinear association between. age and
this variable.
FACTOR Q1 (EXPERIMENTING)
This primary source trait intended to represent experimenting
behaviour, appears to be very much like Eysenck's concept of
radicalism-conservatism. (Eysenck 1969 , Wilson 1975. ) It is
rather curious in that it is the only factor where Cattell claims,
and we have hypothesised, no significant relationship with age.
Our results suggest, however, that this factor is one. of the most
strongly negatively correlated traits with age, both in the male and
female group. We must conclude, not unexpectedly perhaps, that older
subjects in this study tended to be more conservative than younger
subjects, especially if they were male. There was evidence of some
curvilinearity in the relationship for females but not in the male group.
I
181
FACTOR Q2 (SELF-SUFFICIENT)
Cattell has postulated that this variable tends to be positively
related to age for female subjects but to have no clear trend for
males.
Thus the hypothesised differences give a contradiction of the data
collected in this survey. Here we found that age was significantly,
positively correlated with Q2 in males but only slightly, if at all,
for female subjects. In males, tests. of departure from linearity
were in fact significant.
The general trend is for scores to increase on self-sufficiency after
adolescence until about 20-25 years where they level off until about
55 years. From that age onwards up to 68 years, the group means 410
tend to increase.
FACTOR Q3 (SOCIALLY PRDCISE)
The Q3 source trait, representing controlled, socially precise
behaviour, was hypothesised to have a strong, positive association
with age, levelling off in the middle years of life.
The results represented in Figure 15, show a relatively strong
linear relationship with age (r. =. 35) but little evidence of any
levelling off at 40 years. It is mainly this discrepancy in middle
age which is responsible for the male age 'corrected' eta to be higher
than eta calculated on the original data.
182
FACTOR Q4 (TENSION)
The hypothesised slight negative relationship of Factor Q4 with age
was in fact found in these data. However, Cattell's quadratic age
corrections again introduced more systematic bias into the data than
need existed in the first instance.
In the case of females, it was Cattell's postulated increase after
50 years of age, not present in these data, which was mainly
responsible for this distortion. Differences between the sexes
are quite evident from Figure 167 as is some slight rise in tension
level around 50 years of age. Overall, this source trait shows a
somewhat higher correlation with age than the two other main anxiety
factors of C and 0.
aR
R
183
AGE DIFFERENCES AND THE PRIMARY SOURCE TRAITS: GENERAL DISCUSSION
In all, the correspondence between Cattell's graphical descriptions
and the age differences investigated in this study is not impressive
(See Table 2.11. Whilst it is true that five of the primary source
traits (B, F, G, I, M and Q2) show reasonably similar age differences
to those hypothesised, the remaining ten factors vary considerably,
particularly QI (radicalism) and M (imagination).
In part these discrepancies could be due to cross-cultural influ ices
although it is improbable that the more extreme variations could be
explained by such effects alone. Both this study and Cattell's
research are based on a cross-sectional design with data being
collected within a few years of each other, so it is unlikely that
conflicting findings originate in these simple methodological Am.
differences.
A further possibility is that the British age cohort samples (N..: =--2100)
are in some way very unrepresentative of their respective age groups in
the adult population. Because we know that the total sample is
reasonably representative, we are forced to conclude that any such
bias is compensatory in nature. Moreover, the very fact that the
age difference analyses were conducted holding sex and age constant
means that the most likely source of bias involves the social class
composition of the age cohorts. Whilst bias in the social class make
up of the age samples cannot be completely ruled out, there is no
clear reason why this should have occurred.
184
If we turn to Cattell's data on the other hand, we have noticed that
the description of age trends show considerable inconsistencies. His
verbal description of the relationship of the primary source traits
with age are at odds with his own graphical representations and in turn
these are often inconsistent with his quadratic age expressions.
On Lhý' Crucs . ion of the
C; L? adratic relationship of curtain of Catttel l 's
primaries with age, there can be nothing to recommend his formulae.
The amount of additional variance explained by such procedures is
minimal and not sufficient to justify the considerable errors that
they may introduce.
GROUP DIFFERENCES AND THE PRIMARY SOURCE TRAITS: CONCLW IONS
Although there were some inconsistencies across forms of the 16PF,
earlier sub-sections showed that the majority of the , sex, cross-
national and social class differences in the primary source traits
were in line with those hypothesised from Cattell's model.
There is reason for much concern, however, on the question of age
differences in personality. Only two of the primaries demonstrated
a reduction of common variance with age after application of Cattail's
age 'correction' weights. These age results support the general
hypothesis, therefore, that the primary source traits have been
derived from inadequate data sources and are untenable in terms of
the verifiability of group differences postulated from Cattell's theory.
185
Having observed inconsistencies both with regard to questionnaire
forms and the verifiability of group differences, the next section
will consider the important and controversial issue of the
reliability of measurement afforded by Cattell's primary source
traits.
"
i
186
SECTION 4:
RELIABILITY AND THE
PRIMAR YS0URCETRAITS
gý ALTERNATE FORM RELIABILITY OF THE PRIMARY SOURCE TRAITS
410
INTRODUCTION TO RELIABILITY THEORY
The reliability of a test or questionnaire concerns the accuracy,
consistency or precision of measurement without prejudice to the
dimension being measured. Every measuring device is subject to some
degree of error whether it be a simple ruler or a complicated
electronic instrument. Reliability of measurement is a principal
aim in science because it is commonly assumed that reliability is
a necessary, if not sufficient, pre-requisite of a valid set of
observations.
In the classical test model the score of any given individual on a
test or questionnaire is said. to be made up of a true score and error
component.
That. is: -
OS TS +E
where 0a observed score s
Ts= true score
E: error ca iponent
On the assumption that the action of error is completely random,
it follows that:
mo
(a) the mean of all error scores is zero.
(b) the variance of error scores on all tests of the same length is
the same.
(c) the correlations of error scores with true, obtained and other
error scores are zero.
188
Having made these assumptions, we can make the statement that
observed score variance on a scale is made up of true score variance
and error variance.
(i) ö dt + de (basic classical test model formula)
where d2 = observed (total score variance).
0
= true score variance
d= error score variance e
The reliability of a test is oftenexpressed therefore as the
proportion of the observed or total score variance which:
is true score variance.
That is: -
0-2 (ii) rt (basic reliability formula, classical test d model where rt = reliability coefficient)
0
This, the fundamental reliability formula in the classical
test model, has the limitation that we never directly know
the true score variance.
But knowing that:
t (basic reliability formula) r dt
d2 0
derived from the and dt : dö - dE (basic classical test model formula)
then rt = dö 02
d2 0
189
and
(iii) rt =1- 6e
d2 0
Formula (iii) is a more convenient expression because it is possible
in practice to estimate error variance by a number of different methods.
This is achieved by correlating one or more sets of scale scores for
a group of individuals with another set of scores for the same group
on the same scale or variable. Correlation coefficients calculated
in this way are known as reliability coefficients and express the
extent to which a scale agrees with itself in producing consistent
or reliable scores.
Three main methods have been distinguished for estimating the
reliability of a test or questionnaire. Each of these techniques.
isolates different elements contributing to the scores and counts some
as true score variance and others as contaminating error variance.
INTERNAL CONSISTENCY
This method more usually involves assigning the items contributing to
one test score to one of two sub-scores, on some basis which will
produce two equivalent half-tests. Commonly, odd and even numbered
items are the two sub-tests. This method of calculating reliability
has the advantage of requiring only one testing session: the Spearman-
Brown formula is applied to correct for the effective reduction in test
length entailed by the split-half procedure.
I
190
In its purest form, this type of reliability is based on item statistics
and counts as true score variance only that part of item variance which
is common to at least one other item. This is known as the homogeneity
of a test and in a sense estimates reliability on all the possible halves
into which the test could be divided. Statistics developed to facilitate
this process are the Kuder-Richardson coefficients and the more
generalised coefficient alpha (Cronbach 1951 ).
Cattell (1972,1974 ), however, has pointed out that his measures
possess internal consistency without homogeneity. They are deliberately
heterogeneous in content in order that items have a high multiple
correlation on the factor on which they load. Moreover, items of
low homogeneity, he claims, avoid the problem of bloated specifics
which are to be avoided when measuring broad personality traits.
For these reasons, it would seem that conventional homogeneity 4P,
estimates are not applicable when investigating Cattell's primary
Source traits.
TEST-RETEST RELIABILITY
Test-retest reliability is estimated by repeat administration of the
same test. Irregular change in scores between the two administrations
is treated as error. Cattell (1970) has distinguished between the
'dependability' coefficient where the retest session follows too soon
for the subject themselves to change and the 'stability' coefficient,
where two months or more may elapse between the two administrations of
the test.
Stability is a very important concept in trait psychology: in the
191
calculation of stability coefficients both relative shifts in score due
to real changes in the individual and shifts attributable to random
measurement error are treated as error variance. Clearly, the researcher
will want to be able to distinguish between these two sources of 'error'
so as to be able to evaluate real changes in the individual. Attempts
have been made to separate traits which are relatively stable from
shorter-term in-Eluences which are more bound to particular situations
and change in strength over a short period. For example, the StateTrait
Anxiety Inventory (Spielberger et al., 1970) seeks to isolate those
aspects of anxiety which are situationally dependent (state) fron
those which are a more stable feature of the make-up of the individual
(trait). From the point of view of reliability theory, it is important
to recognise that the length of time which elapses between test and retest
may be an important determinant of the size of the test-retest coefficient.
Score variation, due to changes in personality and mood between test
and retest sessions, which will be treated as error variance, means
that great care has to be taken in specifying the exact conditions
or facets (Guilford 1970) of the reliability study. Further, the
stability of measurement over time tells us little or nothing about
the underlying structure of the item content and the very personality
model upon which it is based.
ALTERNATE FORM RELTABILLTY
The third commonly used index of reliability is the alternate form
coefficient, otherwise known as the equivalence or parallel forms.
coefficient. The different names all represent - from different
I
192
theoretical standpoints - the correlation of scores on one form of
a test with scores on another form designed to measure the same
characteristics. That is to say, the alternate form coefficient is
a measure of the extent to which two distinct forms which are designed
to measure the same thing actually do so.
:: eilst difficulties may arise in interpreting test-retest and
conventional internal consistency reliability estimates, there can be
no doubt of the importance of alternate form reliability to Cattell's
theory of personality.. Because of the difficulty. of establishing
really adequate external criteria of Cattell's primary source traits,
the model very much rests on the extent to which it is internally
consistent with regard to replicability across questionnaire forms.
As there was such a short time interval of about 10-15 minutes *º
between administration of the two alternate forms, there are unlikely
to be undue influences in the data collected in this study as a
result of changes in personality over time and for these reasons the
reliability analyses which follow are based on the alternate form approach.
THE STANDARD ERROR OF MEASUREMENT
The reliability coefficient is a useful statistic when a judgement
has to be made on the consistency of scores afforded by a particular
test or questionnaire. It is often the case, however, that we wish
to set a band of tolerance around a given score and make estimates
as to the likely error component.
i
193
From (iii) rt =1 ße
02 0
then r+ (52 e=1
62 0
and (ýe -1-r t
d2 0
Thus: d2 = Uö (1 - rt)
The square root of this value is 'generally known as the
Standard Error of Measurement (SEM)
SF4 SD 1- rt
AP
This statistic is much used in educational and psychological
measurement and can be said to represent the standard deviation of
obtained scores around true scores. It is also conceptualised as
the standard deviation of error scores.
In standard psychometric practice it is often assumed that the SEM
is constant for all score levels and that the standard deviation of
error scores is normal. It is then possible to set up confidence
limits from the normal probability curve, where 68% of observed scores.
will lie 11 SEM units and 95% 12 SF4 units about the true score.
I
194
ALTERNATE FORM RELIABILITY COEFFICIENTS: RESULTS
The alternate form reliability coefficients representing the product
moment correlation coefficients between the primary source
traits of Forms A and B of Cattell's 16PF are listed in Table 9.1.
Results are given not only for the total sample but also for the
sample classified by sex. The standard errors of measurement (SEM)
in n-ste-n units for : ach factor have also been calculated and are
entered in Table 9.2.. But it should be noted that the values are
for the questionnaire forms used separately and have not been
corrected for two form length by the Spearman-Brown formula.,
The first observation is the general order of these coefficients
which are low by standards of cognitive tests. Factors H (social
boldness), F (happy-go-lucky) and Q4 (tense) are the most reliable
primary source traits which is in accord with Cattell's findings (972).
Primaries with coefficients around .5 are A (outgoing), C (emotional
stability), E (assertiveness), G (superego strength), I (tendermindedness),
0 (apprehensive) and Q3 (social precision). Next are two factors which
show rather low coefficients, namely B (intelligence) and Q2 (self-
sufficiency). Finally are source traits in the very low reliability
range of 0.2 to 0.3; these are the factors of L (suspicious),
M (imagination), N (shrewdness) and Q1 (radicalism).
It is of interest that most of the major introversion-extraversion
(A, E, F and H) and the main anxiety factors (C, O and Q4) show tolerable
reliability for variables in the personality domain. The same cannot
be said, however, for a relatively well documented variable in other
195
H H
W U
O
a
a a x U
a O L"a
CQ
4
u.
En H z
H
U U
W
cz N w E
ßf1 to CV rr N. N Ol to to Q1 O CO h co -e to r 01 r c: > tO O Cfl N- O Ln n') IS) q} L2 CD, ; ;
r- Cl Lc r
Nnt r
rLn r
Ci r F-
d
r
m C) M Qn to ct qt N- Ol O lD U) O to N 1ý
m to M cm tO U') LA rý mt r- ;; { C) r-- LO
NM r
rM r
NM r
NM r
rM r
rM r
t7 n O IA N n %Zt r LC) 0) U, Lo r O d' Co CV
CV Cf d tL ) Ln N 110 Ch N- M (ý) L! ) (ý7 r om
rM r
Co M r l") r
U) M r- Ch r
Co M
CM U) M U') U) (F) CV tO Ct LO CO Nr d' tO N- NM
r N rM 1.0121 N N N-O N rO MCQ C: Dr
alt CI) U)N - OM r
Co en OOM OpN
U) al M O N- clM er ul CO a' r (F) N- tO (F) 00 CV
O ýp MM Co LA rN if i Ln LC) LO O 1ý to
r 'ct r Od Oý cf Nd N «z;
(V Q1 ct U) O N- U) C% M O Ql 1- r l0 r %; t C') N
z N rr- to 00 CV U)O cPCD cV N-r- to N.
rM r
ON r
OM r
OM r i---M
F- C) vu r
M 01 1ý O) U) to O M Ida N 00 N U). N 0) (F) N OU) NN- N
- dd U) N- M fam. to Ch (M
CV M I
OC7 r
CVM r
OM r
r. M r
OlM
M Co CO MCo N-O N- im OD LO CO pM Q1M
M Net' UA N- N tD if O tD c' U) «t' O to
(ID M ON OOM 01N Iý. M Co cm
Lr) 0% CM If) tO Co to to to ",: t Ct 00 U) OO U) M tD to N Iý O it qi' r- ch CV M M N- tO CO
rM OM r
Q1(M 00 m M CV r
CJ r
M Mý ýr MO NM
M C`)t- to Co
= 1- m OO 00 N- "
CM to a1 tO N- "
th (F) 1- N-
Nl1) CM to r-
MLA r
MU) r
rLt) r-
" to
r
h 00 0' NN 00 N- Lil O
O a' 00 P- C xr rU) U)O er M NO -e OM N-C)
N NM NM CV M NM NN
d Co it tO 00 ßf1 t0 cf OO ct CD M N N- ýy tG q& O CO N- lD - mr r Co O tO U) Ol Q' U)
N Lff r) (V U) M U) CV ýk M Ct
r toto Nr O0 F N Co to CVI -e- le
W to CO ýt rr t1 f ý"ý M 10 O h ý ct O l0 N.. Oit r-
0-A& r
CM q& I
cm «Z- Ql'- cl
M M U) r OM Co Lin 0) O "-- tt Co O CO 00 rl_ V LA f- 01 N- O U) cV Ql CO Oº O
M r
'ý et r
Ln M F-
tC) ýt r
ýt M r
MM '-
M t'7 t0 r st Co N NN it C) ct' e-- C% I LO
m O F- M Qº N . - «t- O Oº O N C% N- eV N- - N- cV N- N t0 N fr
cu Ln N . - NtD MM a' to N- N0) cm cn 0 -el U'>
Q U) (V O it tD ct c2 O NW Ct O CO ý . -: OM ý
c%M 01M Com rN r
OM r
M in d Ö
d m -i cm Q C
v f C
i
Qr Z: C C N 10C RfC J 11 m MC du(s Z II
1
du C3 tud L N N L
Oº. CL) L+- L A-
iv; LL- 2c L LOi. to
i v)
Ci , o; N ca 9
U)
z ö
Un
E4 0
z U)
q
U W
U)
U)
u) N H
t-ý
O Ul
r v co r} rl N
I Cl I}' Cl ' V Or " "
M C') Cjý N U) LI Lf) 0'
r4 4 14 M ý' C
l0 M H 0 (V M Cl
r-1 -4 H
Cr) N z r r r
rH
r-1 ON
r-1 H r-i
(D o Lr)
H M ý1 lJ
r1 1 H
O O O
un Cr)
H H .1
co r H
N
Cr) M
r-I H r{
O 0
rI H rj
O pp -
Li `" ý"
O N C\l
r Sr ö ö
En En ö
ý 114, Au g2 2 O 7 W ..
40
197
researches (Eysenck, 1954, Wilson, 1972), Cattell's primary source trait
Ql of radicalism.
In order to investigate differences in the reliability of the primary
source traits for the male and female adult groups, the coefficients.
in Table 9.1., were converted to Fisher z-statistics (Table9.3) and
the statistical significance of the differences calculated.
The results of this analysis show that the male group had significantly
higher reliabilities on Factors A (outgoing), I (tender-minded),
Q3 (social precision)"and Q4 (tense) whilst the female group had a
significantly higher coefficient on Factor M (imagination). When the
means of Fisher z-statistics were calculated across all the primary
source traits, there was no significant difference between the two
groups at the 5% level (z for males = . 523, z for females = . 498).
aw We must conclude, therefore, that although the sex of the subject was
related to the reliability of five of the factors, there was no
significant trend for either sex to demonstrate consistently
higher reliability coefficients across all the primary source traits.
To test the hypothesis that the adult reliabilities would tend to be
lower than those reliabilities collected for student groups on the
primary source traits a comparison was made not with Cattell's American
student groups but with data based on a more relevant sample of British
second year undergraduates collected by Saville and Blinkhorn (1976, pp 78).
For the total group of undergraduates (N = 1148) and adults (N = 2007)
the student reliability coefficients were significantly higher on eight
198
H H
N
aw x
Hz
W[z,
Cif
cn
Aw H il Ö 4'
mz cn
cli
ý>ä Co Ä E-+ U4
0 r-I CO
w
t0 M H
r-1 0
M m
lJ
lppp V'
Cr)
H 0 " " " "
N d' i co
/
0
Cy rQ c\j o n
N M
Cr) in
H
o 0
I N °
z N N 0 Z
C\j r-i
CM 0
N O
In O In N
" cyl 0 V3
z rn O
H M co tN , -1 H In Cr)
" r-i
" 0
"
c C
In
C) " " o
t. vn z
N M
O H
N N
C7 un u1 O vl
10 CD in co N 0 Vl
in o
In N
0 co
10 in 0
10 C (Y) ý
u 6f'ß In O fn
H LO
d M m
Pq d' "
O z
i I
0 M
mot' O O
G] E4
vl ýc7 ,
ýý
W
Ü
H V
W 1 N
(W Z0
ý - N H 4 H
Gý Q vn
4s
199
oý w m H
x H H
N M d, O1
"
t0 ýD
"
l0 cri
""
N 0
"
Lt} 0
"
N 1M Q1
M
Of Co
"_ 110 l0
" LO Ln LI
"" 1
a) 0 Ln 0
" N d'
Cr)
Lo
1
I 14; r N co 0
! to to O
NN ""
Ln N
r-4 O
" .
%D "0 En
1 to
Ip tD
uý Ln LO
H r-1
z N N C. (V O" Z
Cr) M' NN d' E N N N1N 0
t
Z
1 N c '(
H ýD 0
1.4 co c n c c4 1 z
O M O
M Cr L
0
H t0 ý 1 ""
Ö
" z
S
O 1 0)
ON MN O
r-f x ao 1o N1O '-+
H
ID M H CMM1 ,... I t7 ýn to d'
0 N
N W CO
N LO t- ID N
N 0) 0
U) O
Oý N1 fit' 0 0 to lO ý1 1 l0
I O. vl
d' 1 d'
0 I rI
z 1N
r- N
N4,1- It) O In pq c+) M ýP d' O
1 O
CO M -4 ko FrT- H I r-4
. eC to Ln I-n 0
Cl) N 11 N
'-" W WU 2
pQ C\j 1 U 0
H W
GH
V
G] tl ý7 (I (s4 H
¢. zz Äz Iý
ý `ý A M
200
of the sixteen primary source traits. These were Factors A, F, G, H,
Q1, Q2, Q3 and Q4. The adult group showed higher alternate form
reliabilities on Factor B whilst the remaining seven factors showed
no significant differences. When the Fisher z-statistics were
averaged across all the primary source traits for the two groups,
the difference was significant at the 5% level. Thus, with the
exceptions noted above, we can conclude that the overall reliability
of Cattell's model as represented by the 16PF is in fact more reliable
on the undergraduate sample than on the general population adults.
On inspection of the descriptive statistics for these adult and student
data collected by Saville and Blinkhorn (1976), it is evident that these
are differences not only between mean scores of the two groups but also
between the scale score standard deviations, the students showing greater
dispersion of scores for the majority of factors. The restriction in
range phenomenon could be responsible therefore for these differences in
reliability coefficients and indeed when restriction of range is taken
into account by use of the SEM statistic , the differences between the
groups are relatively small.
The finding that students showed greater variabilty on most of the primary
source traits is perhaps a little surprising especially as undergraduates
are very selected sub-group of the population. However it must be remembered
that although restriction in range can arise through some influence extrinsic
to the reliability of a test or scale ( for example members of a sub-group
being in some real sense more alike, or perhaps some artificial
truncation of distribution), restriction of range can also as a result of
an influence which is intrinic to scale reliability (inappropriate or
unreliable items tend in themselves to restrict the overall dispersion
of test scores).
201
We should be cautious therefore in making the assumption that correction
for differences in the variability of scores by way of the SEM statistic
enables direct comparison of the reliabilty of a scale in different samples.
It could well be that greater dispersion of scores in the undergraduate
sample is a result of more appropriate, meaningful and reliable item content
for student groups.
201a.
All
10. AGE AND SOCIAL CLASS AS MODERATORS
ON RELIABILITY COEFFICIENTS
202
INTRODUCTION
It is not sufficient that the measurement of personality be reliable
for the adult population as an entirety or for that matter for groups
of students. It also needs to be shown that scale reliability is not
adversely effected by variables such as the age and socio-economic
class of the subject. Low reliability for sub-groups of the
population might suggest that a trait lacks stability, that there is
restricted range, that factor structure is not invariant across
subjects or that item content is simply inappropriate to certain
types of individuals. But whatever the origin of moderator effects
upon scale reliability, they remain a crucial concern to the trait
theorist, especially if his theories make claims of general
applicability, yet had their genesis in small and perhaps atypical
academic groups.
In an earlier section it was postulated that Cattell's primary source
traits would be less reliable on older, lower socio-economic class
subjects. In order to test this hypothesis, tables in Appendix 10
contain alternate form reliability coefficients broken down by the sex,
age and social group of the respondent for each of the primary factors.
Tests yielding only one score are amenable to the investigation of
changes in reliability with moderator variables without recourse to
special techniques. The Cattell inventories, with as many as sixteen
scales to be considered, present a more complex task. One approach to
this problem, however, is to rank the equivalence coefficients within
203
each factor over the classifying variable, and sum the ranks over
all factors. The resultant sums of ranks would be an index of the
relative magnitude of the equivalence coefficients for all factors
with respect to each classification.
AGE AS A MODERATOR ON RELIABILITY: RESULTS
Accordingly, Table lO. llontains the results of ranking the alternate
form reliability coefficients from highest (rank 1) to lowest (rank 4)
on each factor and producing a frequency count of these values across
all 16 primary source traits for each of four age groups (16-24,25-44,
45-64 and 65+).
TABLE: 10.1.
RANK FREQUEINCIES OF THE ALTERNATE FORM RELIABILITY COEFFICIENTS
CLASSIFIED BY AGE GROUP
Most Least Reliable Reliable
anks Age
1 2 3 4 MEAN SD
18-24 6 5 4 1 2.00 . 98
25-44 4 7 2 2 2.09 . 97
45-64 3 3 6 4 2.72 1.06
65+ 2 2 3 9 3.19 1.11
It is evident from this table that there is a gradual decrease in
the chances of a high reliability coefficient for a given factor as
s
the age of the respondent increases. This trend is significant at t
204
the 5% level (x2 = 9.00) when the data are represented in the
2x2 contingency table below.
TABLE: 10.2.
SUMMARY TABLE: RELIABILITY CLASSIFIED BY AGE
"Higher" "Lower" Reliability Reliability
Ranks 1&2 Ranks 3&4
16-44 years 22 10
45-70 years 10 22
x2 = 9.00
We can conclude, therefore, that there is a significant trend for
Cattell's primary source traits to show lower reliability coefficients
for sub jectu of higher ages.
An alternate approach is the use of the parametric Fisher z-statistic,
a logarithmic transformation which enables. the inspection of product-
moment coefficients for significant differences. In Table lO. 3the
reliability coefficients for each age group have been converted to
z-statistics and the number of significant differences calculated on the basis of two - tailed significance tests.
It is evident from this analysis that seven factors, with a total of
fourteen significant differences between reliability coefficients,
gave results consistent with the hypothesis that older subjects
would show lower reliability on the primary source traits. These
205
m ö
H
cn
E-4 t H E-i
rx
U) H W 0 E-1
UI E-f
H
W W 0 U
H H
H L¢n
Ha
N cri t
x E1 O C-l
CO N ri r- a r- ý c ca
0 O
ýp CV 01 U) a 0
" L
" " 0 0
0) LI) co
or d' d d' co 0
N N
"
CN CV
"
CO 0
"
CO ri
" N 0
00 ca
M Lr)
N -4
O V E- 'D 0 O
W C\l
Lo -1 CIO
OD r-
z N "
N "
CV "
O "
0 O
co N M 0
H M Co
N co 0) N
co "
M
" N co
"
oN
."
O 0
M lD N
H M w
LS N U
If) ca 0 0
N N H
X. 01 0l (71 N N 0
C7 ý
" " M
" M
" N O
M CO
1 0CP\ co cl-
H
k+ .
CO "
Ln 0
Ln "
ý") 0
M 0 0
ý) to H 0
N H 0) M
0
W c ý d 0 "-1
° cC ,n
LO U') ý0 Ln
rn CO O N
z co -4 I ý 4 H
" d
" d '
" -73 Ö -I-ý a
ý J (
U C1 4 M
U "d "di
s4 ý0v ., i a Q 4-3
o Icr unul°0 inUlP'
W N I w U r ti 4-t "r -º 4-4
U
r N ý ONO O , -I . ca
ON 41 NN
A 4a O 4-4
4-i E-i 4-4 Z
ý Ä Q-
206
R4 ': ß FREQUENCIES OF THE ALTERNATE FORM RELIABILITY COEFFICIENTS
CLASSIFIED BY SOCIO-ECONOMIC GROUP
Most Least ReliableR eliable
e, SS
AB 1 19 6 0 1 1.56 . 81
Cl 6 6 3 1 1.94 . 93
C2 1 3 8 4 2.94 _85
DE 0 1 5 10 3.56 . 63
27
were namely the Factors of F, G, H, M, Ql and Q2. The three
primaries of A, B and C on the other hand demonstrate a total of
four significant differences in reliability which were in the
opposite direction to that hypothesised.
SOCIAL CLASS AS A . NOD: BATOR ON RELIABILITY: RESULTS
moderator class as a on rý.! _iabilitV, the
coefficients were first ranked from highest to lowest in the same
manner as with the age data. The results of this analysis are
contained in Table 10.4. and in the 2x2 contingency table reproduced
below which is significant at the 1% level (x2 = 30.25).
TABLE: 1O. 5: SOGTAL CLASS AS A MODERATOR ON TI-CE RELIABILITY OF THE PRIMARY SOURCE TRAITS: SUNLMARY TABLE".
"Higher" "Lower" Reliability Reliability
Ranks l&2 Ranks 38: 4
Social Group A, B, Cl 27 5
Social Group C2, D, E 5 27
x2 = -30.25
The corresponding Fisher z statistics are presented in Table 10.6.
In this case, thirteen of the sixteen source traits show a total
of 37 differences which are in line with those hypothesised, (-'-,.,,
Only Factors C, 0 and Q4 do not show significant differences
between the reliability coefficients of the four socio-economic
groups. In addition, the mean of the sixteen reliabilities for
the AB group is significantly higher at the 1% level than that . 11
of the DE group.
208
C;
H
En H
U) rl
N
H Ga O H
zl w
O
H
WQ
aH
W
O
H a a
U
a
8 I)
Iý
P4
N co O o
Cl Co N N N
m '
M t-
I CY)
M CM I O Q N lfl d
LN In
N I
Q' om Co O ci
,
1-4 In C\i
N -4
N O
Co d O
a M d' N N
C) C C) N
C) C) J
Z 0) -
N N N r!
N r-i
r-1 M cM CO
N
N qlo
N 0
x O
Q
c3)
C% "
Co C) oý
"
N I
" N 0
ri
ýO
N
ul M ýP
O d' M O
Gy p N
Co
O, % 0,1
00 N
N 0
G1 M N (Y)
ý0
M
u1 c`ý O
C") M
l0 (Z)
Co LO M
U un Ln O O
N N N M
Ql (
CU 0
O
N Co 0
N O
Q2 d' (D
O CD
Ct A
'O
O 4
U N
ß IN r 4i SI
äi v1 .
r!
WON
vl 0l( Ö
M U)
r-1 0. " 0) 'ßx5.
u N 4-4 ON
4-f O
a v V fý 41 0, G)
NN ab ab
°O .
4-1 ÄQ Ä
41
209
AGE AND SOCIAL CLASS AS MODERATORS ON THE RELIABILITY OF THE PRIMARY SOURCE TRAITS: DISCUSSION
Results give tentative support to the hypothesis that the older age
groups (45-70 as against 16-44 years) would show lower reliability
coefficients on the primary source traits. This was by no means a
particularly strong trend however, but was mainly as a result of
the action of six of the sixteen primaries (F, G, H, M, Ql and 02).
It was interesting that the non-parametric approach of ranking and
using a chi-square analysis on age as a moderator variable gave more
positive results than the parametric Fisher z statistic. Although
the mean z statistics for each age group across all 16PF factors
were generally in the expected direction, they did not reach
statistical significance as did the chi-square analysis. This may
have been because of the large differences in the size of the alternate
form reliability coefficients across the various factors which perhaps
mask the more subtle influence of age. The conversion of these data
back to correlation coefficients shows that the mean reliabilities
across all 16PF factors were . 490 for the 15-24 year age group,
. 495 for the 25-44, . 460 for the 45=64 and . 435 for the 65+ age
group. in view of these relatively. small differences, it is clear
that weilst there may be some slight trend for lower reliability in
older people, variations as a result of differences in factor
reliabilities are likely to be a more important consideration than
differences as a result of age.
We also saw that there was a tendency for the adult group to show
lower reliability than data collected for university students S
210
(Saville and Blinkhorn 1976). This is despite the fact that
undergraduates are a restricted group both in terms of age and
academic attainment. Thus, the evidence supports the original
hypothesis and, to some extent, those who are critical of the
applicability of questionnaires in general. It would seem that when
scales are developed on specific sub-groups of the population such
-is __..; _nt:, that se'_aýcted items are not such a good fit to the
model in the adult population as had been the case with the original
development samples. This is an important point to the personality
theorist who should seek to avoid item content which may be more
appropriate to student groups than adults in general. After all,
a generalised model of personality should provide reliable
measurement not just across specific accessible student. groups
but across the entire population range.
When the effects of social class were investigated as moderators
upon scale reliability coefficients, the results were more positive
than with the age analysis. There would seem to be good evidence
that the 16PF scales showed less reliability with lower socio-economic
groups. Both the parametric z statistic and the non-parametric
chi-square analysis gave very significant results here. In fact, only
Factors C. 0 and Q4 did not show significant difference between the
reliability coefficients..
Apparently then, social class has more effect upon scale reliability
than the age of the respondent. Perhaps this is not altogether
surprising when one considers that social class is some, albeit
indirect, measure of educational attainment and, therefore, academic
211
ability and we may merely be saying here that those of lower scholastic
ability tend to be more unreliable in their responses to questionnaire
data. It is certainly true that the lower social classes will have
had less experience with questionnaire techniques, with form completion
and all the assorted clerical activities which are more usually the
part of higher social economic jobs. Indirect evidence in support
of ß: i0 explanation is the positive correlation between social class
and Cattell's primary source trait of intelligence (Factor B), which
we observed at an earlier section.
The Fisher z comparisons between the reliabilities of the four
social class groups employed here did in fact show significant
differences and when these were converted back to correlation
coefficients, it was found that social class AB had a mean reliability
of. 540 across all 16PF factors, Cl a mean reliability of . 535, C2 a
mean reliability of . 465 and the DE social class a mean reliability
of . 435. Quite clearly these differences are large enough to be of
importance in scale interpretation, even allowing for the large
variations of reliability across the primary source traits.
Certain factors had particularly low reliabilities for most measurement
purposes. It would appear that Factor L (suspicious) especially in
the lower social classes, Factor M (imagination) in the lower social
class and higher age groups, Factor N (shrewdness) in the lower social
classes and higher age groups and Q1 (experimenting) mainly in the
higher age groups, were particularly vulnerable to these moderator
effects. For these source traits, the more extreme age and social
class groups possessed such low reliability as to bring into question }
212
the very meaning of scores on the primary source traits obtained by
such subjects.
As a final question we might ask what the effect of restriction
upon range of scores within certain sub-groups had upon the
reliability patterns reported in these various tables. Although
1972) to e t-L-Tl . te -41e reliäb. 'i1ityr
of a scale based on different group standard deviations, they make
the assumption of the consistency of standard error of measurement
across the two groups. Clearly, this is an assumption we cannot
make because this is the very effect which we are investigating.
What is more important is the fact that 'correcting' group statistics
for restriction in range can often be missing the point. It is a
perfectly legitimate aim for a test constructor to increase the
discrimination of his scales to a maximum. All things being equal,
this will tend to help the reliability of his scale. If, therefore,
certain personality source traits show markedly reduced variances
across a certain age group or social class, correcting for such
inconsistencies would be fallacious. We would only gain statistics
based on artificial assumptions, which will have no practical or
theoretical value in understanding the effects of moderator variables.
Nevertheless, it would be remiss not to investigate the extent to
which restriction in range of scale scores affected the reliability
coefficients calculated here.
I
213
11. VARIATIONS IN THE STANDARD ERROR OF MEASUREMENT
214
INTRODUCTION
Up to this point we have researched variations in the values of alternate
form reliability coefficients. The problem of interpreting such data
where restriction in range is suspected within specific groups of subjects
has also been introduced. The statistic which is commonly supposed to be
r..... ^. nrdeni= of Va iat! iity the sample is the standard error of
measurement and perhaps we should consider this statistic in a little
more detail.
It was stated in an earlier point in this section that the standard error
of measurement (SEM) is commonly supposed to be constant at different
score levels. We have previously defined the SEM as:
SEM SD --r
Guilford (1973) states that the SEM is ablanket index and that if the
score distribution approaches normality and if obtained scores do not
extend over the entire range, the standard error of measurement is
probably uniform at all levels.
Surprisingly, few investigators have challenged this assumption, and the
published studies which do exist seem to be restricted to the testing of
educational attainment. Gulliksen (1950) concludes that the SEM should be
constant for all score levels when the distribution is not skewed and when
kurtosis is three. Mollenkopf (1949) investigated the effects of skewness
and variations in kurtosis on the SEM and concluded that slight skewing
could be tolerated but not departures in kurtosis from three. Lord (1352)
points out that in most distributions the dispersion of errors will be
smallest at the tails of the distribution. Thorndike (1951) concludes,
on the other hand, that the precision of measurement will be smallest at
that point in the score distribution where item difficulties are concentrated
215
which for a general purpose instrument is likely to be the mid-point
of the scale.
The lack of published work on the possible variability of the SEM in
personality scaling is perhaps the more remarkable, when one considers
that the question as to whether certain personality traits in reality
constitute continua is open to more doubt than the more tangible
variables met in educational test data.
A particular problem is one of the dimensionality of the middle response
where, as with Cattell's inventories,. the subject is required to choose
the most appropriate of three possible responses to the item stem. In
such an item format, the middle response could easily become the catch-all
for otherwise unclassifiable responses including item inapplicability and
uncertainty of meaning for the subject. In view of these points and Lord's
work (1952), it would seem reasonable to suppose, therefore, that the
extreme categories of the primary trait score scale will show the
greatest reliability.
Turning to the mechanics of calculating the SEM at different parts of
the score scale, Guilford (1973) has suggested that if there is reason
to suspect that discilnination is actually unequal along the scale, this
can be examined by preparing a scatter diagram showing the relationship
between two forms of the test or questionnaire. The standard. deviations
of the columns and rows at different score levels will indicate where
predictions have the greatest accuracy.
One problem with this approach is that it is more akin not with the standard
error of measurement but the standard error of estimate which is defined as
the standard deviation of errors of prediction from correlated measures and }
is given by the formula:
216
SEE = SD 1- r2 X xy
Comparing this formula with that for the standard error of measurement:
SEM = SD x-r x xy
it is apparent that the SEE isVi + r11 as large again as the SEM. "IV
What is not generally realised is that the SEM is in essence the
SEE corrected for unreliability (attenuation) in one of the variables.
That is to say, that if we take the SEE formula and correct it for
attenuation in one variable by the standard formula we can derive
the SEM.
PROOF
(i) SEE = SD 1- r2
(Standard formula for the standard error of estimate)
(ii) Correcting r, for attenuation in one variable then: -
r=r xy xy
Ny yx
(iii) Substituting:
(correction for attenuation formula)
SEEC SD 1_Y2
ryx ý _.
217
2 SD 1 -r r yx
(iv) But in a reliability study the best estimate of yX = rXY
thus substituting in the formula then:
SEE SD 1-r2 cx
r XY
= SD 1 -r XY
which is the standard formula for the standard error of measurement.
Essentially the SEM is a hypothetical value which apportions error
variance to the two forms in a reliability study and makes an estimate
of the unreliability in the form assuming the other to be perfectly
reliable. Calculating the standard deviations of columns and rows at
different score levels of the scatter diagram as proposed by Guilford (1973),
takes no account of this and if such data are to be interpreted against
the overall standard error of measurement, then the suitable correction
factor would need to be applied.
Lord (1952) has proposed a simple formula for calculating the standard
error of measurement along different parts of the score scale.
SEM = to (n - ta) p (n-1)
where SEMp = standard error of measurement at a particular }
point on the score scale
218
to = total of items correctly answered
n= number of items in the test
For our purposes, however, rather restrictive conditions apply to this
formula in that it assumes that items are scored 0 or 1 and have been
assigned either randomly or in some stratified fashion to the alternate
forms of the test. It is ä1so assumed that one is sealing with
homogeneous tests and not the deliberately heterogeneous item content
of Cattell's inventories.
Fortunately, it is possible to take another approach to the problem
by reconsidering a basic definition of the standard error of measurement
as the standard deviation of observed scores around true scores.
That is: -
SEM = 7- (x - xt) 2
N
where x= obtained score
xt = true score A
N- number of people in the group
Whilst we cannot ever be absolutely sure of an individual's true score,
it can be estimated by calculating the mean universe score, which in this
case will be the mean of the Form A and Form B factor scores. That is: -
ASS xt A+B
2 }
219
where xt = estimated true score for an individual
SA = sten on a given factor for Form A
SB = sten on the same given factor for Form B
By substituting in the above formula then: -
SF1, i SA _ ($A s) -2
2
N
which reduces to:
SEM (SA _ SS)
2
4N
By the use of this formula it is possible to make an estimate of an
individual's true score, calculate its departure from the observed score
(either Form A or Form B) and then compute the standard deviation of
error scores around true scores for each point on the N-sten score scale.
It will be noticed that this formula makes no assumptions about using
binary data or that the test is homogeneous in item content. Moreover,
the use of both forms in estimating'the true score means that the
values gained by way of the above formula are more analogous to standard
errors of measurement than the simple standard deviations calculated from
the rows and columns of a scattergram.
The foregoing demonstrates some of the problems which arise
when investigating errors of measurement. The most obvious source
of confusion is the definition i
220
of what really constitutes error.
obtaining column and row standard
in fact based upon the logic of d
estimating (predicting) the score
parallel test. We shall refer to
an observed score where:
Guilford's (1972) method of
deviations on a scattergram is
fining error as the error made in
on one test from the score on a
this as the error in estimating
si_ std /1-r2
x xY
An alternative but equally acceptable approach is to define errors
of measurement as the error involved in substituting a score on one
test for a score on a parallel test.
In this case we can write the formula:
SE 1d d2 /(XA XBi 2
where SEMd = standard deviation of the error in
substituting a score on one test for a
score do a parallel form.
XA = score on the first form of the test
XB = score on the parallel form
d= difference between scores on the two forms..
Finally we have the more commonly used definition of error developed
here and that is the error made in substituting the observed for the
true score:
221
I
SEM =X _X It2
N
SD Vx-r
x xy
On the assumption that the best estimate of an individual's true
score on a given trait is the average n-sten on the parallel forms
of the 16PF, then we have established above that: -
SEM (SA - SB)2
4N
This statement bears a close resemblance to that derived for the
SEMd (the standard deviation of the error in substituting a score
on one test for a score on a parallel form) and in fact is 1 as large. 4T
We can see, therefore, that the SE14 will be half as large as the SEMd
on the same data base.
There are other possible definitions of error but we shall investigate
how the conventional definition of the standard error of measurement
as the standard deviation of observed scores about the true score
(SD a1 --r ) affects our conclusions on the differential reliability
of the primar' source traits. This particular statistic has been
adopted here because it has the distinct advantage of being based
upon the most generally accepted definition of error according to the
classical true score model.
222
AGE AND SOCIAL CLASS AS MODERATORS ON THE SEMs: RESULTS
SENs were calculated in raw score units on the primary source traits
for each age and social class group. The results of this analysis
are presented for the total sample and for the questionnaire forms
separately (Table 11.1. for Form A and Table 11.2. for Form B). Age and
C l_ ? class were divided Into the four ca tego'ries listed in the
tables.
Because the data are represented in raw score units, no meaningful
comparisons can be made between the SBMs on the different personality
variables. To ensure comparable results, therefore, SEMs for each
social class and age group were ranked on each factor independently
from highest to lowest. Summary tables were then prepared and a
chi-square analysis undertaken (Tables 11.3 to 11.7. ).
AGE AND SOCIAL. CLASS AS MODERATORS ON THE SEMs: DISCUSSION
By referring to Table 11.7 we can see that overall the Form A
scales had lower SEMs in older than in younger subjects, whilst in
the 16PF Form B there were no significant differences. This is in
direct contrast with our previous analysis utilising reliability
coefficients where older subjects tended to have lower reliabilities.
Because the only other statistic appearing in the SEM formula is the
standard deviation of scores, we are forced to conclude, therefore,
that most of the difference in reliabilities amongst age groups is
due to a restriction of scale score variability amongst older subjects;
223
this effect being particularly pronounced in Form A of Cattell's
16PF questionnaire.
Turning to social class, we find that whilst those classified as
being in the lower social classes tended to show lower dispersion,
they were still significantly less reliable in terms of SEMs on the
trait scales of the 16PF than those in the A, B and Cl socio-economic
groups. This trend was significant on Form A but again no differences
were found on Form B of the 16PF.
It would seem, therefore, that a distinction can be drawn between
differential reliability as a result of age and differential
reliability as a result of social class. In the case of age,
reliability coefficients were lower with older individuals as a
direct result of restricted scale discrimination with such subjects.
It is possible that older subjects have a central tendency response
style (Cuilford, 1954) and choose the middle response to the various
questionnaire items more frequently than their younger counterparts.
However, if this is the case, they do not act with any less
consistency and indeed they may act with more. The situation with
social class seems very different on the other hand. Although both
questionnaire forms showed restriction in range in lower social
class individuals, when this lack of variability was taken into
account by use of the SEM statistic, the C2, D and E groups still
appeared on Form A as the more unreliable group in terms of their
alternate form scale scores on the primary source traits. Again,
there was no significant trend on Form B. In this case the smaller
dispersion of scores in lower socio-economic classes and the
224
Q H
w x
x 0 aI
H d` O a0 CJ CO OD Co Co C " . . . " " .
M N CV in CV CV CV N
H C: ) m M V CV N N M M M
N CV N CV N N CV N
w N N in in L G l0 U-1 (X)
N N N N N N N N
N ch N (0 to Co N r1 00 Oa 1-0 N N Q'. N
N N N N N N N N
J N i9 Co i' ltd L) lý ýý
N N N N{ N N N N
Co i «e CM ýD
z pp to (0 l0
N N N N N N N N
h r-f p i H N c,
p O
N N M M N N M M
N d' CV L0 t a)
a o, ao rn w N 1o o co N N cV N
" N (N «i N
N N 01 M lo H to N ttl d' to to
(i CV N N N N N CV
OD (D d' 0 N IN M d`
x pp N N M N CO C) O1
N N N Cr1 N N N N
H M N N Q'. a) V'" ýp l(ý Ln to LO Ul
N N (N N N N N N
N r-1 N H 0) LID In O
W N O "
O O 01 rn CD N M N M M N N ("')
W d' aO
0 O
1-4 OD t0
M O
00 N 0)
N M N' N M N N N
Oý Gp [1' ýO Ci l) CD ý. uD Co N N
N N N N N 'CV N
N ap
in 00 ul
r- 1. D G% N m (n
LI)
r1 O N
O rI
O O
Oý r-1
-1 . -I
a) r-1
r" O
Q N CV N CV N N N
, aM 31
'n ý 4 U U Ä
H N om ( , 0) ü
225
C\l
H
U) U)
U a
8 U) A z Q w 0 Q
0 z H A f1' O UI Qi
£V1 W N
A
z O. G. fsr
w
0 w
lfl H M U) t0 11) rn o) rn o o, rn co rn N N N M N CM N N
OD t0 d' t() d' N d' l0 d'
N N N N N CM N N
[N N CN LN l0 to U) U) N M d' to N to dý M U)
" " " " " " " " CM N N N N N N N
cm CY) o M ) M r-{ L,; U) l p
c M U
CM N N N 0
N N N N
co '. p (p N U) N N N
,ý o rn o 0 0 o o o M N f') C'M CO M M M
C\j 04 to dý ý4' to Co V' d' mal' CM N N N N N CV N
to 0 M %44 ka . ý. H ri N H M N H O
M C') C') ('') M CM f7 M
H N H ko U) IN H N N N N N H C') H N N N N N N N N
CM U) 0 M N l0 a1 H
N N d' rI cß'7 N N M
N N N N N N N N
Oý [l' r1 rI Ql '. p N N
x o ao 0 '. o co m o 0 M N N M N N c'') cM
to 0 H M O U) lp cr H N M N N H H l
N N N N N N N N
H CO CO s r-1 O a)
0) 9
H N
N N (V) N N N N N
r-I O kO m C) l0 N N W l0 00 N U) Ný l0 l0 [ý
N N CV N CV N N N
kD N N O LO kp N OD CO N OD 00 OD co 00 N N N N N N N N
d' d` d' aD raj N
l0 0'% CO H U) %D' O' C') hfl 1l d' M U1 % ul Uf)
N N N N N N N CM
Ef) C N d l0 Ui 8 4.
+ Q U U U) r-i
u) CM
11 d'
l0 O ClU
226
TABLE: 11.3.
SEN RANK FREQUENCIES FOR SOCIAL CLASS
BOTH SEXES - FORM A
RELIABILI'T'Y
SOCIAL Low > High
CLASS 3 4
AB 4 3 3 6
Cl 2 1 7 6
C2 6 3 4 3
DE I
4 9 2 1
TABLE: 11.4.
SEM RANK FREQUENCIES FOR SOCIAL CLASS
BOTH SEXES - FORIIi B
SOCIAL
RELIABILITY Low High
CLASS 1 2 3 4
AB 6 5 2 3
Cl - 5 4 7
C2 5 3 6 2
DE 5 3 4 4
227
TABLE: 11.5.
SUMMARY TABLES: SEMs AND SOCIAL CLASS AS A MODERATOR VARIABLE
FORM A
SOCIAL RELIABILITY (SEM) CLASS Low High
Higher Social Class 10 22
Lower Social Class 2.2 10
x2 9.00 pC . 01
FORTI B
SOCIAL RELIABILITY (SEM) CLASS Low High
Higher Social Class lý 16
Lower Social Class 16 16
chi-square =0
228
TABLE: 1.1.6.
SEM RANK FREQUENCIES FOR AGE
BOTH SEXES - FORM A
AGE
RELIABILITY Low High
2 3 4
15-24 6 7 - 3
25-44 3 5 4 4
45-64 1 4 8 3
65+ 6 - 4 6
SEM RANK FREQUENCIES FOR AGE
BOTH SEXES - FORM B
AGE
RELIABILITY
Low High
1 2 3 4
15-24 2 8 2 4
25-44 4 2 6 4
45-64 5 4 5 2
65+ 5 2 3 6
I
229
TABLE: 11.7.
SUMMARY TABLES: SEMS AND AGE AS A MODERATOR VARIABLE
FORM A
AGE RELIABILITY (SEM) Low High
15-44 21 11
45-65+ 11 21
chi square = 6.25 p . 02
FORM B
AGE RELIABILITY (SEM) Low High
15-44 16 16
45-65+ 16 16
chi-square =0
230
unreliability of their responses as gauged by the SEM, suggests that
it is perhaps the irrelevant item content which is at the root of
differential reliability here. Moreover, Form A of the 16PF seems
to be the main propagater of this bias.
The foregoing seems to suggest that age per se is not as an
tmajor rr; od gyrator vzriab e on the reliability of Cattell's
primary source traits as social class. Any tendency for reliability
coefficients to be lower in older people can be accounted for by
lower dispersion of scores in such groups... Exactly why dispersion
should be lower is difficult to establish. We have noted that for
certain factors, e. g. L (suspicious), that males and females are
relatively close in mean scores at about twenty years of age,
increasingly depart from each other up to forty-five years of age
and then begin to move closer together. If sex differences were
less extreme at older ages, this would tend to decrease the
dispersion of scores. On the other hand, it could be purely a
matter of response style, older people prefering to take the safe
middle response in preference to the extreme alternatives.
In contrast, social class is seen as an important moderator on the
reliability of the source traits. Those classified in the C2, D and.
E categories not only showed restriction in range but considerable
unreliability within that range. An important point is that the test
constructor should be particularly concerned with the applicability
of items with lower socio-economic classes.
In conclusion, we must state that there was conflicting evidence
231
on the effects of age as a moderator on reliability. However, our
hypothesis that the lower social classes show lower reliability is
supported by the data collected here.
RELIABILITY AND THE PRIMARY SOURCE TRAITS: CONCLUSIONS
In this section on the reliability of Cattell's source traits, based
on a sample of approximately 2000 British adults, we have seen that
few significant differences were found between the sexes. Similarly,.
there were few differences between results reported here and those
published by Cattell in the United States. The adult sample tended
to show lower reliability coefficients, however, in relation to
university undergraduates. This was mainly as a result of less
variability in the primary source trait scale scores for the adult
group.
The most reliable of Cattell's primaries, with reliabilities. in excess
of . 5, were A (outgoing), C (emotional stability), E (assertive),
F (happy-go-lucky), H (venturesome), I (tender-minded), 0 (apprehensive),
Q3 (controlled) and Q4 (tense). Factors B(intelligence), G (superego
strength) and Q2 (self-sufficient) were of moderate reliability in
the range .4 to . 5. The remaining source traits, L (suspicious),
M (imaginative) and Q1 (experimenting) all demonstrated low reliabilities
below a value of . 4.
On the study of moderator effects in reliability, it was found that
older and lower social class subjects had lower alternate form
232
coefficients but when account was taken of restriction in range in
the scale'scores of older subjects by use of the SEM statistic, the
older age groups were possibly more reliable than the rest of the
sample. The conclusion that social class was an important moderator
on scale reliability still held after correction for restricted
variances however.
Of course, the problem of moderator effects on reliability is not
peculiar to Cattell's questionnaires and it is useful to take a
retrospective look at the data to see which particular source traits
are most prone to social class influences on reliability. The factors
most influenced were E (assertiveness), G (superego strength),
I (tender-minded), M (imaginative), Ql (experimenting) and Q3 (controlled).
On the other hand, the reliabilities of the three major anxiety
components of C (emotional stability), 0 (apprehensive) and Q4 (tense)
were least affected by the social class variable , none of them showing
any statistically significant differences.
That is not to say that other factors were unaffected by the social
class variable. We should not lose sight of the fact that other
traits with unusually low coefficients also showed a decrease in
reliability in the lower socio-economic groups. For example, although
the difference did not reach statistical significance at the 5% level,
the coefficient for Factor L (suspicious) was . 393 in the middle class
and . 261 for semi-skilled, unskilled and unemployed subjects. A
decrease in reliability which, for many practical purposes, calls into
question the meaningfulness of scores across the two 16PF forms.
233
Our final conclusion must be, therefore, that reliabilities calculated
on studen* groups give a better impression of the consistency of
scale scores across the 16PF than is the case with adult samples.
That is to say, that the power of measurement afforded by Cattell's
model as represented in his two major questionnaires is less impressive
on general population adult groups than was realised hitherto.
This conclusion makes a detailed investigation into the internal
structure of the scales, which purport to measure the primary source
traits, even more necessary and this topic will be taken up in the
next section.
I
234
SECTION 5:
INTERNAL CONSIST ENC YAND
T ,HEPRIMA
RYS0URC E TRAITS
12. INTRODUCTION
I
235
INTRODUCTION
Up to this stage our researches have involved the descriptive power
and reliability of Cattell's source traits as represented by scale
scores. We have noted that certain scales which purport to measure
the primaries at times show inconsistent relationships with external
variables and demonstrate poor reliability, especially in the lower
socio-economic classes. Clearly it is now necessary to investigate
the internal consistency and item characteristics of the model,
an area of much controversy in the past.
There are two basic problems which face the test constructor at the
outset of his enterprise. The first is to discover, choose or
otherwise define the nature and names of the dimensions he sets out
to measure. The second is to develop items which measure these
dimensions efficiently. Of course, this logical separation is not
necessarily mirrored in the actual practical business of scale
construction. One might for instance start with a set of items,
factor than and develop tentative scales and then by a process
of distillation arrive eventually at a set of scales and items
to measure than. Nonetheless, there is a logical distinction to be
upheld between dimensions or traits in a theoretical model and
the device used to measure than.
There is the paradox in scale construction that if one starts with
a set of dimensions and sets out to find items which measure then
successfully, one runs the risk of being accused of producing
entirely arbitrary scales, whilst on the other hand, to start off }
with a totally unstructured approach creates difficulties in that
236
no criteria are available for the selection of items for the initial
item pool. Moreover, if scale development begins without any
hypothetical structure, there is the risk that the emergent factors
will be obscure, difficult to name and of no practical value to
the clinician or other test user. As Nunally (1967) has suggested,
the best approach is probably to start off with hypothesised
'Jimensi, Jns, and refine pese in the course of scale development
by the use of factor analysis or some similar technique.
Where ability testing is concerned, in most applications at least,
the acid test of value is predictive validity: does the test
successfully predict performance on some criterion, and if"so,,
how efficiently(Cffinpbell and Fiske, 1959). ' It has been suggested
(e. g. by Vernon 1963) that the same is true of personality tests.
However, it is a fact that most of the controversy raging around
the question of personality measurement in fact concerns construct
validity. By this we may mean one of two things: either how
valuable the construct (ie. proposed dimension) can be shown to be
as an explanatory tool, or alternatively how well the scale developed
to measure a dimension actually does so. Where factored personality
questionnaires are concerned, notably those devised by Cattell?
Eysenck, Comrey and Guilford, the question of construct validity
in this first sense comes down by and large to the matter of
techniques adopted in the course of conducting the factor analysis.
Argument rages not so much over what sorts of items to include in
the finished inventory, indeed whatever the author, items tend to
be remarkably similar, as how to assemble them into meaningful scales.
Persistent lack of agreement on the question of which techniques to
237
employ has led some to the view that as yet factor analysis is
insufficiently understood, and that it is premature to hope that
it will provide any imediate answers to the problem of personality
measurement.
PRCJLEMM CF RF.? i, ICATING CATTELL' S FACTOR SOLUTION -', F,
In Cattell's (1964) paper "The Importance of Factor Trueness and
Orthogonality in Test Scales", direct validity is defined as the
multiple correlation of the items in a scale with the pure factor
which those items represent. If we accept this as a meaningful
definition of validity, it follows that scale validity is
maximised where: -
(1) each item has a high loading on the factor it purports to
measure
and
(2) the intercorrelation of the items in a scale is kept to a
minimum, ie. there is low homogeneity of item content.
This type of scale construction will also mean that suppressor items
can play a part in suppressing unwanted variance in an item which
loads on more than one factor. This is the concept of purifying
measurement by'means of what Cattell (1964) has termed 'buffered'
scales.
238
A number of problems arise in testing out these claimed construction
methods for Cattell's inventories. First, as Eysenck (1972) has
pointed out, other researchers have repeatedly failed to replicate
Cattell's factor structure. Thus, if our aim is to test how well
questionnaire items measure Cattell's model, then utilising loadings
from a factor analysis as a measure of 'direct' validity would not
seem viable.
Moreover, it is known that no two factor solutions are identical
each is unique either in terms of the number of factors or their
definition by factor loadings. The factor solution is in turn
based on an intercorrelation matrix of 16 variables. Such a matrix
is influenced by the sample, the mental state of the subjects and
many other variables. The concept of the 'pure' factor is rather
like that of 'true score' in the. classical test model - it is a
platonic ideal - not directly measurable but with reasonable
assumptions, it can be estimated.
Similarly, the 'pure' factor can only be estimated. Even if one
were to factor these data and 'find' 16 factors, they will not be
identical to Cattell's solution. But because many investigators
have failed to find such factors, we seem to have reached an impasse!
One way round this problem might be to undertake an arbitrary
factoring procedure (Saville and Blinkhorn, 1976) but even when this
method was tried it was not possible to replicate Cattell's solution,
The only possible solution seems to be the acceptance of Cattell's
scale scores as the delineation of the model; in other words to use
the 16PF scales as our factor estimates. If the scales are inadequate
239
as factor estimates, the only possible conclusion is that the utility
of the model is suspect. It is true to say that response styles,
attenuation and other effects may tend to alter the magnitude of
intercorrelation of scale over factor scores but the fact remains
that the scales are the vanguard of the model. If a theory exists
only within matrix algebra and allows no conversion into useful
eas tr i: E'ý; ý, than we have every right to q' estion The very value
of that theory.
It is also relevant to point out that Cattell's direct validities.
merely tell us how well the scale in the questionnaire represents
the factor in the factor analysis. As such, this tells us nothing
at all about the usefulness of the test in practical situations;
the test user does not have access to the pure factor, only to
scale scores. Any correlations between the pure factor and
performance on outside criteria will usually be mediated by the
scales as found in the questionnaire: hence for the test user, the
scales represent the factors, they must be expected to represent
them efficiently and consistently and ultimately he must judge
whether the assumption that they do so is borne out by such facts
as he has available.
Further justification for this approach, ie. using scale scores as
our factor estimates, IA " Cattell's claimed direct validities of .8 to
.9 between scale and factor scoret. This means that for many scales
the overlap with the 'pure' factor is so great as to suggest that
scale and factor are virtually identical. For these reasons it is
proposed, therefore, to ignore Cattell's strictures on the use of
240
estimates of the pure factor, since in the published inventories
there are provided means of estimating true-factor scores, namely
the factor scales themselves. If these are themselves poor representations
of the factor space, it is difficult to see what
justification exists for their development as means of
measuring personality.
THE PROBLEMS CF RESPONSE FORMAT AND CORRELATIONAL METHODS
Thus, the intention is to correlate item scores with Cattell's
personality scales as the analogue of his direct validities
within a factor analysis. We can view such correlations as item
validities. To quote Guertin and Bailey (1970): -
"The rationale for obtaining factor scores is analogous to
obtaining a total score for a person on a test that has been item
analysed to yield biserial 'r' validity indices between total score
and the items. An item-loading on a factor can be thought of as a
validity coefficient correlating the item and that reference factor
and analogous to the biserial 'r' between an item and total score
in item analysis. "
Empirical evidence for this view has been provided by Henrysson(1962)
where he showed the close similarity between factor loadings and
biserial correlations in item analysis. Unfortunately, this is not
simply a matter of calculating a large number of correlation
coefficients. Questions arise concerning the relationship of items
to scales, which must be considered before proceeding with any item
analysis.
241
The Cattell questionnaires under consideration have three response
categories for each item, generally corresponding to two extreme
choices with a middle 'unsure' or 'in between' category, with the
single exception of Factor B scales which, since they represent an
ability measure, have one right and two wrong alternatives provided.
The question arises as to whether the middle, 'uncertain' response
can be assumed to lie on the same underlying dimension as the two
extreme responses. Circumstances in which a subject might mark
the 'b' response include, besides genuinely wishing to appear as in
between the two alternatives, finding the item altogether irrelevant
or meaningless or possibly unclear and ambiguous. It may be that
in many instances the 'b' response is an escape route -a means of
avoiding a straight answer.
Thus, it might be argued that 'b' responses should be ignored as
representing 'noise' rather than 'signal'. There is certainly a case
to be made out for totally discarding items which have more
than say 80% of subjects responding in the middle category. Having
said all this, however, were we to take the course of ignoring the
middle response altogether, we should not be investigating the 16PF
as it is actually used. The item weights are given on a continuum by Cattell
and are an integral part of the construction of the-test. Furthermore,
by discarding the 'b' responses, we should be not only discarding a
good deal of useful information, but also reducing the comparability
of the coefficients in the final matrix, since different item-scale
correlations would be based on different samples. Whilst the individual
interpretation of many middle responses in the 16PF may be difficult,
Cattell has essentially decided the issue for us by including them
in his scoring scheme and eonsequently it is not proposed to discard
responses in the middle category. 242
It has been suggested that the Pearson product moment correlation
coefficient is not the appropriate statistic for item scale
correlations of the kind where three part data are involved. When
data are entered into a product moment correlation grouped into
as few as three classes, the resultant coefficient will be an
underestimate. Peters and Van Voorhis (1940) have suggested a
for Lh: e ur. &er°_. ^. timate caused by group ng; however
there are assumptions underlying the use of their correction
factor which may well not he met by item data of this kind.
Levonian (1961) suggested that the problem could be circumvented
by dichotomising the item data. Since the instructions on a test
booklet encourage the subject to respond in either the 'a' or 'c'
category, and to restrict his use of the 'b' category to one in
five items at the most, the distribution of item scores will tend
to be bi-modal, thus violating one of the assumptions underlying
the product moment coefficient. Levonian achieved dichotomised
scoring by placing the middle category in that extreme-category
with the fewer case. "This procedure tends to equalise the split,
to maximise the correlations to be computed, to allow analytical,
reproducible results and to introduce no systematic bias. "
(Levonian 1961) However, Levonian was computing inter-item
correlations, using the phi coefficient, a procedure which is not
appropriate at this stage. Furthermore, such dichotomisation does
in effect revise the scoring procedure to an extent which may
produce results whose importance is impossible to estimate. In
the case of item-scale correlations, either the point-biserial
or biserial ccefficients can be applied to dichotomised data: f
since the biserial
243
coefficient is simply an upward revision of the point-biserial,
applicable in circumstances where stronger assumptions can be made
about the underlying shapes of the distributions, it is unlikely
that the rank order of coefficients would be significantly altered
by the choice of one rather than the other. The actual decision
as to whether or not to dichotomise was made on the basis of an
e ,p rl.: al st ud'i with actual q.: estionnaire data and will be discussed
later.
When a scale score is correlated with the score. on an item which
forms part of it, whatever the nature of the scale, the correlation
coefficient which results will be spuriously high, since the item
variance is included in both the variables to be correlated. This
effect increases rapidly as the length of the scale diminishes.
In the construction of a mental ability test, this effect may
reasonably be disregarded, since the number of items in the scale
will generally be large and the effect will be more or less constant
for all items. The Cattell inventories, however, characteristically
consist of a number of short scales and we cannot allow this spurious
inflation of correlation coefficients to contaminate our data.
Furthermore, since the intention is to compare the correlation of
each item with both its own and every other scale, it is important
to avoid such confounding effects. Since in the factor loading
matrix such inflation will not arise, as all the items in the
questionnaire contribute to the factor, swamping the effect, we
must introduce some correction, so that our results are the best
possible estimates of the item true factor correlations.
I
244
The most obvious course is quite simply to correlate the scores
on a given item with scores on the appropriate scale minus that (the item remainder method).
item scoreA The objection to this procedure is that each item will
be correlated with a slightly different scale score and one,
furthermore, which is composed of one fewer items than the true
scale. To overcome these objections, Henrysson (1963) proposed
; Dme now formulae to provide for the estimation of the correlation
of an item with the same total composite of items, with the
spurious overlap effect removed.
For the biserial correlation the formula is:
rb(H) = rbSt - (pq/y)
1S+ pq - 2t rbSty
where: rb = non adjusted biserial between item and scale
rb(H)= corrected biserial coefficient
St = standard deviation of scores on test t
p= proportion passing item
q= proportion failing item
y= ordinate in the unit normal distributions corresponding to p
Wolf (1967) attempted to evaluate empirically the effect of these and
other formulae devised for this purpose. Under conditions of varying
test length it was in fact found that the results obtained by the
Henrysscn formulae agreed perfectly in all cases with the item
remainder method. It, therefore, seems that there is nothing to
choose between the two methods, and consequently, for reasons of
245
convenience, the item remainder method has been chosen here.
Cattell has written in scathing terms of the efforts of those
adopting an itemetric approach to questionnaire structure. By taking
each item in turn and examining its correlations with its own and
every other scale, we may well overlook the fact that the scale may
be 'Hewed not only item by item but also as a structured whole.
As the source traits measured by Cattell's scales are in any case
oblique, ie correlated,. we should expect to find significant
correlations between items and scales other than those on which
they are keyed. In some cases, these correlations may become
disturbingly large. It is, however, a feature of these-scales that
such complications are catered for at the time of construction.
We have already noted Cattell's claims that 'suppressor' action
has been built in to the scales, so that a substantial correlation
between an item and a scale on which it is not keyed is compensated
by the correcting effect of the relations between other items and
that scale. That is to say, where as the individual items in the
scale may show overlarge correlations with other scales, the scale
as a whole is a balanced composite largely free of spurious
correlations.
Interest in suppressor action began with the search for suppressor
tests to partial out unwanted correlations and their effects.
However, as Levine (1952) reports, these proved difficult to develop
and more recently interest has tended to centre around suppressor items.
Lord and Novick (1968) have commented on the fact that few suppressors
246
have nonetheless been found, and Conger and Jackson (1972) conclude
that new predictors will always prove easier to find than new
suppressors. Lubin (1957) showed that suppressor weights are even
more fickle than ordinary regression weights when cross validation
is attempted.
Ra 1e han plunge into te intricacies of whether a rar : icular item
has or does not have suppressor action, the simplest way to assess
whether scales are working in the way they are intended to is to
compute the multiple correlations of the items, constituting the
scale with that scale and with every other scale. If all is well,
the multiple correlation between a set of items and the scale they
constitute will be higher than their correlation with any other scale.
This is the empirical analogue of Cattell's 'direct validity',
viz. the correlation of a scale with the pure factor found in the
factor analysis. Since without replicating in every detail the
factor solution derived by Cattell, we cannot gain access to the pure
factor, for purposes of calculating this correlation, we must accept
the questionnaire scale as the best estimate of this factor.
At this point the importance of using item remainder rather than
ordinary item-scale correlation coefficients emerges. Were the item
scale coefficients not corrected for inflation due to overlap, the
multiple correlation of the items in a scale with that scale.. would.
necessarily be 1.00, since all the variance in the scale score is
to be found in the item scores. However, by using item remainder
coefficients, we obtain the best unbiased estimate available of the
item true factor correlations, which then serve as a basis for
247
calculating the multiple correlation between the items and the scale.
Of course, when it comes to calculating multiple correlations between
sets of items and scales in which they are not included, the item
scale overlap problem does not arise.
It is important to realise that comparatively little weight will be
attached in the ensuing discussion to the absolute size of the
item scale multiple correlations. Our arguments will rest largely
on the relative size of the coefficients. Thus, the error inherent
in the use of the factor scale as an estimate of the pure factor
does not affect our argument, since this error will affect all the
multiple correlations. Quite simply, we shall look for the highest
multiple correlations between sets of items and scale scores. As
in any case the multiple correlation coefficient tends to overestimate
the true correlation, any depression of these values as a result of
the procedures we have adopted is not likely to produce serious
distortion of the relative size of coefficients, or indeed their
absolute size. We shall not, therefore, concern ourselves with
corrections for 'shrinkage' and the like.
Our requirement will be quite simply that for a scale to be acceptable,
it should be seen to have a higher multiple correlation with its own
items than with the items constituting any other scale. This
requirement is lenient in the extreme. Because we will be calculating
the scale score analogue of Cattell's factor direct validities, we
are merely asking that items in a scale measure their own trait
better than they measure another factor.
i
248
THE HSPQ STUDY
We have seen that items of the type where subject responses are
expressed in terms of a choice from three possible preferences
have been analysed by one of three methods:
i1C_h: '.: JiiLl. o. _. CJ ifn}
data .ii
such a 1731 ,%
that i the eiddl e responses \viý. um r1..
(designated the 'b' response in Cattell's inventories) are
included under the smallest of the two extreme responses,
using the biserial'correlation coefficient to represent
the association between item and scale.
(b) dichotomising the item data as in (a) but using the point-
biserial coefficient in order to correlate item and scale.
(c) analysing the item data directly as coded by the subject
without dichotomising data, with use of the conventional
product moment coefficient.
In order to investigate whether conclusions on item adequacy may
in any way be affected by these different approaches, item data
were collected from British standardisation of another of Cattell's
inventories for 14-17 year old school pupils, the High School
Personality Questionnaire (HSPQ). The background to this study
has been described by Saville and Finlayson (1973) and involved
the administration of the HSPQ to some 2000 secondary school pupils.
These data provide a useful experimental base to investigate these
different methodologies, especially as 12 of the 14 primary source
249
traits in the IHSPQ are common to adults and the questionnaire
espouses identical construction principles to those of the 16PF
The second major aim of this study was to investigate the effects
of overlap between personality scales and the various items loading
on that scale. Thus, in addition to biserial, point-biserial
and pr, --duct rnomer. t coefficients being calculated between items
and the total scale scores, the corresponding coefficients
corrected for spurious overlap between scale and item were also
computed. The corrected coefficients were based on an iterative
procedure which involved calculating the correlation between item
and scale with that item removed from the relevant. scale score.
THE SAMPLE
A total of 366 female pupils was sampled by the fixed interval
technique from the British standardisation group (Saville and
Finlayson, 1973) by taking every third record.
PROCEDURE
An outline of the item analysis procedure is contained in Figure 12.1.
RESULTS
Appendix l2contains the results based on dichotomising the HSPQ
item data and includes the corrected for overlap point-biserial
correlation of each of the 140 items with its own scale. In addition, I
250
W
WU O a a
r' ä N
W
(cý H
U, w a ý A H öw HG',
H T' o0
H3 äc1) Q+ý QWU
0
U Öq
w: ý 0u c6 .
EA OH
01. un ü3
HHý ý r :Z
G] xW
kt H
Ö0
Ua cm W a U)
H
w vi HW H ý/1
xö w
aä
cr AH cz x wý
ma > to
E-4 11
z
= E-4 2
HH
ä ýý
zö HQ
HHA +ý
10 O°b Umaa
OH
H. EA OH ý7
UW
ä
U)
ýzq
W Ef) Ö
H ul Or
H HC,
Ll F-! U1
Q fk
. ý? äi rCf ýj EH
C) L4 -4 W HO c71 Wpý
W Ht0 VÄHÖ
(a
"d
ÖE 12 HR
WÖHx
A 14
ý E-1 Z C) ýAI
O2 xO
VAH3CGfý..
dN
pö", ýö
u u2
w Hwo zaH
cý x HÖ WOa
R3. HWö
ýWýC7zýHCi
Q
E--f ul n4 4.1 0
xx HHHäý äaä
w Ö bx
0% Hx H
E-f 3 2:
W Ö
«1
UcH3 fý W
ý-
251
the corrected biserial coefficients and the biserial coefficients
between each item and the remaining thirteen HSPQ primary scales
are included in the table.
The corresponding data based on product moment coefficients without
dichotomising item data are contained in Appendix 12. For both
sets of tables, coefficients under a solid line are negative in value.
A further analysis conducted on these data was to correlate the
corrected dichotomised. biserial with the point-biserial coefficients.
The results of this analysis are contained in Table 12.1. below.
As is to be expected from the mathematical relationship between the
two coefficients, their intercorrelation is very high.
TABLE: 12.1.
Cc*RELATION BETWEEN THE POINT_BISERIAL AND BISERIAL APPROACHES TO
HSPQ ITEM SCALE CORRELATIONS
POINT-BISERIAL BISERIAL ITEM-SCALE ITEM-SCALE CORRELATION CORRELATION r
Mean . 167 . 215
. 99 S. D. . 092 . 117
Whilst the point-biserial correlations between an item and its owrr
scale are lower than the corresponding biserial coefficients, the
point-biserial of an item with an external scale are similarly lower.
It is quite clear therefore that as long as we are not concerned with I
252
the absolute values of the correlation of an item with its own
personality scale, the point-biserial and biserial statistics give
us virtually identical results.
If we adopt the biserial coefficients in that the item data are not
in any real sense a true dichotomy, it is possible to summarise
nu m bcýr of high`r extra t can i. n ra scale correlations for e ach
factor. For example, we can see from Table 12.2. that the first
item on Factor A correlated more highly with its own scale than
any other scale and the second item showed one correlation which
was higher with an outside scale than its own corrected correlation.
However, item 7 on Factor A showed twelve higher extra scale
correlations; in fact, it was only with Factor B that this item
had a lower coefficient than its own scale correlation. The
correlation with all of the remaining twelve scales in the HSPQ were
higher than the correlation upon which it is loaded.
Turning to the second set of tables in the Appendix, containing the
item analysis results using the product moment coefficient on the
item data in its original three point response format, we can see the
very marked effect of correcting the coefficients for spurious overlap.
This is no doubt due to the shortness of the HSPQ scales and their
heterogeneity.
Table 12.3. provides summary statistics for the number. of higher extra
than intra scale correlations for each factor, based on the product
moment approach. We can see that the overall relationship with the
corresponding biserial table is high; Factors A (outgoing), I
253
N
a
m
a
L.
U W cl: O
V)
C7
U-
ýJ dW
ýO O C. ) 4.
LLI
Ent J Q
cy- V
CD t- i- H In. Lu
0' =
1-1 Lý J F- c"C X
LU O Nw a w C-
NJ "Q
N
W JS 00' C7' Qw k- 2
w
ýI S
U.. O
ix 1. l. ß m
. rte..
-41 q7' co r ct CV d- ; f* CO 4 C L() r..
C") O G*' ct' d' r C) N N. O lO (V
CV O'
N r- r N r r M In Lfl CV. v
p ) (
O L[) V-- M O Cl Q1 CV cm -M 'rr d' r--
ID ch o O M aT c LO O N. tD r-. , --
ý-+ Ul d' O r U7 O N O O s. ý r `y- N
L) O r r CJ C3. fit': LC) CV CV N nj N
C3 r O r Cl O O N ¢}' p t, O I-
Li.. r- N N N O O C) -y- to r-. r
ivy
W N 00 . 4- Q1 N Ql CO r r-- Cif M p r '. p
O. LO O cr) cr O r O N . p tO qtr
U M O r e- cd- N. r- tD c, N. CD cat
m w O O r Q O O O C-D tV to
Q O r- r '4) M O N M , ._ ^ w N
W hN
°ý1-4 SW 31I ' N d Z: OCH -j L+-1 '+-+ Li U. 1 IAJ
ca : _C .JC: LU 4a . .., -cri Q
m : r: zv)
C) 4-
0
fu E rt D N 4
f
254
w 0
I-
C
CC 0 I--
U
w
LL
(/) to JZ wo
C3 CC
wJ
rr_ CD C) L)
w
Itl Q WU º-4 N CZC zM
än LLi
I-- H ¢ ýc
w 0
ws a c3
._ <n o C) Ll- NO r-I CJ
"N W LJ JZ co m CD Z:
ý-- sz
C+- kp N cV ' C; r N M N O CV) z: - co r M
M. O r l+) ý- d- CQ CO cV lD O (3 tV C
CV d' CV) (-j Cl) r- r- C. 7 r LO cr. ) C)
O U) e- C') O CO Co. N Cri C9.. .. M i . r-
,
7 d' tC O ýF c M to N% Ln cr) Q
º-ý CV d- O O O N C) C) O Cý) to r N
= %0 O .Y r C7 O Cf' Rj C`) C4 N CV)
(3 O O O N O C) d' d' O r"s C+)
L. r- r r CV O . O C) O c M
W CV Q1 O CO CV O C) e- r OD O O r r"- r-- N.
Cl U) ý-- r U) O O C) N p U") r-
U N O r- ,M r,. U) N In O ' .. 01 CV
CM
m (V O O r O O O O O N L[)
M Cr) CD lfl CV) (V) r- N r- r- M
O N
a+ W ED C
O I L, C) ct V W N
SW3II N =ý-- (/) ý ct: LLJ (n ýw H-)-O
1- C) C)
I
255
C (emotional stability), E (assertive), H (venturesome),
I (tender-minded), J (circumspect individualism), 0 (apprehensive),
Q2 (self-sufficient), Q3 (controlled) and Q4 (tense), all showing a
relatively large number of items with high extra over intra scale
correlations.
Whet: the biserial and product moment methods are compared over both
items and factors, it can be seen from Tables 12.4. and 12.5. that
the overall association is high and we may conclude with a fair
degree of confidence that the two methods of item analysis give
essentially comparable results.
I
256
TABLE= 12.4. THE HIGH SCHOOL PERSONALITY QUESTIONNAIRE: FORM A, GIRLS
CORRELATION BETWEEN PRODUCT MOMENT AND BISERIAL APPROACHES (ALL ITEMS)
PRODUCT MOMENT BISERIAL NUMBER OF HIGHER NUMBER: OF HIGHER
EXTRA THAN INTRA SCALE N EXTRA THAN INTRA SCALE CrýRitýt ,' TLO: Y CORRE1 ATICNS CORRELAýE IONS. .
MEAN S. D. MEAN S. D.
3.03 3.21 140 3.02 3.16 . 871
PRODUCT MOMENT BISERIAL NUMBER OF HIGHER NUMBER OF HIGHER
EXTRA THAN INTRA SCALE N EXTRA THAN INTRA SCALE CßRRELATI01`ý CORRELATIONS -CORRELATIONS - MEAN S. D. MEAN S. D.
30.29 17.65 14 30.21 16.20 . 970-
r
'_I
257
13. ESTIMATING THE DIRECT VALIDITIES
OF THE PRIMARY SOURCE TRAITS
258
INTRODUCTION
From the studies on the High School Personality Questionnaire,
we have established that the biserial and point-biserial
approaches after dichotomising the item data, give very
co"' Arab' e ri sui Ls c the use of the product moment coefficient
on the trichotomised data.
Because of this finding and in that the trichotomised data after
all represent Cattell's original response format, it was decided
to use the product moment technique on the adult. item data,
correcting the statistic for spurious overlap between item and
scale.
A further point has to be borne in mind however. Cattell (1972)
advocates the use of multiple correlational techniques in
calculating the composite correlations of his items with the
factor estimate. It is the case, however, that such techniques
are likely to give the upper bound to scale validities mainly
as a result of their tendency to capitalise on chance effects in
the data. What is needed is not the direct validities calculated
on optimal weights but on the weights as they actually exist in
the item data. Thus, the values '0', '1' and '2' were assigned
to the item responses as laid out by Cattell in his 16PF scoring
key and thlse used to calculate the multiple correlation of sums
between items and scales.
I
259
That is: -
res rci I
dcl �- ii II
Where: -
rci = correlation corrected for overlap between the item xi and the scale (i. varies from 1 to n)
0= standard deviation of the item
rij = correlation between one item xi and any
other component xj, when j is a higher subscript number than i.
In summary, we are to calculate the multiple correlation of sums1
between items corrected for spurious overlap and the relevant
personality scale. These values we shall term composite item-scale
correlations. The scales are taken as factor estimates in order
to establish how well the items measure the underlying model.
This approach, it has been argued, is the analogue in item analysis
terms of Cattell's direct validities and is to all practical purposes,
the only method by which the internal structure of Cattell's model
can be investigated. With this in mind and in the light of Cattell's
claims, a number of hypotheses can be tested.
HYPOTHESES
(1) The composite item-scale correlations, representing direct
validities, will tend to be of the order of .7 to . 9.
(Cattell, 1972) }
260
(2) Because of the method of scale construction used in the 16PF
(Cattell, 1972), which deliberately sacrifices item homogeneity
in favour of suppressor action, these composite item-scale
correlations will tend to be higher than those derived from
conventional homogeneity formulae such as coefficient alpha.
(3) After the principle of 'buffered' scales advocated by Cattell,
the composite item-scale correlations (direct validities) for
any given scale will be higher than the intercorrelation of
that scale with another. (If this is not so, then it suggests
that the items in one scale measure another factor rather
better than they measure their own scale. In such a case,
there can be no real justification that the scale in question
exists as a measurable separate identity. )
DIRECT VALIDITIES OF THE SOURCE TRAITS: PROCEDURE
The responses to each item on Forms A and B of the 16PF for the
adult group (N = 2008) were punched on to cards and scored by
computer according to the keys provided by Cattell to produce
an item score matrix. Scale scores were then created on each of
the primary source traits for each of the two forms of the 16PF.
Item responses were then correlated by the product moment statistic
with the relevant 16PF Form A+B scale, having corrected that scale
I
261
for spurious overlap. ( These data are held separately from the main
thesis. ) The multiple correlation of sums of these values was then
calculated. This procedure was followed for the male and female
groups separately (Table 13.1. ).
DIRECT VALIDITIES-01F. THE SOURCE TRAITS: RESULTS AND DISCUSSION
The first observation we can make regarding the table of composite
item-scale correlations. concerm our hypothesis on their absolute
size. In fact, the data are comparable to the multiple correlations
of scale its with the pure factor as presented. by Cattell (1972),
(see Table 13.1). It should be remembered, however, that Cattell's
values are based on Form A only whereas our item composite
correlations are for two form length (Forms A and B). The only
major difference between the two separately derived estimates
is for Factor I, where Cattell's value is somewhat higher than
the corresponding data reported here.
We can see from the table that the primary source traits of
highest reliability are F (happy-go-lucky), H (venturesome),
0 (apprehensive) and Q4 (tense). Low internal consistency estimates
were obtained on the other hand for Factors L (suspicious),
M (imagination), N (shrewdness) and Ql (experimenting). These
findings are very much in line with those reported in the previous
section where the parallel form method was employed. That is to
say, that the primaries of low reliability as measured by the
alternate form approach are precisely those factors which show }
262
TABLE: 13.1.
COMPOSITE ITEM-SCALE CORRELATIONS (DIRECT VALIDITIES)
FOR CATTELL' S PRLN! ARY SOURCE TRAITS
COi iPOSITE ITE:, i_SCALE COEFFICIENT MULTIPLE R OF " f CORRELATIONS ALPE=. SCALE ITEP; S
FACTOR i
MALES FEMALES MALES FEMALES FAWITH E
CTORU(CATTELL)
A . 601 . 532 . 626 .. 554 . 62
B . 581 . 561 . 599 . 579 . 60
C . 673 . 653 . 695 . 675 . 54
E . 665 . 646 . 687 .. 667 . 53
F . 786 . 763 . 814 . 790 . 76
G . 628 . 575 . 654 . 599 . 69
H . 840 . 848 . 871 . 880 . 79
I . 591 . 454 . 616 . 471 . 73
L . 471 . 506 . 488 . 525 . 44
Ni . 374 . 446 . 385 . 459 . 51
N . 369 . 364 . 382 . 377 . 40
0 . 723 . 706 . 748 . 730 . 65
Qi . 448 . 424 . 465 . 439 . 47
Q2 . 570 . 549 . 593 . 570 . 57
Q3 . 660 . 622 . 688 . 647 . 64
Q4 . 791 . 750 . 819 . 776 . 73
corrected for two form length by the Spearman Brown formula
t
263
low internal consistency, either in terms of homogeneity or by
the 'direct validity' approach.
Referring back to Table 13.1, we in fact notice that the item-composite
correlations are in every case, for both the male and female groups,
lower than the alpha homogeneity estimates. This must lead us to
reject our second ht, -pot; esis that suppressor action in the 16PF
scales would lead to higher composite item-scale correlations
than would be obtained by conventional internal consistency methods.
We are forced to the conclusion, therefore, that no suppressor
action of any marked extent is present in the 16PF scales and
that the internal consistency reliabilities of the primary source
traits are not underestimated by conventional techniques, as has
been claimed by the proponent of the scales.
Despite Cattell's theoretical standpoint on the production of
scales with balanced specifics, the evidence reported here suggests
that suppressor items are fickle in behaviour and that an author
would generally do better to consider the addition of more predictor
items in a scale. It may be the case in Cattell's theory of
personality that the liking for suppressor action is merely a
reflection of the fact that it has been found extremely difficult
to construct simple predictor questionnaire items for some of the
primary source traits. This problem might well have arisen, of
course, as a result of Cattell transferring a model of- personality
from his studies on ratings which included inadequate or arbitrary
traits.
264
The third hypothesis in this section concerned the question of
buffered scales; the principle of items suppressing unwanted
variance due to the influence of other personality traits. The
vital test of this claim is the size of the composite item-scale
correlations in relation to the factor scale intercorrelations.
If a set of items has a composite item-scale correlation which is
ýJ'výý than that scal'ýýS re.. ationship with amothei pYi_rn ry, it
in effect means that: -
(a) the items measure another scale better than they measure
their own scale.
(b) the construct validity of the model is in doubt because
the scale in question is not de-nonstratably a separate
identity from other scales.
Table 13.2.. represents the intercorrelation matrix of the 16PF factors
as scale scores, based on the British adult group, the matrix for
males being above and for females below the main diagonal.
Information from this and a previous table have been included
in Table 13.3-which shows the item composite correlation for males
and females with the highest intercorrelations of the relevant
scale with the other primaries.
It is evident from this table that the items of Factor C (emotional
stability) measure Factors 0 (apprehension) and Q4 (tension) as
well, if not better than they measure their own scale. Indeed
these three factors, although of tolerable reliability, all show }
265
TABLE: 13.2.
CORRELATIONS BETWEEN THE SOURCE TRAITS OF THE 16PF AS SCALE SCORES
FORMS A AND B
A B C E F G H 1 L M N 0 Q1 Q2 Q3 Q4
A -07 08 14 31 02 42 -22 01 -02 -03 -12 04 -43 01 -04
B -04 17 15 07 -05 09 -05 -03 28 -14 -21 22 02 03 -06
C 11 20 06 08 11 41 -14 -37 12 11 -70 02 -12 42 -66
E 10 10 01 48 -15 45 -18 38 32 -44 -16 43 -10 -15 09
F 27 17 12 50 -33 50 -15 30 -07 -42 -09 32 -42 -30 12
G 04 05 09 -18 -30 08 -04 -14 -15 29 -15 -31 -03 60 -24
H 34 17 40 49 58 -01 -03 00 18 -12 -51 16 -44 22 -31
1 08 15 01 -17 -07 10 01 -08 25 05 18 -08 04 -07 12 U'
z L -01 -07 -41 38 19 -09 -01 -16 03 -30 31 19 04 -29 44 äo
w M -10 23 18 34 19 -10 33 18 00 -22 -12 33 08 -11 -04
N 09 -08 09 -41 -42 28 -19 04 -24 -28 -13 -38 05 38 -28
0 -15 -14 -67 -13 -14 -09 -48 07 28 -18 -08 -07 12 -50 71
Qi 07 11 -01 38 32 -23 25 -05 22 27 -31 -08 00 -25 10
Q2 -31 -12 -13 -06 -43 -04 -39 -04 07 00 07 07 00 00 06
Q3 05 09 36 -20 -31 56 09 02 -26 -12 37 -39 -20 02 -56
Q4 -12 -03 -64 09 11 -15 -27 05 40 -07 -24 69 05 03 -48
FEMALES (N=717)
266
M
oý N
Vjl
HI
HQ'.
9j
z H
tn Ö
H I
U
H
U)
l
cr Q ýD 11
C- LD
Cl) Cf 0 U') CV)
N
a
o m
N mo t O t o
11
z
z
a
H
M
U O 'o 110 N
co (f
C\i M 1i' d' M
11
cr1 t0 to to l0 co c1 Ln r-4 11 N N t! 1
" U)
" ýO Q
" L-
" ul
" CO
" Iq d'
" in ýn
" " m
" " d'
" in
" .o
" CD
" Hj V) H 0 E-4 P4
O p N
ºý O N N p)
L- c7 l0
dý 00
rn to
N N M
N M
cV N
to r' in
lo tD N
H
Ö v
4 w u W w 0 x H a z 0 ä
267
such high intercorrelations that their existence as separate
identities seems doubtful in the extreme. This possibility has
been suggested previously by Eysenck (1972) and Saville & Blinkhorn
(1976), where as a result of their researches into student
samples, the C, 0 and Q4 primary source traits seemed to form
one underlying vector.
Two other factors, unassociated it would appear with the anxiety
primaries considered above, but which show intercorrelations about
as high as their own direct validities are G (superego strength)
and Q3 (social-precision).
The picture is a little more confused with the remaining primaries
of L, M, N and Ql. Factor L (suspicious) shows intercorrelations
almost as high as its direct validity estimates with the two anxiety
factors of C (emotional stability) and Q4 (tense) and with
E (dominance). Similarly, Factors M (imaginative) and Ql (experimenting)
show a strong enough association with the Factor E primary source
trait, as to make one wonder as to their functional independence.
Finally, the data demonstrate that Factor N (shrewdness) is a
glaringly weak factor, with the items of F (happy-go-lucky),
G (superego strength) and Q3 (social precision) having higher
correlations with the scale than the items in the scale itself.
Thus, there is the nonsense that the best estimate of this
particular source trait is provided not by itself but by three
other scales.
} 268
DIRECT VALIDITIES OF THE SOURCE TRAITS: SUMMARY AND CONCLUSIONS
We have seen that the direct validities calculated here on two-form
length scale scores are about as high as Cattell's claims for Form A
only. We have utilised the statistical methods advocated by Cattell
with the exception of dealing with scale scores rather than the
more esoteric factor estimates. This approach was adopted because
of the great difficulties of replicating Cattell's factor structure
and in the belief that scale scores are in any event more scientifically
defensible criteria against which to validate items.
The values obtained for scale score direct validities were comparable
to the alternate form reliability coefficients reported in an earlier
section; particularly weak factors being L, M, N and 0-1. We found
no evidence to suggest that the alternate form reliabilities severely
underestimate the reliability of the primary source traits (Cattell, 1972,
C1 Krug personnt . communication, 1977).
When the role of suppressor action in the 16PF scale scores was
investigated, the direct validity estimates showed no tendency to
be higher than alpha statistics calculated on the item data. We are
forced to conclude, therefore, that the suppressor action as a means
of enhancing scale validities is not operating within the primary
source traits, as represented by Cattell's latest questionnaire forms.
Moreover, it was clear from these analyses that certain primary
factors were untenable as separate identities. Namely, the traits
of C (motional stability), 0 (guilt prone) and Q4 (tense) seemed i
269
to form one anxiety constellation and G (superego strength) and
Q3 (social precision), a measure of the extent to which an
individual adheres to conventional social standards.
The position with a number of further 16PF primaries was more complex.
Factor M (imagination) appeared to have'a good deal in common with
Factor E (dominance), as did Q1 (experimenting) and N (shrewdness).
The N trait also had a good deal in common with Factor F, so as to
exhaust most of its meaningful variance. Finally, Factor L (suspicious)
was highly associated with both the anxiety dimensions and E (dominance).
270
14. FACTOR ANALYSIS AND THE PRIMARY SOURCE TRAITS
271
INTRODUCTION
There would seem considerable evidence here to suggest that Cattell's
model of-personality as represented by the 16PF first order factors
is open to serious criticism, especially as we could find little
support for Cattell's contention that the primaries all represent separate
identities.
Because, moreover, no suppressor action of any great influence was
discovered within the item structure of the 16PF and because the
scales were, therefore, very much to be judged by their homogeneity,
there can be no real objection to a conventional factor analysis of
the primary source traits as scale scores.
It is beyond the scope of this thesis, however, to construct an
alternative personality theory to that of Cattell or the other major
factor theorists. The literature is strewn with many one-off analyses
which add little to a more general understanding of the structure of
personality. There is also the point that it is inappropriate
here to become heavily involved in the controversial issues of the
type of factor analysis best employed.
Nevertheless, it is feasible to test whether as few as three factors
are sufficient to account for the 16PF traits; for example, are the
three dimensions of the Eysenck Personality Questionnaire (Eysenck
and Eysenck, 1976) sufficient to explain Cattell's primaries?
For this purpose promax factor analyses of the 32 scale scores of
Forms A and B of the 16PF was undertaken on the male (N = 1004) and
272
female (N = 1003) samples collected for the purposes of these
researches.
Promax (Hendrickson and White, 1964) involves first rotating to an
orthogonal varimax solution and then relaxing the orthogonality
of the factors to better fit simple structure. From the orthogonally
rotated matrix, an ideal factor matrix is constructed in which high
loadings of varimax are made higher and low loadings smaller. This
is achieved by normalising the orthogonal matrix by rows and columns
and taking, in this case, the fourth power of each loading. Finally,
the least squares fit to the ideal matrix is found using Hurley
and Cattell's Procrustes technique.
Because of the nature of the hypothesis. that as few as three factors
will account for Cattell's primary source traits, a relatively
conservative criterion of the number of factors to extract was
adopted here. The criterion was that of limiting the factors to
those with eigenvalues greater than one (Kaiser, 1960).
273
TABLE : 14.1.
16PF FACTOR ANALYSIS: MALES:
EIGEN VALUES AND PERCENTAGE OF TRACE
FACTOR EIGEN VALUE
PERCENTAGE OF TRACE
CUMULATIVE PERCENTAGE
1 5.6249 17.58 17.58
2 4.9684 15.52 33.10
3 2.5615 8.01 41.11
4 1.9918 6.22 47.33
5 1.6511 5.16 52.40
6 1.2370 3.87 56.36
TABLE: 14.2.
16PF FACTOR ANALYSIS: FEMALES:
EIGEN VALUES AND PERCENTAGE OF TRACE
FACTOR EIGEN VALUE
PERCENTAGE OF TPACE
CUMULATIVE PERCENTAGE
1 5.2104 16.28 16.28
2 4.9571 15.49 31.77
3 2.1967 6.87 38.64
4 1.9962 6.24 44.88
5 1.6954 5.30 50.18
6 1.2906 4.03 54.21
7 1.1149 3.48 57.69
274
TABLE: 14.3.
PROMAX FACTOR ANALYSIS: MALES: TABLE OF FACTOR LOADINGS
OBLIQUE PRIMARY FACTOR LOADINGS FIRST ORDER 1 2 3 4 5 TRAIT
A 0.6873 -0.2582 B C 0.8345 E -0.8141 F -0.3226 0.5100 -0.3027 G 0.9366 H 0.2832 -0.4938 0.5041 I -0.7739 öL
-0.4453 -0.5825 M -0.3899 N 0.4395 O -0.7494 Q1 -0.4639 -0.2966 Q2 -0.7673 Q3 0.2976 0.6337 Q4 -0.8688
A 0.6303 -0.3266 B C 0.6988 E -0.8619 0.2033 F -0.3747 0.4240 -0.2852 G 0.7970
czý H 0.4111 -0.3892 0.4500 I 0.2322 -0.7710 09 L -0.4005 -0.5420 0.2089
44 M -0.4364 -0.4572 N 0.2185 0.3124 0.2676 0 -0.7732 Q1 -0.2700 -0.2364 Q2 -0.2132 -0.7699 Q3 0.3739 0.6190 Q4 -0.8170
6
0.8743
0.3121
0.8668
0.2821
f
275
TABLE: 14.4.
PROMAX FACTOR ANALYSIS: FEMALES: TABLE OF FACTOR LOADINGS
FIRST ORDER TRAIT
A B C E F G H I L M N O Q1 Q2 Q3 Q4
A B C E F G H I
OL P4 M
N 0 Q1 Q2 Q3 Q4
OBLIQUE PRIMARY FACTOR LOADINGS
234567
0.7304 -0.7959
-0.7854 0.731j7 0.3094 0.5232 -0.2310
0.9093
-0.2363 0.6126 0.3822 -0.7455
0.4600 0.5066 0.2194 0.1825
-0.2441 0.3360 -0.4630 -0.4677 -0.2577
0.7478 -0.2177 0.2132 -0.2118 0.2531 -0.4180
-0.7436 -0.2496 0.6291
0.8472
-0.6391 0.8278 0.3575 0.4670 -0.2470
0.7847 -0.4227 0.5005 0.3186
0.2273 -0.7462 0.3400 0.4412 -0.2147 0.2018
0.6303 -0.3406 -0.2024 -0.2925 0.2944
0.8231 -0.2052 0.5525 -0.3029 -0.3145
-0.7873 -0.3722 0.5769
0.7935
0.2804
0.3277
0.7948 -0.8176 -0.2734
-0.2187 0.2287
0.25 34
276
TABLE: 14.5.
CORRELATIONS OF THE OBLIQUE PRIMARY FACTORS
MALES
1234 5 6
1 01 16 -08 32 23
-24 03 41 -32
3 06 -07 -10
4 -03 -18
5 -18
TABLE: 14.6.
CORRELATION'S OF THE OBLIQUE PRIMARY FACTORS
FEMALES
1234 5 6 7
1 01 -21 21 -25 18 -08
2 24 03 -42 -22 05
3 -10 -15 -09 15
4 -08 15 26
5 00 10
6 -04
I
277
FACTOR ANALYSIS: RESULTS AND DISCUSSION
The promax factor analysis of the 16PF scale scores produced a
six factor solution for the male groups. Factor loadings are
coma r. ßä in 1abic 14.3. -and the correlations between the oblique
primary factors and the percentage of the variance which they accounted
for are contained in Table 14.5.. The corresponding data for the
female group are contained in Tables 14.4. and 14.6..
If we first turn our attention to the results, based on the male
sample, we can see that the first major factor to be extracted
from the matrix and accounting for some 17% of the variance,
appears to be a general anxiety vector with high loadings on
C (emotional stability), 0 (guilt prone) and Q4 (tense). The
second and third factors appear as two extraversion-introversion
factors; the second factor having high negative loadings on
E (assertiveness) and to a lesser extent, L (suspicious) and
H (socially boldness). The high factor loadings for the third
factor are with Q2 negatively (group dependence) and A positively
(outgoingness). Also loaded on this factor are F (happy-go-lucky)
and H (social boldness).
}
278
Factor four is largely Cattell's I (tough minded) with some
relationship with M (practical). As we might have expected from
our previous section, the fifth factor is largely G (superego strength)
and Q3 (social precision). Finally, there is an ability factor
highly loaded by B (intelligence) with smaller loadings on
M (imagination).
Turning to the factor analysis undertaken on the female group,
we find that seven factors had eigenvalues greater than unity
and that there is a good deal of agreement with the male study.
Factors one and two seem largely to be an anxiety and assertiveness
factor respectively. Factors four (tough mindedness), five (superego
strength) and six (intelligence) are also closely in line with
results found for the male group.
The important departure between the two studies is on the third
factor in the male group. In males this factor has high loadings
on A, F, H and Q2, whilst in the female sample A (outgoing) seems
to form a separate factor.
The results of these factor analyses are interesting in that they
do not support the hypothesis that only three factors are required
to explain Cattell's primary source traits. It would appear that
a minimum of six to seven and perhaps a maximum of nine are needed.
If we are'to accept the trait labels provided by Cattell for these
studies, the following structure evolves from the promax factor
analyses. I
279
PROVISIONAL FACTOR NAME
I Anxiety
II Dominance/Assertiveness
III Group Adherence
IV Tender-mindedness
V Superego Strength
VI Intelligence
and possibly
VII Outgoingness/warmheartedness
HIGH PRIMARY SOURCE TRAIT LOADINGS
C- O+ Q4+
E+ L+ H-F
Q2- F+
I+ M+
G+ Q3+ Q1-
B+
N-
A+
Some confusion exists as to the dimensionality of the extraversion
components of the 16PF but it would certainly seem that Cattell has
managed to break this factor down into meaningful sub-factors, even
if not as many as he would claim. Clearly what is needed in future
research is to further investigate this domain using more sensitive
factor analytic techniques, perhaps working on item statistics rather
than scale scores.
CONCLUSIONS ON THE FACTOR ANALYSES
The results of the factor analyses conducted on the male and female
samples produced highly comparable results to the calculation of direct
validities on the 16PF scale scores. By using a promax technique, six
factors provisionally named anxiety, assertiveness, group dependence,
tender-mindedness, superego strength and intelligence with the
possibility of a seventh factor of outgoingness were extracted.
From all the evidence presented in this section, we are forced to
conclude that Cattell's primary source traits are not the most
parsimonious description of personality but that perhaps as many
as seven to nine factors are required to explain the domain as
represented by Cattell's questionnaire scales.
280
}
SECTION 6:
SUMMARYANDC0NCLUS10NS
15. SUMMARY AND CONCLUSIONS
281
The general hypothesis developed in this research was that Cattell's
theory of personality rests on a system of primary source traits
which is untenable in terms of the verifiability of important group
differences, the reliability of measurement which it affords and
its internal consistency. It was also hypothesised that despite
the existence of some independent, reliable and descriptively useful
factors, Cattell's model, as represented by his primary questionaire
scales, is not as parsimonious as he has previously claimed.
It was also pointed out in the introduction to this thesis that
there is unlikely to be a once and for all test of Cattell's theories.
As Howarth (1976) has stated "what is now needed is not a system
defence but a number of investigations to form a 'concensus buttressed
by co-ordinated studies".
What evidence have these studies provided then on the existence and
value of Cattell's primaries as identifiable constructs? We shall
seek to answer this question by considering each of the primary
source traits in turn, observing how well the model actually fits
the data base of some 2000 British adults collected here. We shall
also take the opportunity at this stage to paraphrase the 16PF items
which best define Cattell's primaries. For the
purpose of this section, items which have reached an arbitrary
criterion of a correlation of . 25 or more with the scale it was
designed to measure are included. Items, which are paraphrased here
merely for reasons of copyright, are expressed in this format where
an affirmative response corresponds to the positive pole of the
primary source traits.
282
FACTOR A (Reserved vs. Outgoing)
1, GROUP DIFFERENCES
I. I. Cross National: - British adults-tended to be significantly
lower on this factor than American adults.
1.2. So,, D rferonces: - As was hyaothesised, the female sample
obtained a significantly higher mean score on this factor.
1.3. Social Class: - As hypothesised from Cattell's theory, no
significant differences were found between the social
classes on this factor.
1.4. Age: - No evidence was found for Cattell's claim of an increase up
to 30 years with subsequent levelling off. A relatively flat profile
of differences was in fact found across. the age cohorts.
There was no evidence of any curvilinear relationship of
this primary with age.
2. RELIABILITY
2.1. Group Values: - The alternate form reliability of this factor
was approximately . 5, with significantly higher reliability
in male over female subjects.
2.2. Social Class & Age as Moderators: - Reliability was found. to be higher
in the higher social classes. Age did not act as 4L significant
moderator here.
I
28 3
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - Values of . 60 ,
(males)
and . 53 (females) were found as estimated direct validities
for this scale. Coefficient alpha was . 626 and . 554 for males
and females respectively.
3.2. Items Defininct the Factor: -
Item Number Form A
51 I would rather be a secondary school teacher
than a forester.
Form B
26 At work I would rather be in charge of
interviewing people than mechanical matters.
51 I would rather be a hotel manager than a
research chemist.
52 1 enjoy selling things or asking for funds.
53 I'd rather be a business manager than architect.
176 I'd prefer being in charge of children than
helping a watchmaker.
3.3. Factor Analyses: The varimax factor analyses of the female data,
where A (Outgoing) emerged as a separate variable gives support
to the view that this trait could conceivably be a first order
factor of a second order extraversion/introversion factor. In
males, however, this primary source trait was absorbed into a
more generalised extraversion measure. We can see from the items
defining this particular trait that they are in the main concerned with
284
occupational preferences and it may well be that the social
stereotype of women being attracted to people orientated jobs is
responsible for differences in the factor solution across the sexes.
It is also possible that this factor is mainly a 'thing (job)'
versus 'people' orientated discriminator. It is not altogether
surprising, therefore, that Cattell has claimed that Factor A
is important in distincruishing h& we n, for ? Xd: i l, sci nt is is
and sales persons. The content of high loading items on this
primary does not, suggest that this factor is necessarily
closely associated with softheartedness, adaptability and
easy-goingness. There is also the considerable danger that
this factor's apparent power to distinguish between
occupational groups is based upon teleological causes.
FACTOR B (Less Intelligent vs. More Intelligent)
1. GROUP DIFFERENCES
1.1. Cross-National: - Whilst results were not significant with
the female groups, the British males were found to be significantly
higher on this factor than their American counterparts.
1.2. Sex Differences: - On this factor males were found to achieve
a significantly higher mean score than the female group.
1.3. Social Class: - As was hypothesised, the higher social classes
tended to achieve higher scores. This was a relatively strong
trend. I
285
1.4. Age: - There was a significant negative correlation between
this primary source trait and age for both males and females.
Tests of the departure of linearity of regression were also
significant, scores showing a more rapid decline after 40
years of age than they had shown thereto.
2. RELIABILITY
2.1. Group Values: - This was a factor of relatively low
reliability of approximately . 44.
2.2. Social Class and Age as Moderators: - Some evidence was found
for lower social classes and higher age groups to have lower
reliability on this factor.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated scale direct
validities for this factor were . 58 for males and . 56 for females.
The corresponding alpha statistics were . 60 and . 58 respectively.
3.2. Items Defining the Factor: - Most of the adequate items on
this particular scale were of the conventional verbal analogous
type. Three items, however, (103 and 152 on Form A and 177 on
Form B) showed negative item scale correlations.
3.3. Factor Analyses: - Despite the relatively low reliabilities of
this factor, it still emerged as a separate identity accounting
for some 4% of the variance. In addition, this trait generally
behaved consistently with regard to group differences hypothesised
from Cattell's model and the estimated scale direct validities
suggest that with the exception of Factor M, this primary is
286.
relatively untouched by the other 16PF variables. Not surprisingly,
these researches suggest that Factor B (Intelligence) is an
important and independent variable within Cattell's 16PF. Its
low reliability, however, only makes it suitable for use with
group rather than individual comparisons.
FACTOR C '(Affected by Feelings vs. Emotionally Stab-)
1. GROUP DIFFERENCES
1.1. Cross-National: - In line with the hypothesis, no significant
differences between British and American subjects were found
in this factor.
1.2. Sex Differences: - The female group were found to be significantly
lower than males on this particular trait. This was contrary
to the hypothesis of no difference between group means.
1.3. Social Class: - There was a significant positive linear trend
between emotional stability and social class in this sample.
This finding is in line with the prediction made from Cattell's
description of his primaries.
1.4. Age: - No evidence was found for a slight positive relationship
between this factor and age.
2. RELIABILITY
2.1. Group Values: - The group reliability coefficients. for
Factor C were approximately .5 with no significant differences. }
between the male and female subjects.
28 7
2.2. Social Class & Age as Moderators: - Neither social class nor
age were seen as significant moderators upon the reliability
of this particular scale.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated composite
item scale correlations obtained for this study were . 67 for
males and . -65 for females. This compares with alpha coefficients
of . 70 and . 68 respectively.
3.2. Items Defining the Factor: -
Item Number Form A
55 1 have hardly ever been let down by my friends.
79 People do not seem to ignore or avoid me.
80 People do not treat me unreasonably.
129 After looking forward to something, I
always, feel like going.
154 My sleep is not disturbed by vivid dreams.
Form B
41 never feel 'shut in' when in a small space.
5I never find it difficult to put trivial troubles
out of my mind.
80 The way people say things never hurts me.
130 I am ready for life and what it demands.
154 I never find it difficult understanding some
people because of their unusual use of words.
179 1 do not want a more sheltered life.
t
288
3.3. Factor Analyses: - This primary was highly loaded on an-
anxiety factor extracted from the intercorrelation matrix of the
32 Form A and B scale scores. Direct validities estimated from
the data were about as high as the intercorrelation of this scale
with Factors C and Q4. From the evidence collected here, this
trait would not seem to constitute a separate identity from the
o, -. her anxiety vsriaUles within the y PF.
FACTOR E (Humble vs. Assertive)
1. GROUP DIFFERENCES
1.1. Cross-National: - Whilst there were no differences between
the female American and British groups, the British males
were significantly higher on this factor than the average
American male.
1.2. Sex Differences: - As was hypothesised very marked sex
differences were found on this primary source trait; males
being significantly more assertive than females at every
age group.
1.3. Social Class: - No significant correlation was found between
this factor and social class.
1.4. Age: - This factor showed significant negative linear
correlations with age in both groups. For the male sample
there was a slight but statistically significant departure
from linearity. This was in line with the hypothesised age
289
differences although the decline would appear to commence
somewhat earlier than 40 years of age as has been proposed
by Cattell.
2. RELIABILITY
2.1. Group Values: - The alternate form coefficient between Forms
A and B of the l6PF was . 57 for the total sexes but somewhat
lower for the single sex groups.
2.2. Social Class & Age as Moderators-- When account had been taken
of restriction in range across the various groupings, no
important differences in reliability were found across either
age or social class.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The composite
item scale correlations for this factor were . 66 for males
and . 65 for females. The corresponding alpha reliability
estimates were . 69 and . 67 respectively.
3.2. Items Defining the Factor: -
Item Number Form A
61 do not hold back from criticising others.
7I make clever, sarcastic remarks to others.
56 In some ways I feel definitely superior to others.
81 Foul language does not disgust me, not even in mixed company.
106 1 am best described as 'forceful. '.
290
156 1 am pleased when I become the one in command.
180 I almost always put forward ideas.
181 1 am good at meeting challenges.
Form B
6I tend to speak out when another person's
reasoning is in error.
31 1 do not avoid saying things that embarrass
others.
32 I would not feel nervous with a loaded gun in my hand.
57 It is important to get your ideas into practice.
106 I usually tell a superior when my opinion
differs- from his.
131 I am more energetic and ambitious than many
equally successful people.
3.3. Factor Analyses: - In both the male and female groups Factor E
had high loadings on a second oblique primary factor, which
accounted for some 15% of the variance. It seemed to have
little in common, for example, with Factor A and appears to
constitute an important primary within Cattell's exvia
second order factor. Group differences were generally in
line with those hypothesised on this particular factor.
I
291
FACTOR F (Sober vs. Happy-go-Lucky)
1. GROUP DIFFERENCES
1.1. Cross-National: -Group differences between the American and
British subjects were not significant either for males or
females on this primary source trait.
1.2. Sex Differences: - Similarly, there was no significant
difference between the sexes.
1.3. Social Class: - Whilst there were no significant product
moment correlation coefficients between this factor and social
class, there was some slight tendency for eta coefficients
to reach statistical significance; the higher social classes
scoring more highly on this primary.
1.4. Aqe: - The trend in this factor is for a very pronounced
negative correlation with age for both sexes.
2. RELIABILITY
2.1. Group Values: - The Factor F primary source trait was seen
to be one of the more reliable variables in Cattell's
personality model; the reliability being estimated at
approximately . 65 for both sexes.
2.2. Social Class & Age as Moderators: - No systematic variation
was found as a result of age or social class acting as
potential moderators on the reliability of this factor. I
292
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated direct
validities for this primary were . 79 for males and . 76 for
females which compare with values of . 81 and . 79 respectively
by the alpha statistic.
3.2. Items Defining the Factor: -
item Number Form A
33 I am rated as an amusing talker.
58 I go out for entertainment more than average.
82 I have more friends than most people.
107 I do not avoid social functions.
132 I spend a good deal of my time talking
with friends.
133 I like doing daring things for fun.
157 I like dressing with eye-catching style.
158 I prefer a lively party to a quiet hobby.
182 I am considered an enthusiastic person.
183 1 like jobs which offer change and travel.
Form B
8 I often tell amusing jokes.
33 I enjoy playing practical jokes.
58 I enjoy the principle of 'laugh and be merry'.
83 I like being in exciting situations and bustle.
107 I do not avoid embarrassing topics when
talking with the opposite sex.
132 1 would like more excitement.
182 I tend to act on impulse.
183 1 eng best described as 'happy-go-111cky'.
I
293
3.3. Factor Analyses: - In the varimax factor analysis, this
source trait did not emerge as an independent factor but
rather was absorbed into Assertiveness (E+) and Group
Adherence (Q2-). This factor also appeared to have some
negative loading on a factor well defined by Cattell's G
(Superego Strength). It would seem, therefore, that whilst
this is a relatively reliable trait, it is not independent
of certain other extraversion/introversion variables within
the 16PF.
FACTOR G (Expedient vs. Conscientious)
1. GROUP DIFFERENCES
I. J. Cross-National: - Contrary to the hypothesis of no significant
differences between the two nations, American adults achieved
significantly higher results on this factor.
1.2. Sex Differences: - No significant differences were found
between the sexes on this variable.
1.3. Social Class: - With regard to social class, no clear
relationship was found.
1.4. Aqe: - A strong positive correlation with age was found in
both males and females.
294
2. RELIABILITY
2.1. Group Values: - This source trait was of moderate reliability
with an estimated value of approximately . 47.
2.2. Social Class & Age as Moderators=- When account was taken
of differences in variability between the sub-groups, no
consis ter: L c-lid: -nc.. :,; es found for age and social class
acting as significant moderators on scale reliability.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated scale
direct validities for this primary were . 63 for males and
. 58 for females with corresponding values of . 65 and . 60
by the alpha method.
3.2. Items Defining the Factor: -
Item Number Form A
34 1 get annoyed when I see untidy people.
134 The sight of an untidy room annoys me.
160 I always refer to the rules of right and wrong.
184 I insist on always doing things correctly.
Form B
134 It is good to plan and avoid wasting time.
184 1 do not like seeing disorder.
f
295
3.3. Factor Analyses: - This trait, with Q3 (social precision),
emerged with high loadings on the fifth factor extracted
from the correlation matrices. The serious question emerges,
however, whether the items which define the factor can
in any real sense be regarded as measures of superego
strength. Rather they would seem to centre on tidiness
of dress, order and perhaps a love of structure. It is
not surprising then that this variable shows a considerable
increase over age; adults, after all, generally become less
accepting and more critical of the dress of others as they
mature in age. Cattell (1972) has claimed that G is
negatively. related with. creativity and it may well be
that the variable is merely a liking for orderliness and
routine in daily life. In any event, if we are to judge
this factor by the adequate items within Factor G, there
can be little reason for viewing this source trait as
representing superego strength within the generally
accepted definition of the term.
FACTO1 H (Shy vs. Venturesome)
1. GROUP DIFFERENCES
1.1. Cross-National: - American females were seen as higher on
this factor whilst there were no differences between the
American and British males.
1.2. Sex Differences: - As was hypothesised, males scored =
significantly higher on this factor.
296
1.3. Social Class: - With regard to social class, a positive
relationship was found in the female sample but no
significant trend was discovered for male subjects.
1.4. Aqe: - The results of the age analysis for this primary
source trait showed a relatively flat pattern of differences
across t tz, six ag_ cohC . rts. The hypothesis of a positive
relationship between this source trait and age was rejected.
2. RELIABILITY
2.1. Group Values: - This was the most reliable of Cattell's
primary source traits with an estimated alternate form
coefficient of . 73.
2.2. Social Class & Age as Moderators: - Some evidence was found
that age was a significant moderator on the reliability of
Factor H; older age groups (65+ years) tended to have higher
standard error of measurements than younger subjects. No
significant effects on reliability were found, however,
with social class.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The composite item
scale correlations for this factor were . 84 for males and
. 85 for females. The Cronbach alpha statistics were
calculated as . 87 (males) and . 88 (females) on Forms A and
B of the 16PF.
I
297
3.2. Items Defining the Factor: -
Item Number Form A
10 I come forward at social occasions.
35 I like being the focus of attention in a group.
36 I like to join a gathering of people.
60 1 do not keep quiet in the presence of
senior people.
61 I do not find it difficult to speak to a
group of people.
86 I am not content to let others do the
talking in a small group.
110 It is easy for me to mix with people at
a social event.
111 I can use tact and persuasion to get people moving.
135 I am a sociable, outgoing person.
136 I show my feelings in social relationships.
161 1 do not mind people watching me at work.
Form B
10 I have been active in organising social groups.
35 I am not unsure of myself socially.
36 I like getting into conversation with others.
60 I am not troubled by feelings of inferiority
in my social relations.
61 I do not feel awkward in company.
85 I am not embarrassed by people watching
me in a shop.
86 I avoid joining in a conversation with
other people.
298
110 I usually succeed in impressing people
favourably.
135 I am usually involved with what is going on.
136 I fit into a new group immediately.
161 It is not difficult for me to start
conversations with strangers.
186 1 am not bothered by shyness.
3.3. Factor Analyses: - Like Factor F, this variable has loadings
on a number of different factors extracted in the male and
female analyses. It seems to be related to anxiety,
assertiveness (negatively) and group adherence (A+, F+, Q2-).
We can see, therefore, that despite its very high reliability,
this scale is unlikely to be uni-dimensional and is related
to a rather complicated cluster of extraversion/introversion
and perhaps anxiety components.
FACTOR I (Toughminded vs. Tenderninded)
1. GROUP DIFFERENCES
1.1. Cross-National: - In both male and female groups, American
subjects achieved a mean score which was significantly
higher than that of the British samples.
1.2. Sex Differences: - This is one of the major discriminators
between the sexes and in line with the hypothesis, females
attained a considerably higher mean score than males in the
British sample.
299
1.3. Social Class: - No differences were found between the social
class groups in the male sample. There was, however, a
significant positive relationship between this variable
and social class for the female sample.
1.4. Aqe: - Some support was found for the hypothesis of a low
positive relationship between this primary source trait
and age.
2. RELIABILITY
2.1. Group Values: - The reliability coefficient calculated on
the total sex group was. approximately . 55. However, the
sex of the respondent was a significant moderator here
with males and females showing reliabilities of . 47 and . 32
respectively.
2.2. Social Class & Aqe as Moderators: - Age had no significant
effect as a moderator on the reliability of this primary
source trait. Whilst there were differences in the
reliability coefficients for the various social class groups,
(the higher social classes showing higher reliability), there
were no significant differences between the groups when
account was taken of differences in variability of scores.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated scale
direct validities were . 59 for males and . 45. for females.
The corresponding alpha statistics were . 62 and . 47
respectively. i
300
3.2. Items Defining the Factor: -
Item Number Form A
11 I would rather be a playwright than a
construction engineer.
62 1 do not have a good sense of direction.
87 I prefer reading a novel to accounts of
hat-11 es
112 Being a guidance counsellor appeals to me
more than an engineering manager.
138 The beauty of a poem appeals to me more
than that of a well-made gun.
Form B
87 T do not tend to be interested in
mechanical matters.
112 T would rather be a philosopher than a mechanical engineer.
163 I would rather watch a concert artist
than a program on new inventions on the
television.
3.3. Factor Analyses: - This scale was highly loaded on factors
extracted from both the male and female intercorrelation.
matrices by the promax technique. Other 16PF variables
with sizeable loadings on this factor includedA (negatively)
and L (negatively) in males and L (negatively), M (negatively)
and Ql (positively) in females. Perhaps the similarity of the
items defining the I and A primary source traits should not go
without mention. Both possess a high vocational content with
particular emphasis on practical versus people orientated jobs. I
301
FACTOR L (Trusting vs. Suspicious)
1. GROUP DIFFERENCES
1.1. Cross-National: - Despite the hypothesis of no significant
differences between the two nations, British adults had a
significantly higher mean than their American counterparts
for both male and female subjects.
1.2. Sex Differences: - A significant difference in favour of
males was found in this primary.
1.3. Social Class: - With regard to social class, a negative
relationship was found in females, whilst there was no
significant association in the male group.
1.4. Agee: - This primary source trait showed the strongest non-
linear association with age; males showing a slight decrease
in scores over age group, whilst females demonstrated a
decline in scores until about 50 years of age where they
began to increase again.
2. RELIABILITY
2.1. Group Values: - Cattell's Factor L was one of the less
reliable primary source traits with an approximate alternate
form reliability coefficient of . 31.
2.2. Social Class & Age as Moderators: - Age had no significant
effect as a moderator in the reliability of this trait. I
There were significant differences between the reliability
302
coefficients of different social classes. However this
effect was largely due to differences in variability of sums
across the various social groups.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - When Cattell's direct
Vc iid±tties were eS L' mate C1 by the composi'e i -err, scale
method, values of . 47 for males and . 51 for females
were obtained. This compared with alpha values of . 49
and . 53 respectively. When these data were compared
with the scale intercorrelation with other factors, the
items in Factor L appeared to measure the C. (negatively).,.
E (positively) and Q4 (positively) source traits almost
as well as they measure themselves.
3.2. Items Defining the Factor: -
Item Number Form A
113 If a person is selfish, I will show him up.
114 1 do not tend to make silly remarks in fun.
Form B
Adopting the criterion of item remainder first order
scale correlations of . 25 and over, there was not one
adequate item in Form B of the 16PF to measure this
variable. The best item with the value of . 203 was
No. 64 - People would like to see me in trouble
i
303
3.3. Factor Analyses: - As to be expected from the direct validity
estimates, this variable did not emerge with particularly
strong loadings on any of the oblique primary factors.
This trait appeared to be related to anxiety (positively),
assertiveness (positively) and tendermindedness (positively)
and as such seems an unreliable amalgam of different
variables within Cattell's personality model.
FACTOR M (Practical vs. -Imaginative)
1. GROUP DIFFERFACES.
1.1. Cross-National: - In both the male and female groups, British
adults attained a significantly higher mean score than that
of Cattell's American standardisation sample.
1.2. Sex Differences: - Contrary to the hypothesis,
males were higher on this factor than female subjects.
1.3. Social Class: _ A significant positive relationship was
found between this variable and social class for both the
male and female samples.
1.4. Age: - Results on this source trait indicated relatively
small differences across age groups. Product moment
coefficients were generally non-significant, as were
statistical tests for the significance of departures
from linearity. 3
304
2. RELIABILITY
2.1. Group Values: - This was another factor of low alternate
form reliability with an estimated value of . 27.
2.2. Social Class & Aqe as Moderators: - Reliability estimates
tended to be particularly low in the lower social classes
and in bigger ace groups,
although both of these trend-
appear to have their origin in differences in variability
of scores.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated scale
direct validities for this factor were . 37 for males and
. 45 for females. The corresponding alpha values were . 39
and . 46 for males and females respectively. It was also
noticeable that the items in this source trait measured
E, H and Ql almost as well as they measured their own scale.
3.2. Items Defining the Factor: - None of the items in either
Forms A or B reached the criterion of a composite item
scale correlation in excess of . 25. The best items,
however, were as follows: -
Item Number Form A
90 I dislike the way some people stare at
others. (r - . 22)
141 One should be careful about being with
strangers because of problems of disease, etc.
(r = . 22) }
305
Item Number Form B
15 1 dislike completing routine jobs (r = . 24).
65 It is more important to be concerned about
the meaning of life than making a good income.
(r = . 24)
3.3. Factor Analyses: - From the factor analyses of the male and
female intercorrelation matrices, this trait had negative
loadings on anxiety in females, positive loadings on
assertiveness for both sexes, positive loadings on
tendermindedness for both sexes and positive loadings
independence for males. The overall picture for this
factor, especially in view of its very low reliability,
is extremely complex. From these analyses, however, it
would seem to be a combination of emotional stability,
an interest in artistic matters and ideas and high
independence. We must question, therefore, whether this
is in fact an independent factor and whether the practical
versus imaginative label is the most appropriate term to be
applied to this collection of very heterogeneous items.
FACTOR N (Forthright vs. Shrewd)
1. GROUP DIFFERENCES
1.1. Cross-National: - British females were significantly higher
than American females on this factor, whilst there were
no significant differences for the male samples.
I
306
1.2. Sex Differences: - Contrary to the hypothesis, females
exceeded males significantly in terms of their group
mean.
1.3. Social Class: - Despite Cattell's claim that this variable
is highly related to social class, no clear relationship
f12 S found .
1.4. Ace: - This factor showed a moderately strong relationship
with age with no significant departure from linearity.
This was in line with the hypothesis developed from
Cattell's model.
2. RELIABILITY
2.1. Group Values: - Along with Factor Ql, this primary source
trait was the most unreliable variable with an estimated
coefficient of . 25.
2.2. Social Class & Age as Moderators: - Some evidence was found
for Factor N to show lower reliability coefficients in the
unskilled working class but this effect was mainly as a
consequence of lower test score variance in these groups.
No relationship was discovered between age and the
reliability of the scale.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated direct
validities of the scales were . 37 for males and . 36 for
females. These values were in relatively close agreement
307
with alpha statistics of . 38 for both male and female groups.
When these internal consistency estimates were compared with
intercorrelations of Factor N items with other scales,
there was in fact a higher negative correlation with
Factors E, F and Q3.
3.2. Items Defining the Factor-- Only one item (No. 42 in Form A)
reached the criterion of an item scale correlation exceeding
. 25. Two other items did in fact exceed . 20, however, and
these are listed below.
Item Number Form A
41 I do not feel like taking part in tough
physical activity.
42 1 prefer polite people to rough individuals.
Form B
67 I get many unusual ideas which are too
many to put into use.
3.3. Factor Analyses: - The major loading of this trait is
negatively on a factor best defined by E (Assertiveness)
on the 16PF. In. females, however, there is also a negative
loading on a general social extraversion factor defined by
Cattell's F and H primary source traits. The results of
the estimated scale direct validities, factor analyses and
the reliability studies suggest that this is a variable
already contained within the extraversion/introversion domain.
I
308
FACTOR 0 (Self-Assured vs. Apprehensive)
1. GROUP DIFFERENCES
1.1. Cross-National: - No significant differences were found
between the American and British female groups, although
in males, American adults were significantly higher.
1.2. Sex Differences: - As was hypothesised, British females
scored higher than the male respondents in this survey.
1.3. Social Class: - Slight negative association with social
class was found for this variable.
1.4. Age: - When Factor 0 was analysed by age cohorts, a relatively
flat profile of mean scores was found.
2. RELIABILITY
2.1. Group Values: - This primary was one of the more reliable
source traits with an estimated value of . 61.
2.2. Social Class & Age as Moderators: - No significant differences
in the reliability of this scale for different age or social
class groups were found.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The internal consistency
of Factor 0 estimated by the composite item scale correlation
on items with the scale score was . 72 for males and . 71 for j
females. Alpha statistics achieved a value of . 75 and . 73
309
respectively. When these values were compared with the
intercorrelation of the factor with the remaining primaries,
they were found to be about as high as the scale's correlation
with C and Q4.
3.2. Items Defining the Factor: -
I ter Number Form A
18 I occasionally have a sense of danger
which I cannot. understand.
43 I feel very down when I am criticised in
a group.
44 I fear I have done something: wrong when
called in by my boss.
68 Often I do not feel in the right mood to
see anybody.
93 I feel down if acquaintances show they
dislike me.
118 I often feel that I will be punished even though I have done nothing wrong.
168. I am not regarded as a solid, undisturbed
person.
Form B
18 I can be overcome by thoughts of worthlessness.
43 My friends do not need me as much as I need them.
44 I worry if people think badly of me.
68 My spirits are not generally high.
69 I worry about an event at night.
93 1 feel overcome when small things go wrong.
I
310
94 I am disturbed by guilt feelings.
118 1 become afraid that my happiness will not last.
119 1 occasionally feel depressed and miserable.
143 When catching a train, I get tense and anxious.
3.3. Factor Analyses: - Cattell's 0 (Apprehensive) loaded highly
o ',
-be first, and largest t_ Factor from -ý '-he
prom ax sC,
This was clearly a general anxiety factor where Factor C
and Q4 also loaded highly. From the scale direct validities
and the factor analyses., there would seem no reason to
suppose that 0 is in any way different from the C and Q4
primary source traits in Cattell's model.
FACTOR Q1 (Conservative vs. Experimenting)
1. GROUP DIFFERENCES
1.1. Cross-National: - As was predicted from Cattell's model of
personality, there were no significant differences between
the American and British groups on this factor.
1.2. Sex Differences: - Males achieved significantly higher scores
on this primary source trait. This result was in line with
the hypothesis developed in an early section of this thesis.
1.3. Social Class: - No significant relationship was found between
this variable and social class.
t
311.
1.4. Age: - Despite Cattell's hypothesis that this factor shows no
clear relationship with age, these results showed a very
strong negative correlation both in the male and female groups.
2. RELIABILITY
2.1. Group Values: - This particular primary source trait was
one of the most unreliable of Cattell's variables with an
estimated alternate form coefficient of . 25.
2.2. Social Class & Age as Moderators: - The reliability coefficients
of this factor showed significant differences between the
social groups, higher social grades showing more marked
consistency of scores across the two forms. It was
also noted that coefficients tend to be lower in the
older age groups. However, when account was taken of
differences in variability of scores by the use of the
SEM statistic, no significant trend was found.
3, INTERNAL CONSISTE2ICY
3.1. Estimated Scale Direct Validities: - The estimated scale
direct validities were found to be . 45 for males and . 42
for female subjects. The corresponding alpha statistics
were . 47 and . 44 respectively. These estimates were about
as high as the intercorrelation of this primary with
Factor E (Assertiveness).
3.2. Items Defining the Factor: - No items reached the criterion
of having a correlation of . 25 or more with the scale..
312
However, five its exceeded the value of . 20 and these
are listed below.
Item Number Form A
70 As a teenager, I kept to my own opinions
if I differed with my parents.
95 I would rather have a 'oh with a large salary
which meant my persuading people that I. was
worth it.
145 In a heated argument between people, I
like to see one side or the other win.
Form B
46 In a newspaper I like to see discussion.
of basic social issues.
95 I disagree with the idea that people should
get together in order to worship regularly.
3.3. Factor Analyses: - This source trait had loadings on the
factor which we have provisionally entitled 'Superego Strength'
(G+, Q3+). Sizeable loadings were also evident on Assertiveness
(E+) and negatively on Tendermindedness (I+). In all, this
seems a relatively unreliable factor which has variance in
common with a number of other source traits. It certainly is
poorly described by Cattell in terms of its relationship with
external variables such as age. Whilst it is difficult at
this stage to say whether Cattell's QI is in any way comparable
to Eysenck's social attitude of radicalism, we have seen from
these analyses that the trait as measured by 16PF items does
not appear to be independent of the remaining primaries.
313
Further, clarification and definition of this trait is
tequired, hopefully with far more reliable items than those
employed in Cattell's inventories.
FACTOR Q2 (Group-Dependent vs. Self Sufficient)
1. GROUP DIFFERE['JCES
1.1. Cross-National: - Despite the hypothesis of no difference
between the two national groups, both the male and female
British samples achieved significantly higher mean scores
on this factor than their American counterparts.
1.2. Sex Differences: - British males were significantly higher
than females on this source trait.
1.3. Social Class: - No significant relationship with social
class was found with this variable.
1.4. Acre: - The general trend is for scores to increase on
Self-Sufficiency after adolescence until about 20-25 years
where they level off until about 55 years. There was a
significant positive relationship between Q2 and age in
males but only a slight association for female subjects.
2. RELIABILITY
2.1. Group Values: - This primary source trait was only of
moderate reliability with an approximate alternate form
coefficient of . 41.
314
2.2. Social Class & Age as Moderators: - There was some evidence
that reliability coefficients tend to be lower in higher
age groups and in the lower social classes. This trend
was largely eradicated, however, when account was taken of
differences in variability of scores. The one exception
to this was in Form B of the 16PF where older subjects
demonstrated higher standard errors of measurement.
3. INTERNAL CONSISTENCY: -
3.1. Estimated Scale Direct Validities: - The direct validities
estimated by the composite item scale correlations were
. 57 for males and . 55 for females. These corresponded
to values of . 59 and . 47 respectively based on alpha
internal consistency values.
3.2. Items Defining the Factor: -
Item Number Form A
97 I do not like to take an active part in
social affairs and committee work.
122 I would rather work on my own than with a committee.
171 I learn better by reading a good book on the topic than by joining a group.
Form B
47 Books are more entertaining than companions.
t
315
3.3 Factor Analyses: - This source trait had high negative
loadings on a third factor extracted form the promax factor
analyses. The other traits with high loadings were
F (positively), H (positively) and in the case of males
A (positively). Unlike F and H, this source trait did
not load heavily on a factor provisionally called
'Assertiveness' and as such seems to be a purer measure
of group dependence or perhaps group adherence. From
our analyses it would seem that Q2 along with E measure
two sub-components of extraversion/introversion which show
only low intercorrelation.
FACTOR Q3 (Undisciplined Self-conflict vs. Controlled)
1. GROUP DIFFERENCES
1.1. Cross-National: - No significant cross-national differences
were found on this factor.
1.2. Sex Differences: - The adult male sample achieved a
significant higher mean-score on this source trait.
1.3. Social Class: - This variable showed a significant non-linear
relationship with social class. Mean scores tended to be
high for the higher social classes, to decrease in the middle
class, skilled and unskilled working class and to increase
again in the unemployed and retired (Group E).
}
316
1.4. Age: - Results showed a relatively strong, positive linear
relationship with age but that little evidence of any
levelling off at 40 years as has been postulated by Cattell.
2. RELIABILITY
2.1. Group Values: - The estimated alternate form reliability
cf this factor vas . 5'_, it tending to be significantly
higher in male rather than female subjects.
2.2. Social Class & Age as. Moderators: - The reliability
coefficients for this factor were considerably higher
in higher social class groups but there was little
appreciable difference within the age cohorts. Little
difference between either the age or social class groups
was noticed however when the SEM statistic was used
to investigate reliability.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated scale
direct validities were . 66 for males and . 62 for females
with corresponding alpha values of . 69 and . 65. These
estimates were approximately as high as the intercorrelation
of the scale with Cattell's Factor G (Superego Strength).
3.2. Items Defining the Factor: -
Item Number Form A
23 When looking in the-mirror, I rarely get
confused as to which is right or left.
317
48 I keep my room well organised.
98 I am never satisfied until minor details
are given close attention.
123 I never have times when I feel sorry for
myself.
147 I do not allow my actions to get influenced
by jealousy.
Form B
23 I do not like to leave things to chance.
24 I go about my life in a deliberate,
methodical way.
73 When I need to make a quick decision, I
rely on calm logic.
147 I do not tend to get over-excited and rattled
easily.
148 I am not absent-minded.
173 1 cannot really be described as a dreamer.
A most interesting anomaly occurs with item 98, Form B of
the 16PF2which correlates -. 22 with the Q3 scale. Inspection
of the item (I make a fuss if I am left out of something by
others) suggests that the score of this scale has in some
way been reversed with respect to the underlying factor.
We are in the position, therefore, that removal of this item
will tend to increase the reliability of the scale.
3.3. Factor Analyses: - This source trait tended to have some
positive loading on the first oblique factor of anxiety
extracted from the male and female correlation matrices.
318
Its major loading, however, was on a factor provisionally
called 'Superego Strength' where Factor G was also highly
loaded. It would seem, therefore, that this variable is
largely Cattell's G with some anxiety variance. This conclusion
is very much supported by the list of items defining the
factor which is provided above.
FACTOR Q4 (Relaxed vs. Tense)
1. GROUP DIFFERENCES
1.1. Cross-National: - British female subjects were significantly
higher on this factor than their American counterparts,
whilst there were no significant differences between the
male groups.
1.2. Sex Differences: - As was hypothesised, females were
significantly higher on this factor.
1.3. Social Class: - With regard to social class, no differences
were found for male respondents but there was a significant
negative relationship with social class in the female group.
1.4. Age: - The hypothesis of slight, negative relationship on.
Factor Q4 with age was in fact found in this data.
2. RELIABILITY
2.1. Group Values: - The alternate form reliabilities for this I
factor were relatively high with a value of approximately . 66.
319
2.2. Social Class & Age as Moderators: - Neither age nor social
class were seen as moderators on the reliability of this
scale.
3. INTERNAL CONSISTENCY
3.1. Estimated Scale Direct Validities: - The estimated scale direct
validities were . 79 for males and . 75 for females. The
corresponding alpha statistics were . 82 and . 78 for males
and females respectively. These tended to be about as high
as the intercorrelation of this scale with Factors C (negatively)
and 0 (positively).
3.2. Items Defining the Factor: -
Item Number Form A
25 I find it difficult to calm down when
something makes me angry.
49 I sometimes feel tense when I think
of the day's happenings.
50 I sometimes feel that others are not interested in what i am saying.
74 I find that I am upset by criticism.
75 I find it difficult to keep my feelings
under control.
99 Small setbacks can irritate me too much.
124 I often get angry with people too quickly.
149 I get tense when I think of the things
in front of me.
174 Trivial things can get on my nerves.
I
320
Form B
25 I often feel restless without knowing why.
49 Certain sounds can set my nerves on edge.
50 I often feel tired when I get up in the
morning.
74 Useless thoughts tend to stray through
my mind.
99 I can be easily distracted from working.
125 If I am annoyed, I must let off steam.
149 A near accident or argument can leave
me shaky..
150 1 often feel that I am boiling up inside.
174 My emotions can run away with me.
175 I frequently feel so furious, I could break a window.
3.3. Factor Analyses: - This primary source trait had high loadings
on the general anxiety factor extracted from the correlation
matrices. As previously indicated, there would seem to be no
reason to suppose that this variable is in any way different
from C (emotionally unstable) and 0 (apprehensive).
I
321
CONCLUSIONS ON GROUP DIFFERENCES
In order to summarise results on group differences in personality,
Tablel5-7. shows the relationships between Cattell's primary source
traits and the sex, social class and age of the adults sampled in
this study. This information has been organised under the seven
factor solution originating from the promax factor analysis of
the female intercorrelation matrix. The factors have been
provisonally named from the Cattellian variables which have the.
highest factor loadings.
It is not suggested that these seven factors are in any real sense
the true definition of the personality domain. Factor analysis,
utilising a relatively conservative criterion of the number of
factors to rotate, was employed here to test the hypothesis which
we have subsequently rejected that as few as two or three variables
are required to explain the sixteen source traits. But the results
of the factor analyses can also serve as a convenient descriptive
and summary device under which to consider Cattell's primaries.
The first and largest factor, provisionally entitled Anxiety (C-,
0+, Q4+) shows group differences where females and the lower social
classes tend to score higher but where there is no clear relationship
with age. On Assertiveness (E+, L+, H+, N-), males achieve higher
scores as do younger subjects. Social class, on the other hand, is
not related here. The third factor of Group Adherence (F+, Q2-)
shows rather conflicting sex differences, no relationship with social
class but strong negative correlations with age. I
322
TABLE: 15.1
CONSISTENCY OF SEX, SOCIAL CLASS AND AGE DIFFERENCES ON CATTELL'S
SOURCE TRAITS FOR THE SEVEN FACTOR SOLUTION
FACTOR Sex Relationship Relationship Differences with Social with Aqe (Higher Class Scoring Sex)
I Tt:?:; ýi . ety't
C- F Negative Zero
0+ F Negative Zero
Q4+ F Negative Slight Negative
II "Assertiveness"
E+ M Zero Negative
L+ M? Negative
H+ M Zero Zero
N- M Zero Negative
III "Group Adherence"
F+ NS Zero Negative
Q2- F Zero Negative r
IV "Tendermindedness"
1+ F Positive Positive
M+ M Positive Zero
V "Superego Strength"
G+ NS Zero Positive
Q3+ M Positive
Qi- F Zero Positive }
323
TABLE 15.1 (Continued)
FACTOR Sex Relationship Differences with Social (Higher Class Scoring Sex)
VI "Intelligence"
B+ M Negative
VII "Outgoing"
A+ F Zero
NOTES:
Relationship with Aqe
Negative
Zero
(1) In order to maintain consistency with regard to the direction
of relationships in this table, group differences are reversed
where a primary trait has a negative loading on the factor
concerned.
(2) KEY:
F. Females scoring significantly higher on the negative
or positive pole of the primary trait.
M: Males scoring significantly higher.
NS: No significant differences.
?: Relationship unclear due to conflicting results.
I
.f
324
In terms of the fourth factor, Superego Strength (G+, 03+, Ql-),
the picture of sex differences is rather confused but the relationships
with age are consistently strong positive with no apparent differences
between the social classes. Tender-mindedness (I+, M+) again has
conflicting sex and age differences, perhaps because Cattell's M
tends also to be related to B (intelligence), but shows positive
rý-=at_ion.; h os v th social- class.
Finally, Intelligence (B+) has demonstrated a negative association
with age and social class, whilst Outgoingness (A+), unlike the
Assertiveness and Group Adherence factors, has no clear relationship
with either variable.
Thus these researches provide further support for the view that
Cattell's primaries are not functionally independent. When grouped
under the relatively co&irse headings used here, the source traits
acting as major contributors to these factors behave with considerable
consistency. And this is despite the fact that a number of the
primaries load on more than one of the oblique factors derived
from these analyses.
We have also seen that Cattell's age corrections had little appreciable
effect on some eight of his source traits (B, E, G, H, I, L, Q1 and Q3).
On six of the remaining factors (A, C, M, O, Q2 and Q4), they introduced
error which was far more extreme than the original scale-age correlations.
There was no clear benefit to be gained, moreover, in using non-linear
explanations of age differences in personality. Indeed, in some instances.,
Cattell'S quadratic expressions tended to compound rather than minimise
error.
325
CONCLUSIONS REGARDING THE RELIABILITY OF THE PRIMARY SOURCE TRAITS
The reliability of the primary source traits estimated by the
correlation between the scale scores on the corresponding factors
of Forms A and B of the 16PF were generally in line with those
provided by Cattell, although as a result of reduced variability
in scores, somewhat lower than those calculated on student groups.
The most reliable primaries were the factors. of E (assertive),
F (happy-go-lucky), H (socially bold), I (tenderminded),
O (apprehensive) and Q4 (tense), whilst four factors, L (suspicious),
M (imaginative), N (shrewd) and Q1 (experimenting), were of such low
reliability (coefficients ranged from . 31 to . 25) as to be of little
practical utility.
The variables of age and particularly social class were found to be
significant moderators on the reliability coefficients of the
questionnaire scale scores. These effects were largely eradicated,
however, when the SEM statistic was used to take account of differences
in the variability of scores within the various sub-groups.
Nevertheless, some tendency still remained for lower social class
respondents to show'lower reliability of scores.
The traits which cause most concern in terms of reliability appear
to be the primaries of L, M and N. which make up three of the last
four factors previously extracted from L data, and Q1, the first of
the four primaries only found in the questionnaire realm. We must
accept, therefore, our general hypothesis; reliabilities as low as .2
326
certainly suggest that these factors are untenable in terms of the
precision of measurement afforded by the model. Moreover, use of
the scales which supposedly measure these traits in the lower social
classes will mean that the reliability of measurement is lower still.
CONCLUSIONS ON THE INTERNAL CONSISTENCY OF THE MODEL
The internal consistency of the primary source traits was investigated
by way of composite item-scale correlations, being the multiple
correlation of sums of the items with the 16PF Forms A+B scale
scores corrected for spurious item scale-score overlap. This is
analogous in scale score terms with Cattell's concept of 'direct
validities' which he has described as the multiple correlation of
the items in a scale with the 'pure' factor abstracted from factor
analyses.
It was also shown that the multiple correlation of sums of the items
in one scale with another scale is directly equivalent to the
intercorrelation of the two scales. It was possible, therefore,
to compare composite item-scale coefficients with scale intercorrelations
not only as a measure of the estimated'direct validities' of. items on
their own scale but also the 'direct validities' for items on external
scales.
The primaries of high estimated scale direct validity were C, E, F, H, O,
Q3 and Q4. Factors L, M, N and 02, on the other hand, showed relatively
327
low composite item-scale correlations; precisely those traits which
showed low alternate form reliability values. In general, the
composite item correlations tended to be slightly below alpha values
and the alternate form reliabilities corrected for two form length
by the Spearman-Brown prophecy formula.
vTnen the scale direct -, validity estimates wer` corparad with the
'validities' of items with external scales (ie. scale intercorrelations),
it was found that the anxiety factors of C, 0 and Q4 tended to measure
themselves about as well as they measured one another. Similarly,,
Factor G (superego strength) items had direct validity estimates on
Factor Q3 (social precision) which were virtually as high as their
direct validities on their own scale. Other source traits which
appear to be made up of non-discreet item content were the low
reliability primaries of L (suspicious), M (imaginative), N (shrewd)
and Ql (experimenting).
We have noted that Cattell has advocated the use of suppressor action
to improve the multiple correlation of items with some estimate of the
pure personality factor. The presence of suppressor action within the
adult personality questionnaires was tested here by comparing the
composite item-scale correlations with conventional homogeneity
estimates provided by Cronbach's coefficient alpha. If the action
of suppressor items was marked within the scales, then according to
the general law of multiple correlation the composite item-scale
correlations would tend to be higher than alpha statistics calculated
on the same data.
328
Unfortunately, no evidence was found in this study for any pronounced
suppressor effects within the primary source traits as represented by
the 16PF scale scores. Indeed, alpha coefficients tended to be slightly
but consistently higher than the composite item-scale correlations.
We must conclude, therefore, that despite Cattell's claims,
conventional alternate form or homogeneity estimates do not
z derestirnate the reliability of his scales.
CONCLUDING REMARKS
There would seem a good deal of evidence here to suggest that Catteil. 's.
primary source traits cannot be described as the simplest, most
reliable definition of the personality domain. It is particularly
alarming that so many glaring inadequacies exist in such widely used
personality questionnaires. Most obvious of these are inaccurate
age corrections, the low reliability and the complete interdependence
of certain source traits.
What is the most likely explanation of these findings? Perhaps it
was Cattell's early work on ratings based on small student groups and
using relatively unsophisticated correlational techniques which was
in error. Considerable criticism has been made of Cattell's method
of rotating to a defined target. As factor analytic and computer
techniques improved, Cattell, perhaps misguidedly, has spent his
resources attempting to replicate the results of early researches,
rather than refine his initial strictures on the structure of
personality. In other words, Cattell's heavy investment in his
theory has blinded him to criticisms made of the primary source traits.
329
Indeed, if this theory has undergone any change it has been by the
addition of new traits rather than by the critical evaluation of
those already existing in the model.
This is not to detract from Cattell's pioneering work on the factor
analytic investigation of personality. His writings on the methodology
of factor analysis have mate ö considerable contrihu is; n to the field..
But, in the case of Cattell's primary source traits, our conclusion
must be that theory seems to have outstripped experimental verification.
AREAS OF FURTHER RESEARCH
There would appear to be little point in carrying out further cosmetic
research on Cattell's model of personality. Howarth and Browne (1971),
Eysenck (1972) and Saville and Blinkhorn (1976) have all shown that
clear inconsistencies exist in the model as represented in questionnaire
data.
Whereas anxiety as represented by factors C, 0 and Q4 in the 16PF
emerge as a unitary trait, the nature of introversion/extraversion as
a variable in the personality domain is far less clear-cut. The
evidence here points to possibly three sub-components of this dimension
and further research should concern itself with the problem of defining
these components by factor analytic investigations of the 16PF items,
which supposedly measure the A, E, F, G, H, L, N and Q2 '. source' traits.
In addition, the nature of the factors which have been provisionally
330
named Tendermindedness (I+, M+) and Superego Strength (G+, Q3+ and
Ql-) need further study in order to establish their psychological
significance and to ascertain their exact relationship with
variables in other personality theories.
331
16. REFERENCES
332
RF. FRRF. NC F:.,;
ABRAHAMS NM and ALF E (1972) Pratfalls in Moderator Research Journal of Applied Psychology, Vol. 56, No. 3,245-251.
ADCOCK CJ (1965) A Comparison of the Concepts of Cattell and Eysenck: Research Notes. Brit. J. Educ. Psychol., 35,90-97.
ADCOCK N (1973)
Testing the Test: How Adequate is the 16PF with a New Zealand Student Sample? Paper presented to New Zealand Psychological Society Conference.
ADCOCK N V, ADCOCK CJ and WALKEY EH (1974) Basic Dimensions of Personality Victoria University of Wellington, New Zealand.
ALEAMONI LM and SPENCER RE (1969)
A Comparison of Biserial Discrimination, Point-Biserial Discrimination and Difficulty Indices in Item Analysis Data. Educ. & Psychol. Meas., 29,353-358.
ALF E and ABRAHAMS N (1971) A Significance Test for Biserial r' Educ. & Psychol. Meas., vol. 31, Autumn.
ANASTASI A (1961) Psychological Testing MacMillan Co., New York, 2nd Edition.
ARNHOFF FN and LEON HV (1963)
Personality Factors related to Success and Failure in Sensory Deprivation Subjects. Perceptual and Motor Skills, 16,46.
BAKKER CB and LEVENSON RM (1967) Determinants of Angina Pectoris Psychosomatic Medicine, 29,621-633.
BALTES PB (1968) Longitudinal and Cross-sectional Sequences in the Study of Age and Generation Effects. Human Development, 11,145-171.
BAREFOOT JC (1969) Anxiety and Volunteering Psychonomic Science, 16(6), 283-284.
333
BARTON K and CATI'ELL RB (1972)
Personality hefore and after a Chronic Illness J. of Clinical Psychology, Vol. XXVIII, No. 4,464-467, October.
BARTON K, CATTELL RB and VAUGHAN GM (1973)
Changes in Personality as a Function of College Attendance or Work Experience Journal of Counseling Psychology, Vol. 20,2,162-5.
BASS AR and ROSEN H (1969)
Some Potential Moderator Variables in Attitude Research Educ. & Psychol. Meas., 29,331-348.
BEALE EML, KENDALL MG and MANN DW (1967) The Discarding of Variables in Multivariate Analysis Biometrika, 54,3 and 4, p. 357.
BECKER WC (1960) The Matching of Behavior Rating and Questionnaire Personality Factors. Psychol. Bulletin, Vol. 57, No. 3.
BECKER wC (1961) A Comparison of the Factor Structure and Other Properties of the 16PF and the Guilford-Martin Personality Inventories Educ. & Psychol. Meas., Vol. XXI, No. 2.
BOORER D and MURGATROYD S (1973) . Personality and Learning -A Select Annotated Bibliography MTM Publishing House (Wales), Caerphilly.
BOWERS J (1972) A Note on Comparing r-Biserial and r-Point Biserial Educ. & Psychol. Meas., 32,771-775. "
BROWNE MW (1972) Oblique Rotation to a Partially Specified Target Br. J. Math. Stat. Psychol., 25,207-212.
BROWNE MW (1972) Orthogonal Rotation to a Partially Specified Target Br. J. Math. Stat. Psychol., 25,115-120.
BROWNE MW and KRISTOF W (1969) On the Oblique Rotation of a Factor Matrix to a Specified Pattern Psychometrika, 34, No. 2, June.
334
BÜROS 0K (Editor) (1972) The Seventh Mental Measurements Yearbook (Vols 1 and 11) Institute of Mental Measurements Gryphon Press, Highland Park, New Jersey, USA.
BUZZARD RB (1971)
Obituary to Sir Frederick Bartlett Occupational Psychology, 45, Vol. 1,1-6.
CAMPBELL DT and FISKE DW (1959) Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix Psychological Bulletin, 56, No. 2, March.
CARROLL JB (1961)
The Nature of the Data or How to Choose a Correlation Coefficient Psychometrika, Vol. 26, No. 4, December.
CATTELL RB (1933) Temperament Tests I. Temperament Brit. Journ. Psychol., 23.
CATTELL RB (1934)
Temperament Tests II. On the Nature of Preservation and the Tests devised to Measure it. Brit. Journ. Psychol., 24.
CATTELL RB (1943) The Description of Personality Basic Traits resolved into Clusters. J. Abnormal & Soc. Psychol., 38,476-506.
CATTELL RB (1943) The Description of Personality I: Foundations of Trait Measurement Psychol. Review, 50,6,559-592.
CATTELL RB (1944)
Interpretation of the Twelve Primary Personality Factors Character and Personality, 13,55-91.
CATTELL RB (1944)
Psychological Measurement: Normative, Ipsative, Interactive Psychol. Review, 51,292-303.
CATTELL RB (1945) The Description of Personality. III: Principles and Findings in a Factor Analysis Amer. J. Psychol., 58,69-90.
I
335
CATTELL RB (1946) The Description and Measurement of Personality World Book Co. (now HBJ), New York.
CATTELL RB (1946) A Shortened 'Basic English' Version (Form C) of the 16PF Questionnaire Journal of Social Psychology, Vol. XLIV, 257-258.
CATTELL RB (1946)
Simple Structure in relation to Alternative Factorization of Personality ST)he re
Journal of General Psychology, Vol. 35,225-238.
CATTELL RB (1947) Oblique, Second Order and Cooperative Factors in Personality Analysis Journal of General Psychology, 36,3-22.
CATTELL RB (1950) The Main Personality Factors in Questionnaire Self-Estimate Material Journal of Social Psychology, 31,3-38.
CATTELL RB (1956) Validation and Intensification of the 16PF Journal of Clinical Psychology, 12,205-214.
CATTELL RB (1956) Second-Order Personality Factors in the Questionnaire Realm Journal of Consulting Psychology, 20, No. 6.
CATTELL RB (1957)
Formulae and Table for obtaining Validities and Reliabilities of Extended Factor Scales Educ. & Psychol. Meas., Vol. IV, 491-498.
CATI*ELL RB (1957) Personality and Motivation Structure and Measurement World Book Co. (now HBJ), New York.
CATTELL RB (1959) Anxiety, Extraversion, and other Second Order Personality Factors in Children Journal of Personality, 27,464-476.
CATTELL RB (1959) The Dynamic Calculus: Concepts and Crucial 'Experiments Nebraska Symposium on Motivation (University of Nebraska Press)
336
CATTELL RB (1960) The Multiple Abstract Variance Analysis Equations and Solutions: For Nature-Nurture Research on Continuous Variables Psychological Review, 67, No. 6. November.
CATTELL RB (1962) The Basis of Recognition and Interpretation of Factors Educ. & Psychol. Meas., Vol. XXIi, No. 4.
CATTELL RB (1963) Theory of Fluid and Crystalized Intelligence: A Critical Experiment Journ. Educ. Psychol., Vol. 54, No. 1,1-22.
CATTELL RB (1963) Methodological Advances in Evaluating Hereditary and Env ironmental Influences and their Interaction Laboratory of Personality Assessment, University of Illinois Advance Publication No. 19, March.
CATTELL RB (1963) Teachers' Personality Description of Six Year Olds: A Check on Structure
Brit. Journ. Educ. Psychol., 33, Part 3,2=9-235.
CATTELL RB (1963) The Interaction of Hereditary and Environmental Influences Brit. J. Stat. Psychol., Vol. XVI, Part 2, November.
CATTELL RB (1964) Objective Personality Tests: A Reply to Dr. Eysenck Occupational Psychology, 38, No. 2, April.
CATTELL RB (1965) The Scientific Analysis of Personality Pelican Books, London.
CATTELL RB (1965) The Configurative Method for surer Identification of Personality Dimensions, notably in Child Study Psychol. Reports, 16,269-270.
CATTELL RB (1966) Handbook of Multivariate Experimental Psychology Rand McNally & Co., Chicago.
CATTELL RB (1966) The Scree Test for the Number of Factors Multivariate Behavioral Research, Vol. 1,245-276, April.
337
CATTELL RB (1966) Further Investigation of Objective Personality Test Factors in Early Childhood ETS Research Bulletin RB-66-42. August.
CATTELL RB (1968) Trait-View Theory of Pertubations in Ratings and Self Ratings (L(BR)-
and Q-Data): Its Application to obtaining Pure Trait Score Estimates in Questionnaires Psychological Review, 75, No. 2,96-113.
CATTELL RB (19019)
The Profile Similarity Coefficient, rp, in Vocational Guidance and Diagnostic Classification Brit. J. Educ. Psychol., 39, Part 2, June.
CATTELL RB (1970) Real Base Factor Analysis Laboratory of Personality Analysis, University of Illinois Advance Publication No. 15. Tentative November.
CATTELL RB (1970) Separating Endogenous, Exogenous, Ecogenic and Epogenic Component Curves in Development Data Dev. Psychol., Vol. 3, No. 2. APA Inc.
CATTELL RB (1972) Tabular Supplement (of Norms) No. 2 to the 16PF Handbook (for 1969 editions of Forms C and D) IPAT, Champaign, I1 linois, USA.
CATTELL RB (1972) The 16PF and Basic Personality Structure J. Behav. Science, 1, (4), 169-187.
CATTELL RB (1973) 1
Personality and Mood by Questionnaire Jossey-Bass, San Francisco.
CATTELL RB (1973) The Organisation of Independent Basic Research Institutes Symbiotic with Universities Higher Education, 2,1-14.
CATTELL RB (1974) How Good is the Modern Questionnaire? General Principles for Evaluation J. Personality Assessment, 38, No. 2.
338
CATTELL RB (1974)
A Large Sample Cross-Check on 16PF The Journal, Fall, 179-195.
CATTELL RB (1975)
Third Order Personality Structure in Q-Data - Evidence from 11 Experiments The Journal, Winter, 118-149.
CATTELL R B, BALCAR K, HORN JL and NESSELROADE JR (1969) Factor Matching Procedures: An Improvement of the s Index; with Tables
Educ. & Psychol. Meas., 29.
CATTELL R B, BARTON K and DIELMAN TE (1972)
Prediction of School Achievement from Motivation, Personality and Ability Measures Psychological Reports, 30,35-43.
CATTELL RB and BELOFF H (1953)
Research Origin and Construction of the IPAT Junior Personality Quiz
Journal of Consulting Psychology, 17, No. 6.
CATTELL R B, BLEWETT D and BELOFF JR (1955)
The Inheritance of Personality. A Multiple Variance Analysis Determination
of approximate Nature Nurture Ratios for Primary Personality Factors in
Q-data. American J. Human Genetics, Vol. 7, No. 2, June.
CATTELL RB and BUTCHER J (1968) The Prediction of Achievement and Creativity Bobbs-Merrill, Indianapolis.
CATTELL RB and COAN RW (1957) Child Personality Structures as revealed in Teachers' Behavior Ratings. Journal of Clinical Psychology, 13,315-327.
CATTELL RB and COAN RW (1957) Personality Factors in Middle Childhood as revealed in Parents' Ratings Child Development, 28, No. 4, December.
CATTELL RB and COAN RW (1959) Objective Test Assessment of the Primary Personality Dinensions. in Middle Childhood Brit. J. Psychol., 50,235-252.
CATTELL R B, COAN RW and BELOFF H (1958) A Re-examination of Personality Structure in Late Childhood, and Development of the High School Personality Questionnaire J. Exper. Education, 27, December.
339
CATTELL RB and DAMARIN F (1966)
Further Investigation of Objective Personality Tests Factors in Early Childhood ETS Bulletin RB-66-42. August.
CATTELL RB and EBER HW (1967-1970)
Sixteen Personality Factor Questionnaire (16PF) Forms A&B (1967-68 Edition) and C&D (1969-70 Edition) IPAT, Champaign.
CATTELL R B, EBER HW and DELJHEES K (1968)
A Large Sample Cross Validation of the Personality Trait Structure of the 16PF with some Clinical Implications Multivariate Behavioral Research, Special Issue, 107-132.
CATTELL R B, EBER HW and TATSUOKA M (1970) Handbook for the Sixteen Personality Factor Questionnaire (16PF) Institute for Personality and Ability Testing, Champaign, Illinois, USA.
CATTELL RB and GIBBONS BD (1968) Personality Factor Structure of the Combined'Guilford and Cattell Personality Questionnaires J. Personality and Social Psychology, Vol. 9, No. 1,107-120.
CATTELL RB and GRUEN W (1954) Primary Personality in the Questionnaire Medium for Children Eleven to Fourteen Years Old Educ. & Psychol. Meas., 14,50-76.
CATTELL RB and GRUEN W (1955) The Primary Personality Factors in 11-year-old Children by Objective Tests Journal of Personality, 23,460-478.
CATTELL RB and HOWARTH E (1964) Verification of Objective Test Personality Factor Patterns in Middle Childhood J. Genetic Psychology, 104,331-349.
CATTELL R B, KAWASH G and DEYOUNG GE (1972) Validation of Objective Measures of Ergic Tension: Response of the Sex Erg to Visual Stimulation Journal of Experimental Research in Personality, Vol. 6, No. 1, March.
CATTELL R 8, KONII OS E and TATRO D (1968)
Significant Differences of Affective, Paranoid and Non-Paranoid Schizophrenic Psychotics on Primary Source Traits in the 16PF.
Multivariate Behavioral Research, Special Issue, pp. 33-54.
340
CATTELL RB and MERLE JL (1960) The 'Maxplane' Program for Factor Rotation to Oblique Simple Structure Educ. & Psychol. Meas., Vol. XX, No. 3.
CATTELL RB and NICHOLS KE (1972) An Improved Definition from Ten Researchers, of Second Order Personality Factors in Q-data (with cross-cultural checks). Journal of Social Psychology, 86,187-203.
CATTELL RB and RADCLIFFE JA (1962) Reliabilities and Validities of Simple and Extended Weighted and Buffered Unifactor Scales. Brit. Journ. Stat. Psychol., Vol. XV, Part 2, November.
CATTELL RB and SCHIER IH (1961)
The Meaning and Measurement of Neuroticism and Anxiety Ronald Press, New York.
CATI'ELL R B, SEALY AP and SWENEY AB (1965)
What can Personality and Motivation Source Traits add to the Prediction
of School Achievement?
CATTELL R B, STICE EF and KRISTY NF (1957)
A First Approximation to Nature-Nurture Ratios for Eleven Primary
Personality Factors in Objective Tests
J. Ab. & Soc. Psychol., 54, No. 2, March.
CATTELL RB and TSi7TIOKA B (1964) The Importance of Factor-Trueness and Validity versus Homogeneity and Orthogonality in Test Scales Educat. & Psychol. Meas., Vol. XXIV, No. l.
CATTELL R B, WAGNER A and CATrELL MD (1970) Adolescent Personality Structures in Q-data, checked in the High School Personality Questionnaire. Brit. J. Psychol., 61,1,39-54.
CATIELL RB and WARBURTON FW (1961) A Cross-cultural Comparison of Patterns of Extraversion and Anxiety Brit. J. Psychol., 52,1,3-15.
CHANDLER RE (1961) Two Additional Formulae for use with Supressor Variables Educ. & Psychol. Meas., Vol. XXI, No. 4.
CHILD D (1966) Personality and Social Status Brit. Journ. Soc. & Clin. Psychol., 5,196-199.
341
CHILD D (1969)
A Comparative Study of Personality, Intelligence and Social Class in a Technological University Brit. Journ. Educ. Psychol., 39,40-46.
COAN RW and CATTELL RB (1958)
Reproducible Personality Factors in Middle Childhood Journal of Clinical Psychology, 14,339-45, October.
COHEN J (1968) Multiple Regression as a General Data-Analytic System Psych. Bulletin, 70, No. G, 426-443.
CONGER AJ (1969) An Analysis fo Ghiselli's Moderator Variable Psychological Reports, 25,519-527.
CONGER AJ and JACKSON DN (1972) Suppressor Variables, Prediction and the Interpretation of Psychological Relationships Educ. & Psychol. Meas., 32,579-599.
COOLEY WW and LOHNES PR (1971) Multivariate Data Analysis Wiley, New York.
CRONBACH LJ (1951) Coefficient Alpha and the Internal Structure of Tests
Psychometrika, 16, No. 3, September.
CRONBACH LJ (1970)
Essentials of Psychological Testing Harper & Row International, New YorkL Evanston London. Third Edition.
CRONBACH LJ and GLESER G (1953) Assessing Similarity between Profiles Psychol. Bulletin, Vol. 50, No. 6.
DARLINGTON RB (1968) Multiple Regression in Psychological Research and Practice Psychol. Bulletin, Vol. 69, No. 3,161-182.
DEYOUNG GE (1972) Standards of Decision regarding Personality Factors in Questionnaires Canadian Journal of Behavioral Psychology.
342
DICKEN CF (1959) Simulated Patterns on the Edwards Personal Preference Schedule J. Applied Psychology, 43,372-378, December.
DUNETI'E M (1972)
Comments on Abrahams and Alf's 'Pratfalls' in Moderator Research J. Applied Psychology, 56,3,252-56.
EDWARDS AL and ABBOTT RD (1973)
Relationships between the EPI Scales and the 16PF, CPI and EPPS Scales Educ. & Psychol. Meas., 33,231-38.
ENTWISTLE NJ (1972) Students and their Academic Performance in Different Types of Institution In: 'Contemporary Problems in Higher Education' (Ed. Butcher HJ and Rudd E) London, McGraw-Hill.
ENTWISTLE NJ and CUNNINGHAM S (1968) Neuroticism and School Attainment: A Linear Relationship Brit. Journ. Educ. Psychol., 38,123-132.
EYSEN KHJ (1947) Dimensions of Personality London, Routledge & Kegan Paul.
EYSEN KHJ (1954)
The Psychology of Politics London, Routledge & Kegan Paul.
EYSENCK HJ (1960) The Structure of Human Personality 2nd Edition- M ethuen & Co., London.
EYSENCK HJ (1964) Eysenck on Vernon's 'Personality Assessment: A Critical Survey' (London, Methuen, 1964) Occupational Psychology, 38,2,115-117.
EYSENCK HJ (1972) Primaries or Second-Order Factors: A Critical Consideration of Cattell's 16PF Battery
Brit. Journ. Soc. & Clin. Psychol., II, 265-269.
EYSENCK HJ and EYSENCK SBG (1964) Eysenck Personality Inventory University of London Press, London.
343
EYSENCK HJ and EYSENCK SBG (1969) Petsonality Structure and Measurement London, Routledge & Kegan Paul.
EYSENCK HJ and EYSENCK SBG (1975) Eysenck Personality Questionnaire Hodder & Stoughton, London.
EYSENCK SBG and EYSENCK HJ (1972)
The Questionnaire Measurement of Psychoticism Psychol. Medicine, 2, No. 1,50-55, February.
FLEISHM. z1N JJ and FINE BJ (1971) Note on Cognitive Factors related to Factor B of the 16PF Psychological Reports, 29,1075-1077.
FORBES A R, DEXTER WR and COMFREY A (1974) A Cross-Cultural Comparison of certain Personality Factors Multivariate Behavioral Research, Vol. 9, October.
GHISELLI E (1972) Comment on the Use of Moderator Variables Journ. Applied Psychology, 56,3,270.
GHISELLI EE and SANDERS ET (1967) Moderating Heterascedasticity Educ. & Psychol. Meas., 27,581-590.
GOLDBERG LR (1970) Why Measure that Trait? An Historical Analysis of Personality Scales and Inventories Tutorial Address to Western Psychological Association, Los Angeles, April.
GOODWIN W, JOYCE J, ABRAMS A, BIAGGIO A and BIAGGIO L (1965) Note: Relationships between Matching Scales of the Cattell 16PF and the HSPQ. Symposium at National Council on Measurement in Education, Chicago, Illinois. February.
GORMLY G and EDELBERG W (1974) Validity in Personality Trait Attribution American Psychologist, March, 189-193.
GORSUCH RL (1970) A Comparison of Biquartimin, Maxplane, Promax and Varimax. Educ. & Psychol. Meas., 30,861-872.
GUERTIN WH (1965) Updating the Internal Criterion of Item Analysis Psychological Reports, 16,376-378. Sourhern Universities Press.
t
344
GUILFORD JP (1954) Psychometric Methods McGraw-Hill Book Co., New York. 2nd Edition
GUILFORD JP (1959) Personality McGraw-Hill Book Co., New York.
GUILFORD JP and ZIMMERMAN WS (1949-55) The Guilford-Zimmerman Temperament Survey Sheridan Psychological Services Inc., Orange, California.
GULLIKSEN H (1968) Methods for Determining Equivalence of Measures Psychological Bulletin, 70, No. 6,534-544.
HARTMAN BJ (1966) Personality Factors of the Cattell 16PF Test and Hypnotic Susceptibility Psychological Reports, 19,1337-1338.
HENRYSSON S (1962) The Relation between Factor Loadings and Biserial Correlations in Item Analysis Psychometrika, 27, No. 4, December.
HENRYSSON S (1963) Correction of Item-Total Correlations in Item Analysis Psychometrika, 28, No. 2, June.
HOWARD KI and FOREHAND G (1962) A Method for Correcting Item-Total Correlations for the Effect of Relevant Item Inclusion Educ. & Psychol. Meas., 22, No. 4.
HOWARTH E and BROWNE JA (1971) An Item Factor Analysis of the 16PF Personality, Vol. 2, No. 2.
HOWARTH E and BROWNE J (1972) An Item Factbr Analysis of the Eysenck Personality Inventory Brit. Journ. Soc. Clin. Psychol., 11,162-174.
HOWARTH E, BROWNE JA and MARCEAU R (1972) An Item Analysis of Cattell's 16PF Canadian Journ. Behav. Sci. /Rev. Canadian Sci_ Comp., 4(l).
345
JAHODA G (1970) A Cross-Cultural Perspective in Psychology The Advancement of Science, 27,1970-71.
JOHNSGARD KW and OGILVIE BC (1968) The Competitive Racing Driver: A Preliminary Report. Journ. Sports Medicine and Physical Fitness, 8,87-95.
KELVIN P R, LUCAS CV and OJHA AB (1965) The Relationship between Personality, Mental Health and Academic PýrforTance in University Students Brit. Journ. Soc. & Clin. Psychol., 4,244.
KLINE P (1966) Extraversion, Neuroticism and Performance among Ghanaian University Students Brit. Journ. Educ. Psychol.., 36,92-94.
LEVONIAN E (1961) Personality Measurement with Items selected from the. 16PF Educ. & Psychol. Meas., 21, No. 4.
LEVONIAN E (1961) Statistical Analysis of the 16PF Educ. & Psychol. Meas., 21, No. 3.
LEVY P (1973)
On the Relation between Test Theory and Psychology In 'New Approaches in Psychological Measurement' (Wiley, London) Ed. Kline P.
LORD FM and NOVICK MR (1968) Statistical Theories of Mental Test Scores Addison-Wesley Publishing Co.
LYNN R (1959) Two Personality Characteristics related to Academic Achievement Brit. Journ. Educ. Psychol., 29,213-216.
McKENNELL AC (1967) The Choice of Correlation Coefficients for Attitude Scale Construction. Research Bureau Ltd.
MANDRYK TR and SCHUERGER J (1974) Cross-Validation of the HSPQ as a Predictor for High School Grades Educ. & Psychol. Meas., 34,449-454.
346
MEEHL P (1945)
The Dynamics of 'Structureä'Personality Tests Journ. Clin. Psychol., 1,296-303.
MOLLENKOPF WG (1949)
Variation of the Standard Error of Measurement Psychometrika, 14, No. 3, September.
MORRIS J and MARTIN J (1972) Computer Personnel Selection - 2: Programmers National Computing Centre, Manchester.
MURRAY HA (1938) Explorations in Personality Oxford University Press.
NIE N, BENT D and HULL C (1970) Statistical Package for the Social Sciences McGraw-Hill, New York.
NUNNALLY JC (1967) Psychometric Theory McGraw-Hill, New York.
OOSTERHOF AC (1976) Similarity of Various Item Discrimination Indices Journ. Educational Measurement, 13, No. 2.
OVERALL JG and KLETT CJ (1972) Applied Multivariate Analysis McGraw-Hill, New York.
ROSENBERG M (1970) The Logical Status of Suppressor Variables
SAVAGE RD (1972) Personality Factors and Academic Performance In 'Personality and Educational Achievement' (Ed. Naylor F D) John Wiley & Sons Australasia Pty. Ltd., Sydney.
SAVILLE P (1972) The British Standardisation of the 16PF: Supplement of Norms, Forms A and B NFER Publishing Co., Windsor.
347
SAVILLE P (1973)
The Standardisation of an Adult Personality Inventory on the British Population Bulletin British Psychological Society, 26,25-29.
SAVILLE P and BLINKHORN S (1976) British Undergraduate Norms to the 16PF (Forms A and B) NFER Publishing Co., Windsor
SAVILLE P and BLINKHORN S (1976) British Undergraduate Norms to the 16PF (Forms C and D) NFER Publishing Co., Windsor.
SAVILLE P and BLINKHORN S (1976)
British Undergraduate Norms to the IPAT Anxiety Scale, Neuroticism Scale Questionnaire and the Eysenck Personality Inventory (Form A) NFER Publishing Co., Windsor.
SAVILLE P and FINLAYSON L (1973)
British Supplement to the High School Personality Questionnaire (Form A) Anglicised 1968/69 Edition. NFER Publishing Co., Windsor.
SCHAIE KW (1962) On the Equivalence of Questionnaire and Rating Data Psychological Reports, 10,521-522.
SCHEIER IH and CATTELL RB (1961) Neuroticism Scale Questionnaire IPAT, Champaign, Illinois.
SEALY AP and CATTELL RB (1966) Adolescent Personality Trends in Primary Factors measured on the 16PF and HSBQ Questionnaires through Ages 11 to 23 Brit. Journ. Soc. & Clin. Psychol., 5,172-184.
SIXTEEN PERSONALITY FACTOR QUESTIONNAIRE (16PF) Review: Buros 5th Mental Measurements Yearbook, Entry 112.
SIXTEEN PERSONALITY FACTOR QUESTIONNAIRE (16PF)
Review: Buros 6th Mental Measurements Yearbook, Entry 174.
SPIELBERGER C D, OORSUCH RL and LUSHENE R (1970) State-Trait Anxiety Inventory Consulting Psychologists Press, Palo Alto, California.
348
SUI VW (1953) The Cattell 16PF as a Prognosticator of Accident Susceptibility Proceedings of the Iowa Academy of Science, 60,558-561.
TATSUOKA MM (1972) Selected Topics in Advanced Statistics IPAT, Champaign, Illinois.
TATSUOKA MM and CATTELL RB (1970) Linear Equations for Estimating a Person's Occupational Adjustment, based on Information on Occupational Profiles Brit. Journ. Educ. Psychol., 40, Part 3, November.
TIMM V (1968) Reliabität und Faktonenstniktur von Cattell's 16PF - Test bei einer Deutschen Stichprobe Zeiteschrift fur experimentelle und angewandte psychologie, 15,354-373.
UNIVERSITY GRANTS COMMITTEE (UGC) Statistics in Education: Universities, Vol. 6.
H. M. S. O.
VAUGHAN D (1973) The Relative Methodological Soundness of Several Major Personality Factor Analyses Journ. Of Behav. Sci., No. 5.
VELICER WF (1972)
Comment on the General Inapplicability of Ghiselli's Moderator System for Two Predictors Journ. Applied Psychol., 56, No. 3,262-265.
VELICER WF (1972) The Moderator Variable viewed as Heterogeneous Regression Journ. Applied Psychol., 56, No. 3,266-269.
VERNON PE (1964) Personality Assessment Methuen & Co., London.
VON SIEGFRIED GREIF (1970) Untersuchungen zur deutschen Übersetzung des 16PF - Fragebagens Investigation of the German Translation of the 16PF Questionnaire. Psychologische Beitrage, 12,186-213 (Berlin)
349
WARBURTON F W, BUTCHER HJ and FORREST GM (1963)
Predicting Student Performance in a University Department of Education Brit. Journ. Educ. Psychol., 33,68-79.
WILSON G (1975) Wilson-Patterson Attitude Inventory, Manual. NFER Publishing Co., Windsor.
WOLF R (1967) Evaluation of Several Formulae for Correction of Item-Total Correlations in Item Analysis Journ. Educational Measurement, 4, No. 1, Spring.
ZWEIG F (1964) The Student in the Age of Anxiety Free Press (MacMillan Co. ), New York.
350
SDCTION 7:
APPENDICES
APPENDIX CONTENTS
APPENDIX 6: N-STEH NORM TABLES
Index of Norm Tables Tables 1- 81
APPENDIX 7: DESCRIPTIVE STATISTICS: PRIMARY SOURCE TRAITS BY SCCIAL CLASS
Tables 82 - 90
APPENDIX 8: DESCRIPTIVE STATISTICS: PRIMARY SOURCE TRAITS BY AGE
Tables 91- 103
APPENDIX 10: ALTERNATE FORM RELIABILITIES OF THE PRIMARY SOURCE TRAITS BY SEX AGE AND SOCIAL CLASS
Tables 104 - 119
APPENDIX 12: ITEM ANALYSIS DATA ON THE HSPQ
Tables 120 - 134
Pa es
3
4 5-85
86
87-95
96
97-109
110
111-126
127
128-142
I
2
APPENDIX 6'
N- STENN0RMT ABLE S
Please note that Appendix numbering cross references with the main text.
3
z 010 1 HÖ
a a
LO Cý4 CO N M tO l- CA
N Co [- CO La le C) N V-1 0 U) U V-1 CV Cl d' U) to l- Co U)
w N M Cf) N U d ' t to
p n
U)
ä (D Ln N rli (IN c-4 M d' Ln 4.0 tto N
o t ßf1 d' M CV c4. C) CF% ýp to c-I N M c' U) in lD N
dt N
to v CV M ý
to h
141
[14 M CM Ir-i CD (n Q
Uj vi CV c') M LI) to N
(ý M 7 C
Ch
M N A N
Lj v-+ CV (NI Il) tO N
d' N
ý N le vi r -I M Ln h
H
W W Ctl ti, Q OD + Q CO t Q fA f "0 0 ý+ w a Q Q
U)
)
r4
C-14
Ln 10 N It
N
j m4
I
11 CA
O Ol
* N
lD ON
ý-1
-4 Q1
M
H N
.4
00 lfl 4
C7 d1
M
M d'
in
-I N
M
N lo
M
m fit'
M
o 01
N
00 -'
mot'
ka _4
m
00 W
M
to M
M
in . -1
tf)
Lý N C'
CA
.0 M
I-
lo ri
In
01 d'
m
0 O
l0
N I
O
d' l
d'
r4 . --1
m
N 7
Ol
-tv r4
N
m 71
a%
co Ln
O
in N
r-4
N Ul
O
-4 r1
-4
00 c
N
U
Q'i W U W W U` x. H a ý" Z. Q 1-1 a
N a
M Oý
cN Ot W
O . -I
O fV
1 I'D r-1
m r-4 1 N H
to N I
M N
ko N
1 (V (V
W N
1 in N
Co tV I
61
lD N
I LA (V
O tV
I CO . -a
O N 1
N . -1
110 N I
O N
O N I
lß -i
l0 N 1
ý-4 N
O N 1
-i
O N 1
CO r-1
O N I
CO . -I
lfl N I
N N
O H
oý
LA . -1 I
dr f-I
-4 . -1
(V (V I
1-1 - N
. -4 (4
ý: p N
co 11 I
N -1
-W N
I M N
CN r. 4
I
1-I
l0 -4 I
in --1
Ol r-{ I
CO r-4
in e-i
O N I
co -
co r-. 4 I
N . -I
N 1-1 I
l0 ri
N i-1
r-1 N 1
O N
a1
co O . -4
O N
O N
OÖ rl
ß'V1 N
r1 i N
l0 r-i
to 1-1
cV N
r4 N
LA . -4 M
. -4
. -4 l-. -4
ýo r-4
d' . -4 M
.4
N r-I
tÖ r-i
ko ri*
in -4
l0 --i 1
to º1
al , -{
CO i-4
00
N
rl ri
Ch r-1
CO I
r-f
iN r-1 I
l0 r-4
O (N
I O1 r -l
et' rl 1
N r-4
O N 1
CO r-I
N r-1
rn . -! I
N r-1
in r-1 I
. -1
N '-1 1
-I rl
in '-4 I
M r-4
in 1
i qll 1-4
T . -I
II
e--1
er ri I
m . -d
N _4 I
l0 -1
h
O r-I co
N ri I
in ri V'
CO . --1
r- I
r-1 -4
C- r-1 L1
ý-4 r"4
ä1
ý-4 r-i
Ö
M r -q
O r-4
N r-I
ri
M -
N
N r-4
r-4
N r-1
Ln -1 I
N to
I V
fM
I N
tG
1 d4
O I
Q1
c4'
I N r-I
I N C6
fV
1 . -1
O
1 01
-4 -4
O I
C1
. -4 1
O
N
I H
U)
Ul t+
'd' co 1
N
ýO tß'1 -4 1
r-i r4 I
OO
M r-4 1
CO 1
N
.-4 ri I
01
ko IN I
lfl I C1
co 1 N
00 I
N
'
1 01
QO 1
tN
ON 1
CO C1O d'
M ko LA
H i
O
ON
00
H .4
m
l0
LA l0
to
-, T
in CO if
in
tß I
co \p
Ln
;
lD
i M
N in
vl'
-w 01
C6
IN
to
CO C_
d' N
to d'
M
l")
N
W, <f' M
c
l-
lÖ
' Ln M N N
" M
(DI M
ý N
ö Ln
b tD
a H
ö CA
Al N
ö N
cl, to
Cb C
CID . -t
C, ý to
0 N
0 V
0 r-I
0 a
w a tja U w w 0 W H a z o ä Öº Cl Oý
a
w
}
w
44
wý ä2
z
o, Un Un OD (n co 0 ' Ir w c, m v' M co o A (NJ O IN VI 00 M tf) t. () u) (N a) Q\ H co A
" C7
" M
" N
" M
" d'
" a'
" M
" In
" M
" M
" M
" N
" M
" C. )
" M M"
" d'
" U)
U) '. o U) 0 M N V' VI ' d' t0 CYl VI O tt' ýi' Z c+ LO C N VI 0 C O S tO 0) M '. o t O N z 0
pl N N -I
M VI
M VA
N -I
(V -4
OA co N VI
ON O I
O i -1
N r
N i c c c C r c v
U 4 co U W ka 0 r H a Fam. " C7 Q
W G.
O M ko to tD O 'O O O %0 O '. o O O O '. D N VI N N N N N N N CV CV CV N CV (11 N 1 N 4 N A
Ol 4
cl, [" d 1
ý ! 1 t N
1 4 i N N N cI N ci 1 N
tD 4
9 p i (P
9 N s c r c c
c') (V c'1 lD kD 0) OD VI 01
-4 V
N I VI
(V U) (0 VI
I -I t[ -I 9 c C N
c N C'4 VI
U) VI VI
r c VI
c 0 N
V' T-4 O 0 N V ý' d' t0 to O'. VI N N N VI N c1 ri VI %-1 T-I
O t A CA ý
Azr L o 'ID 1
IQ 4
00 a
VI cý VI A VI c c t _ r
c
+ 1 V-ý c VI ß
VI VI
tý
VI '
c
tD 4
c
r+
ri
ci to
V N
VI A
r4 -1 c 4
VI!
,4
ci n
ri VI ti VI Ti ri I VI tI ci cI ci %-1 VI L VI
(sý to N M to O O CN r+ N (V rM [t' W rl VI cI VI VI VI ri rI c-1 c4 1-1 c-I
i t cm 4 r r T-4 r4 r4 r4 ri Tq V-1 ri r4 T ri U)
Vý Ln M CV N N O O ri cm N C1 r4 ca ri VI co VI ?I ci ri VI ca
ý 4 qqrAi It (* r! A C', l A A J
1 r4 1 rq ' ri a l C V X-4 T-4
Tq O 1,
OO OO
f
OO
I ý W 0 o0 Cp
m 1 40 N Ö' Oi o Cl) O N O
a VI
VI co I- V) U) N O) 0 N M In tf'1 { Go i co I '. 4 Ol I Ui lp I M
CO W
t- Ln r- q to r- U) N d' d' 01 1 1 j M M rt U) v sf1 N ( N
in tn ý ä ä I ä ä ä ä ä ä ä ä ä ä ý a a
w 4 0o U w w cD x H a z o cr ä ör U
w
en w
H
I
s
ri
z
O A ri O tV Cr) r c') -' vi d' N Ul N LO M ri A
" [11
" N
" N
" d'
" d'
" '
" Cr)
" Wi
" 1
" Cr)
" Cr) Cr)
" '1
" N"
" Cr) N"
" In
" VI
N c-I tN O\ to Cr) Cr) (V to N M N [- In co [- Z c+) c+ 0 O to Cr) cl co PN ri tr CD V N z QQQ
ON N Ln N V Cr) Cr) 01 coo N O O al %- M VA
R o w u w w 0 x H a x z O CY är Or ä
O c r) to tO 'O W O O '. O O '0 O O ko (11 c-4 " f-V N N N N CV (V N (V N N
P A N + 4 N ' ON Q' L-,
o c4 L-, ßl
t ' N N L c-I ri N N CV N r4
t c-I (V ri rI
k c-I
a % c-I N
0 cl c 0 N N N -I t cl c-l
ý (V
i ci 1 1 1 I Cr) N 1 Ln
d' 4
i " ol 1 4 1 1 I r
N cl c
c-1 C vi
C ýp -1 N c
r4 c ýý C\l (V
;; r l
CY) -4 4 i -I
-4
N
d N c
W cl d r
A T
N c
N c
cl cl c c
1 T '9 (\I -I t-l CV X-4 -1 i i
0C ý -ý c c r c c
N (11 lD tf 0) CV ci l0 d' d' N
E- 0 1 o ý N
i d 1 f
i l` c-I
CQ C+
t c-I c-I 0
c c-+ c-i
u c-+ c-1
C ) vi
ý k vi
N t4' f") t1) l0 '3' C11 (V d' to c' 14 H c c-ý + c c-ý cý c-ý c+ rº c+ W ö lD O W L N N t
I ll ci 4 N 10
t ' ' c 0 U ri 0 ri ri %-I C-I ri ri ri c-i t-I
Q
U
V] m c-I Cl cl j ri CO ri ý
r, cl C'l '
ý 1 1 T-4 tn OD 0 co a c- ý 41
d' r 'O tq C), % ON cl
f L- w
CO N op
1
i C0
d' OD CO O o
j C r4
p
I d N 11 co 1 1 1 , ID I- l
ci 1 o l
N 1 Q'
N N L O m cm .
N Cn Ný lf1 C") tt1 f') (V c-I ko d' ca <r d' tN N
`ý d d d d d d d d d d d d d d d d `ý
Q u w w 0 x H a z z O ör c` cl ä u w
z U H e-!
O a
Xa HE
Hö d, ýw
$! a
" Q') N W t- co ON M N c-L -t; f' O '4 t u) ' " A 0 N co a N r. ' Ln N M w 0) a) co N O {- A M N M M V' M U) 1 M M N Q' N (") C4 4 Cl)
0 tN r S H to tN c+ Ln W u-) 0 c1 O tV co ä '; r 0 ON ýO co 1o N O O O t- d' k0 %Zv -zz3' Ci 0 co N ci O 0) c'4 M N
Z
G 7 c-t c4 r4 c-4 vl rý ý
.
a a u < m u w w co x H a .: Z 0 oar a cy a
0 M (. o %D W W O O "O O ' to O O O 10 N ri N N N N N N N N N N N N N 0
tÖ N N d N Ö fl d { + W ý
c-q c-4 (V N CV N t 11
l c-1 (V
pp r N -I
pp ß -1
O 1 c c c N
°' " -I
,I { ýp
om, N -4
N d °'
C \l ýp sa cý N T
ci i i ýa ci c
c c N
0 ß 4 h T Vi c -4 T4 N c 4 cc'4 c i -1 T co 1 p 1
i ' A % I co Cl CA I,
IH c N ( IH
(N W c-4 -1
T-4 T-i ,. . V C
(V rI
u-i ¶4
LI) ¶4
N 0
Ol ¶4
N r4
¶4 ¶4
I c4
Ul ¶4
N r-)
In ¶4
l0 ¶4
N ¶4
¶4 .ý ci ci O -+ h + i n
c-4 ci ¶4 c-1 ¶4 ci c4 . c i C"4 M ¶4 c
r1 ¶4 u ri i "
co N . -4
Cl ¶4
N ¶4
u) ¶4
N 'r-+
O r+ 0) ¶4
N ¶4
Cl ¶4
Nr t-+
w
r-, I
Cý 11, 14 J,
Y ci
r-4 r4 TA r4 ri co ( )
IH T"I
ON Tq ri IH -4
co tO ö (A
Ö °D r- I I ' o o r a o c ri +
d' cý vl N T-1 ON co
chi t" I 1 O j °1 tl-
' t- 0 N T-1 4
O d'
ci u co
e) 10 C
O
! ýi
tc 9 trZ CO {
co tr) '. D
c, OO
1 t") l0 Ct C
In Ln Ln 00 it) co N 1 M QO 1 1 I 1 M 1.0 l0 d' N to [ tV
N N N M N l0 N fV N U) t(1 Cl to to N
d d d d d d d d d d d d d d d d
a a w u W w o x H z Cr cf o
w ý,
ýý d
Hý
0 qw N
. -1 c N v N O I- I- 0 tO to al N r-1 l0 01 N ri CO . -i ( N Ul O N LO Co Co O M CO o- (
M N M c}+ cyv m to l'7 M M N M N mT: cN : Y' (j
r- D N -' I- M qr 1- O W I- CO CO (0 to cI' C' N l0 I- m bt M tf) N tI1 l0 N M N H f
0i O Ul 0 Ol CO N 0) Co N H 0) 00 N cf' r-4 r-1 ri ri H H H r-1 H H
ýC cn U W C=+ C7 W H ý7 H 2 O Oº ä ä Oi
O CO tD lD LO O tO O O tO O tp O O O to o N . -I N N N N N N N N N N N N N N O 4 I I 1 1 I 1 I 1 I I I I 1 I . -i Iý q M Co Cl '-' tO l. 0 O co m to M --I H r-1 N N N N ---4 H N H H . -i 1-1 N
lO O N 01 ON 01 M 10 10 On r- co to co Cl O H r-I N H H . -i N H r-H H r-1 H . -1 H H N
O1 n I I I v ö d rn L N ý co to t ý . -H ri .i N H H ri r-1 . -I r-H
It: N -q to to 00 O d' M I- Lo i-n CO I- co co A 01 N H r -I H N '-1 H - 1-4 H H H r-i H
Co I I I I I I I I I I I I Co CO 0 10 C- o co N l0 I; T N l0 Iý r-I N 14 r-1 - H ri H r1 -1 H i-1
N ON V' CO I- 00 N r-1 lf1 -j M -I In r- to 1 00 H r-1 H r-I 1-4 r-H H ri H H r-i . -H . -i H
n 1 M - M N p ý 0 0 N %. D 110 l ' H '-i 1-1 r-1 H H H H H H 8
-4 I- N H tO 10 O O ' N r-i o d' to ch '-1 N H H r-1 . -1 H H r-1 H H r-4 H r-1 H , -i
I I I 1 I I O l0 -4 O N ON m O 0) M N r-4 H ri H r-1 1.4 '--1 H H
Ln a C rI N H IT 01 to H r-4 Oa r-4 .1 01 CO 1 . --I ON co H r-1 "-1
i-n I 1 I I I , I I 1 6 1 I f , t. f) O1 00 H M CO [ H C - 1 M H
CO H -4 N CO t. f) N N H co N tO [N IN
ý -i CO
I I 1'' I I I I 1 I I I I d' N N IN tp r-1 N o1 O l0 to 0. lfl
H O H to -'}' r-i lO tO "--4 to to Y C co U1 10 oo -4 ü7
M I I I I I I I I I M to O t0 ('1 M ßf2 10 " I- "^
0 ,* CO (A d' N co d, C M to r- CO 'd' CO
N I to ( 1 N M d) C 1 %D N l0 a1 N
N N I, N H N M M N 'V Ln -I CO tn CO H H 1 I 1 I. i 1 1 1 I I I I I H O O O o o O o O o 0 o O O o
r-q N M o ý pq 0 W w t, x H º-a 2 0 C 01 a a a
i
01 M N d' Ol 01 to O' M to N N I, c'1 t, m "ý O O cYý . -1 a1 to C l11 '-I N N O Lf) to N Ä
M N -zr T Ln M LO M M M M M M CO 'n Cl)
. -1 - 'V 00 N r-1 II) cal ICT O lD l0 OD u1 00 cM O N tD to 'r r-i O to (N 61 Lfl O d1 r-1 N
r) O CO to N .1 M '7' Ol O M r-I -4 . -4 I-I r4 r-1 -4
O M Ol _4 M -4 r- i 4 -4
m u w w 0 x H a Z z 0 1Y äN c+ ä w w
o M to to to C) o CD to o to o 0 0 to O N . -I N (N N N N N N N (N (N N N N N O -4 1 I I I I I I I I 1 I 1 1 -4 N N M N d' LO Co to . -. 4 Co d1 tN m N
r. 4 r-1 N N N N "-1 r-t N '-1 H -a -I N
to (N -4 M 01 lqfl N 1; 31 O N CO to CO M fN . -1 N N N r-1 N - 4 N r-1 r-I N
I 1 M 0 Ol Ln O " to t to Ö
N 'N N N
.' r-I ri 01 O CO N l0 m ßl Lf) LI) Vr N Co 01 r1 r1 N r1 N rl N i-1 r-i . --1 ri i4 p-1 r-I .4 . -I co I I I I I I I I I I 1 I QO O Co N . -1 IV CO d' V' c'ß'1 to tý N
(N . -1 r-I N r-I r-4 . --I . -1 r-I r1 . -I . --I
M O ß1 1- to N O M N I- M co N u1 to to . -1 r-I r-1 . -4 -4 N ri r-1
N I 1 1 1 I 1 I I 1 I 1 I i N N CO tf') to CO f N O to .1 . -1 Cý ý ri r-I r-i -4 r-4 r-1 r-I r-4 ri r-I ri
M U) 1- H v) N O o M to ON rl ý4 e-1 -I r-1 rl 0) r--I r-l . -I i-4 r-1 rl O I I 1 1 1 1 I I I ! 1 1 64 q. Ln q CO
1
U) O Lff N r-i M d' M O N M . -a . -i 00 . -i i1 r-I rl . -ß Ol N -4 r-i CAD 01 r-1 . -s . -i in I I 1 I 1 I 1 I I I I I u1 ý N 01 tý C O CO -4 r -I -4 - r -I
- U)
1., -1 -4 ý4 C14 CD CO NN -a CO IN tO l CO tO N I
CO N Q N CO 1 I N 1 [
Ö k. 0 L ) to 01 Co r-
V-4 r-1 1-i H
^ý O tO to '--I CO to 01 I- u1 d' ' -4 N mot' if co 01 l0 CO I M
If) O l0 tf) C'- to r 17 N 00 ":
v
N d' 1 0 Ln W to L'1 M M C) Ul M 111 tD N M N 1 Oý
N "r Jr N 1 N N l
M N er N M M N r-1 N W N ey" qw , -f 1 I t 1 O
I O
1 O
1 O
I t 1 1 H O O O O O O O O O O
a a r-f N M ý tý0 Cý W w C7 pC H z O E a a a o a w w
10
x O
z
U d1
3
" c-4 o% ci C) cI to d' 'H ýp N N 01 Ol 01 ri cv) " (+') r+ CO L! ) LO N 00 LI) d to c-+ p I tT LO to 0 0 «) N M ;, '' 111 C') to M c4 M M Q' M C') C) LI) U)
N (n C) C' ß1 N M -l (V 00 N C-) (V O\ r N tD d' c-i co c14 c-i N U) N O 00 N w tQ 00 ý a%
" [N
" LI)
" fV
" N
" N
" M
" O
" co
" m
" Q
" o%
" 0
" ci
" (4
" ¶4
q W
¶4 c'1 ¶4 ¶ ¶4 ci ¶4 ci ¶4 C'4 ¶4
V
Q ¢ m cý w w o x H a z o `iu- ü ä
C
w w
o « L0 /o w O O o 0 w o to o o to N 1-4 N (V CV N N CV N (V N CV N N lV O O ý
i
ý
N N ¶4 <V
¶ ¶4 N H 9
1 9 all N O
r 1 1 r
cn V-4 41 ; 1, W t-D N ¶ C- CA (V N' LV CV r-i i c t -4 ¶4 ¶4
ON Ln ¶4 ri -4 N d N ýp cn N `
d' Ol o 'Pi A m O N a% c
CV CV (V c
N c4 ci c-4 t c-1
' 4
I ¶4
ri tV %-1 CV N c4 c9 0 ¶ IH ¶
cý d
S J
a O c i 9 " o 4 t + 1 co r-4 CV ¶4
\ ¶4 c (V c-l
c t-'l c t
c-4 ¶4 ' c pQ c c
I D ) r-I
4 d ¶4
N A "i
4 n )
IH V4 .) c'{
l ¶4
l C4 c'i
c %-I ¶4 T"4 ¶4
c ¶4
I c4
f ¶4
W N v Ln dq 010 to ca N NN d' W a ¶ ¶4 ¶ r+ ¶ ¶ ¶ ¶+ ¶ ¶
-4 A A 4 4 f o X W ' ' 4 1 ¶ % q ¶ -I ¶ r % ¶ ¶ L
E-f E-4 U) N N N 0 O N N N (f]
0) ¶4 ci c4 r4 CO T-i %-I 01 ¶4 ¶ c4 c-4 N Gp t- .4 OD t. 1 I C\l GI cc) C1 m O p
n ý- 1 t0 ý d,
; Oi e i '
CO l 0 co CI 0k t- t0 to
8 [ý lp N
1 O ¶4 N ý O LU CO
r- rl- %0 Ln C\ tn 'D r- (71 kD tý to ttl IH
5 ä'
-4 O ' n iD p ) M D U a u
O N
C d' M m m l'ß'1 IN u'1 M d' 4 1 A N
tf
N M lt) N . 41 N N (V to ' N M N) v N
`i d d d d d d d d d d d d d d d
cm 4 as v w w x º-ý a z o Or (Y en ä orr w w
11
rH-t Ö
aý w ä cý
z
O v 0, O N c-I N in rý N ch
Ö ui O rn "
Ct) "
(fit "
N "
Cl "
d "
C "
tý ) "
to "
M "
c ý) "
M "
M "
d' "
M M "
C') "
d' "
1;
In Un Ln un (. c-i in 0 c-+ q O cn c1 to q co 0 C1 M m to L! 1 N CC) r- iý l1) in C') U) N C'
ßý l0 c 0 cI c1 cI cý c c'I c -l t- c-ý cý
O Q E-i o 4 Co U W G+ O x H ý7 ? 0 C7 C7 0 Cl U
O cn tO ýD 'O O O in O t0 O O O w cm V4 N N N N N N N N N N N N N A N N N M ON M
N t 0 I i d 1 1 Q j 0
TA "1 N N c N N c-
. c-i ri
[ I N
ci i
ý ( N r c c c
IH C\j (3) w r- VI N
{ i I N l I
N Lo -i -1 ' d
c T C c-{
O N
pp ca c-I
T c-{
t l0 -{
c s c N
(IN c-I "V d' N U) O\ N
co m i ý d I t-
i d
ci m I .4 in
+ I 4
-i ö CO
c N c-I
a0 cI
c N c-4 c-I c c U) cI vi
c c-4
9
N Q1 l0 N l0 a% N N to fit' C') to N Oý c c+ c+ c1 ci c c c-I VA c-I c-I
h h h h -ý ýý lý
i c p
i -q H c i c i I c
ri I i c
c % T T r
W Q N d' In d' lD O 0 M N N c'') d' [N W 1% 't. 4 c- c-I c4 ci t'i %-4 ci c-I VI c-I ri i7ý O t0 I N N N 4 ß1
41 N
'A A
-It N O l0
U V!
l i H 4 i -1 a 4
% H
t q q q c c r c c T I c r r C
E-4 E-4 C') O 0 O T-4 T-4 vl
ca ri N-4 c4 T14 co ca T-4 ri T-4 x-4 LO co tD '
O1 d N
i c4
O , ) O c
() ? O cý l ol (A r' a% d' t- Ln cV Oý CC) c0 OA 00 CD llýr
CD 00
I
`~' A "Zil ö ri IL
I ýý .4 N
r in ' cý ý I. co
e ()N l0 Cl'. U) It W to d' 1ý N N I
tni N M
A 'A t1 alt
N N
cc) N
N ch h d' m d' m IV N d' 'd' N d' tM l0 r4
r' d d d d d d d d d d d d d d d d `i
ü < m cý w w o x H a z z o a cN1 c CY U
w w
12
a
in O O O O OD O\ lqt' lD lD to O in U) 00 M C O O O ( 'O (N CV 1 CO --I . -i . -I , --i Q
N N d' C -V M Ln m M M N mot' m r') M 'r En
CO N CO ý. O Ql 00 O Ln in cr (N N f'ß'1 C1 dý Q. Ln ßl M e1 tý dý m Lri 01 M . -i v l0 N
Ol ýQ d' N N N ON CO -t O O "-i N N
< W U W r%4 Ci x H a z z O a ä ä a b.
O cn I'D 'o lo O '0 O O l0 O w CD O CD %Q O N . --I N N N N N N (N N N N N N N N O r4 I I I I I I I I I I 1 I 1 1 1
. l M N m 01 Qý N l0 CO N O N 00 01 . -1 r1 _4 N N N . -t N '-1 --I r-I rl N . -1 r-I -1 N
t. l) O N -4 N co Cfl lO in N 10 01 o N OD 0 "-4 N N N . -1 N r-I -q rl ri ri r--I ri N
01 I i i ; r ß
1 '- N -I
N N -4 . r -1
M O N O N H d' M lO in N in lO N m "-i d1 N i-i N -i N . --I e-I r-I 1.1 1-4 . --I r-1 . -i r-1 I O , M ! 1 00
l C M t1 lÖ J'
in CO . -i r-1 -1 r-I -4 rl . --I -4 rd
N co %o N "o co N N d' -V In M "p r-- i--4 co r-I ri r-I rl rI r-1 r-I r-4 r-1 r-I -4 . --I r-4 -1 N 1 I i I I I 1 1 I 1 1 1 I 1 1 N '-1 N ýi' in in lfl ri r. 4 M M -V N M in l4 r-1 r-1 r4 .-I ri .1 rl ri .. 4 . -1 p- I r-4 .ý r-1 '-1
O l0 m -'r d' Hfl O O N N m -4 -V to rl N ri ý-I r-1 , --I 114 r-1 r-1 ri . r1 ri r-i r1 ri
1 1 1 I I I I I I I /o In N M f7 M p> - N M M
Ct] rl r-1 ri r1 rl H ý-i . -I e-1 [zý
8 ß8y
eT 1 N N N r-1 ý4 a -4 C, 4 z 4 r-4 -q ý4 -4 m 00 O 1
z m u 1 1 I I I J
I 1 l
I in Cý7
CO ()) O O r4 C. O o m O U) r-4 r-4 ýl r-4 r-4
C4 C) -4 N ºn H 01 ON O\ 1ý N ON m 01 co -1 r-4 M ý ä O
. 1 1 C w OD N N Ö1
Cl N . -1 r-i
1 l0 Q' OD N CO N %. D to 00 N t0 lD CO 61 l0
("" I I I I 1 1 I 1 1 1 1 I M 61 N tD N %O in ý" N Ln r- CO in
to en 0c) ko to t0 Ln 41 M %D t0 V in t0 N N d . 4 J ' ' N
' [ r L M
t11 ý1
en N to i M d' M ; N in in c4 V' C' l0 N r-I b b b b b b
O Cl) l l l b O O O C ) C ) O C )
a O N M
ý as v w w cý x 1--1 a z o a Cl a C
UU
13
J
0
H
y
I I IV M N z
N tf1
I
lJ M --! ' N 00 01 M C' N C' tip CV "-i M u lD CO N O -: 3' O lo cM N CO 01 co N -i O Q M r-I
4 4 4 ('n1 LA M N M N
4 M M Cpl Lf1
O CD d' O lO M "Y" eM m Ln I- " . -i LA M lp 1ý M N N O N 01 . -1 fß'1 CO N M LA O 00 1, Ln M lD _4 'r l- m -i Ol m Ol 00 O
-4 rI ri rý ri r- I r-1 rl
N+ ýi
4 W U fý C7 Z. H ºa Z Z O OI ä Izz, a
Lý W
C) M to lD 'O O I'D O O C) l0 C) O O l0 O N -1 N N N (N N N N N N N N N N N O -4 1 1 1 I I 1 1 I I I I 1 1 I I r1 t0 N N Ln N to ', q LA 4 1 Ln cl> N Ul C O M
r-I r-4 N N N - N r-E r-t N r-4 N N
LA -I M -i d' lfl -v t! 1 O mot' C LO N N r'4 1-4 N N N N r-4 r1 (14 r-I r-4 r-I 1-4 1-1 N
I M ! 1 CO 00 L) -4 N N . -1 1-4 -4 -4 N
. -T O N m M LA N M C1 N M N M l0 O r-1 -4 N -4 N rl N ri r-1 . --I -4 . -i r4 r-i r-4 (N
CO 1 I I 1 I I I I I 1 CO M O CO N N tf) to N LA O r-1 N r-4 N N ri -4 r-I -4 r1 r-{
N O%% N r-I d' O -4 N N "W M -4 -P 00 , -i Ol ri -4 (V rl N r-1 ri . -i r-1 r4 .i ri '-4 . -I
r 1 I 1 I I 1 1 I I I I I 1 1 N 1 '- CO l0 C D U) 03 O I - . M M N O M lo
r-i -4 r-I N -4 -4 . -i -4 -4 r-i -1 _I r-4 ri
O N LA 01 N N 0 N r-I N -1 N L) Ui rl CO ri r-4 r-I r4 r4 C\ . -1 rl '-. i r-i r-4 Ol r4 -4
ýO I I I I I I I I I I 1 to 01 Ln CA N Ln CO . -t O '-i O r-4 d'
. -I . -4 ri 1-1 r-i -4 . -I . -1 . -1 r- (
ß N %D r-4 -4 CD co N r1 . -4 -4 rI t- 01 01
-4 tT CO
ý r-4
ý I I 1 to r
ý CO CO
CO N a r4 r-I r-1 . -1 r-4 r. 4
Kr CD C) ýO .4
; '-4 m l0 N O N tN l0 00 i
CM if1 O M 01 LA
CO Ln tO N
CO
ri . -1
. -i N ,w LA r-i ßl CO Co C LO N t) ý, - Lp LA tO r-
co 1 ! I i { { t i E I O\ O3 O 1ý t0 l0 LA M LA V+ to
r4
M co N 01 tG LA M lA tn N -v M N 1 i I N
C C A N
C. 4 d' %D i d' M i d' M M M N M M
b b ; O ö
0 o O O a a a ' O O I a
D O
N M "Cr 0 RC m U W W C7 x H 0 H 2 O C Cl d Of O Oi
w
14
0
N
d z
cn = c12 N
Ha O
a ätw
{ý+ cr-i
l
91 4
" c{ t! ) l0 c-1 N l0 co M ul m m l0 c- M ON " R o0 O O o% h CO %t' N OD oa h to o C CY) k. 0
. . . . " . . . . . . . . " . " U) M N d' m T N to M N M N Q' N M M 4 U)
ON N a0 c-+ co d' M c-1 0 tN q in 0 o C t0 Z M h h ; f' l0 '. O M c-+ N lD O co N N N Q FZ
[ý 00 h to in 4
r+ in co C\ O O C' 0'. 0) r1 N ci r-i c- T-q c1 ci cI ri c-I
a 8 d
U ¢ M U W ý, cý x i ,a z z 0 ä a w w
0 in LD to 1.0 O '-O O O 4-D 0 ' O O O N t-1 N N " (N N N CV N N N CV N N N
O 1 4 1 4 I h
4 1 tp
1 U1
I A [i
I Lý t I 0 0
-4 N rI N
c CV N H -H N c-1 %-i N T-1 T-4 r4 ri
pp ci
N
N ri ri c-i r1 CV cr% U)
- TA M N ! M
N C' -1
I LO -1
' 1 I h i
' 1 r ri N
N c OQ t
A Ln r a
V
N Q) cl w N in h C" [' 01 c-ý N c-I CV Ti N T-i cI c-I r-4 N 0 1 ) ) I
i l N c4 N A WA I
N co
r-I O pý O t C v U c T vi CV c-I N ci c-I c-I c{ rI c-I cI
C\l a% r- ý N ý c-I Cl
c %-I T-4 TA c -I c -I r-i
h 1 A
t`
A i I T
T-i X-4 T-4 c-1 ri cI T c T
0 h in lS in tf) N t-I O M ct'
-H c-i c cI tl 01 c-+ c-I c cý c-I a
-I -4 Vi T T
r EA E-4 (n Un c+) d' M O O U)
OD ci rq IH c c-I - h
ON ci c-i ON O-, I co
c'l in N
V' N (fit c -I c CO 0) 0) N co 0
c-I c-I r4 ri cI
en r4 T4 '.
O cl c 0 co %D h o )
I W i ' t O 0a W ui
t- h L ID 01
i CO u) c
d o n
(31 C- m Go M Ij IS) t l I
U N kr,
W M
U1 4
I U) ch
r4 co CO
I'D N N ii' ) 1
4 j h ( M IV U) I') CV d' M
LA d' N
CD
c-i ('n h U') to to N (V M V d' r-4 M lV d' M
`ý d d d d d d d d d d d d d d d d `ý
a ap a co ü w w O x H a z o
ä Or c 0
4
w w
15
z Ly w
LA HZ c)
C\l wD 0
ä cw,
Cl) 94
H°4ä
" iD L' of c+ tN O to m in N N in G' O O c-+ " n (O o co 0) t-o o Co co m M cs, M r1 N a " a " " " " " " " " " e " " " " s "
lh c-4 Q' M d' N to M N M N v C 1 M tf) U)
M N In M l0 ; i, 0 CO M 00 tf : Cr, c-) C') N M C ýn ýn ýý N rn M O N c rý L in C: -N N `c
I (+, ] C) [N Le) M M N C') CO 61 C O C1 CO CO CV 14
o < m U. w w u x H 0 (i CY Ci ä v
w
o M CO 0 Co 0 0 0 to 0 0 o w N cý N N (V CV N [V N (V N CV N N N (V 0 c-)
i to s-I
I N H N
1 c N CV
I Ol ci CV
1 CO c-1
I CO ý
1 00
1 Vý c-)
I Q CV
I LO c-i
I CO c-)
I Cv c1
I N N
O Ni
d N [N tD O. ) t. () ci cý N 0 T-i c-i c-I N
m i 1+ s-I
c-+ ca
() N
O N
1 14 CV
co H
M CV
v' vi
' c-q
I lý c-
I LO T-i
I tO N-4
d' c-I
I d' v4
I 1 J
(V N
(n c-I
0 CV
N c-I
N N
cr) N-1
C') c-
if) c-1
(n c-I
0) ci
co N V-4
O c')
1 ci (V
I a ci
1 O) ci
I to c-I
1 s-1 N
1 N ; -I
1 N c-I
In c')
d' c-I
I 4,
T-i
I N 1H
M c')
N c-i
1 p) c-i
CO
cý [-i
I O c-+
C
0 CV
I CO c+
N cl
1 )n c-I
CO c-I
+ kD c-1
to c-+
0 N
I CO 174
c :1
I O c-I
ca c-+
c-) I
C') c-+
M c-)
C+') c-)
1 N c)
c- ci
I O s-1
(V T--'
I c-I c-I
l0 -q--i
I ul C-1
N
c c-I
W cN '' to -'T N N c+ O n) . ̂'. iD G1 W
ri c-i A
cq 4 4
VA A
c-1 0
c-i
5
c-{
8
cI
Oil c N-1
TA
C\l
OU
U1 lH T-4 c) Irl cf c-ß
E-i v] ýn N cý ýY 0 0
N CO
1 c-t IH
1 c
ci Ij
N ri
74 I
c1
CO t
N
0) I
a)
V-4 I
00
cI I
01
M I
CO
CO 1
N CO 10
1 N
vi 1
m in
c-t cI
M c-I c-ý Q T-i
IT I m to 1 CV O
ci 1 O 1 Cý I CO
tO N r C) U, j
r- l0 CO
T-1 c-ý c'{ c-i
c-)
M C
M
i
O '-+
Oh
cc
m
C)
N
'L
+n 4
CO
u1 tO N
1 ID
ýt uý in 0)
00
1 rý
N fV d' (7I
co tý I
Ln
rI
N I
N cl) d' d' V)
ýI
N d' I
N 1
IL M N
H
d M
d N
d lO
d c1'
d tO
d c-+
d N
d M
d M
d ýY
d c+
d en
d cH
d to
d N
d `ý
ä Q
w 4 co U w w t7 x H a ý. z o
Vi or
N a
M (Y
". p) Q
w
16
Cl) V4
H
z 0 H H
C)
U) H H H CW3
W
c-ý
z
14 In t En ýc
d
W W
" CA o U) a' 0 CO H co In CO Cl a' 7-4 in In 111 A is O' co r CO tN c+ in ul O In N ci Co c-+ A tl1 M IT m 14 in
4 to M N CC) M 'r N M c') to v}
N Co LI) N (n -t' 01 0 O N M co Lo m z co C`) c-I d' 1. O M c-4 l0 Ol N O N CO CO C- CO 2 N Cý
in C N C') Q co U ci Q'ý N N N c'I
c-+ c c-+ c':
Q. Q ¢ O w w 0 w H a Z z o ä O O u w
O M kO CO 1. o 0 1. o 0 0 0 W 0 0 0 N N ý-+ N N N N Chi N N N (V CV N N (N (V O 1 I j ! t 2 t 1 I I I I 1 . 0 k1l N c-I C, i --t CJ -i (N
U) M a' Co Ul IN t- a' c2' N N s-I N ci 'ci H c'I '1 C%
CY) ci 1-1 N
co ci tc'{ N Co
I 4 11, d G'
c-i CV c-I N c-I V
c+ c-I 00 c-I
c c-I
c-I C\!
cI 0 N tj' IV) in N N N Cr. C \l CO M O
Cp I t
C\ 1 I
I 1 I 1 00 (V ci C I c-4 ci c"I
r -i
ý a ' ti Cý
i c c c
O ca C-4 T-1 ca cI c ci c4 c{ m co
'4 LA Co ID
'I A
i 1', 8 " A A O O 1 C`
c c N C) 1 ' , ) Lf. T-i r4 t-i NA
cý C\l W
CYi Oi ri c-1 c'i c-i c"I ri a' 01 %-I O t ca I 10
1 Cl)
1 A
1 1 1 11", -4 '74 f 1 I I ý' LO U
c- ý c-i c-ý cci v 61 CO co ri
c ul
H 0
Cl) Ul) (N N O cat r+ CO ri co in C- l- I (
O { ( co 1 0 N N N ( in
V ri c-I ci
c cc-II 0 c4
OT-4
M O CV Q ri
d ai ri 7-4 co j
1 1 O) .W
1 I Q)
%D [- Oý cD t0 A r-I ý! '
L O
c+ O
CO M U) (S 1 Cý h
t ýO
10 l0 1 1 1 `
4 d' I I CO
T-4 in V co
(y) zp N It l! tfI
CO in V 4 d` !
Co CO 1
I N Il) N to
CO lD V' N L' -' C') Co CO cj` CV N N C` N
° d d d d d d d d d d d d d d d `4
cEq
a CA U iw Cý x H I R". O V4 CJ a
Cy Cl) Cl or Üi
is.
17
H
qH
0
cu
baW 4
" rr C)
O lV
N N
d G*N
N m
Co CO
d CN
m N
rn qr If)
O co
to N
m C)
N U")
co c+
0 Cl "
+ Cr) N M Cr) N if) Cr) N Cr) N c}+ N c*) Cr) d+ + Ef)
in Ol 1,0 t- C) r C) in O N ca N C) C) d Cr) Z 0 N in c-I C) cc- 0 V' Il) C) N N 00 -. J c1 N z
q C) Ul ¶ O M C" C) C) O c-ý C) N Q1 M 0 c+ c+ ci r c+ c-1 c{ c-ý : z:
E-4 U < rQ U W G-+ 0 x H .4 X. Z 0 TA d
C\j O
m Cl 1: 31
Cl E-4 u
w w
0 o (V ch
c1 41o N to N (o N 0 N (o N o N 0 N (n N 0 N tO
(14 O N o N 0 N to N 0
ci cý c-I
4
N ci N
GI c-i
oll c4
11) N
Iý e'"1
4
r-i G1 c-i
OÖ Ö)
ri
4
ri 1i c-I
Q) c-i
Ö
N
110 rl T-4 4
O c
N N
N co C) N
I c a o
to r-,
IH I- ý
! IH N
%D V-1
C) T-4
c-I I Q\
t N c-I c C \j
t c-4 c-I
l cI c4
co
C l- IN cm d' N to C) C) U3 [N
N N to 1 i
ci ci 1
'C M 1 to %-, 4 71 00
ri c-4 r4 N ci cr ci ri ri ci 1H
ri O in d' to O M M M w in ci CV T-4 T4 T-4 N N O VI d' 1H 0 c TA vi
c+ CL)
I
r4 -4 c-I
I
c-4
I c-4 CI
V"'ý t
Iri
VI C\1 cNi
VI I
vi
I
vi ci
N
N Cr) N Z~ ci ri ca cI CV W
("I Co
1 c{
r4
c-+ 1
r4
ci 1
vi
'' 11-1
V4 a
r4
r+ 1 D
T-4
6r vi
1 C) T. 4
m s1
vi I
IH
01 1
c+ 1
T4
d' c+ I
V4
C w 0 U
z E-4 En OD Ln 1-' vi
0 VI -+ cM H Ql c-1 ) r4 c+
0 H
ul t l0 ! N N 1 CO O\ c1 N C) 1 tf) C) c-
cI cl ý
I N
c-1 CO
t0
(N q N
ri c-+ C- CD c0-i t0 `O
co rl,
cýIý
c0-1
tN d'
VI C
c'ý G
V' 1 O
` (D
to
to
-1
O) u'1 ý1 to tJ
l0 O ý4 4
u1ý ý+
c-I
N N In (n
IL
In 4
d' A
(0 d' A
d+' 'I
tr) A I-
cM N (11
CY
'A
CO
9 N N
T4 VI
a N
a LR1
ö Cr)
8 ä! 1
8 LO
o N
ö fV
ý M
ý N
ý lD
a c-I
c N
A N
8 t0
A cI
a V Ö Ö
E-4 ¢ w v w w cý x H a ý. z 0 ä ä Cr) ä u
to .
CO CO 'r 00 Ln N to N In N l0 I, ( W O M CO l0 ýi' O r1 M a Ln CO N N r 1l CC) l0 l0
m -1 C cp Lf M LI M N M ýM lý M M tf) CO C7
. -i --i C) 61 Cl) t. O 14' LO . -i I'D O -4 00 O I'D r- N M . -4 O 1ý 1 O N LL) c ll ý Y) Co lO -4
co 00 l0 M (V N d m m . -4 O Ol C1 00 N -4 r-1 rýl rl rl r'1 r--1 r-ýI . -1 eý-{
FC4 m U w w 0 H a z z O O+ of a s w
O M W LD l0 C) I'D O O W O LO O O O W O N -4 N N N N N N N (N N N N N N N O
1 1 1 I I I I t 1 I I I 1 1 L I 1 r, l0 N -W P-4 dý Co to N lD -i co O N r M 4 rl N N N 4 N -4 -4 N . -1 N r-1 r-1 -I N
tl "-I M O M 1N v lU ß. t1 O IN 01 %D l0 CO N r-I 4 N N N N -4 '-1 N 1-1 1-1 r-4 -1 1-4 N
(14 m 00 LO Ln C) r1 N r-I -4 1-1 . -I I--i N
C1) (N 61 -4 W M LI M Co lD N -: It N Cl
1-1 N -4 N 1-1 N I -q --1 . -4 o--1 -1 -1 .4 r-I '-I Co I 1 1 M 1 CO
r -4 CO Ö1
1 4 N
lp t1) dr M M
0 0 N r-1 r'1 N -i -1 1-1 r-I 1-1 ri . -i r -I
N O O N 00 t. I O N -i LI M N N W r , -i r-1 N ýr-1 r1 r -i N -4 1-I r-I r-1 r-1 -4 r-I r-I r -j
tý I I I I I I I I 1 I I I I l- '-1 Ol lD LO V' Ql IV M N -I Ln In
r-l r-1 rl ri r-I r- I r-q ri -1 . -4 r-1 -4
O co l. f) to ('M CO O O M N "-i C) d' qr r-i Ql r-i . -ý ri ri r1 ri ri . -i rl r. 1
1 ! I I I I I I I I I '. D a% I- V' N 40 Ch N . -I co O U r-q
ri
8 8 p -4 M C) co 00 :4 p4 ý OD O, 1-1 ° U' m U) ° Ln I I 1 I I I I I I I ! I I L1
N V -4 -4 Co O N Q) N CO (fl -I
p ° -I a 1-4
M ° t0 1ý r -I
° 01
° N N a1 01 lfl N l4
4 N
d' i 1 1 ! I I I I 1 1 1 1 1 1 -º' eM N 01 CO dl CO lD U) OD LI lD u) O) l0
-4 M . -4 CO N CO N L lD ßf1 --1- CO to
Cl , o i ;
M N lfl d' \o l0 M N C` 1
-4 In a% WO lC Ln co L Ln U) N mot' Cl l0 N N "
; J 1 " N
! I r
N .4 1 . r N LI
d' tD LI Cl to -4 N M Cl dl 1-1 M . -1 ei0 p-1 .i i . . I ! I I 1
O I
O b I
b 1 1 ! f ý. i O O O O O O O O CD O
PG ý M W P4 !' x H a z o
ä ä ä ä
19
w
w
H
z 0
c: a 0
x EP) H
w
N
z Irl U U,
HU
" C tp d' c+ u) ýt ý+ O O c) cý cý "
a) co c-H (N Lil N "
tD v) Lf) H C') t: CV d t! ) GJ q C)
" (r)
" C'S
" 4i
" Q'
" Il) Cr)
" LI)
" Ci
" N
" r' , ')
" 'v
" M
" c*l
" tr
" "
H N N LI) O Ln Ln N N r-i ,v G\ ('V M " " " " " " " e " f " " " 7 " " "
q
H H H H H co Gl H QQ 01 (J
vN-' cv-i [
H H cN M '' EI
.4 fc U iiý t0 H a Z, 2, Q 0 C Or O U
O M 1, W LO 0 W O O to O 0 O O O W N c'H N N ('\J CV N N N N N N N N N CV O
-i I -
l 14
I 1 1 t
I N
I fit)
I I 1 I I 1 1 O T
i c-i
N i T-4 N
H N
in N
a) c-1
) t (V H-1 H
H CV
OJ H
(T c-H
N H
In rH
01 c-i
CV N
c-i
flo d' d' to O [N co ur c+ ci N N c-{ N H H ci C'J i i I l 1 CO
Lo N N c H C u H co ', 0 N ul 7-1 p
H tV N H H H IH H N
fV ß1 N CV 3' C') tN in tp cr '' N c1 cý N v-I N N c'H H r H rl c-i H c-H c'H
W M 1 4 c10 -0
A 1 CV
I I 1 7 Ö co 1 N N
H I N
7 c-i H
II c-1 c-i
1 1 ci
c c 7
l ý-. 1
N H
N O N 0 in 0 N in d' N tO c N H N H N H 1 r H H c-1 r- 'A
H O1
, 1 0 rä
IH c
1 d'
Cl) (V -1
rl -I
4 N
i c- c-I Ci co H cI
0 H t-I H c c
H c H
0 O in N N 0 Cr) N c`I ci 0 CY) 'c'ý ci H s-I c-ý H H H IH H r1 c-I
410 ý*1 :
( Ol
t 0
t G
M ! i 01
Oi -
1 ý
1 d I I I tW U O
1 c-i c-I c-I
c In c+
c N H
t c+
O H cý
Cl C n c-I
N H O
Ei Go vi r-4 Ti H m M
- 0) 0) 1H -H
H vi to
rt 4 ý N H I
rl l I I 00 I I ul Gi
c cc) 0) a N H ci H H H H
T-4 H T4 0 l0 c{ IH H l0 tO N N H a) '' i to
n 1 t ( 1 l 1 I q OD I 1 1 I (N 0) C a , O'. u) u) 1 t0 rn ti
H c-+ CD CO co r- [ to I CC) to M d' t9 C N
! G N 10
to - O i)
' " ' M
. l t lJ M C N L 1
A N
N I t0 1 M u)
vI u) cf W j N
O
H d! N d' N in M N d' M M tV U) N
`A d d d d d d d d d d d ° d d d d
E5 CO V CO W c7 x H a z o
ä Cf C Q
w w
20
4-
U cý a 00 en
a aý u
i7 N xwu HE VI
ý f4Sý7
H AEU)
O Cl) Cl) O) co N co r+ N' In p In O '1 O c " : LO Ol ON co N Ol Ll r+ N UI o, In ai H-+ (Y) ao n
. " " " " " " " " " " " " " " . " U) In c-4 In In V' CV In M N M N G' CV In C") ý'i'
C`') W
lD c'
Lf) N
d' Cý
0 O
N ýi'
c7 01
N Gl Ql
N N
CT) (V
l0 fn
c' M
L'1 ;o
Q' H
Cl N
. -ý
N N il T4
N 'l-1
V' H'4
N r4
M C-i
N CO C''" c-I
CS C; ) CO N C'H
c4 c-I c-.
O C-i cý N c") G"
O E-i
< 'a U W G C) X H i-=1 ý. 0 0 0 Lý a ,' Q 0-i
0 In G 1. c 1. o O '. o O O O '. D O O O CV H (V CV CV (V N N N N N N CV CV N N
O ca1 c-I
1I t--1 N
NI N
("I)
N 00I c
All N
I lC
i
1 to ýý
1 I H
I
s-1
1
cý
1
c-i
1 H
i N
U
cý O c-I
M CV
N I
Cs
CV (V
N H
In N
c-{ A
c"i
TA
tý d, rH J)
c-{
c-I
N
cI
C))
c-1 4
c-+
N f
p
C') ci
N CV t-I
N-4 N
CV N
N H
M H
L(1 c-H
"o c-i
CV ci
M ri
N c°i
(3) 0
N
c
1 d
CV li H
d
N IH 1ý CV 'A c4
CV H .4 VA
1 X-A c-I H
fV c-I cI i-I
TA cý
1 p
G\ H
ýI c"1
W H
InI
0) H
1h In c
0 CV
I a) c"'ý
O -H cI Iý
cH I
(N
H 1
T-4 ci O ca v)
q-4
W a: Oi
1-0 M Ch
co CO
N c1I
16 H
d' c-+
11) a1
to H 4
cý
d' c-I
A cý
N c-+
in c1
m co
0 ý-1
ON I IH
c+ T-1 rI
ci r1 d
cý
01
0 H
(A
d' H
cý
d' c"+ N
c1
w lam' O
iO
Cl) V)
Ln N
r- 0 4
c-H
c-I
c-I I
tV
H ý
H
H r- 00
H ý
H
GS ß1
I,
CO
[ i CO H
H
H d
c-I c-I c-H c-4 H 7-4 c-I
en
N .O
01
T-1
pO 0
0
00 I N co
il) kA
0
Oll i1 -41
In I M in c-i
r-I Cc) o, % tN C: 1
ý
10 d, I
LI1 U)
h I ý I
.? ' 1
N I
tf) C"')
N N ýi
O ý t- 1.0 N Ln 14
C', Ln 4 in cm (1
M ko CV
`' H
ö N
d [N
d LI
d d'
d l0
d CV
d (Si
d (I)
d C')
d d'
cS c-i
d CV
d N
d to
d N
d `'
o a
w M U w rz, C) H a z 0 T-4 ä cm ä m ä ä U
21
ýi m
a *lo
q
rn M N i!
14
uJ
to -4 "D h Co Ol 0 ßl co ßl N 0 l0 r1 r-i p1 O O [`- to N M c in 1 i`N in to -i M p
M N d' M -t' N Ln N N M N -P N M M m c
kO o- 01 M ß1 N l0 -I O N in N in N N d' M in M t`N 1. N d1 CO in llý . -l N 'P. O 4f1 C'
00 w r-I M N N 00 . CO a1 p O co 00 r-P N
--l clq Cl) Itr ý tY. l U W W C7 t-1 a ý Z O Ot OI Ot ot W W
O M LO LO ýp O sO O C) l0 CD \O O O O %p O N . -i N N N (N N N (N N N N N N N N O 4 1 1 1 1 1 1 I I 1 1 1 i 1 I 1 1 "-+ l: ) -I - d' CO l0 t! ) W L O cM lfl CO N
N N N -+ fN --i -i -1 --4 N -1 "-i -i (N
c } c
in q M O Cfl N M U1 -'T in .D a% M Ln N 4 r1 ri N N N r"i N r-1 r1 c-1 rI -4 r-4 r-t r-4 N
r 1 I I I n I I mot' i 0 0 i N
Ö
r-t (N rl N N r-i r-i N
M C) NN O t0! "-i m M Vf' d' l0 N N CO C1 r-i O N rl N r{ N r p-4 ,. q -4 +-4 p4 . -1 r-4 r--4 ß n O 00
Ö1 lD l Ö1 N M i r-4 Ui C r-4 ci -4 r-I A -1 rl r-I . -1 . -1
N CO In CO in CO N . -i N M Q -4 r- r-I co r-1 r$ rl r-1 r-1 .i r-q ri ri r-1 ri -f r-i 4
1N I I I I 1 I I 1. ! 1 tl- 1-1 N d' N +I` lD . -1 N m in '-I -i r -I ri r -i 1-4 -4 r-1 r-4 P-4
(o f+") l0 M in Q -1 O M ß b N i r-1 .i ri . -I rý .1 0' . -I r
Q0 -A 1 1 1 1 1 1 l l ! ! 1 " ß M rn ON C) O N ' Ql -f-1 r -4 r -1 I . -1 . -4 S
ý r4 ° r4 CNI
CO X) 4 r-4 -1 00 CO 0. C 00 CO in I 1 I I I I 1 1 I I I it) N M O r1 4 00 00 N O r4
-4 <n
- O O I'D in - m i N [`N Q l t ko M
r
t ' i
CO in ; CO lfl CO 0 4
cM 'tt' r-I CO 0" Ql N U) LO lo N to in in CO N M I I n N (r1
I N LÖ i it) 'ä' --w i
M te) oD -, O NN CO in v -T l 0 M M d' N 1 Ö V n
ý, + Y N N N
) 1 t N C ) in in ( )
, --4 N (0 d' in -, O "-i M M N C' N M N d' Chi
+ "- 1 a 1 1 1 c I 1 1 t 1 l 1 a -4 O O O O 0 C D O 0 O 0
O C a
O O
U W w cý x H a z o Or c ä, " °C)l
rz
22
N W 1ý
lJ
i;
MG
. -1
" Q at- 1D
.
"-i O
N M Cl
. -t V
N ri
ON .ý
C) 1
14 M
N co
N kD
1-1 M
N O
cN 00
CO O
N N Gi
(! ) in m N tN co l0 O
ý- 4 l0 in ul ct' M in 'n ý, O C ilk
M O
co
(N O
SST
CO '7'
O
i- N
l0
N 1ý9
N
tN O
. -1
%D L l!
co
m
l-
Gl CO
co
N M
M
CA N
co
N mot'
O
iý U)
-4
to CO
co
N in
. -i
.v in
Ln -i r-1 M N M N N r-I . -i N r-1 N N '4 N N c
p Q co u W W 0. x H 0 Z O 1-1 0 OI
CY) OI -zv O+
9
O --I
O mot' 1
114 N
1
N lt) I
N Ln
I
(N to
I
O d'
I
N in 1
O F
I
O C
I
N LO i
O -V
I
N to I
C) d' I
CD -7' I
O d' I
N it)
1 C) -4 O
M N N
Ln ý'
N ''
C) ý'
M m
N d'
. -a M O M
CO M
CO N
N M
N M . -1 M
v' M
M -ä
0) ON N I
r-1 N
I "t' N2. I
. -i ýN I
CO -W I
N C1
I
W d'
I
O M I
0) N
I N M I
f- N
W M I
-1 M I
Co M I
m M I
N cM
I 01 I- N
0 CV
N ; 7'
0) M
LD ýP
4 M
M d'
CO N
N N
M M
-r M
O M
CO N
N M
C) "V
co
1o CV I
Ol . -1
1-1 "7' I
CO M I
' U) w I
C) m I
N "4'
I
N N I
lD N
I
N M I
tD N I
CM M I
Ol N I
N N I 1
, -i M I
Ol M
I co 111 (V
co M
W M
1-4 14T
0) N
O "t'
-v7' N
to N
O M
; 7' N
01 N
CO N
L1 ) N
CO (N
lG M
lý d' N I
CO . -1 I
tN M
1
N M 1
O 1t; M I
N N 1
0) (')
1
M N I
el' N
1 0) N 1
M N 1
00 (V I
N N I
c' N I
tý N 1
Ln M 1 c-
'-1 N
1ý "4
U) M
r'1 M M
U) N
W M
C) N
N N
'D N
-4 N
Lfl N
t! ) N
N N
it) N
ri M
'O
O N 1
lW r-4 1
d' m
1 O m 1
1- M 1
d' N 1
lt1 (Y)
I
O1 . -i 1
r-1 N 1
Ul N l
c) N t:
ýM N 1
'I' N 1
-4 N I
dr N 1
O M
1 lO CO . -i
in r1
-I s C")
N N
M M
N N
ßl N
N -
C) -4
C ) N
Ol -4
. - 4 N
N N
C D N
N N
W N
ix7 to W
I- .4 I %0
'V' . -I
O M
1 IN
%0 CV
I M
N M 1
0)
i-1 N I
0
00 N
I "t'
l0 "-4 I -: r
CO -4 I t0
N ( I
ri
CO . -I I N
0 N 1
rl N I
CA -i I
s-i N I
U1 N
I NJ
8 . -1 N N N . -1 N . -I r-i N # -I N . -i
0 ri
N r-1
01 . -1
M N Q
st'
tn . -I
ý1i
-I r-I
N
I d'
N
I O
N CO
i Ln
CO
F' in
N
I 0)
s--i I
.4
' , -I I
V'
O ýI co
'D
I \D
1
co
1 lo
i 00
1
N II
d' C) . -i N N N r-1 rl -4 .i --4 r-i
M .q
S . -i
1@J ý4
kO r--1
WO 1
ý1
N . -I 1
N r-1 1
M N
1
Cº .i 1
IT N f
'a' r-1 1
l- r-1 I
O -4 1
M r-1
I
N . -i 1 "
M . -i I
N "-1 1
to . -I I
N '-4 I
U) r-I I
N r-1 I C') 0
'-1 u-C ri
O N
N . -i
0 ('1
M r-4
N . --i
01 N . -i
a -1
N r-1
U) to . -i
O ri
M e-I
. -i rý
N rn c0 I
c
. -i i It
0
. -4 1
110
. -4 1
(71
.ý I
N
u4 i O
r-1
. -1 1 CO
CO 1 N
. -1
-4 1 O
in
' . -i 1 -1
.4 I MI
s I oo
NV
s-1 I O
C)
r4 I tý
N r-4 r-4 ýq ý r-4 r-i r-4 r-1
"-+ N
CI
co I
O
Ln u--I I O
v .i I 0
m -I I
O
Ol I
O
N I
CD
l0 1
O
0) 1
O
m . -1 -k- O
O . -1 1 O
M 1
O
N i 1
O
5 I
O
Ol I
O
W I
O ý--4
a pa
a rý u q w cý x H a 2 o r-i a N of
C') Of
' Of
w
23
N N
M
cm C4
13
C4
O . -i
d' M !m
C) OO
'D
co Gl
lo
d` 1-9
co
Co N
t! 1
O "-1 O
lD 00
to
co -1 LO
[- l-
in
C) LO
O Ul
a' O
in
a' %. o
t11
Ln t
to
-'T r-
CO
CO in
r- r1
N Cl
Ln .i
co N
. 14 M
1-
I'D N
N CO
r- N
O in
N
N N
0 N
O -4
rý r-4
in O
010 r-i
01 ()
M N
a' a'
C1 r7
co in
0l -4
ý. o CC)
01 'A
in in
O N
M 1-
M N
l0 N
C N
Z
a Qs4
oa v w w o x +-ý a z o o a N
a M
a mot'
a w
O "-+
O st' 1
O M
.o N 1
N N
N to
1 U1 cr
N to I
.-4 d1
N in i
Ul Itr
C) W !
it) M
N Ln I
l1) lqN
0 zr I i
M
O W 1
a0 N
N in
I CO M
O -! '
1 d1 N
N tt) 1
LO I
M
O 'd'
I
M
O P
I
M
O w i
X11 M
N to
"I
O
ON 01 Cy I
N N
r-1 N
d' -tv I
N
O -W
co I
M
d' -V I
fV 'cr
eM M I
N ("1
d0 1*1 I
M 14:: p
O M I
010 N
N N I
l0 N
N M I
(V M
co N I
I- N
-4' M I
. -4 M
Q1 N I
CO N
M M I
O M
-N M I
M M
'-1 IýT I
co M
O'
GO
tC N
I U') N
O N I
dl -4
-4 Nr I
C3) <q)
N M I
In M
-4 IV I
CO M
-I M I
01 N
(V `w I
N CM
Ul CV
I f'ý1 CV
it) N
IV
N
-4 ß't1 I
O M
to (V
I 1n CV
c) M I
00 N
N N
I l0 N
ß) N I
N N
N M 1
O M
I- M
-Jr M
oO
N
d' N I
.i N
CO _4
CO M
I if) M
'd' M I
O lM
N M
I m M
00 N
I N N
l0 M
I N M
fV N I
O N
M N I
-l N
C1 N
1 in N
-.! ' (V 1
M N
N (V 1
M N
to N
1 ch N
LO N 1
M N
a' CV I
00 N
M m 1
a' N
1-
W O N I
CO r-1
1- -I I l0 1.4
'-' M I
N M
01 N I
N N
N M
ýI N
%D N 1d '
N
M II
CV
ON -I I N -1
O N
I dl r-1
to N I
C') N
N N I
. -i N
(V N
I O N
N N I
O N
N N
I -1 N
N fV I
'ä N
00 N I
In N
to
W u'1 iz
8
r. . -1 i
in -1
U() .i 1
et' . -1
I M
I Cl N
kD N
i M N
N ("1 I
d' N
tM fV
1 . -1 N
N (V
I N N
%D -4 I
"Ir r-1
0) . -i i
ýG r-1
fV N I
.i (V
O N
1 co "-1
(n . -4 I
l0 .. i
0) -i i
00 -4
O N I
. -i
f) N
1 N (V
q' (V
1 a' uI
ý
ýi
Vý d' ,4 I r-I
1 T-. 4
co
I N
N C-4
I N
In N I
ý
N
1 . r-1
N ý4 1
. ý-I
1 u-
ILO
u- 4
C-4 ýO
r-4
l4
I . -.
uC
I rN-1
[, 4
I rß-1
-4 [-
1 . Ln-4
'4 (-4 I
-4
u-C co
M .i "-1 1 0)
. -4 r-I
Ln N 1N
(14
Gl r-1 1N
OD . -i 1
CO rl 1
Ln r -q 1
'"1
. -I
I O
N . -1 I
u- i
1-
1lD u')
IM . -I
I O . -C
'd'
1 M -i
NN
I M . -.
CO
1 tp -t
to
1N -4
M
N OD i
l0
p V-4 I
01
r4 N
rLI r-1
to . -1 I
. -4
it) r-I I -1
u-I
tf1 . -i I N
. -C
O ri 1
N
a) I
N
O r4 1
C)
lfl . -I 1
C ri
N . -I 1
e-1 . -1
Ol 1
N
N r4 I
. -1
.i
N 1.1 IO
r-1
in rI I
N ri
1-4 . -1 11
C N
In a
Co 1
O
r-4 1
O
C%j
1 O
O
0 "u-4 I
O
w I
O
to 1
O
co I
O
M-I
1 Cl
-4 1
O
l0 1
O
C)
I O
C1 1
O
'u-1 1
O
co 1
O
,
kr-,
w z o 01
ww
24
0 H
ä ýr
a 90 -e J)
N si
Ö W
bi w Q C, F 14 4
" In IZN
ý. p C'
to c-4
to 01
ON Cl
to ct)
to ,
('M N c1
M l0
N to
Ol Ql
p Ct)
O b
O M
0'+ U)
f: i " " " " " " " " " " " " s " " " to Cl) N lD N to In Ci lf1 Q' N ý1' If) to Cpl 0
0 m 0 %x. ) N N (D O c-i c-+ c-4 O N W if) z N O c-4 M N 0 t) N CV (�Si rI M N Q1 N c-I Z
z N ci
to c-+
r+ M
to N
to N
to N
l N
r. tN c-I
m N
c-i N
01 c-I
co t-+
o N
lD CV
M N
O Q
Q q U W is+ C7 x H ca z 0 C7 C7 CJ Y O Q
O 0 d'
tO N
N If)
tV to
N [C)
0 d'
N In
0 v
(D -f'
N u-I
O Q
N C
O %J-
O !' 0 '
N to 0
I OD N
I 1-, N
4 d'
I d'
A d'
16 M
1 d'
I M
1 N
1 Ct)
(ý pý
1 M
U1 N
M M
U) M
Ct) d'
cI
t- C\'
O N
M d
O d '
N d'
to (I)
tO \1'
O M
00 (V
d+ c)
co N
U) M
N N
N CO
to M
N
cri N i 1b 00
4 4 W In
M N l0
Ö 4 MI
O
N d' M Ct) CO mot' CV N CO Cf) N CO Cl) CO
CO N
Q1 c-4
c'i d'
Ln M
N CO
M M
M d
N N
4' N
(V ct)
N N
ca M
to N
Cc N
m CC)
CO M
O NA (A cl)
CO
C,, lj CO of V
(N
C\11 CV cl, N (V N cot
A
CM (ý4
4 co
M (V to CC) M M CV N
N N CV N N (C) M
Ö N LA M
Ö"
N N Oi N
4
co
'
N
Ö
N CV
4 N
4
N ci N
4
N Oli N
OÖ N
ß1 r4
L- T-1
dt CC!
O N
O N
00 N
CO M
O N
M c'I
[C) N
(y) tV
M N
O CV
CO N
cc) N
r- N
1
, CD tU
t-i
1ý
c0
A
N IL N
A
N co CV
CIO c'+
r-, r1
A
N (V N
pl N
d%
c-i
N
N [! N
CO N
W tN c4
CO c-I
O rt
d' N
11) N
W N
N N
tN t-4
Up N
N N
c-I (V
0 ý
qý ý-1
N LO N
C\l N
w
N N o i N
y
tf1 c tý
Ca N
N
N N N N LA c-!
t e CV
ä1
c-4 lÄ ri
1 c-i
ck -H
.4 N
00 c-4 v7
Ei
d' V4
ci
CO c-I
V-1
N N
N
c-4 N
Tq
c4 tV
ei
cO! N
(v
c4 N
4
d' r4
t'1
d' ci
c-4
O N
c1
CO c-I
r4
ei' T-1
V-4
U) 4
H
a) c-I
ch lV
N
N
rf
VI
(C)
cl
Cis I
c-I 1
O cl
N
1 cI N
ýp c-i
1 N r1
r- c4
I C d
O N
! Gc1 c-ý
U)
1M -I
ci r4
t4
N d
c'+ c'
N ßc-4
t,! 4 c-i
LO ca
1 In c-I
O d
p> !
CO c-j
N c't
!N c-1
T-1 N
I cl
N c-I
I
N co
Ct
0) 0 N N
to ci 4
CO ti ý
CU ci A
N VI
CIO
ON 0 cl
lfl d A
d'
ý O
tD
N c'" Ö
l %-4
C
Lý 'ri 'A
C) J) N
t0 3
co 8
c-1 c 0,
cI c-
8
ýHo 8
-q: r I-
8 N 8
co 8
Ol
0
ý
8 cN-I 8
U-) 8
Q1 8
01 8
in
8 d' 3
x aO
.4 an U w rL. (D x ý-! ] z o ä ä ä 4
w w
25
E H
N Ir
z
Lfl
N N
w ý
I 0) I I n lD N N
CO O 111 O
N O
i i w n u r w ko u) O I -v In co cri (-i Ln a) G7
In - "7' O co C) CO m lD M ! -- Cr) l0 ßcr ul t1 N 1 0 0l w l0 co N N N r -1 M
. '
. z : n d' O 'G' r" 1 1ý0 r- ! kp N N O [ pý ýp cr "-4 -4 M N N N N ý-4 r-4 N N N . -4 . -i N N
0 N M
a W U W L C7 x ! -ý a X 'Z O O! (DI O{ a C)
O O N N N O N Cl O N O N O O O N C) cr N If) ºn to 'd' 1f) v tT LO 'c7' In c' C' '! ' IC) O 1-1 1 1 1 1 I I i I I I I I I I I 1 -1 O -4 Cr) C) -4 in t- O 0) Cr) N CO n N Q-0 in
N) N l! ' "Cr d' M "t' Cr) N M Cr) m N Cr) Cr) .; S+
41 O CV 61 O d' "0 O) co N ri ! - lD . -! U) , i' N N dt M '4' M " N N c") Cr) CI) cV Cr) Cr) v
of 1 I 1 1 "I. 1 I I 1 1 I I 1 C- 00 ý--4 'V' co M n 1n . -t O U) C) 0) 0) N d' M co 'f' N N (1) M M N N M
! -- a) C) Cl) n M N %D ll;! ' O 0 'T' N co -W CO N -1 -f' M Cr) Cr) mot' N N C) N M N N (Y) M CO I 1 1 1 I { I I 1 I I I 1 , I co co . -i N ON 't' N OO n Ol . -i In M U) N M M re) (') N N N N N N N Cr) (-)
Cr) CO n '
O r- q N 00 M . -I n W QC) O '! ' N d' N - r 1 el Cr) CO M N N N N N N N M M n 1 I I I I I i I I I I 1 1-1 N I.
' co CO O c! ' O Ol Ln in t. f)
1 N
1 O
I al
n N ri fß 1 N N Cr) Cr) N -1 N N N N (') N
ýcJ O l0 M n n d1 f'') M 00 cP v d' M , -1 al 0) 4 N . -1 M N N N M -4 r-1 N N N , -1 N cV N G] 8 %0 i
00 i I I
' I I 1 1 1 I I I I I I I -O üi
U) '-! in '! C O Q1 CO i- C) M O co O n V' Cl) f4 . -4 M CN N N N -4 -1 N N (N -I N N N
U) rý IV C) qv m r- 00 r- t. 0 N N n O\ Cr) - 4 r-1 N N N N '-i -1-1 ! ! -4 ' ! N N N Cl) Lt) I I I I I I I I I I I I
ý ý 11) M n . -1 O ý N L) In O O t! ) Ii) ! - d' ri
rI /-! N N N N (V . -1 '-! N N r-1 i-f r-4 N N
c1' N l0 O d1 d' ---1 w V' Q% 0) tP -tT lD M CD
- (N N YI Ci C C 11 P 1 1 I 1 I I 1 I I I I d' N r-1 sý Q1 l0 M n C'-) A CO C0 M u) . -1 U)
ri ri N ri .i CV -4 i4 r-i , -1 -4 r-4 N '-!
-+ c M co In cam: v) N '-, I N- r, cv m a .-4 N i r-1 (N r-+ . -i . -i r-1 "-1 r-( . -i (N "-! M ! ! i ! ! ! i i t ! t { { ' I m 0) O In N O C' . -! --i Lo O N N ! - N
N . -1 . -4 N -4 . -! . -1 . -i -4 r-1 . -i r-I . -1
a1 d0 -1 0) (4 O O er U) , -t -4 Co O) . -4 . -1 . -i 4 -4 ---t ri -1 -1 C1 r-I -I i-f . -4 cV I I 1 1 1 I 1 1 I 1 I 1 1 1 I N 40 iß rl Cr- l0 CO (n CO N -gr n O m C') 00
LO CD Ln -4 N CC) r- -4 ko ON 00 t i t 1 I t t 1 ! I 1 i i
ý-! Co O C) O O O O O O O O O O O O O
CO U W W C7 x H º- Z 0 -4 Oi
N 0
M O! i
cya O!
2
0
26
cn II
N
z
in Ulf
j Ql Cl
tf)
-4 N
fM
%D O
N
U)
[-
O N
L
Ol 01
:: M
-1 '! '
(5
LO to
to
N v
mot'
'o U)
Ul
. -i C'
NP
C' i
I-
l0 --
-'S'
Co tf
l. [1
Lt) C
eP
L CO
co N
c '!
i-
Ln M
M
fn 0
O
L) to
1
cV M
O
M O
t-
to C
to
61 iý
CO
CD 'I
l-
N ý1
N
61 N
r1
00 fl0
O
O 4.0
l4
to r!
N
O tý
r-
'r d'
-q
ý2',
.1 r-1 M N N N N -4 1-4 N N r-4 '--1 N N (V
a a CA U w w 0 x H a z o -4 ä ä ä 0
w w CD to r4 N N O N C) C) N O O O O O N
O 1
O N1
N I
. -1 CV
1n 1
d' d'
U) I
01 M
Ln 1
to M
v I
to M
In I
Co d'
M 1
N M
W 1
N N
LI) I
ýM M
-:: v I
lV M
mot' i
Ln M
V1 I
Co N
y I
IV M
-'ZY' I
N M
Ln I
O O '-4
0 O r O 1d) U1 N r-4 to m . -1 W N M lU 01 N CV 61' c') M M 'V' M N cI) M M N M M M
(71 1 %D N
1 01 . -1
1 N V'
I 'a' N'1
1 N f'r1
1 M d'
1 N CV
1 mot' fV
1 N M
I O M
1 O (")
I M N
I O M
I tf) m
1 Ln M
crn _
co Ul N I
V' N
co . -1
ri -' I
ON M
M M I
0) N
-4 (Y) 1
co N
c' M
I N M
N r
OI M
to N
N
M CA
1 N N
. -i cM I
Ol N
C) N
1 a N
Ql N r I
N
N N I
r-1 N
CA N
I a) N
.' (I)
Q' M I ý
f+')
Co
N M N I
r-4 N
N r1 I %b
CO c"') 1
U1 M
CO N
I L) N
IN N i
' N
r4 M
N M I
M M
I' N I
N N
"-i N I
O N
ao N
I U') N
N N I
to N
to N I
M N
c) N I
ON r-4
N N 1
tO N
cr1 M 1
. -1 M
O M 1
N N
N
tO
O N I
CO . -1
In . -1 I sT r1
d' M I
. -1 ()
C N I
N N
Cý7 CV I
O N
O M
OI N
C%' M I
N (N
-4 N 1
Ol r-1
ßl
1 N r-I
er N I
M N
LO N Jr N
N N I1
0 r-4
CO -i I N
-4
to N 1
c'-1 (V
c) M 100
N
to (V
(IV N
to
19 8 U) N I-
I
M 9-4 I
C -4
O M I
N
r-i N
I
v--4
Oh
. -I
IN N I
N
to N
I N
Co r-i I 4
to 1
I
LO r-i
N N
I C N
M N I
i N
CO . -1 I.
to v4
%D .i 1
rv-i
N N I
N N
1
N
-4 (V 1
N Ln
W
-4 I
cl, . -1
. -1 I
cý . -I
N
4r V'
(V
CO
t! 1 Ln . -I
ý-1
I -W . -1
N I
A N
C)
I I r-4
-
t A1
-4
rv-1 M
'-1
. v-1
l- -4
N 1
ý v--1 1
rN-1
. -w-1 1
-4
1
. ý-1
N
1 1 C-4
- I 1
/N-1
mot' -4-
U]
M r-I Ch
Q1
C6
N
., 4
-w ri N
M r-4 Ö
N CV Ö1
11) v-I
fV
N
r1i
N
. I1
tD ü1
CO
N
ý--I G1
N lD -I
N N I
O
. -I
. -i I
d1 M
N N 1
to
N 1
to
0 N I
C o rl
. -1 I c 0
C1 1
to
00 . -1 1
N . -1
.4
. -I I
00
O . -4 I
01
O . -I I (in
'cr r -I I
N
to -1 I
Ln
00 I
l0
C)
I a%
.1 .M I
ý--1
v--4 I
G10
CO I
ýN N
ý-I U) I
O
W 1
O'
r- . -1 I
O
tN 1
O
In I
O
rv-I I
O
N I
O
00 1
O
00 1
O rl 1
O . -1 I
O
Ln I
O
a) I
O
CD
r-f I O
ri
O1
M 1
O -4
a
a w w w 0 w H a z o ý--1 a
N a
M of
C' a
w w
27
H
M
. -1
Ný
r-i c) N N 01 r-1 "-a k. D N co -'r M en N O N N tf) ýY' Ln 00 [- -4 CO O) A N N Eý N
. . . . . lp C') N N
rn 'o
ä LO Ln in ": 4- CO :.
7 lD .C
ö
u1 ý r+ cn O N (14 N co r-f '( O M M Il- Iýv ON 00 m ko N C! ) lD CO N N M l- U) 4 mot' hfl CO l, N O z
t OO U;
4 1! )
4 to 00 00 N to .i m CO O LO C")
0
r-1 fh N N N N -4 . -I N N r-I -1 N N N
ö '-i N M G
W U W W 0 x H F-7 Z Z Ja of CI cv a G, W
O lC (V N N Cl N O O N O N C) O O N O C' N t! ) to to ct+ t. 0 rM Q' to [ -,: P ;r 'i` to C) 1-4 1 1 1 1 1 I I I I I I I I I I I . --4
01 N 'Y -I 1ý '. 0 CO cr CA 01 N S -A V N LO N N -' d' Kt' M d' M (N (') c"1 M M M () ýN
CO . -4 (l O l0 Lf1 N M CO 00 . --1 l0 C) Cl) l0 N N eM 1141 llw M d' M N M M M M M m
O1 I I I I I I I I I I I I I 0) N CO N w . '-4 ko if) O rl 00 N LO a1 . Cl d' d' M N rM co m N C') M M
l- O -q t- rl d' [0 O Ln -W Q) C) l- . -1 V' cu N N '4' (') d' C') 'd' M N (0 <'4 C') N M M ("l
co I 1 1 I 1 I I I 1 1 L I ! i ' CO lC C! :r [0 C7 O ýp M N lý r 00 M 0 N d CM C) fn d N N (0 N N N N M 4 ,
to ON 01 t0 er CV 01 t0 N r-4 lD %D CV l- N " N M M M c"'1 M N N C`') N N CV N CO C')
r- I I I 1 I 1 1 1 I 1 1 I I 1 I rý N l- 11-1 co C) ll1 (V -4 d1 ! f1 (0 r--4 ' CD 01 N M M N N M N N Cl N N N N M CV
. -4 CO ID O N CO -W 4 O CO "w N O (n O1 00 N -4 M M N N M N N N N N N N N N
%D I 1 I 1 I l 1 1 1 1 I l I1 I I 'o 0) (0 [0 M [0 d1 l 0) CO U) CV
-CO a l 4 1 N CO 0-1 M N N N N r-4 T-4 N N r4 r-4 N N N
8 00 N N LI) N t1) Co co N d' . -1 l- 00 O kO (N i r1 M N C4 (V N N N -4 r-1 N (V N
V] I V M 1 1 J M 0 1 8
ýD Ö1
C O c+ 7 1 1 t ) ) V ' N CO d.
CO '-1 ý4 N N N N N 1 N -4 -1 r-4 -4 (N1 -1 R
tO In a . -4 C1 N CV C' "M N 00 M t- M s .I p4 N N r. 4 C4 N r4 r'4 N ^I r-4 r-1 -4 N -4
d' 1 t ! ' t 1 " 1 I I I I I I I "a' [V lw d' O N Cl O1 N c , M . -1 5 r- 1 Ln O cT . -1 -4 N r-1 -4 .i . -1 . -1 r4 CV r-i -t -4 1-t lV 1-4
. -i C') C') co Ln co 'D rl N O [D O (n -M C) c' P4 r-1 N '-1 11-1 r-t .4 -4 p4
M 1 1 I i l ! 1 I 1 1 I S 1 1 1 1 M r-1 P4 l0 (V tf) N Q "-! tf) 0O N M --: f' N. - 4 N -4 . -1 . -i -4 . -t r-i . -i
O O ul f-i v' p-I O r' d' r-I N M co . -1 N . -i rl .4 .1 0% -4 N i -4 P4 [O
N 1 1 1 I I I 1 I I I 1 1 1 1 N [0 "-1 c r-1 co n ß 44 [0 G1 O 0 IV
r-I -1 -1 . -1 1 . 4
. tI as " -l " 44 r
° N N lo "-+ "-i Ln CO rn ON (0
ý I ý -ý a (b ( D 0 0 ý ý ý ý ý ý a a
P4 N M Q W U W W C7 x H 2 O OI Ol OI a
W W
28
Li
(1 Ný
H
N U) O c') fß"1 N '. D m a en . -4 N to rl CO -i M r- O C' .4 CC) -i O O IV w %D 1-I VV O
910 (fl l0 co O l0 O t0 ul lU L1 1'- 1'1 1j; Lt C Ü) ri r1 (I)
M N fM N U) Ln to CO -4 O N Lo N O u) -4 O 41 !n If) C N CO N N l0 to r- N lo Cý N
CO lfl r-4 Ln rý :r f- r- r- 4 O 06 0l O ci' M
. -i r4 M N N N N 4 . -4 N N ri .i N (N N
LY.
FC W U W W C7 w H º 4 O Or C9 Ol Or O
W W
O l0 N (V N C) N Cl O N C) N O O O N O V' N to U) U) r U) it 'd' in y U C d' V' U1 O H 1 1 1 I I I I I 1 1 I 1 I 1 1 "-4
r-1 N d' O a I C) l C . -4 O N . - 1 to O ' N l0 N M N . Cr ' 1' ;P M -V M M to m M M M M cr
O M 01 N --Zr 00 O ßl l0 O -V ßl M in -1 Co N V M . 40 M d' M N M M M N M M "V
( 1 I I 1 1 I I 1 1 1 I ON N N «t' e d' 0) l0 1* . -4 O 'V' 00 N cl' . V. M d' N N M M M Cr) M
"o O --4 CO M N cM Imo- Ul CO 0) O co 0) M r,, N N -W M qtN Co d' N N M N ( N N M M
CO I I I I I I I I I 1 I 1 I I CO I to 01 ßf1 Q1 ri O U1 d' . -4 N N w N N v N M M M e mot' N N M N N N N M M
d' 01 00 IY co O a% d' M O to W Lf) %0 r-4 M N .i M M M CM lt) N N M N N (V N M Co
N I I I I I I I I I 1 1 I I I I I N fV CO %D O V' CO d' '-4 O CO 4' M st' d' Ol CO N -I M M M N Co N N N N N N N N N
r1 N [n d1 to r- M O 0) 1- M N M en CO N N M N M N M N -4 N N N N N N N
%D J I 1 1 1 I 1 I 1 1 I 1 1 1 1 tD m N ý CO U1 0 co CO U) rl Ch Co rl to mot' r4 M N N N CO P-4 H N N . -1 N N (N N
S CO %D . -i in t- -' al N N ýP O 00 O in en Q
U) M N CV fV N -4 ; N N i i ('1 N N Cý i i N M
Ln to dl r.
f N
U) lfl N CO If) N CO M Co . -a '. 4 N N N N N H -4 N ""4 -1 r-4 -4 N r4
0)
MC' C' CO O 4 -4 N cP to -4 N mot' lO N N I- . -i i CV N N lV (V -4 . -1 N r"r -4 -4 N -4
q: ' I 1 1 1 1 I 1 ! I 1 I I I 1 d' t - th n 0) N ' V N M C O L tl (V
4r ' cr C O d+ .
r-1 -4 i CV -I . -4 . -1 . -1 r4 . -1 .1 . -1 "-i r-4 ri '-1 r-1
O N .0 I- l0 r- tf) r-4 fV N d' -1 M M N CCM ri . -4 N -4 . -1 -4 -4 r4 - -4 . -I -4 ri r-I r-1
M i s ! r f w r I ! t 1 r r s M 01 . -t N d' N t' "-a C) r-I ßf1 c'M 0) (N ri Lr) r-t
. -i CV . -i -4 -4 p-i -4 11 1-4 . -;
O r4 M -1 M O O Ln tV . -1 O `V C) CO r1 N rl r4 r-1 ' -4 00 rl . -1 . -i N r-1 '-1 . -1 . -I
N I I I I i I I I I I I 1 I N %D rn rn oo . -+ N N ýr O rn C6
.4 H r-4 -4
co C) 0 M r-1 tC CO 1-4 -4 N 1-4 \O l0 ON "i 1-1 M 61 CC)
-4 N
-' I 1 I I I I 1 I I I I I 1 I 1 H O O O O O O O O :) O O O O O O
Q', Cll U W W a 4 ,' Za O a a a
W
29
C
GL
aM
ý+ U
V) ax
U HO
NHWU
" ýi u'
d' c-I
N N
0 O
O ýD
Co lO
(') O
O1 M
M O
pQ ci
N CO
N lD
M to
CO to
Co in
M C% " o . " . " " " " " " " " " " " " " " to
" Ln M tO N Co In O to to to d N d' In uI 0) " L: cf UZ
0 d' to s, n d' 00 W O CO N tD 0 0 -i 2 Cý ci f1 O l0 CO M l'- N O N to lD N Cl N Z Q " " " " " " " " " " " " , " " " iil r
10 lI
d C1
- + c M
In N
to N
d' N
b N
tD c-ý
N c-4
CV N
O N
Ol OD c4
O N
C N
M N Z
o Ö
o ¢ m o w w o x H a Z Z O O ä 0 0 0
O w N N N O (11 O O (V O N O O O cV O C' N ul U) in d' to d FC tt) C Ln [1' c d u) O
Cl c_ Ul 4
'0' to 01 01 C) O
iN Cl
cV to N r-i
N (V d' C Cf4 M d' N lV M M M N f'-l el
CD N
O N C'
0,, M
M c}'
q
M d' d
Co N
Co N
n! M
01 N
0 M
00 N
vi M
in M
c+ cJ'
'0 Oi N
tN "i 14 cf1 if lfl ci CDI N 9
011 4
011 N c-4 -e M C1 m e fV N Cn (V M (V CV M M
U1 CO to dl M N 'V' Ln O N N W CD M Co cV c-I d' M c1 M d' N N cM N M N CV M M
O en Oi
1 M ýÖ ci 0'.
N v' OD tD W .4 ci
C, M M M M M N CV CV (V N N N M M
N N Co cm N O Co M IN to t- M 10 O M N c- 4 M M cn M m N N N N N N N M M
c1 N
1 tp %--1
u to
CO N (41
Ö, N
4 M
Ö N
Q N
to N N
� N
1 N
N N co, N
op N
m c-1
to c-4
[t' M
N N
O M
W N
M M
ol X
O% H
d' N
M N
N N
O N
N N
LN N
t N
cW p
zW
H t! 1
tO 1 N ci
CD vi
cý
d' s ! r4
1 c-4 M
o ('1 m
N
I u1 N
b' N N
N
1 Cl- N
lD N
N
N
. tn N
(V
1 0) N
N N N
N
I N c4
to %
c-4
I Co c"4
tN vi
ul c4
I N N
c- N 0
n1
I N N
Ö
N
011 c"I
I O N
al c-i
to' c'_
1 0' c'-4
CO c-4
N c"i
1 ci N
0 N
CC) ci
N
In N
f1 lV
N
M (V 0 N
£4
O
H Ui
c+1 t"'I
N s-1
S N
In N
c-I N
UN vi
N N
CO c't
N (N
N
N [t c-i
M c+ N
c"ý
d' c(+
7 s-I
0,1
pý c't
W
N ýr"t
L7 c" N
ci
ID c4
In c'1
N c-_
111 -4
N N
0r (V
cl Vi
uý vi
s-I c'4 c' 0) tN 01 W c-4 N r to vi d' d' CN d'
cr t Cl Q c-'
1 In
! c. 4'
t U
i (N
1 C
!, "'
1 : u)
t ý
! Cl
t ll
t (-"i
1 r
1 ci
CO O N
V' r l
M r'
U1 %-'
c'4 -I M
O c
CP Z
M vi aQ
N c'1
c, vi
l0 c'{
O ri
N Co
c+
1 M vi
N
c1
N
t-t tp
M
C"t ci c
lý ci vi
Ol C! N
to CN
O1
cq vi O1
0)
O1
r4 VA OI
iN
OI
N
O1
CO
OI
c4
O1 vi O1
U) I
O co-t
I O
(p i
O vi
I O
lD 1
O
ä 4 w v W w cý x H a z o öt ä ä oar o
30
Ný
IEIINI
01 N to CO 01 N N N IRV 'a' c" r-4 00 00 N to Q M c") 01 co to "qr r- O l0 l0 M tD ti' O l c) Q
In ff; t! ) lfl 00 to Ol tf) d' ýf' d' N d' tf1 tIl 01 U)
Cl co t- lfl M r- N m Cl O -;! ' co u O Cl U) N CD O Q Cl N 0) O CO H M Qý d' (14 O 0)
1- M CA .
M In C LO co [- 1 ri ri CO O C' V' ri . -I N N (V N N H N N N . -i N N N
m u w w 0 w ý+ a z o Of ä Of öl w w
O l9 N N N O N O O N co N O O O N O -V N in U) U) -V UI -' C' U) to d' -d' wN UI O H 1 I 1 1 1 1 1 1 1 I i I I 1
O 0 d' r-I i! 1 U) N Ol Cl *"i O N a1 O tO fV M N d' w -N P1 -' N N M P M N M M eN
01 al M 0 d' <N to co co O Cl l0 co C1 If) . -1 N +--1 'V' ICN M d' N N M N M N (V (n "W
O1 I I I I I I I I I I I I I I I a1 Iý CO r-4 to O M N to U) C1 dq l0 CO r ß1 N r4: M w m d' N N N C7 N N M M
to Qn N ri to d' CO co M U) 1- C1) co W N O M M lh d' N N N N M N (V M M
a0 N V-1 Nr I I I 1 1 I I I I I I I co ' N Ln 44 M N N c, Ln
( 1 [ M M M N C13 N N N M M N M
O ,M O to M N to to ob N to m mot' N to tO M M M M N N N cV N N N N M
en I I j I I I I I I I I I I 1'- ý r-4 r N f-4 fn N M
.I lÄ M
N M N M
N N N N N fV N . -i M
to 01 1. 01 O O N M Ul O (V to a1 O Q' N N N N N N CV N N N N N N N
to N . -1 M I I I I I I I I I I I I I to I ( C ON N N N r N N
C D Cl N N N
( 4 H -1 . N ý-1 N
M N er lf1 I- CO O '-1 r-1 ca O M UI IZ tý r - M GC N CV N N ý-1 4 C
1 N 1
N . -1 (V t'0 N [s] t11 , r N 1 I I I I I I 1 I 1 I to
8 1 1 1 "i ri N "-i l0 1
l0 C1
r"1 01 �'i
a0 .I
N .1
Cl r-1
N N
rl CV 8 Cl)
to r-1
Cl .i
to N
N N N CV - U)
O O . --I O UI u1 CO co N to 00 H
. N
O CCCJJJ]]]
IV N . -i
H 0-1
!n cV
N 1
N I
N 1
N 1 1 1 1 1
N . 1: d'
A M Go O 4 , r W in ß O1 '-1 . -4 .4 r. 4 . -1 r-I '4 r -1 r -I 4
r-1 N
1- N co to N i+ý1 tO N M. Cr U) CO V4 N C) N 4 -4 -1 r-1 rt ri ri r-4 ri r-1 ri ri r-1
t"1 +-i ý--4 N i i 1 I { 1 I 1 t I I 1 I 1 1 I If) Ifl tp M ' -4 N mot' UI O N M N ý-1 O CO 01 ri , -1 -4 ri '1 -4 . --t . -i . -i ri . -1 r4 ri
G9 d' .0 U) N O r-1 M w . -1 (N l0 O C1 I, - r-1 u-1 . -1 .i r-4 r-1 r-1 . -1 -i ON r1
N N M ý Ö N N N to I C1
N ic 1 N Cl Cl N r-1 . -I r-4 1-4 r4
ý V-4 C%j . -1 M C5 C14 \O It) -1 . r-4 CO f-4 tO CO CO r-1 tp f CO r-4 to
-' in O
I 1 O
1 O
I O
1 O
I O
1 O
1 O
I O I
O 1 i "-+ Cl O o 0
P4
M CJ w w 0 x H a z 0 ä ä ä 'S
31
II
Ca In Ny Rzj
-i P-) Pq to C5 t- C> i 0 Ln w C) Ln C) CO O CO lO l0 d' lý t t) N 61 -4 w N '' Q N N M M M Lo N M M M co M M M cI, U
Ol O M C) N O Cl ýO M cf+ O N to M c O to to C' -4
I [- M O in O N M co 1 O M 01 O O L1 -4 ri -4 r-1 --I r-4 ri -4 rl ri ri . -1 Iq
a ca v W r%4 0 x H a z O ä ä Ol ä
a
w
O M to lQ l0 O to O O to O t O O P to C) N "-1 N N fV N N N N N N N N N 414 N O I I I I I 1 I I I I I I I I I I --i N N M 01 in 00 M 01 to m co N N 00 N '
. --4 -1 N 1-4 N ri N . -1 rl r-4 r-1 N '-4 r-I N
tO '-1 N Co . ýr N N co u) co N -4 l0 N lO (') 1-4 r-I N r-1 (N -1 N e-i -i '-1 ý4 N -4 ! -1 r-4 N
ON I I I I I i I I I I I al 4 to N w -4 N tO O Ul to N N -I N --1 N . -'I . -4 N -i . -1 N
Ul O O LO +--4 LO O N d' to Ln 01 to 10 .4 - -4 N -i N . -1 N r-I r-1 . -4 -4 1-1 ri r4 1-4 N 1 I I 1 I I a0 Vi co ýN CO fM LO QO C) m4m O
r-4 1-1 r-1
('') N M C) M N tO N IV N N -'r m ßl r-1 01 -4 .1 N r1 r- rl r-I . -1 r-4 r-1 r-4 r-I . -1 i-4 N I 1 I t t 1 i I i N 1ý Ol l! 1 ul "-i M L n (Y) co
[Ia
8 rd -4 ri 1q r-1 -4 1-1 --i -1 f9
Cl) it) (N co O M 'c! ' i--1 N N (-
( r-q r-4 0 c 1 I I
I I I to C34 -4 lfl -4 N m N M c 1-1 . -i tQo
. -I M O U) O . M r-4 O O rl . -1 r-4 . - . -i -4 CO r-4 -4 '- ON
L1 1 I 1 1 1 1 1 1 1 1 Ln O C'j a% C1 O t- > O C; m M O -4
. -1 N O ON lD -1 CO CO 01 - to 0) d1 . -4 CN 00 co (
I
O a ý 0 CO ') -4 1- 110 1-0 L" N N- 61 to to N
M d 0 o Ö M N
ä1 lÖ
0) ' ) t l CO in to
Ö1
'D C 00 Ln 00 to m 00 M LO iO 1, v to LO 00 N
V1' N I N
1 d'
-J N
T .0 r N
r- t'n to e r--4 N N W hfl M M c'M to Ö Ö I I I I I Ö Ö Ö Ö l
O Ö Ö Ö
C ) Ö
a' LO U W W C7 `; 'i H I-a X Z O +-{ O!
N Of
M a
IM a
32
I--
OM 0% N
J M
WZ Z W
M
I- u
01 NmN
W JUQ
co %0 "u'-
lY co Li'% I- CO O% N V O Co N Co - CO O OD 0 OD 0 %, O - - 1- M %, o O CO - v c - O
Cn N N M ýt ýt M ttl N M M M M M M N d u;
v oo N ao %0 - v N %z Co Co orn r- - %o r Z M O N N 0 N 0 M Ln %0 1'1 ml N M Co M 1 Q Q W - N pý t') - N M I- - - N CD - - W
fY O O
U Q Q] U W LL CD Z - -ý Z 0 d d O' d U Q Q LL. Li
O M %0 %0 %0 0 %0 O O %0 0 '. O O O O co N - N N N N N N N (N N N (N (N N N
O 1 I t 1 I I 1 i 1 I i I 1 1 I 1 O - N N v O' M OD -t Oý U\ 0 Co O u OD 0% M -
- - N - N - N - - N - N - - - N
- N OD N N1 ct Oý 1ý Oh 00 N - N - N N - - - - - N
O% 1 1 1 ' Iý I ' I\ C\ 0 0 t OD e Co - N - N - N N - - - - N
O 0% cO O N M co O N - - - N - - - N
CO u1 I 1 1 1 I I I- N 1 ul r- 1 '. 0 I 1 CO - Oý M 00 u1 CO - - % - - N - tcl pý
-; t Co N t` 1N CO - Ul V - Ul 00
N 1 O' 1 I 1 cf I 1 1 1 ý- i I I - I N M tý - %0 - to
- u1 -
O -
ct -
- f! 1 -
O - -
- N - - - ö
O N
ýD O ul M M M M M %0 c/)
Z ýp N CO 1 1 1 1 I 1 1 1 I Oý 1 1 t ýD Z
- to 0' ýt N N - N N K) N N t1l W
cn N
9-1 rj
u1 1- Co 1 - - 1 N 1 tc _ M 0 '. 0 O O 0 - M
O N O O (N O O N _ _ _ ýf 1 ýD 1
1 1 1 1 CT 1 %0 1 t I.; Y
0% - '0 0% O\ N -- ON 00 O\ -
0 Ul Ul 00 a7 Z V Co CO Co ü1
M 00 1 t I 1 I 1 0 f 1. I I t N- CO h fit' O' 'ý7 ýp N U\ - M 1- N N- C- CO
1N Ln d 0% '. O \o I- N 1 n CO t% I %0 1 1 N cc %0 1 M 1 1N 1 N
%0 4 M Co U1 U'% cc
Ul N 1N N M Ul N N - n n N ý7 %0 u
0 0 0 0 0 0 0 0 0 O
Ö 0 0
< M U W LL- (D M Z O O" O" G7 3
LL- W
33
Fr
H E-+
0
d'
O Lf -Ir .O 0 c m co O N (n c- 0 N m 0 " :n Ol O N CD ri u) LN cl m c4 ci t- Ln (; ) O "
" N v-I tý' d d' M L) N M M Cl Ct' N Cl N t! 1 "
c- to O q% c d co (V 'V' Ln N lU O1% co 2 r+ O %-1 c+ M %-I N LI to v- co 40 V' '0 Q1 c-A 2
.ý k7 " "
N "
d "
O- "
" "
N "
c-1 "
M "
C-- "
N "
-, 0.
N "
N "
" " 4 "
V
Ei
Q ¢ m U w w C) x a : Fl z O cl C\j Cr c C Q w w
0 M k9 tp %D 0 W O O 1 O to O O O W 0 N N-4 N (V N N N (V N N N N t4 N N N O T-1 t 1 ! 1 l 1
` i I 1
O1 "I
M 01 t ý ri Tq N
c N
ýp T-4
t J N
ßl cI ca
O N c-4 fV ri ri
Cp T-1 N
tC rI N c-l N N W LO O\ co O N a) N Cl c- N c-4 N cc- N c'4 ; c-I i C\1 c-i i r-4
(TI 11,
t 'p a1
A N N N
N c cd ri ["i ri N
in r4
O T-i
O N
to ri
CO to c-I
O C\j
[N `.,
N c i
t0 r-I
01 c-I c+
lD vi
l0 rl
TA c
03 ; 1
-i
l
0
I
I ýpI -I
i N
rq ci c c t
c-I ' W O W
c-I ý
%-ý N O U')
T-A Ln IC-4
ON r-
N N A I II
In Ol 4 4
lI N
r4 c1 ri cI c-I cl N ri c-i
(V %D ci d' v Cli Cl) d' M c+
ci' a0
M cl
th Tq
(0
tD Ok1
1 N
A j A
.4 C\l A A 1
in ýD
vi r-ý W
z ci c-1 N
Cl e-I C4
c-1 N ci
c-I '
Cl c-4 t-
cl rl
N vi
N ci tN
T-4 c-I
N ci
d c-I
(rj E-4
A N N O1 c11 O'.
Ö
vlý c- 8
tä N
H
CI) cý c-I ri c-'1 c-1 c1 t-1 vI IC-4 Vi
ON tD %-I t0 CO 0
0) C\j
U) O O
ý ý O'H
'
Ill N (1 ts ca 01 Ik Ok c-ý
CO to O c-t d' l0 00 Ln
O v 11 CO 0) Co. d' W CO co
M t r t-I I
cý M
N d' N cn d' t- r) 01 !V l0 l0 c"') U) N lC N 1
l"
to N N
1
In Cn to N N N ca d' to Cl) N d' CO dý
ö ö of ö i
ol ;
O d d Cl) O O d ö of 1
O . 7.4
v d Co O w W U' X H º7 X Z O ör
or ör C ü
w w
34
)
M
W
"'l
2 H
C,
H
4
w
4 w
0 In N
Ln
w 4
d' CO
CO c-+
rl O
N c-{
O N
N [-
N N
co at
CO : fl
CO M
CO T-.
O ON
N 4.0
114: ' c'
co N
(V 0 "
C; " N N d' '' c' N U) N C4 C4 M M N M N [t " tI) Ui
V4' CO d' at M O CO M CO C N N at CV LI) pl 2 Cr Gý Ul N CO O U) M N r- in CO l0 CO ON N z
O O to M OD O ý4 cI M N -1 N (N CN "I CI , 14 ts %-4 IH c-I %H cI cl CI c-I c-I r4 t"1 c
O O a Ü
w S W U w ; _., 0 x H 4 z 0 cr O '
o a Ü
w
o cn i) CO kD o tO o o tO o 9 10 0 C tD 0 N cl CV N N N N N N N N N N N N (V O N-4 I
CLi (N a 4 8
all 0) A I
al A
CV 4
all oll 4 V,
r c-I N N c-1 c-i N cl c'i 0 N t-4 ci c-1 N
N ci
c-l c4
O N
0) cl
CO cl
N N
0 c1
In t-4 cI
00 cI
c'ý N
M r
CO rI
CO t-i
C) N
I I kD ci
CO 0 (V
M c-4 c-l
Oi ca
Cv T-1
N T-1
(V N
to cI
O cI
01 cl
CO Cl
N c4
N c-I
01 c-I
co c-I
N IH
t0 V-4
N r4
Op c-I
c-I %-I
tO c-I
N c<
c- N
OD .4 Is r-, C-1 1-11 L11) 1ý c
ý
t-1 c-1 IH t1 c4 cl c-I N
M cl 01
Cp c-'ý
cm c-I
U) c- I
to c-I
to t0 vi
O c- I
ý4' cl
CO tl
to c-I
O %-i
U) c-I
U) vi
0) c-I
tÖ 1- J
M LÄ In In J
.4 <` N
cl c-I C'l ti 0 c-+ c-I r"I
p
c\j CO
!!
ý
' T-4
c0-+
ö' T-1
T4
cc-I 'rI: r -4
c
cl co
r"
M r+
c\11 r-I
d' c-i
cl
d' cI CO
M r+
r+
d' c-+
V4
tD IH
1 wa O En
p en 0 M c-I Cl) c"ý N N cI N Cl) 6
EI U)
fl c-I tý T-4
1 (V c-I
CO 1
N
T-4 1
Ql
Y+ 1
(V cl
c-I 1
Ol
c+ 1
cV c-I
tO c-1 c-I 1
Cl
c-I !
cA c-I
C- c-I 1
O t-I
c-I T-4 1
cI c-I
H Ui
ON U) rl to
u; CO
ý T, OD
kb cl U7
4 c-I
011 T-4 -
a1 l0
ý Ol
CIO c4
O
0)
co v. 01 V' tO -H to C: > T-4 M co at co N
0 cl co
N to en cc
ll N1 N d,
cc l-
Cl) N
ON [i
N to In
N! 1 to d' to CO r
N
%-4 d'
a N
ý is
d a eI
a to
ö c-l
ý lO
a O
a d'
a 14)
ý ul
ö M
ý to
ý r-
8 M
a ý Wo 4
G4 4 GO U 67 Cr U' . 7C H º-7 F. 2; O Cr
cy cm m cr a 2 Q Gy
35
Q Ln r
N
l0 -i
N
! - ý
M
Lf1 ['ý
m
r- C0
m
O . -I
m
M O
tf)
co lý
fV
U1 '-t
M
"I er
M
61 CO
N
L N 4
O 01
N
C1 N
M
fV lfl
N
M U1
C! 2 U)
)
H
N m
tV
tn
o -r . -4 m r- Ln c) co rn r O rn 'n r a0 1ý O C1 1ý 00 CO I--i O 1C1 9 r-I O M c
Ln C' 00 m M O M 00 '--1 N N [-- N M C )
. --I -4 - 1-4 ri
M . 4v a'i M U W (X4 (9 x' H ýl x ' O ä ä ä ä
oI
rn
Co 1
11
10 1
M
N
O (h 'D kp %0 O to c) O l0 O 1.0 OO C) l0 N . --I NNNNNNNNNNNNNN
M, -14
M M, a1
N N' -i N 00 mM (\
Ü') I
r1 .iN . -1 r-I N "-i . -1 N r-I N r-i . --I r-1 N
%D O. -I IN co (71 Na C1 00 ON 00 00 . --I . -I -I CV ri -$ ri N r1 r-4 -i -i N r-4 . -4 -4 N
r-I CD C) -I-I 010
Ir ON
lqM C1 --Zr' Lt) 00 '. O MN lG co . -1 r- [-- O
-1 a% rl r--1 'i . -1 -i P--4 . -i r-I ri .qN 111I11III1II
00 00 MmNNt! 1 N l0 %D Ol . -1 r-1 -4r.. ý. -4 r-1 r-1 . -i rl r4
M 1, NN lD Lf1 Lf) '-I -, r If, ýo c, t. [) In 00
MÖMJ Cl lÖ
ri rl ri r-I rl -4 r-1
N to O . -I Lfl N d' 'd' mot' Lfl '--I CO -1 "1 . -1
rl 1 ä1 Öu
u-I tý
M VI'
ri r4
O qv e' 0 C') r-I MNNMM
r-i to ý--t CO ON u-I rl ri 00 . -4 u-1 ri l0 -1 r-1 r-I IN
NNIM O1 NNON u-I i-4 N 1-4 - P-4 r4 r-1 rji r-I
. -I N rl r-1 OO '-i rl Ch U) rl l0 N r-1 co r-I ' d1 ri ; Ln i . -I +
'D C-4 tÖ to co mM
C) C> OOO co M r-I LI) to Lf) . -i d' 00 P--I N 'cr CO -I 00
lo (N CO C') -tv C% Co MM l0 00 L r) MN 01
t, N
Ln rl 00
O
m
00
N
Ln
qw
M
N
to r', Iý N f9 l0 N t0 "-' "" \O "4, ri "O t- -r 1 1 1 1 1 1 1 I 1 1 1 1 1 1 1
O O O O O O O C D O O O O O O O
a' 6ý U W W C9 º't'. 1"'ý ý"a ý+ 'ýZ. a O a a a w
36
H H
Ü a
M Co ý-1
z
H ýä
M tO CO if'1 CO -4 pn CO O O O N Q " C h N l0 to 61 CO Ol 1-I ßl ýN Ql CO
N M to m In N M M M M M M M 4 Q
to
al CO LO N -'M - O 00 O Lr) N to (0 CO N 00 (Y ßl r-1 r-1 ,+ N CO -i to ! . to
cf' m N ro M [- M . -t 4
N- ý4 N M
rý v w w x H a z o ä Cl Cl öl w
Co rn lo to to O to C) C) lD O tO O O O l0 O N r-I N N N N N N N N N N (N N N N O ý-1 1 1 I l 1 1 t I i1 I ! 1 I I ' co N v o l d' 1 Q N --T c to o 4 - C A O tO O M
-l . -i N N N r-1 N r-i r-i N -4 N '. -! ý--1 N
l- lM r1 PM 00 M DO LI) O CO o 00 0) N N N N - Cpl -4 N . -I 4 r-! r4 '4 N
CA I I I 1 1 I l I I I I I a1 l0 -1 00 -1 CO c ;l Ol N C tý CO . --I N -1 N -4 -4 r4 r-1 -4 -4 N
LA -1 0 N O N- N N N co to h CO ýp N O -1 N -4 N -I . -1 -4 -4 -I -4 r-4 -1 N
O I I 1 I t 1 1 I 1 1 00 dl to ON O r1 In l0 CV to c r-1 r-q r-I N -4 r-i '-1 ri -4 ri
dý O Co CO to 01 to CD N LO r-I tf1 to OO . --I i-1 r-1 r-1 ri r-i . -"I r-f r-1 r-4 r-4 r-1 ' -I tý I I I I I I 1 1 1 I I I I I tl- M c t- N to tf1 I- Ol to CO lqv O r-
4 r-1 r-1 r-I ri . -1 r-4 . --I r-1 '-1 r-1 -4
N to 1 to -ý: f' to to In N CO M ý: N to . -1 cc .1 - . -"I . --1 . -i . -d 00 r-1 .j _4 61 -i rl -1 lO I I I I I I I I I I to
ß 6 M M ý CO N M ß -1 r r -1 r 1 r t r I r
8 U) r-i d' N N N d' CO r-I O r-1 N tit
r-q ON ri '--I r-I r-1 1, - 1-4 h r-I -i r--I "Z Ln I I I I I 1 I I 1 1 I 1 1 1 *ý E
Cl) O CO Co O ý-I I
O M to N C) 01 tO 0 -4 r-i E { r-1 . -4 .-4 -4 r--I -I .-I C/)
(V O (N ' O O m l0 '--1 1- 01 1-1 CN r-! to .1 Q1 CO to a- r-I r-4 a' 1 I I I I I I I I I I mot' Co to cc (7, 00 r-q d' ON 00 o) al
r-1 O I'- t! 1 r-1 u1 N CO h -1 c'M 00 00 N 'V N 00 00 C'`) I I I I I I I I I I M O to t, tO N to ýp 11
to V' C; ) M ºn lD d1 N w N 1 1
L) N M CO N
113' CO 00 1.4 c! ' ul N N F-I U) Ln V' r-I CO t ' '-1 I 1 I I I I I I I I I I
o V O O 0 0 O O O O O O O
J> I O
I O
I O
I O . -1
V tý U W W C7 x H i-o z z o Cy O I ä Of
2
U w
37
C
0 w rn tn aN
4Ü N U)
wa ä
H°w ýn
In N N c-4 M d in l0 0 O CN M N M c+ co " C) aý O Q) (l u') N as a, N N 0 0 to C'4 t0 " CV T-4 ' M to M Lý CV 1V [*) M 1; [h M C1 c' "
a)
z N M as c- O O
07 a)
O T-
O lo
c-I M
N c-1
M M
h Q)
N N T 0
U'i 0 2
" " " " " " " " " " " " " " " " N '$ al M c-I N M N N ci c N c N
cý c-{ cý c-{ c-{ C4 C'4 c-I cý c-4 c-1
O
Q CO U w C:. O x H Z z 0 Cl ä 0 ä
Q
c: a w
0 O N
() c-4
(-D fV
to CV
0 (V
O fV
O N
O N
O (V
0D N
O N
to CV
O (V
O (V
O CV
to N 0
TA r1 X-4
+1 1
N
1 c-I
) fV
i oil r-1 (11
j -H "c4 o1 [U 0:
1) c-i r+ CV LI n c4
I a, ri
I rn c-I CV
0
lA c"ý
m N
kp ý
M (V
Ua c4
M N
co c'i
CU c'{
a1 vi
S c' {
O N c-I
CO 10
CO c"i
N N
i -I N V j M ý ý c CV N ( la al L C CV N s-i N V-4 c'l cI ri c'" c-I
U) t'l
0 74
N Lr) ci
%-4 N
W c'l
vi (V
E- vi
T-4 c"f c-ý N-4 ri %-I vi N
co of V-4
d' %-I
al c-i
a1 -I
to c-i c-i
1 c-i
Oi (V
co
d' cl
cý Ol c-1
r M
.1
CO t. n C'1
co t"'l
"1 to Tl
O vi
I u1 c-l
d' c'l
Ln s-4
.1
O c-i
Ln cl
4 Ln c-I
4 C) cI
1 r- T-1 T-1 T-4 cl ri T-4 cl
W 08
O
N c"1 CO
W ri
<ý
O c"I
In Vi
tý-I
.1 c'i
t-I
d' c-l
ci
t1) r4i
l
co '' cl
ci
en c-{
ri
M V--,
c-I
(71 M c4
T-4
M H
l0 c-4
O
19 Ei
H ü)
ut
T4 cl 3
cl
[N d' cl A
cý
O.,
CIO
m c-I
'A cl
N c-i
c s-l Ö
c-i
M t"l N
1
(V
c,, Vi
c-{ c-1 Ö
ci
l -H Ö
\-
N 1
s-I T-1 Ö
c-I
N H
cý l
d' H N
ci
U UZ
Et Ui
Oº tý c. j ci
t
N1
lA
cl
CIO Ö\
a%
l- cN-4
ci cý
U) ri O)
CIO
O1 LO O'
ap
0
A -H Ö
T-4
10
0o Ln 4
Ln N w LO 0 cl C" aO N CD ci' :s rý o,
Cl) I M N a; in Lö
c ö cý
N S
l0 d' W M I U) ( 00 N 0 l0 u') M la to a) N
NI fV U') LA N LO N
TA u1
I O
CU I
O
W I
O
N I
O
CU I
O
41 T4 1 O
EN 1
O t O1
d' 1
O
d' 1 O
c1' 8 IO u1
1O d' I
O In 1
O
u a ao v w w 0 x H a z z o cr c cr Cl H u
38
u)
m
L) H H
O a
4 i Ö
W
W rý
[w
N ON
N U ul N4
v
8
[N 0 L O tN c-I N d' 00 V1 01 c-i l0 U) to c-i s N O CC 5% U cV ci tN Ln M TA O'. 0 M 01 to " "
4. CV M M C; Cn U) N [h (*) !f M M f'') CV vv "
L: M rn ci O N 0 00 t0 0'. N U') N co d' G 2 CO N N ("J : +) CU C') O ()' c-4 Ol cc) c-i N l0 : t' Z
O UD M O', 1V 4
c-{ cM N ri CV C) c'! c'i C' . e. -{ c-I c+ c-+ c1 c-4 V-4 c-I r c-I c-+ c-t
ix P-i
w Q U) U w w cD x H a z O
0 O
N a
(fl of
d of
C4
0 w
O O CV
!
c-I
M c-I
I
c-i
UD N
1
(V
"0 (V
1 r
i0 N
I
N
O CV
I
1D CV N
CV
O (V
0'. rý
O cV
t'i
l0 (V
0k ci
O fV
ß? ri
i0 CV
(V
O "
u1 c-1
O (V
CO c1
O (V
00 c-t
(D CV
N
0
10 ci
I
0 c-4
c+ N
1 ý
S c-I
I.
H N I
O
cli
l- c-I
c-! N
I O
co rq
Ul t-- 4
00 ci
I
c
N ca
O N
I
T-4
d` c-i
t- ri
t, ri
N CV
T-1 01 a'. c'i
pp c-+
u) c'"1 4
lý
rn ri I
Tli
ko ri m c-A
t- ri
r- VA
U) r-i
M c-4
CV c'i
ko c-1
Ul ci
17 0 U)
9 0
rn r!
CV c-I
c-a
t( %-4
c-I
to 0
c-+ CV
0'. c-I
CO
M c-! CO
N -H
IH
C') cI
T-4
N ci
U) ri
ID cI
t! ) c-
ri c
,I
d' c-!
r-I
In T-4
T-4
UJ ri
cI c"!
d' IH
d' ca
W 'H
W S O
(V ci
ci ca
S tf) c't 4
c'! ci
O c+
c-i
cM c+
N
ct
C'i
ci
c{ Ol
"
CV t
M c--!
! 11 c-1
c+
Ol
CO
r") c-t N
cý
Cfl
N
S s -t
Lin c+ W
Ö
U
H
Ui
N
0
I
c-I CI OI
cl
1
t*.
N
c-I
rI
Ö
cI
N
ri
M
r"t
ct
I
c4
T-1
c-1
ý H (J)
ON U1 c'i ID ON
CO t0'+
O; co
10,
V-4 c-t l0
In 0'.
co O ri
O; O c-!
(31 l0 C,
ý O Vq
0) Cl c-t
ol
1 r. 11
t 1 C. - zz,
i cl
3 M
t C-
t .
I i
V^ !
OD
ID (+1 OD M 11) U) C°1 CO N U) l0 tD (f) N 5 i N
t) In In N
ti) N U) N M ü) N N c't c In U ' !
O i
O i
O 1
O 1
O 1
O I
O I
O I
O 1
O 1
O 1
O I
O I
O I
O I
O
4N U 4 to U 14 54 Ch x H a . -. Z O
0 Öt Cl Off'
O U
39
x
cý 1.0
RE,
N
z
ul
O M q o0 t- -w l0 mot' CT N Co tf) T O , t- f- N [- Ln w N p CO 00 Ln m N
" " N N M M IV M 'ci' N M M M M N M N zi'
"O -, zr N Co l0 LI) M i m "o r) CO /D O q. N O M N /D ri O O N do -4 L) . 4zf' V co a0
\O M C1 1- N r4 M O6 O N M CO 14 ý4 4
0 U ýa w cý x H a z o ä ä Of öc
w
O m 'o lp lO O l0 O O l4 O l0 O O O l0 O N -4 N N N N N N N N N N fN N CV N O 1-1 1 1 I I I I 1 I I 1 I I 1 I I I "--4 o .i . -i Ol m al ri 00 lJ OD CO N hfl 01 ý M
r'4 r-i tV -1 (V rl N '4 -1 . -i r-1 N r-1 _4 N
0 O Co N 00 O I- 1- N -1 co I, N r-i N -4 CV r-1 {V v1 ''-1 . -1 N i-1 ri r)
aý 1 1 I 1 I 1 I 1 1 1 M (7N %o c) -4 a% qzzp .
N r-1 N -4
ul co L) "- ß1 tO Co l0 CI) Ln to 01 N 1- to ri -1 01 -4 . -1 -4 ri - --I ri r-1 r-1 r-I ri r-I r-I N
I CO Co dý [I-
Q' N lÖ to
Ö
v-I r-I ri rf r-1 rl r4 N
M t0 M %D LI) to Ln N V' LO N . -1 to to al 1-1 Co - fq v-4 1-4 '-I r-i ri 1--I 1-4 - r-1 . --1 r-1 rl
tý i i i I I I i i I 1 1 1 N in N Q' v CI) 1i' tO O 1. Z14 di N
r4 r-i r-1 r-i 1-1 ri -4 1-1 r--I r-1
N V' ý-1 c'') d' tM --: r O N M in m m tD ri IN r-I r-i '-1 e-I l r-' '-i '-i r-1 '-i d1 vI v-i r-II
to 1 I I I I I I I I I tO O N m 1.4 dl . -i N (V Ul . -i r-i -4 r- 1-4 ri rI r-1 rl
r1 M rl N O M O N (I) v1 r-1 cf' r-I tp -I m 4 . -i r-I ri CO vi v-i ri 00 r-4 r-1 r-I
LO I 1 1 1 I 1 I I I 1 in N co O m N C- r-1 N l- O M
jz] rl r-I . -I ri 1-4 r-1 r1 ri [ý] IZ
i O
In -4 ° ý ° 1 O
in -4 . N a1 co -i r k. 0 in -1 . lc 01 1-4 d' I I $ I 1 C 1' I 1 I "'
++ tt ON CO tO 00 O CO 01 ý
CD C) 00 d' tO N 00 ßf1 -4-4 t1) N ßl 01 Lfl N Co r-i M M o Jr d 6 i I M ßl
t to o C p p
N ý M Ö N
f' N N f 1 l
co to to l 1 N
ko r-I l0 N N Ln N N -4 -ýQm Lf1 tf1 co d' lo t. p 1--1 1 I 1 I I 1 I 1 1 I 1 I I I I I -I
0 O O C D O 0 O O O O O O C D O O
a a o Q 4 w 0 w w 0 x H a Z z o öi ä of ä
w
40
)
_)
r- cr)
6
ýI ýý
ui N II z N
ßl in o lc' In (- M N ß1 l0 Co Co d' co Co CO N CO m N Co Ol 00 N lD N Co N O Ol in r-1
M . -1 M M N In N (4 Co N cY M N Co lcM U:
M O lfl d' /O N lo N l0 N LO N N 00 ßl Co ! - l0 00 N [ý cM r1 CO u0 O lm r, 1 l0 d' lD t! 1 Z
O N Co O t! 1 r1 O Co O ßl M 00 00 m in
4 W U W W C9 x H F7 2 O 0 OI O+ Of W
W
O Co \p lD lO O ' O O lO 0 %D O O O l0 O N -4 tV N N N N N N N N N N N N N O
I I I 1 I I I 1 1 1 1 1 I 1 1 1 H t, CV N O d' CO : t- to a\ to N "o in l- dr -1 r1 N CV N . '1 CV r"1 r-1 r-I r-1 N r-I . -1 r -I N
l0 s4 r-1 4G M N Co l0 CO iÖ 4 In d' '. 0 Co rl , -1 tV N r-4 (V N . -I r4 r-1 cV
I I I 1 1 I C1 O N 'p N in Co "o O mot' Co N N - "-I fN -4 -1 N r-4 -4 N
In a 01 %D N If) r-4 :' N If) -i' o Co N Ln -1 -4 . -I r-1 '-4 N -I N r-I r-4 ri . -I r-I .4 rl '-I N
co 1 I r I N d Ö O CO 1 . CO l C O p r-1 r--1 N ri r-I r-I ri 14 . -I CV
N Co O d' [N M r-4 Co N N -4 rl Co d1 rt 01 r1 -4 N r-I r-4 rl r-4 r4 ri r-I -1 1-1 -1 . -1
N 1 1 I 1 1 I I 1 I 1 1 N M N CO Co N %D O O N CO r-I rl . -4 r-I r. 4 r-1 . -1 r. l rý r-i rl
(V l0 . -4 N N Co N O . -1 r-1 In r-I [N r-1 CO '-1 r4 r-1 r-I r-I . -1 r-I r1 p. 4 r-1 Ol Ol '-1 r-1
(x] tD 1 1 1 1 1 1 1 1 1 t 1 P4 r+ ýr %D r-+ H Olt O O 'r Cl to Q . -I . -1 ri r-i -4 '-1 r-1 "I l1
m CUJi 8 M ° ° M °
N -1 . °
ý4 r-4 00 M a1 " -4 00 CO Oll r-q I I 1 N a d' ä
0 0 -Jr r-i
l r-1 rl rl U)
r-1 Co O 01 l0 r -I co rl . -i N 0) 1'- CO CO . -1
d' 1 1 1 1 I I 1 I I 1 I 1 1 O N N 01 l0 tD [ý r 01 l0 W l0 N
N in 01 lC "-F CO to CO in W I'D CO in N in -q Co I - M I n I M
\Ö (V
to I N I N (
in d' d' 0'. l0 N '. 0 'N <M in .o d' M d' CO N
d' I CO n I N
w i N Co t-
R) r) in Co t- ul r-i Ln co M d" N .4 Ö t Ö Cl)
Ö Ö Ö Ö Ö p p ' p
N Ö
tC) Ö
FO U W W C7 x H ý7 z m ä a ce)
w w
41
O a4 Ln r Cv z
qM 1
NO CKII 1-4
9i H
" ,^
" U)
! c
" 7
!
Lý
ci
Co C
" ('()
Co Oý
" M
in d
" c"
Co GO
" CV
O
" LO
Co O
e M,
CO
" N
Cl Lý
" M
Cl if)
" CV
M Co
" d'
CT C7'
" N
' c1
" M
N Cpl
i 14
in V
f (ý
V
<;
X
N C'
O c-+
ý' H C')
rt
[ in
i! 1 Lo
.' c-
Co N
N c-ý
0 WO
ci c-ý
V' lD
c1 ci
cH O
0;
0 co
(31
O u)
O s
in to
N ca
c-{ N
CO
co C')
C)
Co Ql
O c+
0 Co
' ci
o E-+ o w
Co t,, w e, LD z H a Z , a, 0 0 CI
N CY
M Or
ýS Cr
p
Ü w
O r'
O N
1 co rA
Cl) cI
1 di ri
Co N
1
N N
Co N
1
J c-+
Co N
N
0 N
CIO ci
CO N
ý,; N
O N
oll T-I
O (V
M T-i
Co (N
C) H
O N
Co H
Co N N
N
0 N
ci
0 tV
ui c-I
0 N
ap T-i
Co N 4
CU
O
C)
N c"H
I CO ci
H H
ci N
I O N
or) ci
I LD c-i
Cß'1 fV
I N N
N c-{
N N
1 ri N
co T -I
I 1ý %-I
(V ci
N ci
1 Co T-i
in c-i
ci N
I 0 N
I! ) t-I 4
e-i
NT T-i
IN cH
i tD H
M N
i N N
ON
00
tf) T-1
O T-i
rn ci
co H
In c-i
VI' ci
c+ N
Cil c-I
t0 T-i
O cm
Ol cI
CD
if ci
V" H
Ln c-
H
4' ri
0) VA
CO ci
C') H
H
C') ca
(V 0 H
U) a-4
ci N
N
co
c3'
11, Irf
N M
N
c-I N ri
in q
t) H
d' O M
N
N
M
N
1-4
[N
CH
H ci
Ö
H
N \-4
p1
I
T-4
CO
p
N H N
CO c-H
I in H
H c-4
I O I
Co 7-4
I to H
d' ci
I C') ci
'' H I
CV c-H
M H
I N c-+
0)
I 0)
c-4 H
I O cl
c-H H
if) c-H
I cý) c+
O tý
I Ol
01 CV t--1
t vi
I CD r+
CO W Ö
Z
EWi V)
H I
c-I
cN r+ c+i c-H
0)
CO
c
A
cI
C\j
cI H r-, Oi
H r C)) H H Ö
rI
CO N
9
%
OI vi
H N H
ý}
6 in
H
U]
0 Co N c-i
ci H
N
CO
N H
rL c-i
r'I c-I Ö
c{
00( I
O H Ö:
CO co [!
Ql IO
c
0)
o
t-
ýj
W Ol I
N c-H
I
T-i 0
R
N LP, 0 H
c
in 0 ci ) lt)
ý"
C) : '1 \P CN 5
COQ
t1)
c
N
10
C)
c, II
CM (D %T OD Cl)
I N cl
N M t1ý C-
N
ýý C" 1 o
Cýl I
O r-
1 o
CIJ 1
0 ' ö
C
0 ö 1 0
1 0
1 0
I ol
I 1
o cV
O (y) ö
to
0
oHý
Q
w Q m u W W t7 x H a z o
ä a C\j en
0 ä Q
ý,
42
rn c)
w
H
0 H
äo
z
L, Ea m H
H ofu
wä ý3
Ir) II? 0 C' a) d O O Q to v4 lD c+7 N co " c4 a? O N U) to Ol 10 l0 l0 N- d ! - M l0 l0 " " M c- U) N N M (V d' N M M c}'
f
S O c'+ Ci 7-1 M t; M (T CV to l0 C) lD QA G*' lQ V' N N [S N -i CO L- t-I h In Q\ N lD d' `ý.
1 " " "
' " " "
C .
(v 4" "
N " " r "
h " s
4 "
.4 I Z
O t
h C c-I
0) M c-+ cý c{
O cl
O ci
(N ci
O) ri cI
K
Ü Ü
o M U w , c9 H a Z z O OJ O a a u w w
0 C k0 tD t. O Q 0 0 1.0 0 tD 0 O O to O N ci N N N N N N N N N (V N N N CV O
1 1 I 1 1 (N CIO 14 pp!
1 1 I 1 I N
ý -i V'+ c-{ N -{ -I i C\j H N N c c c c c
lfl c-I 01 M C tN M N h c-{ M d CO M c i N rA (V c-i N H ri r'' r N ci c-I c-4 N
N I N 0 in Ö N lD
N 1
cI 'N N T -i cI T-4 c-4 N c-i
: 17 O O In c-4 Ný c-4 In c-4 tf 01 N M 10 C \J c4 c'4 N cq N ci N cI C"4 0 c-I v-I c-1 VA t-1 fV
(0 4 ' 4 l I 0 4 L, 1-11 N j co c-4 , O c-4 c-4 , o T -i rq N ci ýI c-I
M a) M a) In Ol d' 0 M M lD O U) 00
h 1 ý
1 ý
1 I 1 N
I u1
i O
4 1 P
h t c
c cý
N cI 0+ c c -cl ri r1
CV lD 0 t0 In M 7-1 N d' c'7 l0 r4 OD Ti ca H H rý r-I 0) r -i c-4 0 O\ ON ri rq
O I 4 m N co Ö c4 () 00
I Ln
(4 c-1 to ca c-1 cI c4 c-4 v-I
ý O
bi L Ul
0 c N c 0 N h t4 - Q) cc-I c-I -i c h 0', c -4 h co ri -i H E
N 1 I N l V V4 N 0 co 1 [ - p C E-+
(! ] c-f v-I t-1 rq c-I -(-4 En
V4 0) l0 r-4 co T-4 r4 Q) c'4 0 h m l0 l0 at c"{
1 I l I i k o 0 0 0 c (; , l C t aQ
t'i CI
0 h In M t0 0D ri O 0) In 1-0 CO CO If) h O-
C I I I I 1 I 1{ ý^" ýfl u) ý" C t l SI
In q; r ao d' h O) co M I) lo Lo qr l0 lo N t
' ý
i
N h 1
' N
i N
d [ a N
M M to N 0 CO r"4 t. O (V M in M N c-+ to to 1 8 1 1 1 1 1 1 1 1 1 1 1 1 1 1 O O O O O O O O O O O O O O O
ýo Ö Eý U Q as u w w C7 x H a o
(Y ci cf Z5 u ý w w
43
O ýi'
W
H
z U H
a
w
w
m N
z
Ln
0 F9
4
cý N to Ln M 0 O7 to N co cc N N O N C' " C% co 0 in O 0 C (n tO N L') in (; -' CV m to " " N --i Q' M
4 M Lt) CV CV C^, N N M M d' "
Erl
O\ O O 0 c- O 0-1 C- N 0 In to N N [N lO c c- c-i V' N to N l0 Ol N ci c-+ O N c- r4 s- Z
" O
" N
" Cn
" C-
" N
" M
" cý
" c-ý
" N
" O
" ra
" M
" 0
" 0;
" CV
" ýt
EA v ¢ ca 0 W w (D x H a Z z O cJ C 0 C w w
O M to lO Wp O to O O tO O tO O O O tO 0 N t-4 N N N N CV CV N N N N CV N N N 0
ý
TA N
V-4 N rl c-{ -I i N
-1 i N c T r
to O [-i a) O OA m CO N co lO si in i) N C \j ci c4 N ci (\I c-I N s-i ci ci 7-1 (\I ci c-4 c-i N
A1 to ci N 12)
4 A C- C'i T-1
" N c4 %-1 c-1
d' -
CO d' -4
o% I
0 CV
to -'A
r4 I
in -i
to 0 C'ý) -4
tV i
lO 1 ca
co c c c % c N N V c c N
I
TA vi TA c+ ca 'I cr- i
N CV tO N Co d' 0 d' (') [N O s-i In C) cý al s-I ci ca s-i c '4 ci t-4 ci C- c' ri t-I t, s-i
'
'A A 1ý cl
c-4 VA co
to c+
O c-+
d' c-i
to %-i
to c-1
M c-4 01
c-H VA
N 4 to
%-I Q- 0' C c-I
tO c4
94 4 A 4
C"i CD O OD A
A 1
r4 ri . a Q
-I ý° N m V4 C)
C\1 -+ t 0 %N '1 c- ON IH
VI C\j vi tN CO
ý ý+
E-4 C
N dl cý Ö I', [lý cý H
c-i ci si vq c-I s-4 Vý
C) to V4 N T4 r4 '00 cI W OD M 0
l0 tO 0
c-i Ö
It oll "I
IL oll elk Ln o A ''
to in - it) Co O
In CO l() to co W In ti co c-, M, ß , A
u'1 tN CN M
O t
in d' N d' tD C; N cV t-- Cß'1 N N M to to N A I A 1
9 Ln N o N
i t
d' N Ln cm ýN l0 si tD CV M to d' m ci ýt M `i
8 ö ö ö ö ö ö ö ö ö 0 ö ö ö ö ö `-,
4 cn v w w 0 X H a X z o ä c ä w w
44
z oa +
ýN
'Vol m Eä
H ý°
U) r-i O lD O N in N lO t. f) N c in N o CO O r- O llý C1 Lf 10 L! M L1 rý V' C1 N 1D
; M N -, T M M N to N N M N .4 N N m d' U
N M r-4 CO in If' M IN l0 Cl) co m lw N M CO CD N to r4 C". 11.1 111 Co N in N r: O tO M m ri M N N rý M r-I N l- 00 N N
I
4 W v w w C7 x H a z o a1 No M a s
w w O M l0 ko %D CD lfl O O C) O kD O O O l0 N rI N (N N N N N N N N N N N N N O U I I I I I 1 1 1 1 4 I I 1 I 1 1 1 -4
. --i 11 . -1 (V) co Al O II) co CO 00 M ' ý t! 1 ß 1 M r-1 r-1 N r-I N -1 N r-4 /-1 . -1 . -1 N '-I r-i r-4 N
%D co N N Cl CO qzr N M l- Il- N M d' 00 N r-I -4 N -4 -4 -1 N 1 . -4 N 1-1 -1 1-1 N
C1 1 1 I 1 I 1 1 1 I 1 1 I o" to -4 in r N l0 l0 O' N M 1-1 N /-1 . -4 N r -i u r1 . -i f -i . -i N
O eM l0 N N l0 4 in in 00 r-I N N O ; C1 N . -1 .4 N 4 i r-1 r-1 . --1 . -i . -I N
CO ' i
co M 00 cM C1 in d' in l0 CO
lV fý N to lo co -ýr O cß'1 cN S O '-t if 1' r- -4 co r4 1.4 r-i 1.4 e-4 rl r -i -4 r-1 -i i-i
I ' Ü N l -4 1 -W n lÖ
61 lo
. -. 1-1 1--1 r-1 r-I r-1 1-4 rl
r-1 u1 O M tl' d' f'M I C+') to O -t' in rl S . -1 .i ri rý r-4 ý-4 O\ . -4 . -4 . -1 CO . -I r1 r-1
V Ö N Y M ä1 lV f ( )
Ö1
M ý M N .4 N N ° %D ' -1 00 -I . . -4 r4 -4 00 01 r -I t- CO -1 -4
1 I I S I , 0 C) CA N -f in CV CD C) v Cl
00 r-I -4 0 0 ö
l0 CO 1. N r-1 10 1 N ON
1 6
1 v I-cr C kD ö ý
N M ° in tN , -4 "o M in to o'. N in u) . co Ct) 1 1 1 1 f 1 I 1 I 1 M l0 01 l0 O to 00 in ct' O tD
N N C in I 01 "Cr CO in v :r M I I
06 I I
cal 1 ;
N kO -P
M r-1 %0 M ' l*l en %D N V S N M C11 M
-ý 1 1 to
ý, b a ý O (b Q ý ý ý ý ý c C:: C) H N M Vý
Q
a m V W w cý x H a z o a of a
45
l0 f- NMN d' OO -e M 01 Ln Q' M r-1 Ol 1 00 ßl ON -q OONmm It) rV co N0
M r-i MýýM l0 MNMN d' MMMý (fý
)
).
N
ä aý
En N
MN d' N co 0'% ýP ýN O r- Om -q to NM NNQ: r-1 L. r) N OO N llý Ql NMNN l0
O 00 d' OMMMN t- . -i ON al co . -i m
p ww cý xazoäää öi
O M tO kp w O l0 QD O W O l0 O O O W O N r-1 N N N N N N N (N fV N CV N N N C H I I I I I 1' I I I I I I I I I I N N m -4 C' C; ) LO O1 Lfl r4 N N CO l0 ON M
i --i N N N . -4 (N -1 .1 N . -i N -4 . -I -1 N
l0 . -1 CV O M 00 d' co d' C) lD r-I N tf1 CO N r-I r-I N N CV r-I N -1 r-I N r-4 N . -I -1 -1 N n 1 I
L co N N CO Ö If N . -1 ri (14 r-4 r-I N r-4 . -"i 1-1 .
d' -I [N "-1 N M N . -4 I-- In Ql cN M lo '-4 rl N r-I CV ri N r-4 r-1 '-4 '-1 . -i , -i r"i -I N
Co I I 1 1 1 1 I i I 1 I CO ON 01 ri W t0 N N ß r1 r-4 r-I CV ri r-1 r-1 r-I r-I ri
M O co M co W O Ln O u M W M r-1 111 CO r-I . -i r-I . --4 .i r4 N r-4 r-I r-1 -I rl -4 1-1 N N 1 1 -
W 0 0 V ' Ln . i a h
. -1
. -1 N -4 In to N d' M N dT O et' l0 r-I O1 "i -1 r-I r-I r-i -4 C) . -I i-1 e-4 '-1 01 r-1 -1
Lo i I 1 I I 1 I I I 1 I I 1 ýO \O 0 d' d' (") CO N . -1 M p1 M dv Cs] -"I . -I . -1 '4 r-I r-4 r-I r-I r-1 r-I rl (ý ]
p Ln m M m N -i O N N M r4 CO i ßl ; . --1 i . -i N 1-4 ý--I .i 00 Co '-i r-4 0
M .a r-1 O N I I
ce u1
U) "-4 r4 r-i . --i -f
cV O N O 1- i co -4 -4 j C1 (n 01 I- Wo d1 "-4 1w
Co I
1 CO 1
01 1
C6 O 00 q6 i
W i
lf1 I
QO I
C1
O C CO -4 r- r- n `r r- 00
c { 1 F ! I { I C%. Li', m tf) a' ýD 41ö klp N
L t'1 CO Kr (i CO d' CO M UP) W U) -v M U) kD N I N cr Q' Lf1 w M r
dý N V '
r
M M N M mot' to N l0 N V U M M '-1 m M - , I I I I I I I I I I I I I I 1
0 O O O O O O b O O O O U O O O
a a oa u w w 0 x H a '4 z 0 ä ä ä
w w
46
a aN
C7 pU
H týsa
Ü
ý `a wa wwýý
H ei ý,, i7
tD Cl N N in q l- co m fei i
" z t0 in in Oý co co N tD N o- in cc r-i q t- [may " " e " " " " $ " " " " " " " " ý;
" M TA M c") c`, ' N U) N (N M tV (\J M M 4 " U)
rn c-+ r q to O\ an N -31 0 in d' to CO O M Z in to Co O V O r' O vO ý' N [r in 0 to c-I 2
o (y] G Lý d' 'c Cý1 N N N O O N q co c-I c' fu c-l TA c-I c-4 TA c-4 c-4 vi c-ý cl Z
U Ü T-4 C"li CY) &4 Ü
G ý7 U W 44 io x H a Z Z O O r CY Cl C7 U
O in tD t0 tD O t0 O O cO O to O O O to O N ca N CV N (V (V (V N N N N N N N CV 0 V-1 1 I I 1 1 1 N ! I i f I I +l ý
c-I N
1-4 N V-4 (V c-I cl t"l N c-I [-1 TA CV
tN c-I N d' co M CO N CO N c"i M d' q cal r4 N cI N c+ N r 711 c-+ ci N c-I l c-+ N
ON I I 1 I 1 1 1 i 1 I I I C) O tO .N (V [` N to 0 M fl ci N N N t-i ti r- N c-1 cI N
to 0 0 u) cI N cI t cl to ºn a- N N l0 O IH c-I c- vi N t-1 (V c-{ ci %-I c-1 ca c-4 -H ri N
CO l 0 Ö uD LI)
Q' [N
k CO U ci c-I N N cl TA c1 c-I
O ri
C q i 01 l0 ch d O d' M l0 c-I c-I to co c-1 Ol ri Cl cI t1 c %-4 VA c-i c-I c-{ c-1 1-4
-, N ý A I
Oi o 0 Ö
rý t
T-I r ca
L c-I r-4 T-4
t c'4 c-I ci cI
N %D rN W d' in CO N N "j' C') (N- ca CO c-I c-i Tq cl H H CO ca ci c-1 0) (3) ca ri
ä 1 Ö J c+l
N 1-11 -11
I I W c t 1
W W
7 Ö
C, 0 +4' IZI, M N r+ O O N Li) ci N c-i Ol ci c cl c N 1-1 rl -N co co rq c-i
o
r Lfj
in I
CV N O
Oi c- N I 1 Lr)
z a T4 c-I cl V-4 c-1 c-I H
N c-1 ri 0 ý C() l0 cI [`N H 0 ON c' to CO ON tý to 01 H
lq; t N 0 8 N CO C) to c cl ri
c1 0 l0 L') c+ to C- N" to C) u7 cc r- CC) u-) v tý O)
Cl; I I I I ! I I 1 I C, i G 1n Ca C' '. 7. 'IT iý -- Qu c- r+
in V' ) 0 d ' IN co r`') 0D i d' to to [f' M in W N J
' t '
+ f
lI N
NI Cii ý N
d co t ý lp in j
M M IN N to to cI W ri (I u: d i N N Ln 1 1 1 I 1 I 1 I I I i I I 1 c-
O O O 4 O O O O O p O O O O O O
U 4 W U W W Cý x H "4 Z O Cr ci Y u W W
47
14, ll; r w
4
0 H H
L1. O a
a4
w w a 4 w
0 to m U
z
U m a4!
U
U)
T, C- [S 0 0 15 ON CO N U1 Gl CO ci co CO CO
" 0 00 rn CO "V CO to N Co M In L co o M " " ^'3 ¶ -4 C7 CO N U') (N Cv CO (N J} N (N co S! '
ý L7
N N N (0 r -A 0 { c M d O) d CO 1N c-I pp CD N M c-( cý N to L(1 0-1 0 N O CD Z
" " " " " " " 'i
" " " " " " " '
" is G3 U r CO Ol d' CV 'r cc) O1 O CV co CO r :T
p O Ff
Q C U W [, 0 X H -1 } z 0 O" Or 0, C! U u ý4
O cM C to %D 0 tO O O (9 C CO O O O U7 C) N -j N N N N N N (V CV N N N N N N 0
1 1 I i I I 1 1 I S1 1 1 I i l I N r4 N ON M O) CO N CO N tf) L( ) l C O rý r4 N c-( N c-1 N c4 T-A T -i r-i N 0 c-I " -4 N
tO O co N CD N lfl CO lo u') N ';; V N CO ci c'I N cI N r4 N ri c-4 ri ri N c'+ c'( c1 N
1 + i 1 ON p tip C- c (f 0 (' N
N c c-{ N N T-1 N
to 01 tf) r4 to O to N V Ci ) N CO CO c'I c-I c--I c-( N c-i N rq rI Tq ri T-+ i-{ c-( c-I N
co 4 1 Ö ) 1 ci c1
1 { N to
Ö CO
c( O cI ri
0 0
C c9 cl ri cR
c 11 1, % N
CO N M pp In N d' 0 N CO N O ri d' 0 ý-! r1 ýH c-l ri cl c-4 %-I c-I ri ri c-( 1 1
i I' 1 M1 c-I{ Ili to
I I O CO
1 N
r, 1 c--
N r1
N c-I
Q cl
c' ci c-4 c-I ci r4 VA c-1 c-I
N U) ri to m M N 0 'H to N CD ci C0 c e'1 tI c d cl O) -. - c-4 t'! 01 Q, ri ci
' 4 1 1 I 1
°0 1 to
c` -+ d ° cl c T 0 0
0 c C\j © 0 h i5 N -i a rn r 4 cc-I m ON ° c N O i c c-1 zz
4 N CD' °° (N 11 °° °
(N Q
0) tp i
N ;
ýO i ; O, ID N O) H
to t0 a1 C\j
11 (X ;
d co - 0 tp o 1.0 CO cl
co O o I i d In CO
co
CO d' OD d' N N N tN M d' lD to M M to N N ý 1
CO CA In N
l l0
'4' N %D N u) tD r4 tf1 N M Ln (N (11 N d' 11
c-4 1 1 1 1 1 I 1 1 1 I I 1 1 1 1 O (A O O O O O O O O O O O O O O
r-I N c Q CQ U W W 0 x H a 2 Z 0 d U C7
48
H
in in
N
z
W
CA Q
Ln W d'
REI
W ! - 0) Ql N ! - W -i ! - ('M 00 M r1 W 00 m O : , 10 to 1+'1 Ol '" N 10 M l0 O1 M M
Q ý
M -4 M co d' N Lf) (N N M N V N N C*) d+ Uj V1
N N N N N l0 L -1 O O -I 0) --1 '. + to !Z -, v m in N N -v r-i l0 U) N in Ol 00 -4 j
O w N cl M N O r4 O Cl O M N CO O L
O rH N M Vý
ýC !ý U N t: 4 0 x H ý-1 2 O O! OI O! Cl
O M w W lo O O O l0 O l0 O O O l0 O N -1 N N N N N N N N N N N N N N O I I I I I I I! ' I I I I 1 I I I 1 I . -1
N O O Ol M 0% N N mot' N N M ýf' If1 CO m -4 r-! N -4 N V-4 N -1 r-1 N r-1 rl N
%D ON CO N co . --1 W M lp W N M qM N N "--1 . -i . -i (N ri -"1 N . --1 -1 r-I N
I) Ö O U1 th to b N lÖ r-i N N r1 r-i i N -4 -i
U-) CO -W Q'. N Ol d' N ct- 'M 0) -4 M Lo 'i r-1 0) p-4 -4 r! -4 r-q '-I pI r-I r-1 r4 '-1 r-I ? -i N
CO I I I I 1 1 I 1 I 1 t CO N f") CO l0 lC 1-1 fM CO N O
. -1 . -1 r-1 . -1 . -1 r-I i-1 r-I . -1 -1 N
m t0 N N tf) If) co O N M N O -I ,: r 01 4 CO r-1 r-1 r-1 r-i ri ri -1 ri -1 r-I 1-1 r-1 -4 r-1
[ D M N Ö J M I N L 1 --I % \ , C 1-1 rI r-4 .i r-I r-I 1-4 '-4
N d' O It) d' N N O r-i to O N r- r-4 N '--1 1-1 . -4 . -1 . -1 r-I 0) r-1 . -1 . -1 00 . -4 r1 rl I I I I I 0) I I lD
- r-I . -1 r-I r-1 r-I
S 1 ý
ý r4 Q -4 8 ° ,D ý ý (3) c C, N ý ° ý r 0
N 1 b 1 I M l 0 0 o 0 0 0
Z r-I 1-1 '"I r1 r-1
C) 1 1-4 N 0'. if) N 1 -4 -4 tN 0) I'- 1'- -4 -1 ID N G1 -i '
b Ö 1 Iýv I
b w 1 lÖ
ÖO 1 4
r- I , -1 1-4
N eN O'. i) 01 01 it) 00 ID ID ! - 00 u') LI) N M 1 t 1 ! ! 1 ! 1 ! M
l'J IL !" C" 1_ L') LI ! ̀_ l0 CO
It M N d' ID co M l0 'V' et+ ID l0 -w cM Lo N N e N
10 "0 N
L) N N Ü)
Ct) N ilo
N U) N d' t i It) M M t') 'i' M N M lt I b I b
ý
c, O 1
0 1
0 l
O O O I I
OI b
04 as v W w 0 x H a z 0 ä ä ä ä
49
6 H
H H}ý
t1ä t'1]
x
ýv N
N
L1) .i
c kc -, r co N in N tN . -e ýo CO c) O %0 "o
I
Q . -1 N %0 d' -4 "T dv r-i O l0 O 00 ,: 1, 1,0 M 6
In M l0 l0 CO ul O ý1 In l0 a` N V t1) t! 1 N U)
OD 9
co 10
(N CO
N CC)
N O
Ol OD
11) "-A
it) -4
61 N
C! ' M
m W
- M
M r
Q0 6l
('l O
r-i r- C
ri Wit' Iý O "-a . -1 M ý* N ri p c; 1ý [ ý7 rq -4 N N M N N N -I N N N N M
0
a3 U W G4 C7 x H 7 Z O Ot U c Ot
O O 'IT
lQ N
N in
N If)
N in
O moW
N IC)
O mot'
0 mot'
N to
O q'
N EI)
O V
O d'
O d.
N in Cl
1 1 1 I I I i I 1 1 I 1 i I I - N N
O mal'
dl Cl)
N N M
in C
in Co
OD N
iO M
O M
H O M
O M
N M
lf1 W
1 -4 N CO w r4 Vw d' N to 0 O Ol Ol .4 Vw M N V M Ct' M C M N M N -, v N N M
O\ 1 Ol
1 Cl
1 CO
I . -i
i V'
1 N
I '-i
I U)
1 H
1 co
1 I,
1 W
I N
1 Ol
I m a\
tV N M M - cr M N M N M (N N N W
01 N
N I
N
0l 1-4 I OD H
N M I
to M
O M I
r N
M -W I
O mot'
O M I
CO N
H C 1
In M
O M I
a1 N
d N I
M N
c. M 1
N N
N N 1
lfl N
"o M I
Vý M
in N 1 M
N
W N I
t[) N
CO N !
[- N
N V'
Ol M
tN l0 (N I
in N
tN -4
d' O
I N M
l0 YV
I cP N
01 M I
l0 M
N N I
In N
d' M I
dl N
00 N
I N N
N N
I 0 N
%D N
I :r
N
in N I
N
M M {
'-I M
N N I
O N
-. 4N N I
M N
l0 N
I M N
co M I
[C) M
Iý
[ý7 IYi
8 N
OD C'1
'p "1
Ul -4
- M
01 N
M N
-4 CV
to M N
M
dr N M
N
co N M
N
l0 N
U) N
61 H
010 H
OD N N
(N
M N N
N
O O
N (V
Ol -4
co 1
-i
N N
0 r-i
N N
O N
N M N
M 8
tr, U)
0 N N 14
CO N
CD N
-1 M
N N
N CV
d N
N -4
. -1 N
. -i N
i. p (. 4
N . -4
is -4
01 -1
'-4 M
to u1 I O N
I if) N
I co r1
I N N
I M . -1
I l- . -i
I M N
I Lf) r1
I 01 .1
I Ol -4
I M N
I W rl,
I N '-1
I N H
I CO N
LO U)
QA 4
t` .i
m r-i
N . -1
CV
-14 CV
I- . --1
Ln' "-1
l0 (V M
N
CO "-1
[N -4
%D r-4
-4
N tV
O1 N
zf' "-1
M -1
CO '-I 1
10 r-1
00 . -I 1
1ý r-1
N N
1 O N
it) . -i 1
"dý -i
l0 -1
W
t0 r-I
to
N (y
N
(C) l0
I in
r4
1 c,
O cN I
CO
-IV -1
MI O
N N i
1-4
l0 r-1 I
1-1
N . -1 1
4m
Ol . -1 1
00 14
N
1 O
U)
I to -1 1
"W
0)
1 l0 r-I
fM
('1 . -4
M
.4
. -i
t1'
N
M N
OI M
N '-4 1
01 I
Iti 0'4
°
N r4 1
00 r-1 1
f') r-1 1
CO 1
N r-1 1
Ol 1
(") . -4 I
M -1
1 1 . -4 1
4
I r-l '-I 1
. ýi
1 N . -i r-1
co in ) r-4
01 L'1 r4
r-1 ri
mot' l0 rl
CO . -1 '. 4
r-1 .4
N -4
C) . -4
ßl 0) tt) -4
O N
ß- r4 OD
ß r-1
-1 M
in N
C) C> H 4 0) 00 c -1
1 1 I 1 1 1 I 1 1 1 1 1 1 I I 1 O O O O O O O O O O O O O O Co O
a a W U W f4 0 x H rl Z O
r-1 O+
N Ck
O Ot
V4 OI
0
W rs,
50
N
H
J
x
yw x
0.4 ýv
sr N N
In <Y' ri N l0 N co C O "--4 .o O In V l0 " Ol cý? . -i 01 '-I m %D 10 O to l7 Ol l0 N tý A
v" cn lD "o CO to O , qC V' .O mot' N d' I11 L1 N U] H
Lf l0 mot' -4 IJ 00 .( N 0 In 00 . -1 -, . -1 N 00 M LI 1fl '00 1D ', ' OO L 110 rý cl Ol rý C O
. ! %
N d' co Co t- d' co In If) "-i -i 'cr ko C1 N m N - N -4 N N N N -4 N N N H N N
ý+ Lei
.ý ca U w w 0 x H a z o Of Of a s w w
O lO N N N O N O O N O N O O O N CD -! ' N to LI) Ln V' t11 d' mot' to d' in Q' V' -` It)
1 I I 1 1 I I i I 1 1 1 I I 1 -I M (N -4 d' Ln W to to tO N r1 (n 0) N V q U1 C') N 114` M C' ('') mot' ce) N M M C7 N m c) V'
N -1 O M V' m d' 'r In tO O co co j cM c. (Y) N "1' C') ý' C') d' M N (") tn (V) N M IV) IRW
01 I I I I I I I I I I I L I I I I C, O 0) . -1 -1 . -i (`1 co N O r ßl . -i -4 N N M CI -t' C' M N M N M N N C') d'
ON Co C) O N O N N "-i CO to 'V' CO C) O N -a f"1 c' e' r) -i' r) N C") N M N N r1 'd'
00 1 1 1 1 1 1 1 1 1 1 L 1 I 1 1 I Ca 00 00 t0 lD l0 O Iý r-1 r-1 00 r N to 01 N N rl M N m C') m (') N N N ('r) N N N C'')
N N to U) In Q1 l0 O O N tO N . -1 in 00 to N "-4 C"1 (14 M N C') m N N N (n N N N M
N I 1 1 1 I 1 I I I 1 1 1 1 1 1, In N N N O DO CO to In a 01 C') t0 V' N C") N M N M (N ---4 N N (') . --I N N M
t! ' lý '-1 . -1 , -1 N 61 N N fit' -V CA co N LI) M N H r"1 N M N N N H N N N -4 N N M
1 , o 1 1 Q
C) Ll 0 1
M cc in
M to t "
N lÖ
N Ö M p
8 (N r1 N r-i N N N N H N N N H N N m U
c
N d' 00 U) N C N in LI Co -4 N l0 ON N (31 N "-4 N . -1 N N N N H N N N -4 .1 N (%i U)
ÜÜÜ]]] in I I I i I I 1 I i I 1 I 1 I I 1 in . -. 1 M 'o ý 19V e 00 [h M m O N U) ('N , -. 4 kp N -4 N r-I N N H (N H H N CV . -I -1 N N
O (N i -7' C"1 N N N N CO 01 . --1 -W "p O Lf) N H N -4 N N H (N --4 -4 -4 N -1 N N r I I f I ý Ö Ö ) 1 "4'
00 c 1 N
L 1 º l n U M b' 0 0 cy
. -I N -I N N r^I N '4 '4 . -i N
tN .i N r-i 01 OA d' O O Ln N N N f') N r-1 ri -4 N 9-4 .4 . -4 "-i N "-i . -ý r-I ri r-1 N
M Ö M 1 1 I M ý1 (5 0 to N 0) CT Ch 110 C1 4 -4 rl N . -4 14 . -I
-4 -i 'i -4
c' dl in to CO d' N N -4 fit' ýO "q m . -I O r-I '-4 U) -4 CO . -i . -4 . -1 01 "--1 . - 4
N I I I I 1 I I I 1 I N N N rl '1' \O l0 N '"'1 d' O
Ö M N
1--4 tO O ch If) C) en N . -i . --4 co A N r-1 -I Ldi ri tO "4 '-1 ON 00 M r-1 -I
H 1.1 1 1 l 1 1 I I I I I 1 1 H O O O
a O c ) O O C D O O O O
J) O O O
x a U w 0 = H a ,Z o ct (DI a s
w w
51
ii
a
Co
er
c '1
" U)
M Cý
C'
N N
m
CO Cý
ko
w l0
tD
O C
a)
co co
M l-
O
t0 'V
mot'
ßl m
d'
Co m
l0
cM Vr
d'
fN l0
N
r4 N
ct'
N O
lfl
01
in
Lf W
CO
q
U)
C4 1-
, -4 N
'-! to
-' -4
tl l l' CO N
C1 NN
00 -I
Co N
Ln N
CO M
LO N
d' I
cM N
O ul
Ln N
N I
In
m N
N N
-4 In
N N
CO . -1
Ln N
u Cfl
ltd . -4
-S- N
a1 -4
Ln N
M N
m L)
co N
w u w w cý x H a Z z o ä ä ä w
(D . -4
O ct' 1
. -1 m
tp N 1
. -1 N
N In 1
N -M
N U) I
C' m
N in 1
in m1'
O d'
I dý M
N U) I
dý IC'
O I
I() M
O -'M 1
LA N
N in
1 l0 M
O dW I
N (n
N If) 1
N
O -v 1
in N
O 4
I . -i M
O VW
1 -V M
N in I
Ln t: P
O "-+
a%
O t'7
O (V
"-1 W I
O q.
M M I
O M
d' -tv I
O '
M m
M "W I
-4 ;: P
'q: N M I
M M
If' N 1
M N
N M
1 M M
-4 M 1
O m
. -I 1
00 M
V N 1
N N
O M 1
0, N
M M I
N M
'. IT
OA
CO
ON N I
N N
ON /--1
-4
p1 (y) 1
t0 M
01 N
1 to tV
C) M 1
(")
N M I
. -i m
O "W
I N m
N M I
O M
N N I
O N
N M I
0) N
ON N I
N N
I- M
M I
M
-4 N I
O N
Co N I .o N
. -1 M I
O M
M c' I
M
Co
tý t0 N I
in N
N . -1
in M
I M M
-' N
1 (4 N
M M 1
cV
O m
01 1 N
t0 m
I N M
61 N 1
a0 N
a% -4 I CI -1
00 N
I '. O N
l0 N I
in N
N m
I (5 M
al . -1 I co -4
in N
I V N
6N N
1 N N
r m 1
m M
tý
%D d' N 1
N N
lG _4 1 in I
N M 1
01 N
'--1 N I
6% r-1
CO (V 1
l0 (N
00 fV I
N
4 C')
1 d' N
N N 1
to (V
N -4 l w -
in N
ll N N
N
M N
01 N
I lD N
N . -i 1 lC -4
C') N I
. -i N
l0 N
1 to N
N tM 1
M N
'O
ul Ls7 {Y+4
cn
,4 N I
O N
'V' -4 1 A
. -. i
CO N 1
l0 N
00 -1 1 110
in N I 0-4 N
If) N I
M N
M N 1
00 ºi
in N "
+w" N
to r-f I
f r) i-1
. -1 N I
Al -4
fV N I
N
to N I
. -1 N
in r-1 I er . -i
c> N I N . -4
c' N 1
"i N
co N 1
NP N
to R
8i 'j,
uý,
l! d'
(A . -i I n V-4
N r-4
Ln N
I N N
LO "-I
"-1
O N 1
N 4
fV N 1
. -I N
[- -4 I
M . -i
M N 1
N N
N -4 1
"-1 r--4
co . -1 1
l0 r-i
O N 1
Co +-1
O N I
CO V-4
M r-1 1 N . -1
. -i I
. -4
O N
1 CO . -1
M N I
. -1 N
En
Nt'
M '. O I
'a' -1
r4 I
O . -1
. -1 CI
CO -t
N -4
Oi . -1
%D _4 I
M -4
O CI CO -4
N
OI f-4
. -1 rI
-4
O 4
I CO
in r-4
ýI'
N r-4 I
N --4
-IV
rl I
O r-4
M
OI -c
N
i[I? . -i
O (I
in -4
m
N ý
G1 1
CO . I, - -41 1
t0 r4
0) 1
t0
C%l
1-14 1-1
Vý -4 LI ''-1
M 1
l0
co r4 I N ri
N -4
I N
in
I 'IV -4
. -4 I N r4
Q1 1
CO
01 -4 -4
r-f
ý -1 1
N H
N
"4
N1
1 O
N 1
O . ý-4
I Cl
in ö
O
I O
". -1 i
O
in I
O "ý-1 I
O
ýC 1
O
V-4 .-I 1
O
M ri 1
O
'-1 'I I
o
N 1 O
CO
O
N H
O
I '-i
O
I OCC M C) W P4 C7 x H a Z Z 0 ä ä a öl
52
)
H
Z er Cä77 Ln
W
f'M Lfl mot- lO f") Ln "-I ON ! - ON to M U1 (°) to O " O 0 00 "r [- . -i lfl l- m m 410 00 r-
Q . . En M N %D N V' O C t11 to C N
4 L. ä GQ U)
co fV a0 O . --I -c7' cV C If) Ol Ln N to wo lfl -T (l -! 0i O M 0O N ß'r1 O ko N O LO N 0ý
O ýr w o0 m k. 0 m m Ul N M Ln lD Ql Ln [- N (N -1 N N N N --I N N N -4 -f N N ".
a a o
V Iii V W W U x H z o ä ä öt a om, w
O %D N N N Q N O 0 N O (N O C O N O d' N u1 to En Wit' 1n It' "V' LO V' En ! J' %r v, In O
i I 1 1 1 1 1 I 1 1 1 I 1 f 1 -ý .ý -4 (N q1' 00 00 'Cr LIl N t. 0 M N CO lfl l! 1 CV) N d' M M M d' M N M M %T N M Cr) II'
C) C) . -4 M N 1- M cl' lD En N N O E l-W M N 'i' M CV) CV) %V M N M M ' N (Y) (V) eil
O) I I I I I I I I I I I I I I I 01 m CO O En En O ('1 M N r-I 00 M C M . -I (V M m l'') M IV M N M M Cl (N N (w) ei4
co 01 N C1 VV d' 01 N N r-1 O N (N 00 N O N Cl N M M M M N M (n M N N M mot'
00 1 1 1 1 I 1 I I i I 1 1 I I 1 co t0 In d' -4 N LI) . --1 O N Q1 cN N -I N CIA M N M m M C') CV N N Cl N N M M
to CO 1' M O r-4 d' O 01 lC OD Cl O w a ko N r-4 M N M M M M r4 N N M N N M CM
N I I I I I I I I I i I I, I I I I N eN l0 N ri r' O O CO L n N M 00 (41 0O M N . -1 Cl N N M M N r-1 N N N -i N N M
M En . -1 O %Q dl a1 N N c7' lD co N N t- N N M CV N N fV N ""1 N N N -i N N M
1 t I I 1 1 1 I I t I 1 I N d' co N rl N d' \O U) CH mot' w l0 O la Ol N N r-I N N CV N r4 N N N r4 N N N
. -1 M N '0 (% lD M LA qf' '-4 M Ln N C En co N N -4 N N N N . -1 N N N 1-4 r-4 N fV
8 Ln I I i 1 1I 1 'l. '
I M 1 1 1 1 1 1 t tft C6 M O l U O m mot'
Uj rl N N N "i N r1 N CV N -I r-I N N Un
- N r-4 M N Cl ) - t N N to Q% r-1 1 1 N
c N
I . N
M .4
w r4
N N
M (V
d' 1 1 1 t i ý1 N I 1 1 I 1 1 1 I d' ý N N C N N r ß , -q -1 " -1 " . -1 r -1 '. 4 '1
N
1. O in O C1 . -I t- (V V-1 O O . -1 CO N r-I CV r-4 01 r-1 r-1 r4 r-1 a-4 N "-1 N "--I i . -i ty '-1 M 1 1 1 ! t I 1 1 I "4' - t l M
ßl ID O C' O oC) CC NJ 1-1 i lÖ N
o 0 Iq sl ri r. i . -i ri N . -t
r) U) M 0% N %0 N C) a-I N Cl . -I CO . -1 a% rl a-1 N r-I N r-1 P-4 . -1 . -i ri -1 . -I
N , ` i I I o Ö M 1 t N . i t
dý N O l '. V ' t l 10 01
r4 rl . -4 r-1 i-1 . -1 . -1
C> M a ý C-4 -W rq . -1 %D a ' 00 -1 r CV) -1 . tI P-4 H '-I o%
ý H OD
I 1 . i 1 t I I b I I I I "-1 a Cl O O O C) W-4
a vo v w w Cý x H a z o Öl ä Öt ä
w
53
0
Eý+
I
cQ ä w 0-4
z
In In
M N -4 lo V' M N CO N E) '1 ) (y M O " ?1 in r- N 00 co 00 M ýi' -I .1 . -i dl rl (71 M Q
V M I %0 l0 ýi' dl ýY' ýf` lf) C 00 (+ 1 Lf) d' CO U; U)
d' lD N ll' N [, d' Ln C) \O N -4 co C) N O O tI O O ßl O M OD L') LI C 1ý %9 O N
1 N 00 CO O r- rM Ln ul -4 mot' :r -tr O lD l0 N . -4 (N r-4 N N N N . -1 N CV N .4 N N N
O q U W rZ, 0 x H a Z O Ot 0 CX Ot
tu Lww
O LD N N N O N O O N O N O O O N O Zv N LO t! 1 to %r to C' d hfl w Ln -W mot' C If) O H 1 1 i I ! i I I I I I 1 I I I I -4 O O M --ýr OO N l0 --zr tll ct' (h r-1 M . -1 t. f) (Y)
M N d' M M cI ýi' M N Cl) M d. N cM M lqr
ü1 ON N P t, W U1 co mot' c"1 N O N O ýP N N . -4 d' cr) PD P V M N C) M w N M C') dq
ON I 1 I I I I I I I I 1 1 1 1 0 00 O co r4 LI) 'j N M O r1 'O CO M 61 -4 mot' CV M M c! ' M N M C' M N M M
00 tl- 01 N O N O . -4 N ) O U) -4 N N 00 N -1 M N M M V' PD (14 N M M N (N M M
OD I 1 I 1 I I I 1 I ! I I I I 1 OD N to dr 00 m d' O A tý CSl m O l0 -i LP) N M N N tß) M M N (V CV M N N (V) M
kD to d' M N N t'M O1 O ko 00 N 0) ß. f1 O It: N N . -1 M N (N CV) M N N N N m ý--4 N M M
r- I I I I I I I 6 1 I I I I I N '' to CV rl c" O co C d1 ý' 01 t, r 01 . -1 N . -1 CM N N M N N rl N N 1-1 N CV M
g C+) d' . -4 O (I) 01 tN [N OD M t- co %D C') CO O
8
N M N N N CS CV i N N N ; N N M U) I I I I 'o
N M CO CO 1-4 C O " lfl to r-1 LI t! ) u1 . -4 N N [ z j N N . -1 N N N N r1 N N N rl N N N W [U ý
i H
,i N N t- O N e4 to Ul O 'a' v d' O lfl l4 N r-1 N -4 N N N N -4 N lV N -4 N N N
! 1) 1 1 1 1 I I I I 1 1 1 1 ! I i ý O r-I In U) CO "D 00 : ' 'ci' O! c+1 .4 M CO to M cV . --I YV 0-4 r4 N r1 N r-1 -1 N N . -i -4 fV N
0) O d' e' N to t- M M CO N O N N V N f-4 r-4 C -4 l N NI rl j N N q .I N N
C' 1 1 I
1 I I N O1 m (I) to C') d' N N Iý .I Lt') -1 W CV 00 .4 N '-1 r-1 N r-1 N ri r4 N . -4 . -i ' -4 N . -1
l0 N [V "7' N M . -4 lD O ä O tf) -i N . -1 00 [V r'1 -4 N '"4 cl, -4 11 N . -i . -1 4 (N r-1 en 1 I I 1 1 I i I ! i 1 I 1 1 N O O N O O) o l O ý1 o l M 1 ) . -1 N r-4 r-1 N "-i . -i . -4 .i -1 . -1
ý r -I
N CA -4 01 co 'd' CO N N N CN i-i l0 . -i f IN rl . -4 co rt (A ri '1 r-1 01 r-1 l4 -1
N 1 1 d 1 1 1 1 1 1 t i i I ! N r4 110 C rn r- 'n co O 0) '
w LO CN in In r-4 N 00 .4 l0 -4.4 l0 CO I`N 0) '-4 OD
1-4 1 1 1 1 I I I I I ! 1 1 1 H O O O O O O O O O O O O O O O O
ofý . -C N M 'AT
04 CQ U W Cs4 0 `4 H f-7 : - Z O a a a of W G4
54
er
hT ý N
H r+l
f -T l9 tß O ýv Ln (n uD tD N 00 CO ;7 O t. f) " r) t'') N N O O a Ol N Ln r- d' t11 d' L! 1 N Cý
L n M ý10 t- ON l0 O d' L'1 l0 qr [- LC W l0 o Ui C! 1
c') r1 C`1 Ol CO O Lil Lf1 ') ct' "ý ri . -i M M M M O 61 O r, t, CJ r- r- C' N qtl M p ýy
-4 ;) O Olt 110 to W l0 V' 4
N M r: 01 '! N N N . -1 M ý-4 N N N N N N N "
a oa _ N M d+
u W 44 c7 x H a 0 8
a Of Of 2
O lD N N N O N CD O N O N O O O N ,!: r cV u1 I U') .:: r U') -1' d' Ln eT to zr d' ;r t! 1 O
I 1 I I 1 1 I 1 I 1 i I I 1 d 1 -4 C 1 M o
M N , in ('m -rr (1) cV M M N: r M M
O N M co N IIl '. D O N r' N C1 O N tn M
M N cV (Y) "r ('p) C M N M M M M M M dý I I I 1 1 I I 1 I 1 1 1 I I I I a
cy% T% -4 C', 'r C, M Ln M m' C) LO co m CO C> N N i. M er M N 1 ß m'1 M N N M "zr
co CD -1 m . -i M M d" (N to 01 Cl N CO N ON N N i. M 'N C+1 -ýZp m N M N M N N M M I I I I I I I I I I I I I I co
CO Nl- CO IN %D (V C) N O M CO . -1 mot' tN ID C14 f'7 N M M ct. M N M N M N N M
t0 O'. 1- l0 tf1 "-1 O, '-1 01 fV I- O M kc . -i U1 (N i CO N M M M fM . -t M N !M N N M M 1 1 1 1 1 1 1 1 I I I I 1 1 1 1 rý
N LI) U) cyý cý1 . -i O ' I' O CO 0) ul CO O r') C1 N ' cr i M N m M M cM . -1 N N N N N N 1'x 1
.V tom- M N O a% M 41 N N wr N ON N 00 ý--1 N H M N M N M N H N N N H N N M
I I I I I I I I I I I I I I I l0 N O O'. to N Imo- r 1I) LI1 M cr N O t! 1 co
N I"1 r4 CV N N N '--t (V N N ri N N N
-4 to CO N S In CQ rr 4
N N
N N N ' N N ri r-{ N N
' I I I I I I I I I I I I 1 1 I u 1 U) O Lft N %. O N d' O u1 N r-t O O V' N N C") tV -1 N rl N N CV N -4 N N N -4 . -1 N N
M d` lD [I) -4 M d1 <Y -q O ON ON M "p -4 N -4 . -i fV r-1 N N -4 N -4 N --4 '"'4 r4 -4 N N
st' NI 1M
C" 1')
d I'
aI 01 1 1
l0 CV I 1O
CO i 1QO
l0 1 1N
M tfý Cb *. -r
. -1 -1 CV rI . -4 . --t i-1 CN "--i r-1 r4 r-t -4 . -I _i r-1
M ý co U-) +-1 N I- tF ri N %o 1'- A N ý4 . -ý "4 N 01 ri rý r1 . -i . -1 . -i , -I
C") I I I 1 1 1
1 1 1 I 1 to
1 M
I I I 4') , -1 -4 N r'1 -4 N . -i -i -4 -4 .. 4
"^. O C O lzr O C' ;, 3 u) c' CN M M i. 4 rt N r-4 -4 -4 Ql r-i . -1
N 1 m
1 Ql
1 00 [
i O C") N lD lÖ
ý' M Oll CO N N
4 14 H -4 r-4 1-4 -4
r- C%4 tI M N ri 00 40 ri l0 M H %D "-1 : It . -I -1 00 0) N -4 CO
-4 1 O
1 O
1 O 0
1 1 O
1 O
I O
I O
I O
I O
I O
I O
I O
I O
I co
I O
H
a a H N M c!!
+Qi AIi Cý W W C9 `"1: H F-7 . -`ý'. '1Zi CJ Oi OI a a w (z+
55
N Ln
11
Eri
.i N
u
O M Ql (N 00 O 'd' 0) Ul 00 T N O) 01 (C) ßl Vý ß1 in M N- O1 N '1 O N X11 to 10 ý--1 N A
Ln N l0 ko 01 in O in [- Cý Ln l0 iJi CA
O 0 C' 00 to M to --1 co cN r- 00 CO N O --" d1 O CC) 00 O ý-I 01 Ol N l0 p1 M M M %. p (y
-4 Ln ON 00 r- In Nä in 4
N -l 4
w o M 00 (N -4 N r-4 N N N N 1-4 N N N -1 r-4 N N
a a -4 N M "
4 W U W fL4 0 x H Z O OI a a a 1 WG
O l0 N N N Cl N O O N O N O C O N O d' N in in U) d' in Zir 'M U) ý: v in w d' in O 4 I I I I I I I I I I I I 1 I I
fM . -4 M N N- U) N- to to 00 (''1 r-1 00 '-i t11 to M N 'd' (C) M d' CC N M M d' N M () cr
N O N I t d' to Ln -'1' IN N O I- O mot' cP C') N d' (Y) ýi' M (C) N CM M ItT N M m d'
O1 I I I I I I I I I I I I I I 1 01 O O O (C) cM 'V' "cI' N N O to d' O0 N . --1 m .T M I' () "14' M N M M M N N M
0) Ol 0) 0) N N M M r-1 _ 01 U) C') [N ri O N .4 M N d' m w M N M N M N N m ds
co I I I I I I I I I I I I I 1 co co t- I DO N -I Ol (3) ob N r-1 l0 O N- N M N M M M . -1 N (N M N CV M CM
N- co tD in r- r-1 l0 O co co IN '-f O in C) lfl N -4 Cl N cM f') M M . -1 N N M N N N Cl
N I I I I I I I I 1 I 1 1 1 I I 1 N- in r- m N co 01 . -1 01 N to in cn C7 ,
M N- M N r1 M N M N M N 1-1 N N N r-1 N N M
. f' %D N r-1 N CO O O tO in C` CO 00 CV tO N N 1-4 m N M N M N . (V N N -4 N N M
to 1 I I I I I I I I I I I I I M O 41 C P% . to to to U) M (C) LO N 6) to M N C+) r-I N (V N N -4 N N N r-4 r-1 N N
8 N V) a% OD co in v Ul Q' N N mot' to OD d' CO f! 1 N . -1 N . -I N CV N N r-4 N N fV '-4 rl N N
in 1 1 1 t 1 1 t I I 1 I 1 1 I I ºn U) 01 to dO CM C") C1 d' M O O '-I L11 N N r 4 ;4 N '-1 N N . -1 N N N N '-1 '-i N N
co f'') in ill N N CO Cl N 01 a% O 'a' to r-4 f') -4 N -1 N N . -i C, 4 -4 -/ N -4 r-1 N N
'd' 1 1 1 l I I I 1 I I I 1 1 ý -4 c , N 1 N C)
N -/ 4
N "-1 r-1 -4 -4 -4 1-1 . -1 -4 '-4 -4
U') . --4 (n N t-- Cl N tO lfl r-I N tD 0') --i _4 N ri . -4 -4 . --! N -I . -i . -i . -t . -i r-4 . -t -4
t''% I 1 1 1 1 I 1 I I 1 l I i I I <'1 ý1 Qý O C N O1 C% OD mot' 1! 1 C ) O -1 . ' d' In .-i -4 . --1 ri V-4 . -i -4 rß-1 . -4 -4 . -1 r-1 . -i
Cl O 00 f"1 to co M N' N O M st' . -i r- 4 ri . -I rl co "-t N- -4 r-1 .1 p . -1 -1 r-1
N 1 1 I 1 1 I I 1 I I I 1 1 N p 10 r cl to rl- t° -11 - Co ao O -1 -4 r-4 -4 r-4 ". 4 -I
r-4 in cv) W C) C14 9-4 DA . -1 to ON "-I in .-4 U1 r-I . -1 01 t- 01 a% 1-4 t I I 1 1 1 1 1 I 1 1 I 1 1 1 1 '--1 O O O O O O O O O O O O O O O O
a a N
RC 011 U W W C7 W H O Z O a a a a Q
CL
o 9 Io 11
N U
CJl
lr)
O . -i o CO N to O) CO w al co l0 ri m 'V' 61 Ql N in N O -4 O O N Ln l0 W N '--1 00
, av cri %D .o ao U) C) -cw
4 in
4 r 4 in cri r: U;
rt M O U) N CO ßl N M U) ý' . -4 N q M . -4 Cpl 1. CO 10 1 c', ' r1 (V r- c'7 C% N mot` D
ri M (N 00 lp ,v N C l0 O N Ln lU Cl N Ol N -i N '-1 N N N N 1-1 CV N N rl r-1 N N
a
V FC W U W w C7 x H 1.7 X 2 O 31 a Of OI w w
O 1ýD N N N O CV O O N O N O O C) N O V' N U) in 1n d' U) mot' V to mot' U) -' "Y' V' U) O r-4 1 1 1 1 1 1 J I 1 1 I I 1 1 1 I I -4 N . -I r-I d' % V r M M N ' V Pl O N ý-i M Ui
M N d' M d' Cr 1 :! ' M N M M -N N (r) fM 'T
. -1 O O M M M N N lo M O O', l0 O (V cr M N qr m 'd' (vi d' M N M M M N M M d'
O' I I I I I I I I I I I I 1 1 1 I On 61 Ol co C) CD N Ol -4 d' O ß1 CO N -4 M tM 'i' M M fM N C') N M N N M
00 00 N Cl 01 . -1 OO O M M 00 kO C'') N O O CV r-1 M N M M M M N N CV M N N f'1 qM
aO 1 1 I 1 1 1 1 I 1 I 1 1 I 1 1 00 N to In to O -W 01 N l0 N v . -t U) O l N N M N M M Ci N tV N (V M fV N N M
kO F- -W `C d' a-, M 00 --1 U) l0 M O d' 00 'p N .1 c"1 N m (14 r') N N N N M N (V N M
r- I I I I I I I I I I I I I I I I t- Ul lD N N co co 61 (v1 LC) . -i al M lfl d' N . --1 M N M N CV N r-i N N M -4 N N C')
C' to . -l . -1 O LN N l0 co N d' O co N In M N r-1 M N M N N CV '-i N N M r-1 N ('4 M
tD I I '
I 1 I I I I I N
I- 1 I I 1 I to K,
8 N N
H r--4.
CO N
m rl
N (V
U) N
N N N r-i
1 r N
M N
N iV
N . -1
O N
M N
O m Wi
ý
8 U] . -4 fn N co l0 C' -4 d' '. O O N ' t0 Gl N Ol E
En CV
I 4 N
fi . -1 1 N
I N M
N l
N i
r-q 1 1
N I
N I
N I rl
I r-1 1 N 1
N 1 En
O d' O ý M r - M ' 1.4 . ßl . -1 (V ) c N t'ý 1 r- to N (V . -I CV N . -I N 1"1 -4 N (V ri 4 N fV U]
ß 1 N CO N . r -4 . -1 N -1 . N 1-1 N r-I C14 r 1 -4
N N
l 1 M N o 6 1 1 I i 14 t m 0 , al -4 t p N
. -1 . -) N ri . -1 N r4 N -4 '-1 r-1 r-i -i r-1 I N
l1 C) C~ ('t CO O? (N Q --i V) r'- i vi N r-+ . -1 cV rt ri . -. .4 N "-i "-d . -t ri . -1 . -4 r-I (N
cri I i I I I 1 0) J I I ci Ln -4 (n ß O "' l0 r
Ül N
4 . -4 _1 r-1 . -4 -1 . -4 . -1 . -4 rl r. 4 r-4
d' CO d- un N M Ln M O i dT %D . -4 41 r4 Cl r4 rl co r1 M r4 . -4 4 f-1 r-$ -4 r-4
N I 1 ý
1 1 1 '
1 I ý1
1 Iý
1 N
1 M 1 1 1 1 4 1 N
. -1 l Iý N d l0 O 01 O C -1
O V' . -I lß) d' . -4 N tV O ri 'O . -1 '. D . -I -4 U) .' l0 -4 r-1 co i
1 I 1 1 I 1 I 1 i I j j . -i O O O o O O O O o O o o O o O O
a pppa
U W w U' '. I+' H ºa Z O .4 a
N
a M a
ep a ýw w
än
W Q
x ui
Lr
H 4"1
. Nr In ýy+ M C l4 M CV <3' O I'D M N ON al in lD M N O r-1 C- M M N -1 N d' lo
cM lO l0 CO in Ol P C' U-) 'T lý ýI' to tf1 N t/ý M
-4 l7 l0 6l QC N Ol tN tO to c' N r-1 tO M -l Ol ýI' N CD O fN M r-1 ri Ctl it) CO : 7` d` in r-1
.{ N M tll '7' "A ý2' C- O N l0 lD O N O r N . -t N 1--4 N N N N -4 N (N N 1-4 N N P1 ý,
a a
cv U W w cý x H a Z O ö' ö' ä cz w
C) tp (N N N C) N O O (V O N O O CD N O '1' N Ln LA to mot' lf) mot' C, Lfl mot' lt1 "M V C' I) C)
-1 1 1 I 1 I 1 I I I 1 1 1 1 1 1 1 ,,.. t co M 7' tO N (V ý4 N l0 ri mot' LA M .1 M M dý M d' M N M M eM N M f' ) Vý
O CO tN N M LO O l0 O -4 in O M N M M 1! ' N Ct) (1) '
O1 I I 1 1 I 1 1 I I I I 1 I I 1 o' M M P M (3) r O's IV N M
c M N N N Cl) N N M
CO N , 4' O% 00 N ýD Ol M CO CO l0 N CO O e--I N '--4 M N M M M N N N N M N N M "V
Co I I I I I I I I I I I I I 1 CO l- M LA M O (V N to r 411 -i to c 7N N M CV M M M N CV N (") N N N M
l0 t0 N d' N 01 . -1 co . 44 to tD 'ýv O in CO CO N . -'1 M N M N M N N N N lM N N N M
N I I I I I I I I I I I I I 1 I N t(1 Ui O N 0 CO "0 l, O M -4 al V' to to N r-4 M N N N CV N N N M r-4 N N M
": r 'r M "-i Co l- If) tO a1 N LI) O CO P U'1 'IT CV -I N N N N N N ' N N M --i N N P
tp I { 1 I l I I I I I I I 1 I I I M C) S C) 111 to . -1 ý N O M CO N r-i M . -4 N '-1 N 4 N N N N 't N N N F-i N N M
8 8 N N to Co tl' to O 'ý7' N Ol N N to o N O 8 N I N N N N N +q N N ri N N M
LI) I I I I I I I I I I I i I 1 I ºn r-O . -1 r-1 tý1 to r-1 (7 N M t! ) QO I eý v r to CO O N N F-4 N tV N r N rq CV N rl -i N N
O O N in O N to N d' N O ' 1ý G1 to N .4 N r+ (N N -4 cV -t . -+ N N ri . -i . -4 N
mot' 1 1 I 1 1 1 1 1 I I I 1 I 1 I 'd' Co CO O rl N l0 O o M In n m i-I N "-1 . -t N "-1 CV -1 -4 -1 N . -4 . -1 ri N
N O M N 01 N 0 N Lfl CC) 0) N ci %D N r1 Co N "-1 r4 "-1 . -1 N -1 N
M i I I I 1 I C 1 I I I I 1 I it) N '-t ' Co 01 O 'd' l0 . -' M in Ol
ef' ýO O cal N t, M u7 t11 O N r co I- I N . -1 '-4
N 1 1 1 I 1 1 1 1 1 I I 1 N { N N to M 01 O tO Ui W N M 0 ý &. 4 r-i r-1 r-1 ý4 ri r4 -i ri , -i '-i
CSI i M -I (n N "-i M ' 1 t) rr q CO C) . 4 CO . -1 . -1 r- 4 O1 O's r-4 r-A I I I I t 1 1 I I I L 1 1 1 I I r-4 C) o C) O C) O O O . O O 0 O O O
a ao N M c
Q W U W W C7 x H ºa . ý z O Ot OI d OI
w
SR
ÖM C
Öu
W LL Z
W ýT JN
W- I-- U.
in co Z
W U. a
CCD J Q. w
ä F- = 7- Q
t- 0 '. O rn \0 M '. D O M '. p co v d - in
I
C] O\ O Oý Lf 00 19 u - M Vl M pý p
u? N N M v M Itl M M Nl M M N'1 M C cr;
V M M - N O\ M CO - - u'1 Z M N ü1 O tt) N - N O t` N co L", U'1 N CO Z
W O N c* N to 0 M - rn - O - O O O M
Of 0
U < m cý w lý Cý _ - -ý Z 0 Oý O? O" d a L i. LL-
O M %D ýo %0 0 'D 0 0 1-0 0 '. O 0 0 0 N - N (N N N N N N (N N N N N N N
O 1 1 1 I 1 1 I 1 I 1 I 1 I 1 O - I N M - ul 0p O\ I` O t`ý - CO 00 C
- - N N N - N - - (N - N - - - N
'. O N 0 d' M CO '. D O\ 0 t- 1- N N - N N N N - - N - - N l I I f 1
u - a' - N N I ' 40 - ON '. O %o kD - N - N N - - - - - - - N
O co N %0 - -r r- to Co u\ 0 N - N - N - - - - - co d
- 0 -
I C'
1 N
1 -
1 u'1
I O
'. O -
I N'.
I '. O
1 1 ul ul 1 co
M CO %0 0 -1- m u1 N U-) M ýp eT ýT co N
1 I
I I 1 I I 1 1 M 1 1 1 r- C14 N d' O' Kl 10 4 - "t N ul - M M 1-
w '. p Nl co N n Pr1 0 M N N N '. O w O %O - I I I I d I I I I I 1 to O
N %0 O ' N - ý
Z Z w w
(1) 0 v ßt1 O r'I = O O O O M u~i cc) 111 1
p% N 1
M 1 1
ON 1
O 1
1- 1
O 1
ON I
p1\(17\ 1
M 1 tIl
N M 0 -
d CO %0 1 - 1 co 1 - 1 '. O O\ CO
Oý 1 CO
co I CO 1 V
- co r- co co 1
0 0 N - N - ýO N uý oo tý N tý ýp r- co M 1 Lrl 1 I 1 I I N I I I 1 1 t I r1 ýo a'. 10 c'. Ln 10 v N 10 10 'o In '. o w
00 %* v ko Ui to Ln N If1 a0 I M f I I %r Lr% 1 N N'1 to l l M M
Ict M iN of %D N N M N Na- M N qt M r: t N - a a a a o a o o o o o o o ö l a - )
0 U < CO 0 W 1- U' M - -J Z 0 O" O p ä
U- LL.
59
t-i
Ca
W
aa gö uw x .ý H Lv H
Lnn z
, wa cý. w ma
ä
1
iEl
U) N
0 6
" C" _' , .. :f LD :; -T -H c- Cý S cn 5 H cß'1 m r N 0 CD UI N N 01 °O , zT ;' O 0 `ý- c. P O Lo U; 01 N M
4 4 M ul M c7 01 M
4 01 c+1 M " 1!
?_ c1 N 4J :J 0 07 It ('`) r H N T M
Cl M
Cl) N Q ri Cl . -i r-1 M W
x O O Q Q CO U W Z O
0' 0 O
0 v
C-. {?
O cri W 0 0) O 1 O O 0) O u O O O 00 N r-i N N N N N N N N N N N N N N 0 I
t- I i N 1 I
Gl 1 1 I I I I 1 I O
.! N N 61 ID O N 0) N 0p (p CV c rl ri N N N ri N r-1 H N H H H H H N
to N H N co M co t! 1 O Lp
1 I I H N r lý
.4 1 (D CO I [ý I-- . -1 m
L H H 1
a r! r-I ý-1 N ri N H N r-1 N H r-1 H H
0 H-1 H
N
0) H 110
O1, w H
N Ö 11) 1-0 i . CO l r-H H ri O H H r-i r-i '-I r-I r-i r-H r. {
CO UI lS U) CO Ln H L(1 LI) M N if) Q) M rl
ON ti
i
10
r! I r-7 1 i
V ri O
r-j1
d' M
Hr
'
'i N
ri 4
Hj i Ö N
H r-1 r-1 H r-i r-q r-i r-i d r-1 r-i r-i
dý H
t r-1
WA tY, N lQ M in cn Ul m M N fM H M M Ln O . -i . -i r-i q . -I rl r-I M r-t H H H H H r-i O UU) r4 Ln
I N N N CO
N H N d N 4 L N
E-4 0 d' 0 CI) N , -4 r-I H M
Ln I 01
N I M
I ON
I -1 H
d d N rI H O -
`ý d H d H . -1
' r i CO H C%j '-1 H H
co CO O r-'
O 0 M ON C- M
O H %D C\I
H 1
0 !
N a) to 011 I I o d
CO OD 0
ri
CV) l0 U1 Oi
A t0 L-I LA lÖ CA
co 5 I LO N 00 I `
M t
LO Ln C j4 l0 d ' l0 lL lC (N 1 d'
d' CO Cr 1 d' l0 i
Cý1 d' N
tL
ui .4 F
M n L N L I to
cn 01 !5 M M U1 N (n H Ill d' M N V' u'1 cl'
d d d d d d d d d d d d d d d 0 ý pp Ö Q
6 CA v w t=+ 0 x H 4 Z z 0 5 cl Q
60
N Ul N
0 Q-
E-4 Ha
is.
.¢
w w
CD Or)
z
a
, -1 lJ) H : ]D C. (r% ßl 0 (\J v' (V N L l0 r-'1
Vl N N 'V' d' d' M LU cM Cr) cr') M 'd' N M M LU C")
Z CD r-i
M N
in Cý
N l0
N if )
0 LU
[`N l0
Ol '
M (10
N IQ"
ON N
M IQ"
0 Q
I N
N l-
O O 7
C CL O N O ra N (N ri
r-I tý N
1-1 H H
ri r-1
N ri r-1
N
r-{
M H st'..
x x
< m w w 0 x H a Z o ä O c 0 U
w
o r, W to LD o CO o 0 ýD O Uo O O O to N r1 N N N N N N N N N N N N N N
lI N C
rý N
O1, 4
O, J) d
Ol d
in CU 01 M 0
r-{ N N N r-1 N r1 r-1 N H N H r-I r-i N
<'') O H m CO tN O d' co N N N N N H H N H H N
CU LU H
H r--!
I N
I I al ', r
1 N
I tý
ICZV r-{
O1 H CO
1 CU
I M
I lý
OD r r-{
O' N H ri N r-I r-1 r-1 r-I H N
co E- r-i co 0 N r-1 H N H
ý H
Co r H
0 . "U
1 Q1
+ lp
1 N
CD 1 Ol
'C H
1 A 9
Ln r-i IL
N r-1
C 1 r-1
U) 0'
On
r-I r-1 . -1 ri r-I r-i H r-H H H
M
C' CO
1 LU
1 CD
II LU CO
1
1U J
O to
I 1-11 ' LU
H LO
I L. n I
N N [N M .4 H N ri Ul H H
4' H Cp
H H r-I H H r-H r-i r-i r-1 r-1
w aQ
t0 t7j . -a
r-I cU
lp ri
I LU
N e-1 1 r-1
fß'1 r -i 1N
d rl
cU
U) rl
M H
1 N
ON
tb
d r-i
1 M
M r-1
1 N
M r-I
N
Q
rl
(31
M H
i N
'Cr H
I M
LU . --I
14
ß', O m 10
r-i H ri rl H H . -i H H H r-I H
47 E1 E+ En
0 r-1
IV H+
M H
0 rl
oll
r-H r1 0
N i-1
r" 1
H
N H
O H
r-I H d
H
1
N H
-11 e-i
r-{ H d
H
r-{ r-I
0
40
N r-1 r-{
M r-I
rl H
r-I
r-i
Op I
OD
N ý
O'
9
0)
CIO
I'D
In
0
Ol
01
0O
N H r-i CD
H H
Q)
C
M I ýD
to I ON
LU 1
LO co I I
in I
(U OD N I
ul ul cN I
00 I
LO M
N N [i A Ln N Ö ko I N N l M 1 CA
v tU tp N C') in N C r-4 Ul to N lV It lfl N
`ý d d d d d d d d d d d d d d d d ý'
ü
w ¢ w v w w t7 x H a z O
ä ä 9 o
Q
61
)
°ý
a4 H
z H E+
0 a
H E-4 H
4 r4 Ö W W a ýi w a z 4 N ik
cý M
z
ö ý3 W 4
" '1-
N C!
CO r--1
a' Qa
00 N
in N
M 0
ý1' V
Ol N
ty' C; r
-1 to
Cr) rS
In M
in CO
' Cr)
Cr) (l
(V co
" Ci
U) (*1 N M d 4
Cr) LI) Cr) Cr) M M ' fJ M N d in
Z N H
N O
CV M
co M
-' N
N M
O Cr)
CO C)
00 to
VT 0'\
N (-
Ol in
d to
r- in
r-1 N
61 0 Z
O H
N 4 H
O H
O
. -4
M H
N 1-I
O ri
N ri H
H H
H r-1
CO rl H
Cr) s -I
Cr) s-i ..
O O E-4 "ý q U W LL4 t9 x H a X Z O o ^ý a
Ö Q
W w,
CV r-I to CV
W N
W N
O N
to N.
O CV
O N
,o N
o N
to CV
o CV
O N
O N
WD N
N ri
N
r-i (4 N
d
N
d
CV r-1 -4 N H
I
H
J
H 0,, H
N
CV LA, r-i
ßl H
A
H
M
N
o
W 0) ON r
m C\j r-4
H -q
- C\j C ON
U) r-I
H H
r-1 N
1
Q) r-I
1
co H
CO H
1 . -H CV
1 tý r-I
s ri
co 1
1 H
t co r-I
d' r-1
H
9 H
00 H
I
H N
O j
N N r-I
0 Iý
to M N to r-I
Cr) H
O N
co d' r-1
O rI ON
I I in
I N H CO
I In
I fV
I 10
I In
N H
I (V
L0 H
N '-i all H H r-i -4 H "! r-i H r-1 14
M 4) in t- t1' r-1 in to in l0 co
N [!
A A -, lSi ýTl O C7' r-1 " .4 U1 lD
H rl H H H r-4 r-I r-i r-4 r-I H r-I
w
O tD H co I
to r-I
1 '
N H
1
CV r-H I
I! ) i-+
L r) H
N H
ON I
OD (')
Cr) H
M ri
I
0 H
1
m r-I I d'
in r-I
I
W
O ID O C1ä
r-1 N U ) r-I
H H
H r-I H H
H (V ri
tV ri
dl N H
H d' H
U
W vI H
0 r-I r-H eN-1
r-4 -1-1 n
C\j H r-i
r-I CO ri r -1
M H
Cr) -t
Cl)
(ý1 Ql d d
Ol rl
rl d
lý ýJ O N A LA
r-1 ri r'1 r-1 r-. 4 H H r-1
d' co u) 1 QO f kD ON O
I d' 1-4 10 N r-i o 0) 5 H r-I
ce) r-
0 LI)
0 r-4 0\ U') l0
M C) S 00 to Lo
CX) N
0 -i I
CD I M
Q1 O l to
N 4 c'1 co
r) 1 I 1 N In tL
, " , " 1 N
H Cr)
d N
d lO
d CV
d N
d tO
d N
d Cr)
d H
d d Lo
d d"
d M
d in
d t-
d M
d H
4 ao v w s14 cý x H a z O a ö Q w w
62
)
0 In W J CIO
0
äl J
aD O a J
W
W
2 N H_
ti a I'D
Q
0 w
w J
w U-
0
w J
I
M , Kr
N
Z
. 67 a-
OD rn M ýo N in Co co ao 0 mt ct v 0 O «' D 0 - r 0 - - M 'V' - Ltl O, N 0 M Co ( L1
to M N M d ýt t lfý M M M N d M CV 't c! )
Co - 0 Un e M 0 %1o n ON r - 'o Co co Z O N r CO CO O\ t- 0 N1 00 - OD G\ M 1- M Z
W O O p - N O r N M N w
O 0 - N M cY I-
U Q Co U w LL- c S - -t z 0 O" d O- Ch. 0 Q Q W W
O M CO CO O CO 0 O %0 O CO 0 O N - N N N N N N (N N N N N N
O I 1 1 I 1 0 1 1 I 1 1 1 1 1 O 1 0 - r - Kl 01 O N d CC) u1 O CO 0 to 0' (N N -
- - N - N N - - N - N - - N
%0 N 00 O' ßr1 O' m ýf' - - N '- - (N - - - N
m O I N 0 o't I mot' N m M CO I (j% - - - - O
- N - - (N - - - N
0 CO to O \o M r r r Co rn W d O' 1 1 1 00 I 1 1 1 CO 1 N 1 1 I Co
[O ul N - CO - %O r Co
M 00 M M r r cf - Ul n CO - r
r 1 aD 1 1 1 1 1 1 I 1 I I 1 CO Ip 1 r N r N N '0 try M 0 d 't M O - - 10
W ýp - - V N M N d CO CO W r l N M - ýO O Ü
- I i O - N M M
z z ~
(n O d' d' - N O N N CO
M Ö Ö - r 1 - t11 (7% Co O t41 > - N -
a 0
in N-
r r N - Co 0%
O Co %0
O N (IN le a0 1 I e ' 1 0 0 d' 1- 0 0 % -
$
O r CO CO - ýo r tn 00 rn "U \
'c7 O' to CO P N 010 CO ct CO
-
I M
In a' w'l 00 d UIN Co r v r rn e N N M N ý n M O i M 1 1 I N d t r gzr C NÖ t l \p Co M
M - r N - %D N M N e ul N (N CO r N
O O O O (DI 01 O O O O O O o
CY- Cif
Q Q m U W u- CD S - --t Z C, 4 C),
M) O
U- L¢
63
0 0 w J
Q F
Q Tg
I. L
0 LL
J
w U-
O Z
w -j Q
C' 1ý Oý N Oý N ýU ýt O Co h"1 tt e M N D - O CO CO - CO tO - t r-% N N M t- Ln - 0
U) M N N LC1 M Lf, z; M �l Nl M M M t! 1 u;
ttl 1- 00 'ci' N e U l0 M N Co ul to %0 r- Z ttl rn N N O Iý ul 00 ýO ltd O - tO U CO tO Z
4 w O 1- it1 - N N M - 1ý M - O - N N LL!
O O Q
¢ in U w Li. - CD S - --i Z O d d O" ä
Li 1L-
O v', w O tO O O tO O i0 O O tD N - N (N N N N N N N N N N N N
O I 1 1 1 1 1 I 1 1 I 1 I I I O 1 O - 1- (V t N 't ON U1 O\ tO - O-, O 10 O' N M -
- - N N N - N - - N - N - - N
M - M lY u1 CO p' to Q\ N N N N N - - - - -- N
O% 1 J ö 1 1 1 CO 1 O I O\ - N 1 - N tO N - CO
(V (V (V (V - - - - - (V
UN - C 0 Co N M rn N 0 CO 1 - 1 1 1 1 N 1 1 Ul 1 M 1 tý I CO
v - O N a) N O - N oo - uý - tO - rn
Oý tD 1- 0' 1,0 - tý e -e N ul tO co N M O i 1 1 1 1 I 1 1 I 1 I 1 N
- - cc ý- ul - N ul O tO M M - d' u1 tO
w N I" M Lr \O U'% N O M ul W er - 0% - - - - - - - - Q O %0 U
1 -
1 CD
1 tD
I -
1 N
1 M
1 ýt
1 M
1 OD
1 ýt
N -
1 O
1 !T
1 N
ýt -
1 to
O U
to - - - - -' - - - - - - V)
W W
() O O _ N 1'1 N M _ M N If) r _ t 1
O' N 1
vý 1
O 1
- 1
- 1
O 1
tO 1
N 1
O 1
a0 1
tý 01 1N 1 ifl
M O O - - ý _ 1 _ 1 1
et CO ýD 1 cV
I N N
I a' Co (0 O CO ýO
ýp t CO
1 O cO
t- - t Co '- r- u1 r- ~ M 1 ul t 1 I 1 I 1 t C N- um 1 t
i j tr ;
ýO U Ln N- to ýn r1 v ýp CO to
rn le to Wn (o v N - N W I d I d N I 'p M Kl ul I N CD
M U% nt r- e
M
' M N (4 M d tO 10 - to um CV N d- W N
- a a a a 0 o o a lo o o a 0 o a o -
ö ö < CD O W U. - (D M - -1 Z O C), ä
d o Q
LL.
u_
64
H H
Gaa a °9
99ON
z rj w N u) aÜ
tn
w l: i
ý4a äwS U)
O "
(V O ül <i' d' 110 co 0) N lD CV CV Ul lD d Ol Li 5! )
" rn
" N
" m
" d
" U)
" M tll
" M
" M
" M
" m
" 4 M
" s-n
" cn
" ' : r)
z N m in 0
H N
O O1
O r-l
M l0
o d
U Ol
H [
O N
d 0)
H 0 to
rn lI)
(o Gil Z
O O CV N N r01 N N O O Ol H CV (V (z7
o .ý co r-) W 0 H a z O ä 0 0 0 0
0 N rl tO N to CV O N to N o N to N o N ko N o N O N o N 10 N O rl
1 Lý
1 N
1 4 N0
4 1 Ol
O N
I r-1
I CIO
1 tN
! O1
1 (71,
1 N
O r-I H H (14 N r-l N r-I N 1-1 CV H r-I r-I (NJ
r) 0 r) cfl 1ý
in O N m IO to ON W
r-4 H rI
1 N
I H
I N
co +"{ N
W r-4
A
r{
1
H
1
r-4
1 ý
I
rt
1
ri
CO rI N rn
to H
H CV
N r-I
H N
lS r-1
H N
N H
00 H
t- r-H
V ri
N r-I
0 N
OD [1' 4 0
rf 1
O 1
IL 1
Ol 1
l0 1
0) S rl
I rI
1 S
In ri
1 tD
I 'O H
1 l0
1 ß\
OD H N H H rI rl rl . -i rl ri r1 º-1
t'n r-1
0) H
to H
CO H
CO H
ID rl
d' H
I[) H
N H
ul r-1
CO f-i r-
CV oll
CIO A I U')
r, I t1r)
rl O
L In rl 4 LO
rI tÖ r-4 r-1 r-i 1 H r -l r4 r -f r-1 r-I rH
lS r-I
N , -4
Lf) , -1
d' r-1
ºn , -1
M H ON
d' H
fV H
N r-i
0 s-I
C') r-I
[N , -I
t. I . -,
14 x
a
7 r-l In ri 1 -I H ri rl N A
.4 t. 0 0
t ] H . -1 1-4 - r . -i r-1 r-I H H r1
En Ei
0 d' 0 (V N N r-1 N O O H r-4
N 'i
cy) q
v)
ck 'A A d ý d d ý ý 0) (3; N
O r1i " Ln r-1 r-I H r-I r-i H r-1 r4 r-l
11 OD ý N
c0 L
a% I
A
of
rn I
0'
CIO
W
ui 0 co
co t°
ON I
O i
O
l t ý ' C O C rn co
m p Ln r.
1 ýi t0 co
I L: ) -l
tO to LO r
N In
.4 .1
01 d! '
A
d'
(A
l0
if M
to 4
M N
N Ln
[i' A
d'
CA Lf) W In
N Cc) ! to
ý' CM
d [fl
d C-
d N
d N
d d-
d N
d cßl
d r-l
d Ul
d -'T
d C \f
d N
d [j'
d CY
d f')
d `ý
u < cn u W w o x H a x z o El cf a ü w w
65
ö Fl
9 a
ä
N ID
ri Hý
c0
N U
N
4a äý
O O O
w co
ý M
0 N
cN C %-1 M
0 N
d N
cA (M
C. (1i
O (1
O Cl
rn V
Ir 0 co CO
"I CO
Cl N (*l Ci ýJ M tI} M C*)* M: M M M M d' l1
Z N C
d pp
r- in
m CO
m G"
O N
r4 (1
N co
in M
0 to
r4 (V
to to
CO N
CV d'
cM CV
N r4 : 7-
cý s7 lD V'
cI C; c-I
(V ri
N [-I
N c-{
0 c-I
CO cr c-i
ca c-I
rH rl
01 c"4 c-i
tV c1
M r-i Z
o < as O w w CO x H .a z 0 C C\l
CY CM7 CY Q,
)
O CV
M Vq
in N
in N
tD N
O CV
to CV
O N
O N
in (V
O CV
to N
o N
O N
O N
tO N
o I I r+ 1 rn
I co
I ýo
I a, 1m r1 1
co I
rn CA o
T-i T-1 N N N ci CV c-1 c1 ri (V T-i c-I c-4 N
C\j N
N
.N (
OI l
(IN to c{
0 N
CO c-i
1
N
Lý c ci
tý jpp I
-
[-
i
j
N Qý
C\j c c-
O N
N c4
rn c-
t- ri
O (V
'. 0 ri
d' r4
'O c-I
to [-I
N ri
d' c-I
0 N
c-I Oi c-I
lÖ T-4
ýpI c-I
tÖ T-4
Qi c-I
ul t'i ri
ui r4
ui c-{
tp cl
il
cI cc-i c'D -I
I 0' c-I
co
T-4 TI N-i 0 TI tý 04 cc) ILO--l
I ci
Iq 1-4
A
14 4 4
IL N
c-1 IH c-I ri c-4 vi T -f % c-ý c-+
W Q
t-i
ý
w ri
N -
[1 c4
d' 'r,
in c-i
M ý
O
I
M
1 N
M
I O
Cl
c' 1
LO
[+7 Ö
l0 H
N ý0 d [ Uj r-1
A A A CV 01 CV TI (V t-I
! N c-i
4
r-1 c-I cH c-I r4 cl r-I cl r4 c-I `Ei L I
in E in
d' c'l
O (V ri
N N c'I
ri c-l CO
c"1 ri
c-i ri
ri Q
c-i c' 1
N T-i
M c4
01 kO ck Ol tý '
O O '
O CO '
4 4 O O
r4 %_4
CO U"
c-+ -
co 0 ?4
i 0% N
co [N
kO
ti
O, I
0' N o
01 I cl k 10
N t N a) O t OD t C D c
Cl '. O '1'
0 r4
I ko
I C-
I , I W
I ýG ' or"
I NN 0 in I I 00
I M Ol in in 9 to N lfl CO l0
N In
( M OD CA
N 4
to
t1
d'
CA
tI
4 ( N
l0
Ln
W
to
in Cl'
EN in N
cl
d lV
d N
d N
d Cl
d dr
d (V
d Cl
d V-4
d d'
d d'
d tn
d en
d -0
d in
d N
d ''
a 4 m U w w o x H a z o r o a ä d u w w
66
)
0 W J
F-
o_
J
3-
J
W Z wI
I
N
CD
U-
CL 10
Q
tL
W J
Lß. 1 U-
O Z Q
W J Q
00 Ln
Z
W
c2 Z
O
cn N
J U
J
U)
0 trý CD Co ýKI- Ln c, r-- n -e 0 N t 0 N Q Z; a) O N O C O M M N -. u N N M C M ül M M M M - M M M ý- cn
%O N N O CO ' N - N - OD 00 ,O r- o M Z d N Oý 00 O - 00 M 00 N ßf1 N U1 - ýD Q ¢ w O %0 M O N N - - OD - - N Oý - N M w
(Y ý
p O Q [a O w u- -1 57 Z
Om K Q lº LL.
p M \0 %0 ýo O %0 O O %D O '. O O O '0 N - N N N N (N N N N N N N N N
O 1 1 1 1 1 1 1 1 I l I I I O I O
- '. 0 - N - M ON M CO ýO c o 00 a0 - Iý 00 N M - - - N N N - N - - - - N - - N
p N N LC) O % N N 'N N N - N - N
Oý Uý O I 1 1 Co I r 1 1- N I 1 N CO C% - - O aO - - - - v - - rn - - -
N - N N - - - N
ON r- 0 r- o 'o %D %Z ao nr p N
C0 Oý 1 1 1 1 1 1 M 1 1 I 1 ýO 1 CD - 00 U1 cc) '. O N ul - t[l Ul 1ý M - - Oý
M r- r- N Ul Ul% OD
r CC) Z 1 le U - N ý K'% WN - Ln - M - N 'ct -t IN
n N ýt d M O M ýt M %0 w W O - N 1 7 1 1 1 1 I N 1 I O 1 ºý1 ýO O
M N N O\ - N M - N - U to - - - - - - - - - - U)
Z Z ui LU N
M O - N - - - - N - N M U)
- - ' - N p1 p O O O O m O N
O
t- a10 Ö+
r-, m' Ö ä0 12, 1 1
,0 N- CO C; Iý CZ %z Co CO ND I- le 1 r n 1 l N 011 l 1 1 N 1 M
0 % n
1 0 . e u -% 1 - t0
%0 M CO n Kr ýo N ýo 1, N to M tf1 C Iý N
In N N M M Ul '. O u1 d
sT - %D M N l! 1 N e t N d' LID d' M 'e %0 ý O i O O O ý O O
M - O O O O O O O
tY ý O O
ä ä ä Q ¢ m U w 1i- t9 S - -i Z O O
U LL .
67
d1 kD W J an d F-
O
I-- Q J ZD CL. O G.
J
W
W
N
I -- º. r
Co
LA. C i- %0 I-
C) U- W
aJ
U.
O Z
a W J
O M Oct n
z
v N 1
L
C
O DG CD
W CB Q
'5 cM l0 %D d' -t' l0 F -i' al to CO /O O u1 00 r-I . -I '1 O 01 M I- co co co r-1 O o r
t: m -I c. r M U) C'') N c N ý1' M C) cý ) ýf
N U7
O Ln W t- 0) 01 -I co N O N tD l0 N N M F- l0 LO %D -4 N l4 ;r a l0 Ln l0 -i d' '-1 M
LJ (n r- -W -4 10 ß Q1 01 Ol Ol co O m 1-4 rI 1 -4 r -1 -1 r-4 1-1 -1 ir
cr_ a O
Q PQ U W W U' w H F4 x z O d ä ä a
Q
Na: LL- (sr
O c'r1 %D l0 t0 O l0 O O O l0 O O O l0 O N 1-1 N N N N N N N N N N N N N N O I- 1 1 1 1 1 1 1 1 I 1 1 1 I I I -4
N N Cr) r 1 In N U) N U) O l0 r-1 l 0 U) 00 M . -1 N N N H N -4 . -1 N r4 N 4 . -t 1-4 N
l0 -I N O d' l0 k0 d' ON U) O U) N N r-I N N N r-1 N r-q -1 N r-1 r-1 N
I ; i
I CIN N N . ß
-I . r -1
U) O r-1 co N In r-4 d' M ýp m 01 d' M Ln r-i r- rl N . -I N r-1 N ri r-I r-1 _I . --I . -1 "- I . -i N
co 1 1 1 I 1 1 1 1 I 1 1 1 1 a0 Rd' ON l0 . -1 O M N U) N M N ItT O . -4 -4 N CV 1-4 . -1 -4 r-4 N
M CO to O -' C1 N -4 -' N lC N "4 Cr) al "- (3N "-4 . -I N . -1 . -1 . -i . -1 . -1 . -1 . -I .ý ^ý . -1 r-1 d O M ' [, - N N ' N C ) L)
( 41 O (N U)
1
l0 c'') 00 N l0 ' O N '-f . qr O . -i N LAJ r- CO ri -4 -4 . --I r-1 .4 -i -4 r-i Ol ri r-1. cl: to 1 1 h' I I I I I I I '. o O U) N S f'M O . -4 O M 01 lf1' U . -1 . -I . -1 0-1 r--1 r-1 (ý]
N ýi
8 LAJ N C) ß r-q O N N
a ý U2
Ln I [N - i1
I .. 1 I
. 1 I
i- I
-i 1
O1 1
01 1 1
--1 -4 1
co 01 -4 1
-4 1 L1
rj F ] 00 O 'n o O 00 ( m (a O lý p1 H 1)
CNI ý . -I f-ý l0 r-I 01 " i ON al t, N CO N p' N %D CO 1 tD ( N
N O
O LO Ln . -4 [- -4 co l0 lfl kO t-- lD In lfl in l0 ß1 C") 1 I I I I I I I I t i I I t r` cY' m l0 O N d' ß. l1 Ln Ul -, Ln LC) N
r-I
cM d' CO U) dl ko M V ' ýf' w t ) M R: r M .0 l0 N 1
MI ,
f N N 'P N I )
N M ko d' 1- in H N M M M ri M N M d' , - 5 . I I I I I I I I 1 I I 1 1 1 O O O O O O O O O o O O O O O
o 1 Q co 0 W P4 C7 x H a ,Z o ä ä ä ä ILL- w
68
)
uý ýo w a as H
Z1 C>
a
OCr ! Sa
tý.
vi H
t4
0 cti w a ýi
w a z a w a V.;
0 Ln u
fn j C%l
S (3) CO m In a) l0 tD N cc) lD co C\ N
M N W
to M r1 M 4
V N En M N M N d' N M M U)
C rn U) a) Gl rr P7 H lý ýIl N ch V"
d' O 0 l0 u (V (V a) O a' M
N U)
m M
a J)
¢ 6
L, 0'. N c'
H H H
d H
N r-I
N H
O H
cD O H
O r-4
H H
CO 0) H r-1
M H
u .y 0Y1 U W G. O X H ý-7 Z Z O , --{ C7 O
m O
Irv CJ
0 O N
I
ch H
I
c N
a) (N
I 1
w fV 4
O CV
I
0 N
O Cv
I
O CV
1
a) N
I 1
O N
1 0
O cV
O N t
O CV
I 0
O N
1
0o (V
1 CY)
0 H H N
r{ N r-{ N
' 4 N N
CO r-i N
CO r-i
U) H
G H
l H H
l H
co H N
ON to H N N H
0 (V
1
M N '
M (V
N I s i
lD
;; r H
'A
CO i
9 V) 0 ;
Q
LO H
4 N H1
'
(V
C\j 1-4 H
N 1 0
H 1
(-. j H
CV r1 H r-{ r-I l
H H C} H
r-1 H (V
Co
U) H 4 H
O cV d
CD H Ö
H N d
H
r-+ N
Ol
ul r-I I Q'
c i
H 4 H A
co
N1
Cl)
I N
cf)
1 N
Lo 1
In
0 (1
Ol CO r-i
, H
l H N r-1 1-4
r H H r-H r-I r-I r-1 H
lý cM H
1 (71 CO H !
LO r-i
01 r-1
i
to r-i
CI '
CO -1
I 121)
M r-1
1 N
r-I H d
M H
1 N
n1 lfl H
1 H H d' 00 t-1
1 S N H
C H r-1
CO H r-I r-1 r-I r-i r-1
1., H
r1 r1 r1 H
to H CD
W
I N
fn
I CV
r-
I to
M r-1
ßf1
I1 A +-4
1, a
H 8 d r-i
-H
M
I H
0
I (A
0
I 0'
M
1 N
w 4
.4
a: OCý
0 H
c7ý H r
Ln
Q
0 H H
0 cr r+
A N r-+
H N H
d O H
all CC) °`
O r-i
k O H
°° co
N r-+ r-i
cn H
H Lo C
d' tý lD ,I I N
CO
[
H
H 0
ý
r`
co
r -I
N co g OD
CO I
N I
OA
Op
H
Oý d'
M { `n -
o', in
O
Cl')
co
Ln vi `n , ,
-j Ln
j r
ýr`I M
N N %" 1 dý d M d' a) 1 cV
'ý H
d M
d [1
d M
d In
d to
d CV
d M
d M
d M
d q
d N
d M
d N
d 1'
d cn
d
ö U
w Q ca U W w O x H a z 0 Erl ä c ;5
ö u
w
69
V
H
n
gu Cöz
40 xg1 H1n
a 8
NQ 16
to ýQ
" l0 V` O H N co rl M M 0 M :. l) l0 CV H l to 0, - 1 r-i to N Co M rN rN lD co (' IT H
" " " " " " " " " " " " " " " " " " U) M H C d V N l1) ) N M N d N m M L J]
N cn
Co Q
N Ol
a) cý
a' lD
a' 01
lD .i
N rl
O V
O N
O Co
a' G1
d' N
a' u')
O N
'. o O' Z
C1 N C H M N M C; " O" O" O" Co
" 00
" " "
L. .1 r1 . -I r1 ri r-I H
Ö
Q m cý w w o x ý-+ a z o a c ä ö 0 0 U
w w
o M tD lD ,0 O 1.0 O O tU o u9 0 o O l0 N H N N N N (V N tV N N N N N N CV
0 H
1 N
1 N
1 rý
1 M
! Ql
1 Lý 1 t11
1 H
1 1 1 1 1 O H r-1 fV N N H (V H rl -I N r-I A -, N
01 j
In 'I r -i
N N
j
a' N N
pp A
Cl) N
Co r1
H H H
to
C\l
Ol H
rl co
CV N
d' H(
rl
O r-1 CV
CV
CC) r-i
rl
r-i N
01 H
N r-4
tý r-i
N N
N
IC) r-i d' H
ri Ln rl
.; T H
co rl 1 w ri
CM r1
m rI
IN r1
r-1
r-i (V 0
rl
00
r\ °ý m H
r!
ul r-i
4 Cn r-1 c H
0% r-i r!
(1) H-i N
H H d
d' r-1
M N r -I 4
H H d
N r-i H
q) rý
'ö N
w (Z o r-1
rl
OD N
lD H
CA H
M r
O
l0 i r
N
r-i
d' H
('1
lfl -4
J O rd
° Co rý
N H
a' ccl)
N r-i
O
_q
(') r-i
-4 CD
0 H
CO
d r-1
C\l -i
r-r-4 1
I
CC)
H H
w O
d' lD
CV H
H
ý1 OD
' O
H r-i
I 00
ýI N
'. D I- rn
uj
CC) '0
O
Ok
O
ap
M ul G' L"
O ri (31,
t
tj
Gý 10
c
C ri
Ol
r-
t0 Lrý
'D
uý trl
Ln u1 co
[1
C- M
N M N I
co
[N
tl 4
N
'L I
U) N
Ul
VI'
'i' A "
kD
U) M d
M N l0
U1 1 N
'ý eI M l0 C') u1 N
d r-1
d M
d N
d c+)
d 4' N cri r-1 U) N
H d ö d d d d d d d d d
ö v
w ¢ ao V w w C7 x H a z o o
ä ä cr p
o w
70
p L
O en °4 co
cv
p wz
En L;
tO zö
ww4ý 14 1.11 W
" rn LI)
rn CD
() ri
U) O
M 0
' 01
H ßl
0 Ul
'0 LO
u) t0
r- CO
w co
C) co
N -+
' M
io O Cam.
V1 M , Q cj [i N ti) M N M N Cf N M M Ll (I)
Z M C1
ul N
H M
Ul Q1
d LU
CO In
d O
to H
l0 M
l0 O r{
0 CN
d O
ri O
0p d
H a) ý:
ýu .1l Iý ý7 1-1
0
ri N r-i
M H
M H
O H
0) O H
rt H
H rl
OD O (N H
CV H
i, ]
O o
4 x' U W C 0 x H a £ Z O 0 Ö 0 0 0 ü
0 N
lÖ
M H N
U) N
'. p NJ
W N
CA
0 N
Gi
LO N
0 N
0 N
'. D N
al
0 N
'. o N
1
0 N n
0 N
'
O N
' N A g
ri "--I N N N l
H N rl H l
H CD H
1- 1 N
I r!
U ) r-I
CO r-1
C N
OI ýn H
N C, J I
G` r-q 1
N -C\j d O
M CM I
(V i M -I
CO I '0
N r4 I O
N
1-4 I CA
d' 14 I th t -
N N 4 01%
ri H r-1 N CO , -i N
H N
r r H H r-1 r-{
r-4 - N
d' r-i
(A H O N
(A N H
ý m r-i
CO H H N
3 +n H
4 , U) H
Un r-H
' 0) ri
NH cN , -I
I H H 0 N
) r-i r-i r-1 H N H r H _q `4 I I H a H
C-1
00 H
Ell
ul r- i
14
N r-i
11 Ir
I
01 r-1
ri
r-I H
M H
C\ 'i
tD H
r-4 H
r-
Co r-!
v-1 r-I H r-i r-i H r-1 H H
OW r-i I
co
lD I
fý1 I
d' I
U) CO I
N H
s 0)
H
OH N
Cl I of
O,
d' I
M
tf) 1
M ID
a0
rH-i Hr-I H ri H r - ý r-4 H ý-l X71
Un o0 r
d' 0
r-4 N
I 1
`ý I
N O
1 [0 rn H
`J d
O
I 0,
N ýI
N N
N
d
W Ul
Ln rý ý1 r ri H e e1
d' ýO
N
I r-I
00
0
1 01
N
1 . -i
t I N
co l N
I op 1 ý1
W I v-I
I 8
ON 1
d' H r-I r-{
M in Q
r- ul
cG tÖ
113 H ßi
:0 ül 1G :f
u I o0 S
.) v M N
N ýp
m
N ýj
N 1
IL d'
14 co
ý1
N l0
Vj
M 11* N CO
t1l,
N 4 N
rl
d C7
d U)
d c7
d c/1
d CN
d r-1
d to
d N
d M
d U)
d N
d Cl
d H
d V
d Cl
d 4
Q Q M Q ca V w w o : H a £ Z o
är a u
w w
71
°ý W J
m
a- J
W
S N I. -
m
Lº. a 10
m
0 LL w J
w w 0 Z Q w J
I
un 00 M
N
z
0
(9
tO r- mot CO Lfl M j- M ýp Q tt\ M Oý ýO D t! 1 - N 00 0' CO tN M n N N ! Il M - Qý p
V) M N M N Lc M N M N d N Kl M ýt U)
O' O' O tO M O N N aD - tO - - O' tO N Z Qý tr OD N Oý M N 10 N Q 1- tO U) - p1 M
ui to to - O O M M O N O O r rn N w _ _
cl: O O Q CO U w L- C9 = - w Z O Od
0- O"
0-
Li.
O M tp 1.0 to O '. O O O to O 1.0 O O O '. o N - N N (V N N N N (N N (N N N (N N O 1 1 I I 1 1 1 I 1 I I 1 1 1 t I O - 10 -e O O Q. to co v co CO 0 v h rn N - - - N N N - N - - - - N - - - N
M O' O' ýP N h O' M ýp - N - - N O Ö N ( N l', r- - M %0 - ýO - O L ' -
pý N - - N - - - - - '-
- %0 N ul N ul %0 N 00
O i to V l ' Co rn O u - O ' e n tO - tN - N - N - - - - - - -
N 0ý e u1 '. D O' M Wl N to to tý 1 O M t 8 M N M O i 1 tý
N l If' I
W %0 N N 10 N - M (ýJ M w O O h - 1 1 1 ON O N - O O
O - to
- N
U
Z Z w w V) ý1 O M O M - U)
in I
%0 M 0% 1 N1 1 0 I O% _ _ I I- I 1 1 wl
Co m O> to N Q\
N N O 0 - d 1 to 1
Co I
to 1 -
CYN ýD
CO j - CO N - N
%0 :: h h _ N co jl) %0 'ö 1 0
1 %0 `t
N 1,0 ýo CO "'
LCl lfý cr\ ýj
Kl , T' c% Ul - V_% '0 CO `d
M Co u1 0' «t u1 t[1 M ct r- M N N M 7 ` M e h N 1 1 N
t t M ýO N
- (N u'1 M M %0 N t) M N1 %0 - M (-4 ul -
- 01 CI: ) b 0 c 0 0 0 0 0 1 1 - 0 0 c O Q
Q Co U w 1i CD S - O" O" C), U-
72
Z O
H
J
L". º tfý
U vL
Oz
ZW 03 W U' W"
J.
Vf ;N
W V) LL Q
J
110 WzU J WS äW
N 2- - M v -- co ON N a J D
.D N N M Q\ M (T N (T - cj d o
N r, 1n
C/') M - t; ý; ' v M Hfl Nl N C M C "ý1 M rr1 X11 _
r J - 0 0 \D C o\ v ': O, ! r, 'O G' r^ iI
(N ( \! :n tC N 11 O C`\ v CO CO M in ` R_
O o
- N M C. ) Q CD 0 W U- U) 2 - - 7_ 0 Od Cr Cr Cr U
U-
0 t '. O %a 'O 0 1ý0 0 0 '0 0 'O 0 O 0 %. 0 N - (N N N N N N N N (N N N N N N O -
I 10
I N
I I I d'
I C'
I t11 C7'
I .1
1 -
I 00
i -
1 tN
I 10
t G,
i trl
G - - - N N N - N - - N - (N - - - N
CO '7 0 iý 0 '. O (N N N 1 ON ir\ - N O N N - M , 1 L'I CO
1 G^ N - t M 'p
J, . ) N N - - - - - -- N
C\ - t+l 00 U1 co v 1ý O
CO 1 1 - I N I r-
N 1 1 N 1 1 I d 1 1
11 CO M N I- O\ - - ý1 - 10 d 10 - M '. O of
0 '. D CO %0 0 _ u1 u"' t N rN h N O 1 1 1 I
" t ýt 1 ! M 1 1 1 u1 i h co ýt %D U \ Q) - O d - d N 10
_ I- Kl ü1 ct 1N tn V) N M - O '! ' Ln w
I + I U O '. O N t l ul N N 0 0 t l d- Ü
U) - - - - - - - - - - - - "- V)
z ui Z 1- Ul - N O N N
LU v)
n o
. CO - I
- 1
- I N
- I
- c0 - f 0
- O\ CC) 1 1 c) 0
0' N o 011
N n
M O - O NT N I I I CO ' (IN
I CO t "
2 O\
10 (-4 ý 0 3
CO c0 '. O I
'ID $ ü1 p'
1,
= c ti, i C-3 C ý.
N CO d d ýO
t d
M ifl
M C N to pn N
d t- M M ul N M N d d N M - d N
- O a a 0, a a a a 6 () 1 6 - 0 0 0 0 0 o 0 ¢ m U W LL O S - z O - G Ör Or U LL- I
73
zi ! 0 H H
M
H O
2 w
ý7 Ü
U) w
Au 4
U)
" n
U)
rý N
M
O OO
ci
(Y) C-%
m
CO N
Sr
Ul N
LI)
N r1
M
N CO
U1
N (9
M
O) CO
N
Co Ol
M
Ln -1
Co
ON N
V'
V4 O
Co
k-O M
M
O N
Co
CO G`,
ý(' :!!
Z ., Zr iO
" rn
0 0)
" N
Co (N
" ,n
Ol Sr
" +
CN Sr
"
O l0
" N
Sr Co
" M
H M
" O
Co Sr
" OD
Ul Co
" O
(, j CO
" O
C) N
" O
v-l O
" (
M u)
" Co
0 01
"
V' N
" N
yý W
N C:
w Q w O w w O X H a Z 2 O 0
N a
M a
Sr 0 V2
w
o
0 N
c-I
Co ¶-3 N ci
l0 N A N
to N
I .ý N
l0 N
1 ui (V
O N
I ON c-1
W N
Ln CV
O N
I co r-i
O N
Lin r
lO N d N
O N
I oo c-3
Co N
r (V
O N
'o c1
O N
"c1 c-I
O N
C) c-i
Co N
N
Oo
ON w vi
ri 10
N "
0 N
01 c-ý
Sr CVt
M N
d' N
c) N
t- r-i I
t'1
S' r4 A
c"4
0) t-I
ýpI ci
N ci
lÖ ri
O (V
011 ri
ul %-I 4
e'1
cn cj 4
V-4
ý
N CV
cA N 17,
co Co c-l
1 t-i (E: l
c-4 N
q N
(ý cý
c-4
N N
d N
N V-4
ý s-I
N CV
d (V
N
4 c-4 r
I- c-1
LA l
Lf1 V1
14 ci
co cl
1 T-4 c
M c-1
N c-,
N c-4
to %-i
0 N
0, c-3 00
C- rl I (IN %-i 1 T-4 Sr V-4 CO
cý c-ý
V4
CO c
c-" vi 0
H M
c-I V-4 %-I
cl r4 r-I
W a
Uýp Vý
.l
En
d V4
CTl
`-i tl V1
Lo c"4 4
IH
M ri N c-I
T-4 c-I d
Iri
[ý
to t'-
%; r cl (N
ri
V4
c-+
[N rA
1 d' VI
M t'3
%-I
N %-A I
r, VI
0 cl Q1
C)
N %-I I
cý C+
O t'1
Ol
N N-i
I r c-+
O c'4
Q)
M TA
I c-i V-4
O C"I
d1
O V4
I C))
O 1-4 Ci cl i
CO
N
d' r-4 c(l) V
N ri ý
LI) %-q
I A
c'i
N ri
8
W
cOQU CI
ý
E
`r t- to
r r4 ON
co ci of
Vol ö ýq
0ö 00 r- 00
N GO to I
[- I
l0 ýn
01 co
C" U) to cl c-i lS CO rn N C^
U) N N
f
N fý1 d' J 1 Co l0
N 4 Co
N Sr M 4 4 N
N
d C'1
d NN
d Co
d fn
d Co
d N
d N
d N
d t+)
d Sr
d VI
d c+1
d N
d C+1
d Co
d
I 6
w .ý ao v w w ED M H a X Z O ä ä d Q
w
74
z C I-. N
co N a igz
äü HN
öý. w
" ch O N cc 0 N c1 d' c-' l0 cc 'D " O M ýD " R v Cl O u7 Gý CU M r- N cc cc
" r+
" d
" co
" " " C! )
" (n
" c-4
" p'
" :'
" -t'
" N
" fl
" M
" N
" (Y)
" N
" Q' CV en M ý' :)
Cn Lf . -I N 0 cc cO N rn o al rn to 0 0 rn N t W N c-1 C) : f) to Ln 0) Cpl 0 in 1.0 cj 4
" " " '
" " " " " " " " " --
s " " -,
" i Q, iN Q
n ci c-1
C C1
N c-i
N c-4
C co Q-. 0 c iý
00 CO c C'{
M c-1
L cam...
y .r
C3 `ý Q ca u W W C) .' H ,. a Z Z 0 (7 O I Cl o
0 m co cO to O cO 0 0 c9 O 0 0 0 0 0 CV r4 N N N CV N N N N N N N N N N
O 1 ý
1 I 1
N i I [ý
I A 1 1 I 4 AA I I 1 0 -1 V-4 [
a-4 (-, j c N N
OD cI N c-I
L c-4
00 TH
lo cI
1 CV
t0 c-4
0ý c-R
M C'!
7
to M. O e''n c d' [N 0 d' to N ti N N N c-i IH cI N c-1 c-4 N
0 N ( N CV
E- I
N "I
t M
V) i Ol
M 4 -1 ' in
-I N pp 1
c- fV ci -{ -A
c 0 -I -I
c 1 N c c c c c c
IT T-4 r- rq TA NP Ln N m vi N c-q N N c-I r1 c-1 I - l CM M 1 ý d N C,
d ý N -I N
I 1
c T-1 IH O T-4
c N c- 4 c-I
c c-i
t c-i s-I
c 0 cI
ß -1 c c-1 r cH c"1 cI c-i rq r-4 c-I ri c-I
N C. rl' 4 9 d' lc\lli d N N ä (N cI c--i v c-l ci c-ý c-{ c-I vi %-I cI r
e-4 tD C) tO in c- c-I M M to W
rq c c W. co (Y) ON ON 0
E/! Cý ti v c c c-i c c cý t U)
N 0 O CY) H N
c-I i c
1 1-+
I co!
Cl) ON 1 A ý CO
ö
L cI
rn cI c c-I
0 % C r c-I cI
N r' C)o OD
O mal' l0 to
v CD t. 0 1 O' d'
tD
in
"' C"' N O' 0' 14)
in In '-0
'- in
"' L') .C cc
s
V' c, to cc co c .4
N el dt CD 11)
.4
i N I
N '
to
N
M
N
d' N M
l0 N
Lf1 4 N
N M t- In
N cc en q tp c-I M M M d' c-i N N d' C11
r' d d d d d d d d d d d d d d d
U Q CO U W Ci, C7 x H ,ý z O Ör
a s
U
w w
75
N E W J
J M CL 0 Q.
J
W Z W CD
N
h-
U- d
ýo
m
0 LL- U; J
W
OZ
W J
v rn c
Z
W
0
N ¢ U)
U
J
U O N
"o a' V 1- ',; t ON O I5 M - 01 CO O', t"', N N O :t 0' M co qt co ; - O u1 M /0 t0 X11 0 M M O
V) M - M M d N U M N M N -Y N M M "t to
Z CC) O M 0\ U\ N N 0 T '. O CO 0 N N %0 ; - Z < In to '. O q; T M U'% %D 0 I- v NT I- 0 I, - cc) < W
rn ýO M O M N - O 00 CO 06 - M
cr- O F-
- N Kl -' < co U W U C) S - -i Z O d O" d O"
Q LL
O M %D ko 0 O 'o O O "D O 'O O O O 0 N - N (N N N N N N N N (N N N N N
O 1 1 1 I 1 I I I 1 1 I I 1 I 1 I 0 - N - N O M O\ M 'o U I- 1- N to co M -
- - N N N - N - - - - N - - - N
%0 -- ()N N CD CN or %0 to - (14 N - N - N - - - N N
0 , 1 p M Iý I 01% to -
O Iý 1ý - M ul ttl ý -
- N - N - N - - - - N
d O' '. O 0 0 00 N M % ý 1 I -O I N r I I N U O a0
t 1 a 0 u - r l - N
r- d' 00 N M - N M %O c- C' Iý N 00 1 I 1 Lr1 I 1 1 I 1 I O - 1 I fý
- '. p M '. O - - -
O - N -
- - M 1-
N Ln d ct O %0 LU W O %0 1 s 1 1 1 1 1 - ON O - I 1 I N I %D 0 U O - d M N - V) OD C' - Ul U V) - - - - - - - - U')
Z z W
N1 O M - O N -
0 1 1 1 N $ CO ON I 0 t-- co I I
cn tr%
co G% cI O 0
- co - co co Oý tý Oý - I O Iý
O -
I O $
l N N CO Iý %0 CO
0'
ý %o Llý ko %ID %. 0 r % 0 M I rn I C ý 1 1 1 u; u'. N I rr)
v ao Ln IT 'n L(''. u-1 ýo
CD N1 to I; r qT 0 WI N M M %O N N "Y I I lq: r I W I (N
W I
N N WI M Ul 'O - ýT M M -7 N M N Q M
O O O Cl) O O O O Ö O 1 O O
fr 0
in U W LL U' S - Z O O' O' O' O' ä W W
76
0 H H
a 0 a
M tn
aD +
0 a N z "r N
Ui) sý
4
Cl) u) Orn Qý Gý rn G) ýn c+ "+ M rn ' cq co " N 'c i C) M [- lD M N O N N c'1 l0 Gl t0 "
" Cl) M W rN cc) Lt) O '. O in l0 d' Co u) in u) co " V) c-ý U)
7 (N O
LO C%
U) O
M tD
01 S
c in
0 [N
in N
d 0
0 C-)
CD (p
u') In
0) N
C" Ql
r- CN
0"1 N z
- 0 t, O) 0) c-I c4 U) O OD N ON M O) 00 O CD N r-' N N M N N (N cI N r-i N c-I ri N N
x x U O O . 0) U W v+ C% H r] 'F. Z 0 O' G C7 0
Ü
0 l0 N N N O N O O N O N O O O (V O T N U) Lo u) ;: Y LI) d' d' Ln C in d' d` d' u) 0
c+i tN 1
O 1
CO 1 A) i
rD 1
O 1
[ý oll Qk 1
ci 1
ci CA 4 M (V d' d' d' C') d r*) M c') N M M M M d'
O c'4 fM 01 N N in N ON l0 00 co O O N M M N d' M d M Kt M N M N 0) M M M
CY\ c) o I 0) I
in .+ ci a ý
cat 1ö oli co r+
.ý °i
N N iP M [}' M V M N M N M N N M
co c-I M M d' cn d' N N M N M N N
O M
O V
OD 'A N
!N M
N
M cä d
O (N
ODI M
00 N
4
N CO N
ui N
M, M
W, N
A
N CIO N CIO M
CO
Ln N
pý c+
l0 cI
v M
O IV
N (N
N M
CN N
M N
[- N
d' N
N M
U) N
V N
N N
N M
4 tI
A t- L
AI N 1-1, to A CIO
M N N
N c'4 r'1 N m fV M N N N rV N N N N M
94
'. D
M N
1 r4 N
to c-I
! U) T-4
N M
I O M
l0 N 1
r') N
to M
1 tV (n
[t' N
1 N N
r M
rV
d' N
1 N (V
0 N
W c-I
d' N
N
N N
1 O N
N CV
N
N N
I O N
c-I N
1 O N
M N
1 c-+ rV
co (Y)
I Ql N
kD
Ö Ö
z N of
0 N
c+
cI C) (V1
N
rV N
N
ca f0
N
ca CV
V-1
in N
C\l
c-+ N
cß-4
N c
t1
c4
(N N
N
O' TA
NI
r+l N
I N
Oi r4
0) t-1
Iri
O N
co
0) (V
N 111
1'
N
T-4 1 ate-
M 1-4
C\l 1
\1 to Iý
Cj (V N
at r-a 1
r- %-4
r- C\l
4 rV CI
0) 'H
i cI
N r4 4 c-C
Ln ci
M s-1
0) c--
r ci*
N ca
rn c-1
M c--
to c-I
I- V4
vli 5-s
to
ri
ci
1 cc-
N
N
t" c+
- cý
t- rv
"o c+
M N
V7 i c
M CV lC r
14' rý
ul ' c
N .+
, +
Cl
t{ C1 c-I
ý' r1
01 H
G ci
cý c-1
ci c-+ c-i c-i
c-I c-4
M
c1 'I' c-+
N
%-i c-+
O r4
O c
M l4
00 r4
N t-I Ol
O c-1
O ri
M 0
N c-i
O c-I
N Vi
O ri
T-i %"-
M
N Qk
o a Ul) ol r! a0 Oli Ell 1-11 C)
C) cy"
N
Tq CD
a CD
a ;
o C)
a cV -I
o at
a '-O
o r-
G ü)
O N- ýI
a i
0 '. O
o 0
0 co
a CO
a 00
a 0 u w a w U w w 0 x H z o ä C\j a ä U w
77
o ca a+ c'+
ýd z ZGcý w Cý
M
VH! ZN
H d' Hw ý'O
4W f. ri
CO 10 in 0 N M OD 0 Q' 0 0 M C\j Oh CO in " Oý M L') O 0 N N 1O O O 1 O N 10 d' Ll " "
ul L4 0 1O cl; 0. l to O
c-1 10 to 10 4
CD Lo Ln un CJ " Cl)
Z al CO ": 4" N O O) d' W d 0 lz: i' N ci N
N Z
z O N
d' %-i
Ol N
N N
N N
d N
ct N
c N
10 c-i
N C\j
O N
N N r4
O N
0 N
10 N
r Z
O Ö r-4 C44
Ü co ü W W C7 x H a Z O d G Ü w W
O l0 N
N in
N to
N in
O o
N in
O V'
O N in
0 d
N in
O V
O di
O d'
N to 0
VI N
cn
N
(1i
A
d' O 'C'
A
d'
A
0 u, d'
4
0
9
N
9
(+) O M
L5 (M)
O (. )
A
m
4
Rl
A
q
T,
r c-I N Ql d' C' [' M to l0 OA 00 01 N O (V O N d' M q' M % M N 0 N O N cM (y)
0 I N
1
N C
I
Cl)
,
d'
.
Cl)
.
d'
1
n1
. N
dI
rM a)
(
N 1
c ar)
1 N
. N
1
O
1 M
N 4
M O
%I 0 r4 C\J d' r4 co 'I N N-4 cO L" N
COO) N N M M
CID N N
[N Cl)
1-1, Cl)
r! M
O
Cl) tN Cr1
CID CV
C) N
j
CV N c1 c)
4
C\j kÖ N
Ö
M 111 Cn
co
iO ýp 10 O 1O O\ l0 cN N CO to O M Ui Q' U1 N c-I M r1 M CV M N N N N C') N Ctr l N CO
C- } 4 'A
N [t cä
! tD
8 1 4 dl 14
t M
N r- N cI M N C7 N M N N N N N N N N M
0 N
to c4
C\j M
in N
v4 c')
tO N
O ('')
in N
0' c'4 (V
cn cm
l0 N
O N
N N
tll N
r M
W rý N
in 1
cI
Ö C)
N N
co 1 N
-4 N
Ln, N
N (V
C1 v4
N N
ci C11
M N c4
r1 N N
CIO N
4Z W
O
0 N
d' ci
0) N
ca N N (4
ce) N
d' CV
c-I N
0 0 c'4 N "O' N
N N
[N H
O N
n') N
tN N
O
H M
In I N c-1
4
N N c1
4 N
c! 1 H
IL
ejl N
A
Cl)
, qr H 4
c+
ci N
cý
N 4
N to t--
N 8
(V
N
OA v4
v4
U1 cý
ci
to cý
ri N
cý
11-4
l-, cý
s-I
l- cý c-I
r4 4
cý
c--ß
to rý
N
cý
N
pp1 cý
d'
M cI
vi c-I
C-) N
c}' ci
Ol co ci
C' cI
rY c4
-! c+
to ^{
ID c-i
M T-1
M c-1
C. r-i
- c-{
N V-1
M 1
CI
8
c-I
"1
N
N
cl It Ca
NI ri
d1 11-11 cl
Ö
c'
Ü1
ci Ill ri
14 ti
cý c-I
N
c-I c4 C4 c-I
0
0 O'
0 C\-4 4 Ln
c-i w OD
C) H Q1 cl
cl r' ý c'r, -l
in c-i
(n cl N
cl L" r4
all 11, Irl
L A I
j ýN
C\ll T--l
010 OK Ö VA r4
Ö N
d co
0,
ý
ö co
Cl)
0
ö cl
d in
0 00
d 1o
ö cl
0 cc -I
d iN
ö co
ö 0)
d cl
d ON
d
R
uý I P'. -
w u W w 0 x H Z O ä cm ä y Q
4,
78
z 0 H E1
I ä °+4
cv
ýwa
xä n' HgM
W t7 ýä `ý w
to e C) ul c% N (N G' C fh N lD N 00 M 10 " lL M r-ý N ý"ý c") N cP Q`; O L7 M In ü) lD L)
" oZ
Ln cI r N- Co un O ci
to d' -D «"ý to d' to to rn U)
c r, N-
t0 C5
M O
L- N
" l0
r CO
v. tD
tS c-I
e N
N Op
Co N
w N
N C)
C% Ol
N 00 2
Oý i
a1 N if ) U) l! 1 ca ýA N c1 N lD O Cý Ui c-i _f N N [V CV CV N N N N r4 N N N , t,
O O Ü
Q 03 U W W C7 x H ,ý 2 O O
0 0 d
U 4 ý
O lD N N N O N O O N O N O O O N O N in LO If) d' to d Ln d d' LI O a
c ci M O f1 +
C' O ci
i
Ol lý N
tÖ u! M N ý' d' d' M ýi' M N m M M N M cal
M N V' Cl d
M e M N M
OCM co N M CAM de'
1 a) N
I U) (+)
1 O M
1 N
I M
I N
I () M
I 01 N
1 l0 M
1 Lf) (N
I 01 N
I (V,
I c-i
C
I1 Ol vi d' pp cI c-I O C'1 N U) If) d' Co M O N c ý' M cý1 M CP ti cm m N ei cm N M d'
(N Co M
Ö
(I) M 31 M
(X) M
ßl
N c-, A N
J
N N N
c-li M
1
N N N
ci
en
L
M
Co
U) ap N O' cM O 5 0) O CO '. O O O t0 O u'1 CM c-I M N rn m (I) N N N N rl N N M M M
[ý Oý O1 lÖ U1 t' C1 1 4 J
cý N
N c-) M N N N rI) N N N (N c-+ N N M
N N
tD vi
M M
LI N
M N
CO N
N cn
U) N
W cI
U) N
d' N
l0 N
ýp c1
M N
CO N
O M
w
Ö
N t11 c')
Ö
f") N lÖ N
li N
t- (N
N
N H
( A
N
r lý N N
tý t-f
c14 N
lb (V
tD N
10 w
O x
o ý pý , -i c-)
pý N
i N N
ýp N
tp N
c-) N
l0 N N
N N
N N
l0 c-I
O N
Lt) N
u) N z
H U1
tn [ý cý
tý N c)
Vi N
t M N
Q1 c)
0% cý
d e
Ö N
(D1 N
1 CE) r
LA c-) + N
N N
to W H U1
l0 ci
c1
ý1 i
cý
tp (N
M N
N i
Ln cl
O N
[ý c-)
N N
N
pp c-)
%-4
CD v4
Ln %-f
tý! N-1
1 N rI
Q) X-4
1
c-)
OM i
[ vi
N r) 4
ri
cam-) N
rH
ca 4
rl
N
Ol cI
N
c-)
(I r
r+ r+
N N
-e r{
%t`. cl
O N
cn c+
r" c-ý
, c-a
I. ̂ c-;
Lr: ; -t
cc- V-+
CJ %
ýn r`
in c'
Ln c-)
1
M c1
Ö Ö N V' OÖ c4
cI
N
r'{
O If) %c-4
In, c-)
Ö
c-1 c' c-i
c
c-i t0 c')
CV c"1
M
O c1 vi
cn N- O vl
vi ll 01%
d' -r-1
V' IM 0)
O O U) r+ N
41 %-4
011 tý v4
U1 Iri
Ql W M
K-4 t [ý Oý 01 t`
N
0) 1
0
Co 1
O
tf c) 1
O
W I
O
O ci
I O
ct' c-) I
O
vi I
O
CO 1
a
N I
O
N c, 1
C)
(N r,
1 O
l0
O I
00
O 03 I O
M
I O I
O c-+
4 ¢ ra v w w x H a z o ä a ö a Q >ý w
79
}
EE+
7- 0 H H
0 Cý +
4
O w
a
w "n w
N co
tl z
d' in
a
0
LýJ
Q') O Cl CO N N M V' IN in C )'N c-I M O [N CO " ` ýý N N Oý r, O m W N c-t in M O + Q . . " " " . " " " " " " " " " " []
to t"7 N N C- Sr O 10 10 In d" co Sr tri to (7) N a
z c-I ci
C m
CO in
[i' m
Co in
Sr CO
co in
N c-t
Co o
l0 N
a) Co
c-i co
CO in
L to
N t--
c") O Z
ß1 d' CO N
ca CV
Co N
LO N
in N
cI N
l0 %-I
N N
N N
N N
l0 c{
Cl t-I
I) N
l0 ("4
E-4 Q ¢ m v w w cý x H a z o O O c ä v
a w
0 0 Sr
c0 C\j
N in
N U"1
C\j in
O Sr '
N to
O O d'
N in
O Sr
N in
O d'
Sr O Sr
N U") 0
cA N
(n ci 10 I M
CO 4 A c,
LN N
lD u7 .
M N Sr Co C}' Co "'i Co N Co Co V N c')1 Co d-
N cli C\j C) N 1-4 M am N Sr CO
Sr (') V' M N Co C" ) Czt Co ' d
Oý 1 Co
I
Sr (A Co 1ý Co CA Sr
1 Co
1 N
I
Co 1
Co Co
I
N
i
N
Io
Co
1
Sr
N Ol 01 CV U) Sr N O Sr 0 a Lt') N co m Ol lV c-4 Co Co M M Sr Co N Co N M N N M M
O In Ql N (311
N 'A N c)
r! W Co
N 'A IL C\13 tD
N (0Y N Co Co M N N N N C') N N Co M
Sr N 19
w M
CO N
c Co
(N Co
t0 Co
IN (N
0 N
LI- (N
N (V
c- M
0 N
to N
cI Co
t. 1 Cl)
Co N
rý c-I
N
CO LA N
N
N
I
M 'A Co
to N
Oý N
1 cb N
00 N
(11 cI
'A N
Öý
(V cä Co
W ö
N N
I
tV
Wp
to r1
N Co A
N
Sr N
T4 N
tN N C'
N
Qý N
co N
N M1
(0 N
Sr N N
N
00 c-i
to c-'
b' N
M N
to N N
N
[N N N
N
Co c-I
t- 'r1
N lV Ö
N
Cp N I
N
O Mt
lQ N
Co Ö
. bi m Z 6 01
\-4 Sr H
CO (V
O N
c') N
N C\j
U) N
c-4 N
Ul T-4
N N
N N
N N
W ci
CO ci
to N
in N
Ü
, Ö I ý LI &'I t! j
tD 1-4 c'4
10 N
0ý c-I
Ö
N ýi N (V
Ca 'r ca N N
ýp ci
LA ci
l ri cm
N
N 0
Sr in
T-1
N
1 T4
ßt1
1 N
N
! T-i
a
1 V-1
Sr N
N
Vq I
c-i
N T-1 I
c
M T-4
I tr
Ol ri
ýp1 cr
O CM
1
N j
c-I
Sr `i
ci
Lb V -f i
c+
en N
N
c-I N
tl T-4 -1
rn O Ti
ci Ni c4
ýn ri
cv N
ý' c'
C T-i ca
N \-+
N c- +
z r
(' 7-4
Y t-3
O CV
y ra
M O
cI co ca
14 c4
N
cI
Ö
CV O cl
N
c; Ok
4
c1
( l0 r4
N
cI tI 1-4 I
N
cº rI
N
c-I
M
N CO
CO
ON
CO
N c-S 14
V-4
O 'ri
col
c-S 4
0O
O) c+
T-4
C) i
0
Co M cI
u) ci
c-i i
CO
O Cý
%-I ;
0
N ý .4
N ý
pk N
T-4 N 1
o
N t
o
fn T -4 N 1
0
[ 1
0 V4 c
8 in 1
0
CO 1
o
in
O1
N c-I 1
o -4 cc I
0 CO
0I
01 8
M f
0
V) c4
1 0
CO i
0 -4
ö g Uý Q as v w w C7 x H a z o
ä ä
Öý ä o
W 4
80
tJ -V O VýI' N cß') : -+ 01 J' N u() (IN ßl c+) " CQ 'D O O O Cl M 0
. fl M M O ri (Y) C Gl "
" " " " . " " " " " " " " " " " D " U, M t- t- S
4 O tD V ' tf 4 N d U, t7 C) "
U; ci y
LU LO CO co N lo v, . ýý a ßl lD O Ui (V Cl tUi Ol CO N co lfl co 111 F..
co N C1 O O N 14 i a-ý U1 c-: l7 c4 c-{ c-+ N N N N N CV r1 N N N c-i N N N
er) o 4 C1 U w iL, t7 X H +-a Z Z 0
a Ör a
a U
w w
)
)
N N
w
0 H H
ao
Ö v'
sa z
in a 2ö
aw 4
O W N N N O N O 'O N O N O O O C\j 0 41 N u1 U! ) tf) d' I d' d In dl to d' d' '' U) O
Ö O CO
I M t W [-
M N d' M M c*1 d' c'") N M M M (V M c*1 C
O1, 01 rn N tf lO N I'D Cl) N to In (V Q0 O N 'r d' M M U C U`1 N M C*) C) N M re)
0' 1 pp N
1 ci d'
1N ef)
NI M
1N 4' cI{
r''1
4 N
I c. M
1 O rM
1 d' cn
I M N
1 O rn
1 uý r
1 n m
N 0 ci c-1 d a-4 O U O ON C') N O') 'T t0 N lq M M M d' (In N U (V c+) N C\j M M
CO N ll,
C rt CID
N [N (n
N Ö Ö N
N N
N T-1 M N (V U M N N N M N N M M
V to N l0 N c. 4 '. 0 [5 O t" CO ON 01 l0 c-4 N N ri m N N cy) M N N CV CV (N N cn V)
N N to1 d'
4 O1 O1 Ln! ollI
1ý I r--
' 1 1 1 1 N
N 0 M N N M M C\1 r-' N N N ý N
m N
N d' (q en f'7 Qs Ol d' 0) ¶' l0 Ln N ;; N 01 CO N ri M N N (v N N ¶4 N N N T-1 (V C\j N
W
ID ' Ö
N c1 t"4
O U
O N N
a) N
to N
(1) N c-4
N
C\j N N c4
N
(V r-, N
l1 N
ILD
O
0l c+
lV ci
01 N
01 ¶
0 CV
N CV
d' N
cV (4
l0 ri
-4 N
U) N c4 C\j
to r4
c. 1 N
0 (V
q CJ
ö
CJ1
In N
V4 T4 "
Lot N
[N t. 4
OÖ ¶4 N
Ö
N Q1 c,
Lf1 c-4
Ö
N
N
N N ['H
4
¶'l Os c-4
Ul N
01 ¶4
N In
10
t0 ¶4 N
c-1
O H
to N N
N
l0 c. 4 4
at
N ¶4
,A ¶
U') N N
N
C') ¶
c-,
CO ¶4 I
'CA
¶7' c'4 N
c'{
O') c-4
r-, ¶
c-+ N
O N
t0 t-4 A
c-1
M c4 N
TA
co ¶4
[l, cH
d' N N
N
00 c'4 4
c-i
(11; N C') 'z? (\: C' L) 'c kO (,. N Vl N Cr)
U ¶4
'
V4
Gl 1
CO
N
1 N
¶ 1
VA
¶
1 c'\
1
r4
¶4 1
-
c-{
c-I
^{ 1
'
c4 (
cl
cr !
¶
¶
I %
" c-i 1
ý
N 1
ci
¶4
I C''1
N O') NI
l5 c¶-4 NI
O
Ia)
O
CIO
ca IN
co 1CO
c-4 1O
O')
CIO
¶4 1N
cl
U1 ')
01 I N
ý
aIl
%-4 1
¶4
H
1 m ! N
%-4 0 t in
i O
'- c4
1 O
N 1
O
N 1
O
¶
1 O
N 1
O
m 1
O
NN 1
O
T-i c. 4
I O
14. ¶4
I O
ko I
O CO
1 O
O r{
I O
t0 ¶4
1 O
l0 1
O r1l
aO ä
w Q as u w w 0) x H a Z z o GF Cy (y c( Ü
w
81
00
4
Hw NHA
L Al
" to C)
co N
r- ll
C1 cH
LC) N
d' M
CG QD
It '
r1 %
N (V
0) co
to rH
O : f)
cc) M
w LID
to L, "
a in M N co C' to O N if) tC d' CO Ln 10 10 C)
Z 0) IN
tD N
cH cc)
CO CC
c-1 N
N d
C' d'
lC CD
M r-I
N cH
qll 0)
M O
d' O
c-H "
7-1 O
CC) O 2
" " " " " " " " " " " " " " " " Q' %D O N to Lo N N LC in 4 H ýp O If ((I
6
cH TA M N N N N N ci N N N cH N N N
E-4 T-4 C\j cn U Q U W G Cý x H i4 Z 0 C% C7 Cr
CI U
O to N N N O N O O N O N O O O N O q N In in ul -' in d' ,: i' u') d' Ln ". N d V in O
co (f) N d' d' Q' M
w M M M
M M o M
o am C\j c '1 N d' d' d' ß'r1 [N M CV ()) c) N M
cN m v t A o 1 cl, 01
N N
N
t}' [ M
N
Iq
4
q; r M t N
lÖ M
O co
N
M 00 (V
1 4 Cr)
4
M
a
c}'
N N d' m V M d' M N M
N M N
N M C-)
co rt Ck
A tÖ
N 0 'A (V
N r-I
O N N N 11
CO N C') M r7 M d' M (V Cl) N M N CV CV) M
to Q1 (ý lV to N 0) O cH ý to 0) ; 1' to v-4 in N H 0 M M 0 M M N M N N N N M Cr)
N rn cl N
I
C"1
o
M ui c4
0I N
OI c'H
all N
1I N
11 N
ci N
4
N
a
c*) 1- 3 W
H
V1
lý
t'+1 N
O) N
N c'H
to c' ,
(4)
t0 N N
N
C' N 4
fV
O, N A
N
w M
W N
r- N 4
N
CO ý
It cH
CO CV
u1 N
g N N
N
Ul N
cli N
O N
1 ý
(. 0 N
ci N
0) N
N
O M
lÖ N
t0 Ef
C1 -$
to ci
O m
H N
M N
tD N
N N
C') N
3. O cH
";: f' cV
N N
O N
N rH
O (V
If1 N
U) N
N
ci P-4 N
ýR C\1 cm N
N N
ýv. N N H
ý
H
ý
(ýV
N tH
4 eH
'4' rq A cH
N N
4 N
[N cH
tÄ cH
O N
r-I c-H
(V N
d r4
O N
[, cH
01 r'
If1 TA
M cH
1-1, cH
" N
C1 c'H
Cl cH
ap cH
l0 H
A cH
to V-1
A rH
to rq
4 rq
N N
ý rH
O cV
IL cH
(4) cH
N sH
C4: N
-I t-H
to !H
Q1 c+
t0 c-{
C' -
C) cH
00 'H
N T-
N rH
fV tH
M c-H
N r{
to ri
M
T4 cý H
J.
N
N
c-H 'A ca
U1 r4
N
r4
N
rH 01 Cý
cH c-{
a
c-H ci cH
r+
r-I cH
ca
to
m
N O sH
O eH
O N
rl' ca cH d,
N c+
CA
d' r'
1 T"H 't-4
1 00
c'H vi
1 O%
co n
to `
un C, t`
O e-1+ OS
O
! C)
c"1
I
O
U) N
ý N
a ON
a s'-H
a W
a GO
a r'-H
a N
a co
a 3. O
a T-4
a T-1
a to
a co
a l-
a 'H
a '
a ri 4 4 to V W W C9 x H ri Z O Cy O (Y aU
Ü
w
aý
In H
hH
A
f1 CD N L') Co to o1 m N N N N L7 lD cn c-1 " N c{ r- C' C' C% Ol C) 6' c-+ m - d p1 cal 0 "
" tD M tJ N ß'. in O ýS t0 in CC in tf) tO 01 " N in
Z r4 G1
M ý7
0 ['N N
N tO
c-4 N a) M
cl if) N
ire ýp
d' N
0 tO
C' O
cl O
c1 (`'1
N N Z
I Ch Ln 0 M
N N
N N
1 N
l0 N
c'ý N
to c'
M N
vi N
vi N
CO cl
O CV
Hfl
o Ö
O ¢ U =ý w c7 x E-+ a L Z 0 vt ä (M c Cl
a
o O d'
O N
N in
N cn
N in
O e
N to
O O 2 N in
O -r
N in
O *
O e
0 'cr
N in o
N 1
N 1
t7 {
Ol t
l`ý 101 t
lý I
to I
O I
r-- 1
N 1
ßl 1
O N l
tt) M N d' M M Ci' fn CV M M m M M M d'
c-ý c N Co to fit' tJ C [` ýG c Co Cý cV d' (n r) N d'
c+ C')
tC
V' '(r) L'" C") N
N
il)
! 'n N
M Ö
M
t)
N
CO
M
Q1
Cl'. m
[t' o 1111
N d' C7 cJ' M d' t) N M (+l Cl) N N cM d'
N N d(D ' Ln ri c114 .4 (M ('I) ti -e M
N M
N M N
Co M m olý
l4 N
Q'. ti
Cý r7
cl M
mI M
14 (r)
CO M
j
(Ni
ý ,
C'h tý N
Ö
M to N N
cý (n
t11 M
Co
u) N
ý) c1
N Cl
o C
N Cl
O tM
IN m
co N
O N
ü1 N
to N
o , N
d' N
un N
O M
[r Cl
r 4
N d Cl
C-' N
ý'
M
J
N
A
Cl N Ci ra
N N
uj N
to N
d N
C
N 00 N
c m
Cl N
N Cl to N
m Cl)
Co N
N Cl
in N
W cl
tO N
«e N in
N O N
Cl N
tN N
p Cl
(rj
3
(V t c-ý
d
fl) (N fV
a) N
to
N CDI N
N
N W cl
14 CV
N
N
N
N CCI) di
c+ N
Ul N
tD N W
C1 VI
to
O cm
N rl
ºt1 cl
O Cl 1
Cý N
ca N
1 di
t- N
1 N N
trl N
N
r- N I
-, CM
N N
cl c-a
ifl vi
rl
Cl (V
f. "
N
c-i N 1
c ci
vi N
1 N IM
P t-l
Vi
O cm i
Co -I
d' N I
N N
ýn N
N N
N
[-t U)
1r4
tý-ý
cc-i
rýº
N
N
N-4
Uli
N t-' di
N ýý
N
%. n-l
di
ýý
di N
c
c
1101 di
Lný
T-4
t- c1 r4
TA N
1
vý N ße-1
a
r; %-
N S.
(') N
C' VA
l! ) v+
O'' X_<
C' n
' c-
r' c-.
" c'
t-^, -'
C') v-i
N N^
^1 ci
[N t--)
it'. c-
<'ý 1 ! ! 1 I i I 1 1 1 I M cý cý
N N c-i c-I
to c'I a-i
ti d
Q di
to ri
C c-
O %-i
ci ci in c-1
(4 TH
Cn Q N
'rVi 4 ri c tai Oý Cl
r %--t (A
ý 0 d 'C I
N 1 to
1 G'.
1 N
1 O
1 C'
1 N
1 tý
1 O'.
1 [
i N
{ N 1 [ý
7 CA
1 1 ß
1 O Oý
N
t- OD
C1
m
01
t-
01
OD
O Ö
(D 0
OD 01
0 ä
w
01 cc) 01
cc) 01
c% Ö
Co 0
M.
4 t10 t! W G, c7 x H a Z H O Ö'
a Cl O U
yy
83
x 0 H H
R
5 N
o1
HW s1
w a 4 w
4
a N
Z
N U N
qV)
U a
8 U)
to t- c4 u) N M a lo In N N N lD v'1 tO "0 " N c-{ dl if) M ' M O N lD N t- to `Z3' Co "
" V)
In M %D t- Co uo O ci
O tiro U1 d GJ d ul in Cp ti
2 Ol r:
CI 0
N mot'
'j' 0
%0 to
(Ni in
O O
%. D (''l
Ol O-
M v'
M in
0 t0
N d'
N CU
O CU
M Z
a Ol
j c? N to d' in O '. D 4 4
N N cl th '. O cý N N N N N N C\j VI N N N c-I V-1 N N Z
O o
w < rý o w w C) x H a Z z O (Y ä c! ä Q
w
O O d'
l0 N
N Hfl
N u-)
N ul
0 d'
N U)
O V'
O d'
N If)
O cl'
N to
0 cl'
O d'
O d'
N tr) O
Irl ri Qý ni
N üý ci
ül c, 4
C') N d' M cj' M d' M N M M M N M M 92
U) N d' d' C') 'd' m N c'
UM Co Q M -e
OD CV
0% [T d'
to M
O d'
A
M c d'
c M
tf, N
O M
0'. N M
to N
Ci N
(N M
O
c'
O
N ý1
in N
ýp ri
O
1 N c'1
d' C'1
1 o M
QA M
M
N M
i M
O d'
I M
O) N
I N
d' N
N
ON N
I N
C) N
1 N
ý C')
IN U)
d' N
N
CO N
N
¶ c+1
1o M
0) M
1w M
00
d' N
N c-I
to c'
0% tV
Ln cM
Q) CV
lD M
tp N
N N
l0 N
l0 N
c-I M
N CV
l[) N
OA N
in M
r- 11 IL 1-1 a0
N 14
Ö 4 t1
Ö A t,
N t- N c-I M N U) N M N N N N N N N N ('))
w
w
N N
O (N
Ln c"1
¶4
N M Ö
M
LA N N
N
O cn g
N
N N i
N
ri M
X11 N
I'r) (V
cä fV
C) c1
ll cI
M Cu
c+ N
d' N N
N
tN N
M N
01 c-1
CZi vi
N N
O N
l0 N
N
VI M
tý N
to w
Ö O bi
Z 1-4 VI
in
() ci N
t-I
cM vi
O) N N N
c-+ N ý
c-1
l0 N
N
d' N
1M N
e N tr
I
c-,
0 N
1 O c-I
O C1 N1 V
O N 01
1
ci
ci N 1O
N
N N Ol
1 VI
N c-' 111
I1
ci
0 c-4 [Iý c-
M CV N1 N
lp cV
I M N
ýn
2
"' Ei H U)
to VI
14 Im
cy c4
14 -I
to N N
t1 cm N
cm N cm
ýp ti
vi
N
ý
"'
tI
CO
cI
C) N
vi
W cn ý
V' ci c1
tD c-ý
c-I
ý"i N
c-C
(V N
19
c- ¶4
0 ¶4
N N
V' vl
N c-+
O-, ci
C' ¶4
th -
v-, : -4
tD ¶
'. 0 M N -e ci
r c4
r %-i
M
¶4 0 N
N
t', Ifs r-' s'-
o
r4 l' r4
O vi
4
c-1
"
%-4 li
IM
! tp Ti
N
c-I
M
N C1
c
0%
C&
c
v4
V-4 j
c-I H (3)
IN C)
O)
010
¶ r+ (N
9
H
Irl
Ö
%-4
c1
4 Ti j
`l Nd [S
a rd
CO
d d Ti N
o tO
a aD
a S
a a ý
ý t0
a Q)
ý 0)
ý r4
a Co
a ý
w 4 co u w w C7 x H a Z 0 ör ör 0 ä
0 w
84
CD
z 0 E-+
4 0'
F
m
4Q
u
" N v
L') 1-7
d' n
n 00
UO r,
c-1 )
N l0
t" %
O N
O O
a rl
0% CO
t7 0 c-+
C to
UD Cý "
" LIl M ko to CO CC Ul r! ' cn d' N V' I I) t1) CJ "
;z N 0
C0 n
N 1p
0 cl
O LI)
M to
M ! fl
M N
'r-4 U)
(N n
t0 a'
V N
a' M
M M
co N
n Ln Z
Cl N n u'l G" C) ri n O c-c p` n 0 M n G7 .Z c'c T- N N (N N N N c-c N N N c-i N N N Z
O Q
s G+ U W G., c7 x H a z O ä ä C!
CI , zp
0 c4
0 c' I
O
l0 N
1 O
N u) A
N to 1
q
N U) 4
O a f
to
N to
LA
O ,*
14
O cl'
CIO
N in
c-1q
0 d'
1- I
N in 1
0)
O V t
co
O V' i
O
O ;r
u)1
N cn O
M N d' M C C) -z; ' M N M M M N M M "; r
ON ON N n M c! ' cl' 0 S O O CO E-- C) cl' M N ri d' M d' N) ct' M N M M M N N M d'
a i N
f p ppI
M
1
M
i PD I
cl
I N
I
N cl% N
1
N
10
c*1 Ull)
I
N
I N
1
M
1
v
ý 0
n N
N a-i
N C")
(N M
a' M
N M
ON M
CO N
d' N
0) N
co N
tf1 r')
'q N
n N
N ('C1
O Cº
co It N M CIO N .:
f, n C) -A C') LIn M
! ý N N
I co N n N
A cM NI CV
1ý N oý
1 N M
o0
E-
cf) N
'A CV
lD cI
ea
M M
14 M
n N
t1 N
d' (C) Ö
M
0 M
CIO N
d' M
CO N
l0 N
L) N
N tV
01 N
L17 Nt
V' (V
lfl fV
to N
CV (C) Ö1
CV
rl CV
O N
In N 4
N
CO N
l-I N
to CM N
r')
r,
4.0 N (4
1 V-1 N
'd' c-+
'A T-4
O M 0
CV
d' N
14 N
0) (V
t) (V
n N
i) CV
n N
Cl N
V' N N
(N
Q\ v-I
W 10
M N
I'll CV
d' N N
N
CO N
to N
O' cI W
c-I
M N
ci N
kp CV ý,
N
Cy M
1 N
lD 94
j 4
0 i
H U)
Ln
0 "
I CO c'c
(N cI
N N
I to CV
O CV
I CO V4
'' CV
I t-ý C4
d' N
I N N
N N
1 01 t-i
r+ N
1 0ý ci
n c-I
I cf) ci
O N
i 0) cý
ri N
I O N
d' N
1 I
(N
n t-c '
cI
O fV
ml v-I
CV) N
1 '4 N
r- N
(v to
[H /)
N cI
ti c
c a-I Ö
t-c
d' (V
ci N
n ci
to r4
O N
QO c2
C \l \l O N
a) c
t+ d'
40I c -
ýÖ c-c
d'
t-i
0ciQ
cc
c0)ý '
Cn t
Rl
'4
U) rc
ci ri
cODi -
tp c
O N
I
ý-1
(C) N
Oi c1
c'" C) N t- 0'. (-', c:: r" L) h CO
T-4 '{
'.
'4 Ti 1) ci 1 r1
r4 'A
c"c r4
M
c-i c+ n
n c4
N ci
(") c-+
l0 c-+ Ch
N c4
O ci
M 'H
LO c-+
(N c-I N H
t! 1 c-I
N c-+
N CN ti
U) O o
.4 1 Ö ck
.4 . 31,
Ö p Ok
N
CO 8
u1 0
c
8 0
0
01
01 8
u1
01
O
01 CO 01 01 01
OD
01
0)
01
01
01
ci
01
() 0
. W [i, 0 w H .7 Z 0 O Oc C7 C7 U W
CQý+
85
APPENDIX7
DESCRIPTIVESTATISTICS:
PRIMARYSOURCETRAITSBYSOCIALCLASS ------- - ---- ----- ----- -----
86
N
[ýs, ]3
H
Cl)
NU
U)
U)
U
C4
i4
0 w
U Q W U W f. C7 x H I 2 O
C1 O G G 4 t:.
" rl t- ýn '. O 0 u ý` n co q CO Cý d' cý V) 0 tD cV N N O N ci N M
0 N N M 4 4
M In N M M 17 M f7 14) N 4.
c+ä N
U 0 CV CO (V ' fV 0) '. G (V CO lD l. 7 N 0) lD N
Z 2 cM tD N -4 CO Lfl (fl Q\ N N CO N th m CO CG v Q " " " " . " " " " " " " " " " "
zC3i -i Sý M O O\ M ¶ N 0p c' N N N N N ? ri N ci N r1 N cý N N N c-i c-+
" N 0% N N N N Co 0 w Lfl [i '. 0 d' N a\ LO A to 01 d` N N r Zo ul d' N O CO N W O' ri V) N r: M M d' M V' N M M M Cn N 14 N '4'
N A
u l5') U) 14) 0 t0 OJ d' 15) O N N Ol N T-4 CAD In
2 2 N V-4 CV t- O CO O" O N CO O LD tN In t5 0 ,
0
c t0 C Y) N r O M 00 O N M co c-1 c-a to VA N N cý N N ri N N N VA N
" N p In Q N H N d' co to a% cq W to 1. ) c-4 A N O W o'. W N ci [N If) rn c. - 01% 0 cal (A In n " " " " " " " " " " . " " o " " " " N V! N N M M 'd' M ifl N m m M Cn M N
4
N 01 M
O p un cn
r, M N
r+ 0 m
N c)
O 0
00 O
lD )
Orn N to N CO .1 N; r ' .i
ap
O l0 M
N
Cl N
c
c c M
(3
N c1 4
CO
N
"
pp T-1
l0
N
V
V z N t N c1 c-I N ct c-+ 5-I s- N ca
(5) N N c*1 d' to 1, O 0 0 N to N r+1 c-4 CO A co 01 O co M In N ON (n N N O O U) N ko In N ca ý4' ch U) M Lo N N c4 M Ct M M M CJ'
c5) V4 N
d eN M d' 0% O co O O s-I N cn r N 'N rn tO N M l0 t- O o' c-' kD M rt M rn CN d' O O Z s + ' - 4 i " " " " " " " i c N '4
Ö to r cr 9 N N ý
T-1 N ý-i T14 c-i N -1 T c--i r4 11-4 -4 c-i
" tý dý N s-i c-I d' in 01 N CO CO C") t0 00 0 A 0 c+ co tD r+ 0) 1. D i0 m o N O M O U) In M N M d' N M u1 N C4
4 M M C7 M U)
CQ
Cp ri cý g l0 c-
h c-1
N rn
l! T-I
) Op ý co z z O r O r in aA O c\t 4 4 14 r: t4 c! C4 C 1; 1ý C4 4 r: ; C4 14 V4 T-1 V4 T4 ri r4 T-4 T-4 Tq c-4 N N
" cv G^. Q) C' co ci 0'. C -, V cý cv t- r A r, rn N a c,: C t-0 Lrl T c-+ ca V n ü
c "+ N CA M 4 C4 In 11 M ßf1 14 M <h cal C") C') M M
4
4 in
A N 0
2 c-i to a) M co to in ;: r CV ' d In t. 0 V' 0'. U) " " . " " " " " " " " " " " "
T-4 N V' G1 CV fV fM M N U) ri 4
rz s-1 e4 r4 t4 c'l N vi s-I ri r4 ri s-i c"1
9 t oo v w w 0 x H a X Z O .a en N: r a cl
87
CV)
E-+
U,
U
HUý
4I U,
H
U) W
w
A
a
VA
En
° Q U) U w ct. 0 x H a z o ä Ncl ý
w
" o
ci to
t! 1 0 m
O c-i
(n tO
V' in
S7' Co
0 W
"d' M
(V l0
00 CO
M U7
kc) t0
cA M
c+ h
to r
M V) N N M t1' ;r n1 L? CV CV 14 (V 4
N M N c
fl
Z z C) M (V
M 01
h 0
O h
r' Hfl
<J h
UO Op
01 h
In CO d' co
LO co
CO to
(0 O,
lO "
O ýLD N 01 %-q M c c op (n c+ CV N o) ('3 V
U') to c' C. N M Ol U) O O O co M h to cl,
n A s-4 Qi to V" fV O O\ to 10 d' t0 N 1.0 OD M N (V " " " " " " " " " " " " " " " a " t-{ N M ý-i co co d' M 4' N N M N CV N m V N
A
z z
ý
Q h
O s-'
ltj d'
tp
(13 dl
N 0
h M
Co tl1
Co t'4
0 lO
cv c1
01 c
O c-;
l0 M
O "
W u')
C;
tV d'
Oi
M M
O c'4
co C')
rM c-I
N Oi
h
CO N
co
Cep M
p c-I
M
in C1
A c{ O
f- CO
h m
0 CO
0 cC
N OD
O\ lD
c"S h
N to
to M
Ci in
c0 h
r4 co
OD Ol
c0 M
(1 C3'
o 1.0 U) M c-I M M d' N t r; N M N
4 N CV M d'
M N U
Z Z N N
O
'd' c4
h
tV Cep
M "
tV t0
Ok
'0 CV
11 vi
M (0
N
c4 c,
c v
%D v4
O N
M" W
ca t! 1
"
c-4 u'1
"
M Ol
"
d' 0
co "
M N
" Mcc) w
c' 0
"
ri
CO (D
"
-. 1
ý. "
A t0 C
QO %D
(V U)
N un
to 01
0 co
h 00
N 0 h 00
Om 1.0 H h
h . " " " " o . " " " " " " " . " a r1 (! ý M c-4 M m d' N (V N M N ý3 lV M C) Ci N
cý U If
z z i) : ý.,
(7\
u, A c+
t4 a) ýo tý
h to s3' c
CO . fl of
to
%41 3 c+
oN Q o cri c+
N
co c-4
0 cv . -4
ttO N
d OO
N
" OT-4 ý' q
"
'. 0 to
" aD
O
" aC)
C) inn "
cc-I
cn c+ "
r-I
" q
CO p')
In 0
l0 c-{
s-I h
Q') N
It) O
(V -Ci
CV CD
l0 O
CM O op
'T' ý om' V O
N
cri N
" U) co
" 4" V'
" 4" d'
" " M
" lO
" N CV
. " " ci'
" M" M"
" d'
"
eI
Z v
Z h C0
" O
th CV
" oD
d' t-I N
cý
O (0
c
Ln h
ý, ý c+
co Ol
N cý
W h
M cý
O h
N c+
- h
z h
to d'
r 4
c+
-P 01
" v4
rn co
"
ra
h M
" a%
d' N
" °c)
o N
"
. -I
o to
" v4
" U)
W CO
CO to
w ü1
h kp
CO CN
N d'
c'1 ß')
CX) -1
E- 1.0
N (0
co GO
l'- lD
Ci N
N M
CV r:
o tO
C\ U) M rl M M c3' N U') M N M N 4
M M tM ý3' t1'
4 0 V4 0 ý. {
ýR
OD 1n 11
'T
h to 0ý
ýO
M s-I
M
'IT T4
C4 r4
(''1 1-1
to h
, r, c-4 c-i
0) O c-I
O
M r-1
T-4 pN
cN co
m
r4 ci
rn M C-i
Ü 4 [A U W W t9 x H ºa Z O V-1 C\1 (4) O v v
88
D w
N
cr,
U
8
U, wl tlý I
wl H
[s. ä
to
U,
ra
ö w
C H < C) U w cL4 o x H :. z O n N ý N C
" d C In N IC) 0 In Ol "ZZI Cý Q c-i In In O M lJ (N Q' c-i in fN In N to In
N Cl) c' cM CO N a) 4 C d' 4
to 4
N 3' In lz7 N W
II N N d if) tC ý4' l0 ID N N co N N -;: r c-+ T tO N V' O" % N N U1 In cl ul TA l0 co N N ri l0 Gl 44 N M V' N O M Lo Ul r1 tf) CJ N %- N T-1 N N N N r4 N N N r1 N N N
" d' O N f'') ON NT LS r1 l0 l0 Ol N '1' co N (y) A N M r1 OD O C4 CD OD co N N c'1 0 N d' to
T-1 In V' M 10 I) CJ In CJ c'ý1 d' to ý4' N C7' If) U1 N
A N
Q to N lD fV CV O N In 41 c-4 to m to m ,o Ol
z z CO to C4 O n M O O O M N c+ to cm co 'i'
t N '. O ON In 4 4
N O N N tD O O N c'I N c-I (V lV N N r1 N N N c-I N N m
" 0 r as. co (V In CA CO In dl cx In ci 0) V' O1 ý-. A 01 N to tN O , O M ON (V lf) In l0 cm ci co 0 " " " " " " " " " " " " " " " " "
In 4' M '. O '. O 0; Ln O d' d' S d' to to N N
M
O t-i M O t! l l- OD 01 S C') In d' N-4 t- c cy) c-4
z z l0 ()N to CO l0 r+ '1' N N M N l0
W c4 c4 S C x) lfl C" N C4' lq O N In l0 Cl fV CSI N r4 (V ri N N N N c-' (V N (V c-i N N
" O M ON N CD 0 d' O) to co d' N O1 0) M ON "" A cP Q1 in c7 V' N O) LS V 0 c\t IC) lD iD vi N
" " " " " " " " " " " " " " " " " [-I dl t! 1 N 1p In Ol In O d' d' tp U1 N l ul tD 00 N
T-4 U 11
0 ON d' CD U) CC) to C4 ap d' IN N CO N O d' z z ps. O c'1 CO co 0 O N '. D Cs. rM M c'"') l0 N
" " " " N Ln v N ri lp 0) cM CO y N T4 N vi (V N N N c-, N N N %-I rl N CV
" to ßý 0 d' N N [M N d' Ol to M r1 lD t0 Q
S d' m co Os. d' 0 vP m 00 N LO to N tN O
In L4 M tD N CO tD U) l0 'i' N L 10 t0 Ct
PO h
N U1 L- lD M O In h Ln r W (V z z d' 01
A N d o kO rn r r d M c O ýO
.r " . " " " . " " . " . " " " " " c+ u) O O W un tD to d' d' N M tN m d. 'D N c4 m N N N N N c- N N (\I c-4 ci N N
" t- In In CV cc) It) N ', L '. co c'i CC' N 0 C.. ) CD 0 O c"i L- v'{ IC) co 0 c-i co N ('f: M' N co [7
rn vý ýO M N Oý d O ýn to ýt N uý u1 to N ý
co 14, %-1 0 (7) co LO ý Z O r4 d ci N Ur In N M
CO L' Cl 0
1'1 LD C1 0) 11) N l0 ýD V N d lt) Oý d' CO tV cI (V c-I (V N CV N c-I N N N c-i s^a N N
a o 4 m U w w O x H O H ö c
r
89
co w
U) U)
U
8 U)
U) EJ zU Wa
Ii
114
U) w
d
w
C4 0 H Q `= Ct W r, 4 (D X H ,ý 2 O v1 N M G O' C C
" c Cý Gil N In CS' tO N N a) t0 Co r- fý CO co %- Cl N N r-1 0 ON Co O r- GW
Oý (1ý N 11 M M 1Y' N to (fl ('7 M N " N c'f) N 4
N
u LU N N c-i (fl 0 CO Co Co V' m Ln Lr) d
Z Z In LU 1. D M O Ql N U) C T Co LI N l. n c-I G4' is
Co LUU LU 0 O (0 N Co N c-- s-I Co 0) N V) O . z. c-I t-I cý c ci s-I c - ci
" N Q' 0 crU 01 c'{ 0 Ci' to N N (V tJ ; i' 0 G' O Co O c-4 CO N cM c-I N ra co c-4 11 c+ T-4 W
M U) r1 s- I 4
.4 J' rl t! 1 M M M N 'C' M M r) 4
N
a n N ON -N Co In 00 'd' IN 0 0 M -t %D -V (-, i C\j
Z 'IT L d cý %p Ol CO M ci 1. CO Lr) ('4 M M tf) v Q: " " " " " " " " " " " " " " " "
r 7 Ql l0 d' N N c1 N O" O' t1 O c-i c; c-i N N ia , c-i c4 s-I N-i c-I c'I t-I c-A c--i c-i
" 1 TA VA M N Ol %-4 l0 N 01 a) O CO O N c-i ^ Q Q\ O N c-ý N LU M N co Q ( c-i Ul 0 a' Q " " " " " " " " " " "- " " " " " "
N En N (V M d' d' c'M to M M M M ICi Cl) M M C
(V d'
U u Lo Ln Ln un LO c-+ cn O ci+ op 0 M c' to co co
Z Z 0 al m CM L(1 LO N CO N N U) M CM to N Ol
Ol tp Ln N fV N M pp co 4 O O O c"4 N v-I c+ C+ c-I c-I c+ c'i* V-i
" Vi Co T-i 01 t-i U) z1' -q 1 N N Co Co Co V CU A (y) c+ co Ln w N CO Lr) 5 un VI 0 d, w I'D O
d' Vi M N M V' U) M to M M M M 4
Cfl M M to N
U 11 N M ON (31, VI O N (fl s-i N co N M N ON C-
: 7. CV l0 '' c-I CO ci c-4 N U) (V O 00 (\i tD lD co v Q " " " " " " " " " " " " " " " "
( ý ] Ol N N N N N M ßl M O Q' O VI N VI CL 7y
c-I l-1 VI t-1 VI C-I C'I VI c-i VI
"' cH l0 c-I t- ri LL0 0 5 ti' '. D c-4 N T-4 N u'1 A N Oo vi VI r+ CO 'd` 0 '' N T-1 N O tD tD c`1
N U) M 4 1f) tM In CI' M M M d' M M M U) cI
il l0 m M to lO V M qi' M co l0 V' L(1 lD t! 1
Z z CO ' i' u1 d' N .I OD d' q N to m t0 0) (V co v Q " " " " " " " " " " " " " " " "
( = ] Ol Ol ýL 7ti
c'i VI c'ý ci c-I 11 %-I VIA
" O) LU C- r+ 0 Cl Co N LU lV N ', ' 'v M (IL) n Cý 0 O G' ci q, (, -) C') N N p' '4' rU 0 cc) CO C.
I! ) C/A M N ch U) Ln t LO M M N en d` en M M G"
ýr Q Q
CV V-1 OD ' C; ) O C\j co CC) 0 .; Z d' d' O' d' O) O) 1. O CO 1. o r i d N O M Q1 -c: r
O N to cn cI N '}' rn N t7' O m 0) ci cV r+ t cý ci c c'ý VI r-I vi ci ri c'i
LoY.
C9 x H Ü ýt ft1 U W W ýi z o d (7 C7 cr
G4
90
CD W a H
M U)
z H
i,. rYý Li W
f-i
S
c:.
cn
C c:.
U
O C7 C 0
C r1 C, In M Jl 0) V L') -, z' Ln r, l0 c"i ON Ili CC
O t1) <`'1 N G` M C'? (V (Il 1'1 CV M CV '' N N M ., W
ii
<' [ý
M (V
" C)
0 M
" to
N 0
" ýt
N N
" U
G (V
" v"1
Cl) if)
" (V
M 00
" M
N lf)
" Cl
M M
N"
N tf)
" Ol
N c-1
" ca
CO l0
" Cl
C') lD
1 Cý
CO N
" c)
CO to
e N
[N Q)
C)
^ "
C: i C- V'
1l a) C)
G) LN
(o l0
c-a CJ
0*1 CO
0 C'
co tf)
; i' VI
ko h
a) G+
m t))
q VI
to (N
c' G
O 1] Cl) VI (Y' rl G' N tr) N C M N C CV (4 M Lr) N
F., co M
CO Lý
to M
'i Q)
CV CO
N N
I-i' co
CO tD
' O
VI In
I rl
C" tV
M M
O lp
Cl) CO
G) CN
v !ý " " " " " " " " " " " " " * " " j l+ Co '. 0 q: ci b rn rn O O co co
. S.. . -i VI VI c+ c4 c-+
" O M CO cl 0) i- a) c-+ ý' M O M 0 rn rn G' ^ v-I
Ca a
tD "
Gý "
C) "
co "
N "
ON "
Ln "
T-4 "
N "
In "
Cl "
M "
ý" " "
a.
Q] U) CO IV, M CO a N to r1 N M N ý' N M M `. M
(V O 1
C"1 iD to C+ O N CO N [- t- (71 l0 c-{ Ol d' O'. CO TA [N Lý O ri` 0 al 01 (V N CO r'1 0) VI N
O
. ý'ý.
N N lf) v1
Cv VI
e' N lI
co CI
tN CO 0 ri
0 v-I
ON co co N C'{
c-+ C"I
" d' i'D d' -, v c-+ to "' % O O M cn d' m ^ C. co co c-# N Ul N UO tf) lf) vi C') i'D N d' LC) co
N V) CO 4 4
4 mot' CV) 10 r N ý' (4 c '' M M M ci' (V
U tp 0 w N C) N N 0 L- O V vi cm w l0
7.. Z N c-1 CC) N U) 0 11) to tV N U') tI d' ON N co ý. / '. " " " a " " a `" " " " a " " " "
u. b O L) CV) D' N G co Gý N O Cl O '. OD CV V-!
^ U
" (3
"
0 Cl
"
U) l0
"
N U)
"
r{ C)
"
M M
a
0 N
"
G" Qý
"
c-1 11)
a,
N pp
"
c-I M
"
V M
"
Lt') O
"
O1 qv
"
v-1 U)
"
M C'1
"
Ul CC)
" Cil cn cI VN CO tt) M 11) r'1 N d' M tl) M M M ul
Z z c+ (V
O co
in CO
d' N
co CO co
Co co
In N
In CO
rn ()
r+ Cl
Lo d'
to M
O 0)
CO l0
co r+
co 0; ICA r{ c'4 ci C4 -+ c
C(' ! 01 ri
61 Ql co c1 v
('1 tl'
a
V)
C
M
t
VI
c 4
. ý, (Z) ?
M
t
L)
'i?
M
cX
N
Ch
M
C
M r1
c\i
M
10
ty
r; Q
t7
't `Z
X
VI N
CO
in M
CO
m [-
LO
N lQ
c", r+
CO to
C\J r1
O LO
ci
N O
to
M ON
Ol
IT CD
CO
CV O
C\J
lW CO
O
p r' i
m
[- v4'
M
(') Ul
CD
O lp
(' I
O M
T-4
O w U w w c! x H 1 z o ä cl ä c`'1
91
N GD
w a r7
Cl +
4 f=. a 0
0 E+ E u W L U w H . s. ý. 0 r4
cl i'V 0
M
cY t
o'
" V Lý ý" c-4 M ON ui to CN N M N M a92 co Cl ü ti C (N O CD Lr) 0 d' u) M to C Cs, M to u
! CR U; 'C' M N l0 U) L" Cý if) CJ' LI) (N C- 4
d' Ul a)
W N
Z N N O 61 q (r)
v) tO N 0) Ul 0 N d' M C')
l0 to lU O O
Cr) Cr) CO C 4
tN c-i
0 VA
O (-1
:i N
4 N
ti N
ýU N
CJ) r4
11 ci
C N
Cr) N
G) c-4
to r4 N
N N
G CV
" N v) 00 M T-4 Cr) r+ ýý ßt1 G) N V-1 N Ol C) a U N C) CD N to a: ) 0 LU In M En Ln r '' c,
O 11)
i to M to lid 0.1 In Cl Ln vý V V N "4' to to a)
N
v Q W :4
ON "
r
Clý CO
" Cr) .0
Ch tai
" co CV
O Cr)
" ci' N
q If)
" tO N
m Oý
" CO (V
LU "
If) N
N C)
" r c
O') 'v-'
" co c-4
to c-I
" T -i N
c-, c-I
4 c-{ N
N co
" r1 CV
N tO
"
v-I q
to 0
O" N
C) ý
" Cr) N
0 ý
" U') N
" i) r+ N Q O
O to
() to
CO O
(T) Cn
M O
pp r+
N tO
N to
M t0
q tO
M ºn
M ß'i
T{ " ül ("1 tD N CO I) O V Ll lf) 11' S 4
to l() C;
M (V
2 F
W
c-, Cl
l0 %-i
O c+
C c-+
d' ci
v-i M
Ln O
uo N
Ln tD
tO N
q
'i' N
co M
iý N
tO tý
U) c-4
O N
S vi
CO Q
N N
r t-
O CV
to ul
G) cI
O tD
CO "-I
d' N
O N
O Ol
N
r
(N CV
CV V
Lr) M
O r
M O
Cr) 'Z4"
N c-4
tO co
CO v,
0 O
t0 O
c'i N to
to tD
vI c+
(0 Ci 0
O " to M tO co O t0 "D Ul t0 to N L tO tD Ol U 0
(V Tý
u 11
Z v
k
to
Cr) O
q v-1
CO C)
Ln s-1
co to
c-+ CO
N LX)
L N
LO V'
r'! N
U-) N 4 N
V) co
N N
Co N
N 'H
r-1 N
N T--'
0 Co
c' N
N Ln
0 N
LI) r-
q cl
N t-
rn c-I
0 to
0 N
111 m
v' CV
c'i
M N
" c-+ N c-(
N tU
r O
to C
c-s N
O LI
r r
d' M
vý M
0 N
r M
O ';: M
C) O
ý' ''
Ln to
O CV
" If)
to cn N N C1) u7 O t. ) 4
tf1 to CO In W tO O
11 Z Ln O kD I-
tO c-1
ui ev
c-I M
O N
U) N
M ev d' N
r d'
In CV
r en N N
CO eN q -4
co to N c-1
d' 0 U) (\I
0 u1 ci N
q co q c"l
'' o C) 'r-4
r co O N
(; ) co tl) CV
C) 0 M N
(v rn cn Cr' C) C) ,:, Cý C' C) C) a7 C'! " N (' ý N U; Ul Lrl ("1 V' C) N l; J C\
"" " tD M N q C) N 1 t0 LI) U7 S U) lD r a) M En V'
It
z z is
,£
tG
19
C
Cn r-i
l0
N M
c-'
tO N
d' 4
CV
4Y
c' N
M
O M
CC)
C" V-4
U)
tO ca
Lf)
tO N
d'
ri CV
t0
t- ci
; ý'
q T-1
Hfl
O CV
;r
to N
Ql
N N
Ei Q CD u w 9z4 O x ri v ýN ov w
92
ca Co
H
U
a
8 F4
DL4
ID
U)
H cý
Q z
w
U C U W [l. O x H .: Z O c+
O N C7
rt C
ý* C7
" r tL O c C1 00 U1 C (V co c1 co u N C) r-- co to r+ F. M N c O N N O Q) ( :.
Ol t1: N N .4 C' G 5V) L-1 5V) (*) M [h 1; [V) (V) N '
i.: :I t0 Qý tD ON V-1 'J Q` tV V' Cr) CD 00 c ý* O N
k O ýn v O 0) C'`) r-I c co co N Z c" c-I " C+ c-+ c-4 Irl s c -I T-l N -i --s c-s
" %-4 (ý cc c-+ c-4 co co (0 m [N M in C) l0 0 OD U1 O N ci CO V' O M M N Ol t-{ O C) O N
N (f) N N M 'Cl' L" M In M M M N d' c*) Mi M C
A It
N h -Ii' N to m 1- lfl 0 0 Q M N M ý} c'I 2 Z f7 m CO a% e Ol CO C\i Ol N -tci lU [2' -t C) 00
f O ýD m O N cI c-+ c+ ap ra c-I N al r+ c-+ r) It-4 c'4 %-4 T-4 c'-I c'i N-4 N-4 V-4 c-+ V-i C'1
Q Q to 0 N
N C'
1-4 en
O N ct' co
M N N
O cn
O Cn
M '
ýr O
CD Go _ CJ U CO en to
N V4
" U)
" M
. CV
" M
. d'
. V'
. M
" U)
. M
" M
. M
" cn
" G2'
" en
" en
" M
" C
co N U 11
N `., S M m O c"I N tf 0 N co N CM N 7 z cc co N Co C N M CO M Ln N tf) N I N c4 v . cC
GJ "
Ol "
l0 "
t}' "
O "
N "
N "
N "
O "
Cp .
c-I "
c-4 "
c't "
C) "
cI "
N "
e'") .i Y4 cI [-+ -q c-I c-1 c-I 0 r-+ c-A c-I
" in C) '. O N 1S In cI O In If) In O O ci N co A N O 0) C d' I'D CO Gl N '. D N N In t0 q. C'.
N " 14
" M
" N
" M
" V'
" LI)
" C`1
" LA
" 1.4
" M
" M
" eel
" d'
" M
" M
" M
" ý'i'
U p en N u'1 rc O C <'') O O c-+ 0 d c-+ d' Qc co z 2 N cr 0 N m c-4 ' rn N N m 0 tf) r) 0)
u C Ln 0 r ß 1 tz 0 O 0) c c--I c - c-I O -i c c-I c4
) w Irl Ln ý 00 M in co r- r- ý
O Ch -i
CY) (+l ;: r r-4 f qT (N n A t- l O l0 i cu D r CO t ) T-" to
" C!
" M
" N
" en
" Ci
s to
" e1
" U)
" " M
" C4
" M
" V' M"
" M
" CM
" to
N
u In a) O en c+ t. ' w co c-+ m .' 1n oý 4 cý 'o
z z-. 'i' O c"1 r+ 0 tO cý N S C O N N u1 co In
C cc to c-t N N M N N M t-I O QJ 4
N (N Z r" r4 c-+ c+ c-I c-+ t-+ r c-4 rq V-+ l-+ r-+
" G
N N
rJ C
"n V:,
r-- (",
c CV N G i
L- 11)
Azz. M en c cn
tl tv
. ', G, fN i
N C! M N en N In M 1ý C en M M 4
M M 14 Q' 01
11 N Il) to M co
-+ m Ol
V O
'r V 00 N OD O co N
Z Z U) l0 N LA c co N Ol O) M s 01 Q\ v Q
(yam]
w
" Q s-I
" N
" If) cI
"
cý T-4
" N TI
" N 0
" V c4
" %-4 c-I
" N
" m t-I
" 0 c'I
" 0N
" CO
" T-4 c-C
" N t"t
+ N c-5
cý x H Q m U w w a Z z o c-i C m or
93
m co
1-ý
F
F4
to W
Gc4 [+.
P4
ci
U)
, Oil
rn
Q: 0 w
` ca L w v a 0 cr
" CO C! -, N Ll Cl M N Cý CD in c-n N i- c q c'ý U N C) c-I to to Cr) (+) N tD Co
M U) C`) N ýi M C N to M N C`1 (N c N C') (q-) J) n
11 N ifl Cl Lý O ýO ý; ' C O N l C^, [N C'
Z Z C- Co (0 In V' c-I lf) O\ C -e Cr) u) CN N t- C 2
GR L1 m cl M CV C Cl [ Oý N z ci vý c-; c-+ c-i
" vi to CO If) N M U-) O to N Q1 In vu 0 d' c q if) C cc QD d Ol ('n O U') N C0 W l-0 O M 03
CI N
. L7
" M
" ci
" M
. M
. 'V
. N
. If1
" m
. N
" Cr)
. N
. Q'
. CV Cr)
. C)
. t'
C' A
U') N Cr) in N cI v-I M 0 l0 V' N N Ol CO Oý ice. Z N CO CO to to V' Ln CO CO e Cr) Co c-I l0 Co O v W
Ol l0 M O m N v-1 m " "
O\ "
0 " )
CC) "
CO "
O "
Z v-+ cý c+ rt c+ ý
" c O r tn O N 5 de H u In Io c-, 0 M .. Q ýO rn O vi (D rn co Cr) N mot' N OD co c-I e CI) Irf ý' Cl) c+l c-1 V' c V' N L'1 C'') N c'') (N CN N fM m d' N ` (V
ü u m Ln c+ N O CO to cm C- O rn G% Cj O O Gý Co N t-; Co U) in lf) Ol m 0 cý Ln w vi
Q W
" Ci
" N
" C'
" c-I
" d'
" N
" N
" 01
" ap
" 01
" O
" " 00
" GO
" v-I
" ('`)
c-I TI VA t`{ vi Ca C'4
" 5 C Cr) Co U') N (V N Cl t! ) L) 0) c-I lJ 0 CO q N q O-, N N --I Cj) l0 lO O) �- LN O m N O%
(V ýw
" U)
" M
" rd
" M
" c'
" to
" C')
" In
" M
" N
" M
" M
" V'
" M
" M
" M
" 'd'
U a ý, O M rn O ýr c-+ Cr) U-1 N Ch s -I Co O v Z Z l0 O) (V C' cl' lO CO M d' Co CO N O tf1 Ol N
(: 7 rn N uý r+ v" N rý O Co 0 0 O O, Co c-I N , e, cI cý cý c-I cý c-I IV-1 cý ci tý
" l0 N (y) N N O CO Co (V Ol O O N d' OJ
I
4 C0 cn N m O M rn C- O\ N c-+ c u') 'n N ýY
M ýi Q I() M to M N CN M V M M c7 º1 N
p O cý. O O% N e Ch w c-I CO CD Ol tý N
,Z Z Co N In [N Co to N IS in Zr C) CO M in c-I m
iä 0: cO L1) vi Cr) N M O CD c-I O O Ol Q) N N c-4 c'I vi c"I 5-4 c-1 c-; c-i
a C- 4 G M C^ M C. N C^ 1ý Cl N N (: ) N C-'
N [n Co c- ýN d' ''M In M (N cß'1 N d' ý'ý) M M '' C'
Co 9 ý rn ;Z Z N N In O CO c-I tO r N Co f'ý) N N
`ý t C) m In c-I cv m d' r+ C) c- O c-+ Gi " N c-4 -c- c-i c-+ c-I c-I t"'ý v v-I v- c 'I
aO C U Z M ü G C1 ü 7 W ,a Z 0 vi
a ci m d'
Q a cf
94
0 o, w
(h cn
U
H
t'?
H
EL)
iL
1
H
rý a
0)
X:
a U
4 WC w w ý7 x !i a Z 6 Q c' N U) 1: 111 c a a a LEI
" in it) CV c-i it) N N C' N 7-1 M 0 c-+ vi Q CJ 1"N NN (-I c-. U7 N Ln C ;J M 'f` c-{ M [J 0 a' C: J
u) ýt t*; 9 CO N C' Ol t r) Q' u) cf CL) [t ul V U)
11 in in M Irv L) N Ol in N Cpl a) O in Ol CV 'Zr
ý. fn M (V 6' in a' r in N l7 Lll Q1 C? % U) Gl
O cý ap Oý rI l7 V N w O M N '1 " l0 z-' . ý- N H N H CV CV N N ci N N N H N N N
" M Q co O s-+ C' 0 in M N O 0) m N N "ý A in N LU (Ti (^ U) N' N C (n C C') c-4 ß. I) it) N (Iý trl M W Co CO to Q1 Lý d d N Q in in CJ
A 1 1
q O c-l d !M iý Q c" CM Ol Cl to -t' N c-I z Z. co 0 U. ) tD c-+ H M 0 (. o N in 't "J V-4 N O
z (yi Ol c4 S H tD N' M c-I [
d 4 4 N O N U)
TA c-{ CV N N N N CV %-4 CV C\I N c-H N N CV
" Ul [N c-4 LI) N M C' kD U) t- N N .0 ci ko k-p A N s-1 Ol L7 ('') d' M H O N l0 N N Ln d' OC
v) to M tO r U) LO O in un in d' c; ) mot' Ln in C0
(V U II
O\ ci CV N' Cf) N O W C% M M O b V- 0 co r-+ O "r 0 tD In O M rn d' Ln tO N' co CC) M
Iý C1 4 (ý N '. D d' u'I U CD r+ c-{ N N O\ C') '. G
. 'ý.. ' H H N CV cV N N N c-I N CV N c'i r4 N N
" in (ý r- U1 co ttl Ql M to N C- N in to m c-a ^ A C- cI Ný Ol Cl Q' 01 OD Ol H M c1 d O\ M 0 IN " " " " " " " " " " " " " " " " " M U) l0 M lD t- Ol in 0 l0 kp in Cp LP) in lD Q'.
'i c O tl
ci C') O N N H CO H L7 to 0 U) " H CV ý. (Ti to d CV lD i ce M C) CV lD N l0 0 Cl C'`i N
v ., z " " " " " " " " " " " " " " " "
W U) to 0 N N N in H lD C') c-I c-{ a) O d' in e H s- r N N N N N c--I N N CV c1 CV N N
,
to cm (IN 0., C') Lr) -x LD m k-0 r-' t- Ln LO ^ Q M N a, N c') N H M '1' 0 N
in d' l0
0 9
U) in M N lz C-, if) O cN Lf) in Ln 0) u1 tD t0 0 N
A 11
cy1 OJ N Q w c'-4 CI N 0 w t- l0 ci c-i in in I- ('') co CJ c') N O in M O1 Ol (7'. N c-4 dl a)
O ký Ol lD O M to Ui i N "C ) V' 4 O N O d' v' ý., c-I c-I M N N N CV N tl CV N CV H N N C''
" in N S CV N' V-4 N: ' U) '" U) C"" 0 0 M c-' U) (V cl CO [- M M cV Q C (- in c-I M N N
V) ýLD M LO U) 01 l0 N in to ý7' CD in CD to CO 4
II C7 IN CO u) O co S N' lD c-I H H O EN
Z (7\ C1 tS to cc) Oti U) t0 LI) Cp N CO c-i Cv v> " " " " " " " " " " " " " " "
O'. in O N d' u') C) M to to c-i s-I tN Q Ln Cn H H M N N N (V (V c-i N (V CV 5-I (V N N
LL' o
4 CO U W C 0 X H 4 X z O ý (N a
N' 4
95
APPENDIX8
DESCRIPTIVESTATISTICS: ----------- ----------
PRIMAR YSOURCETRAITSBYAGE ------- ------ ------ -- ---
96
ä Ri a
T-4 I rn a
N p4
w
w
m a w w 4
W
a o Q Co U w cL. tý x H a z 0 ä ä ä Ü
orn c-' u, Ln tD kD O M pp co ON cr to to ' 0 A CO M (n co 0 CO M co :r co O w Cl) V
0% aA N N cal A 4
N N N M M N d' N c) N C'
W 10 U' lD u 4 N N c-r Oý c ci Q d' t0 cal t0 Oý Oý
' rý h
Z z l0 (V OD (NJ 0) O (IN t'4 l0 0) tD d M ' v Q
Gcý O In d' W Ql fml c-I N Co c-I N 1-4 a, cat Cri (V Z r! c-i c1 r-I ýi cI c-1 cI ci t-I c -i
" in cn ý CU c- Co tV in ß1 N 0 '
a O % 4 t-
N Q -ý A 'D O U. ) Zo h cV co tN co m 0 % V -
r U) N N M M M M 1 N V tS N M cal N
uni mot' O m co d' N N r4 q M ;; r ON 19 Z Z co 0O ON o0 0 U) co to m c ICT
. cý ö I m CD I a, 1; 0; 4 C4 C4 tz ri
rl rq T4 ri T-1 TA Tq IH r4 r4 r4
" d' tl) cI N 0 h h 00 l0 In c 0 N qi' cc N 0 q a, 'ri 0 c-l N N N 01 to M ON to "T h N
. d' it))
" Cl)
" N
" CZ
" " " " N
" In
" CV
" M
" CM
" C")
" (4
" N
" N)
" N ' V
N Ln 'r-4 w LA N N M Q CO ON C' h N ON N to
Z Z Oº ON W) w V (0 M N V. t! 1 a t0 ul 01 r-
0; C4 a h am (4 4 h rl cl t -, c - 4 c -4 l-4 e-1 rl
" a) in d' '. o 0 d' o1 c0 N
rn h M N +
0 h
r (0
rn 0'
g ý" A in ON O N o0 cI u1 I M 0% c
IT Omi "
U) " N "
cf "
d' " d'
" ý! ' M I! 1 cV M (v) C') d' N M N V;
M H tn O cr+ c-+
(7% m
ei c-+
V N
OD in
N to
ý vi
d' c0
to to
h d'
l0 t0
O'. rn
U) Z
" " h
" " " . N c-1
. " co
. s
" N cI" N" h" i" c c-"
" d'
%-I c-I ri e-1 c-I t-I Ti c-I t-4
0 O A CO 0 00 0 "o cl h m a, co M
IV ON "
Cl) "
N N" " M
" d' d' " M to N co M M co co M N
yMN
, 4 cm W CM N N
O (0 10 a, c0 ' °M `4
N m M co c
z z O Q t-I h d' O\ M c4 cV M h t-I c-I N 00 c4
l d'
I r4 ci cl cl c. [-4 c4 c-1 ca c-I c T
" h cv) h l0 M in ri A ýA U) C O N ON N 10 in f cm N %-4 t o N
to V
0 ER N
" e1
" fh rn d' (0 in N M e4 M d' 1"1 co M
U) N e'I
i9 p QN Q c N M %D N CV
( pý 9
cm Cm Ln t- 0 '
(; % N CD T. O p" in m , 0 V e M N 1n 10 vJ Ln
t4 J ý O ; e-4 N eý1 W
4 4 4 C1 O O
T-4 c. + It cI e-+ ei 1-+ r4 l T-1 O O
h M
tO d'
N M
O Q\
(ý t0
N c
N N
CO N
c4 0
t0 in
h N
M M A to O o
Cý Vl N N
m C4 4 A
W; N co co M co A N1 M d'
N ri In 0% ¢ý ` CA 0 0 0\
cn r4 ýp ýp co i c"I z N ` a% M Ln lý v c
< co U w W 0 = H 64 z Z O V-4 C-4 c d !i 94
Or cy
97
U)
N O`.
Aýy
G`1 is7
<1 0
w c 4 r-4
N
w
ao ö
G.
I ö Ü < m U W P. 0 x H a z 0 ä ä c (
e-1 to to LSl r-I lD 0 to N. N 0 N -4 ýp 0 N N it c) cn N M m . -a to 0 to O co co
N to M N M M e' N V; N N C') N to N M N C'
W t 0 ul II < '. O M ri l0 (0 co lo O O CC) O r-i [r) M m N WT
Z Z .ý N o, CO H (N 0 w N to 0 C) d -I r-H G) "ý " " f " " " 6
O '. p M U) ri M (Z N
CO O: N 11 h Ol M . -i H H 1-4 r-H H H H r-1 H-i rl
" Q oR rn O C' N N 00 co Q Cl) O O ICP A . O H d' co C% v co W Un 0 to 'IV CO En (n N CY M c') N U1 N N M (V C' N N M
4
U) to W ýD Oý C7 J
co r-4 cn N. C\j N 00 t- 0 ' LI) Z %T N N m N Q N N co O) H v 7 r
.. (c] O U) 1+1 Q) H M N N N Q H N N N M H r-i rl r- i H r-i r-1 H r-1 H r -i
" -. N tD U) n'1 0 0) l0 N. cc) co N LN Q N Q) 9 O Ln O O W O) to N. ul U) rn N M U)
0) Uý N ri ý7' M d' M U) N N c') N d' N M M v
W al -4 Qip 0) O O O r-1 O G) (n cV O u1 U) , -4 N N '. p
d Z H H V' N W 00 W ) N H r-H O N i-i r-I r-H ... (ýi O Iz PA1 0) (N ri ri C: O ri M CO N V'
,2 r-i r-I r-i r-4 ri H r-4 r1 r-4 H-i H
Lo ul O d a0 d O O a0 CT u, d' rI - t. 0 M N OD A 1 9 O N (n "o "1 14) ýo N dý N M 410 w p U) (") r-f V" M c3 N Lo N N M N V4' N M M .4
'a) Wd QLne
N. 0 H co H M d C'*) Ol N in LD 00 W Ch '7' Z U v N N N N cj' co N H N ül Ol N l0 b'
z M M " " " " " " " O N 4 Qý N 9 N O O N °0 H
rA H H H '-1 H -{ -t H
0 t- C- tD co tn OD 0 0) U) m 0) H N Lr) A H a 'i' 00 d' co O d in to l 41 r-4 M
U) V) c+1 ri M M Vi N Ul M N M N C N M M d Ea %1: 1 14
Q LO II N v ., 'cr Ln %D rn d' H ON 0 in '-I co '. O M N ; Z. 01 d' M Lr) in r-. .
l) l0 O 00 U. ) tn N M 01 lp Z v O N d' 0) d' N ri H m 0) O N "
c0 "
00 " O "
e, . -I r- A .ý H H . -i # H H
" pý 0 N W 0 N ON N N ON t- In N r-1 ! f1 C"1 N N O LI) M CO 01 d' l0 U) l0 H H tD C+) d'
U) M H{ 4 to U') N N M N M N M c 0 v
N N C) C-1 M y S O) N ( to d' 0 U1 l0 [N Ol tl' tN H N
z 4 z " " " " " " " " W O t c4 0) ut H H O rn rn H N al? O .4 ri 1-4 H H r-i 1-1
H
0) t! ) m l0 00 () C') 0) to Op U1 co Q1 N U) 0) A N 0) Qý Ili M v S CO N CO 0) N ON N lD G)
E Vj M I M (`l d' m to N N M N d' N M M M H
W
N N p r-1 Z N Ln to O N: CA co IN CY) In U) M M N
" " " " " " " " " " " " " " " "
O N M O H O O co O rn M O) co C ßr1 rl H r4 4 ri H H H H
a 1-4 Q ra v W w 0 x H a z a c c f Q i4
98
ON
H
4 Z O N
a
U, zA Waý
H! i W t7 4
W
01
w, al'. 4
Ö G4
a p F < CA V W C 0 x H .7 Z Z. O
O O Lti
iy% N. Q " O N N tom) N Cr)
Q ýn O tý rn c*ý <" 7
N " Ln rt tD tO N ;7 O ý' V ýr W to ýn C Co cn +N W to
U) C'ý lD h
31 tD ca N N N C r-i Co ' N 4 CO tD 1. ,
M
O O N Co N N r-i UV H 0 OD '0 O CA e uo [. 7 .ý N ao N O N cn .n tO cn c . -i t c
(N H N rA N N N N ri N N N r-1 N N CV
de Ol Cr) N lD CO --1 r-I N r-1 U) O ri N Cr) CO " Co Cr) O C' V' N u1 lD c+1 d' N O CO 00 N Ol
C In " c' M N to lD Ul Ol d ýS to ci CO Cr) c' tf) N
Ww c*I vi
in tl
ß ý -i N l0 c' In cm N 0' 4 N r
Cl N
Cl O
t- d' 0 in Co N O t0 rn r-1 in 01 ch O c ) t
ri CV N C r-1 N N to U) e-4 V in in Ol L N N rI N _A N N N CV rl (V CV tV r-i e N (N
en to d' lo Cr) LO . -) Ol N p1 l0 rn Ln Cr) to O
.. . O r d' to O Co d' N r-t to t- Cr) cn tQ cD b rn
a (rj c1 N (Ö N d' Ö
cl in Lri d' N d' In d
c0
U1 N adu N 8 r-1 c1' N O' l1) c% to N to t0 %O d'
z Z alb ý 0% m M CO (3 Cr) Q ri -. O N O to N O'
O V' to 0) Cl l0 M to in N Cl l0 Ol in N (V r-1 N ri N N (%j N rl N N N e-1 r-4 N N
C"') N CO (0 O Cl Cl w 0) CO c+1 lS r4 N a% u1 " 01 N CO lD d' CO N cl' (l1 Cl d' lO N 0 lC to
" d' Cr) tp l0 o d' O d' d' t0 V' N d' tD u'1 Cl) ie H
<eN N e-i in O'b CO Co
+ 'd' O
' IN 01 '-i Co in d' to cn1
Z Z C- in «e N N c 1 r4 u 1 r-1 N U1 r-1 Cr) N N V) "
r-f ".
CO " "
00 to ü1 Va' In tf1 M N in U) (3) M 0p N N r1 N H N N N N ri N N N . -1 ri N N
Ln d' r-1 N W N Cr) O\ 0% -1 Cr) l0 1 U) '4' l0
.. " rn ri H Cl . -) H cn ýo 0 0 N to O' w N N 14t
; q
" " "
eý "
l0 "
ýD "
CO "
it) " O " d' " d "
ýD "
V "
N "
d' s
t! 1 "
ýn " N
Vý t nN W
V
4NN tD d' C) 1 tý to N
cN in t! ) to Co r Oý C1 Co v cu d' [D QO N d' fi lA U) e-1 e-!
i l0 Ol N 'cl
FF.. N ri N ri N N N N ri N N N r-1 r-1 N N
(A Co ic2 Aj O
N N ' 'mo ' . -. " c% Co N N in l0 Cr) N O C) M V d
-e S q"
d" ' "
N "
1. O "
If) "
CO N" "
O "
V' "
d' "
in "
V' N" "
d' "
ul "
u7 (" tIý Nr ta t
t0
11 0G QN M ßN ci q Vn CO Gn ý, C N ri e Co C% CO
z N l0 C Oý rn V O Co Co 'a' vi Co N (?
r-4 CO C') rn N cM ýO C; ri tn r! cý O O N 1-4 (4 (V (V CV N rl fV N (V rl r-4 N c(1
M r-4 In in w to u) U1 t0 r-i 0 n "
N tff U) C N (1 l0 Cr) d' Co O p
o% Kr q"
in cn tO t0 N ut O d' u1 t0 V' N ' In ýn N r4 y H
ON 9-1 4 Ln t- to (V CO l0 O' 0 d' en d' M r4 N O
Z 0 tD en u1 00 fit' CO N lO N r4 O-, Cr ( v r4 M u1
N , -ý N N (V N N r-4 N N N
r"4
9 ¢ on u w w cD x H a Z Z O CM ä Ict
99
H
ON cýa o
ww
w a
O Ei q U W [-, C7 X H F7 ý. Z O ( H C 0 G
" ('4 GO W M O LA CO O\ CV O M fV N. O Q N rl to e-i [i N in 0) Q' d 0) O C` M O) Q':
v) M N M 'J' d' M LA fV N M N cT N M (V ý* +H n L U tD
Lo iD C C' M VT M M . -1 to 0 V' 0 O'. Cl) fV U') CO VI CO "1' V' N M M ý3' C' l0 to O)
o o c O rn (0 rn OD m co O H ý r-4 ýj
CC) 03 q r-4 C\j -4 Ln t- C) -4 r- ON Ln O -P ^ Q N H H N l0 r- 4 V' tD 0) N O M (.; 1
0
N u M C 'd' ,4 ül CO M m N (ý CO N in
i 7 s J
Q Ln co Iq 0) r-I %I, co 04 N 2 Z '. O 0'% C% CO O 'S' CO
Ö M Co r4 H Ol N
W m ýD H C rn H H rn cD c-4 H rr H H
" M CV lD N CO M CO N H r-1 d' 0 r-1 l0 i0 C' ^ ß 0 N CO Q' CV H to (V C7 D Ql Zn CO N O N N " " " " " " " " " " " " " " " " "
'4N co [fl M N M M d M to M M M N d N M fýl C
Q ýD r- t- t- r-4 Ln r- r-4 tn W LO 0 -4 0 (\J 00 2 Z d' O a, tD co to co O O O ri V' %o . i' V'
Q'. N V' rH O CO N C1 pp N ri O Cl H M N ri r4 H H T-4 ri r-t r-i H r-i r-I
" 0 to M to . -i -4 tD ON r-I l0 CO N (V 1! ) l0 fa C1 H 0 N M c7' M d' ri S1' N LA L- U) ß1 H
"3' O'. VI N N V" d' :i CO In c0 M M M '4' N M N lt)
12 Q l"1 !N H N 01 to m CO N l0 hfl m N S In (ý 5
z z r-1 C' M H l0 O tD fM H 00 N º-I N q [1' N 4
W N L[1 (V H M ( 0t 0) N O O 01 4
4`1 H mot". r-I H-1 H H H r-i H r-l H r4 H
M ºn to CO v+ CO 0 1-0 d' '. e ON rn t4' CO a0 0 N 0 r H d? (fl in to u') N CO O) d' M r-f CO M N M ti' d' (0 N CO CO CO CV CO CO CO Cpl d'
NMN E
it Q N to iD In O M N 't' H '4' '4' l0 M H O
z e-1 to d' CV H 0 Ol 0 N W Ol M CD [- O N cý (il 0' N Co Vol N N Cl 0; (V (J') O O H N (, 4 . i'. rý eI r1 rI H H H rI r1 r-1 H
" a) a) M CO lC N 0 V H d' 40 t0 CO in 0 O ^ A m 0% (1 H tN r-I c0 co i. O U) OD d' H to in r1
U) (V H CO '4' in c1) (*1 M fV Cf c^1 (0 CO in NNH
L" Ch C:; l. 0 O? 0 *, ̂ V OD l. 0 C 0 cv Gt H Ci
M N d` M Ui O C3' O1 Q. N O') O 1-4 tV . F. H r-H H r-I H r-i H H rI r-1 r1
" N O\ LD LA 4 0 1.0 01 CO 01 N "ZI. (V fV (M H ß N ap pp CO In LO W [t' N r-1 01 In N N H Ol
n ()N C
" cn
" cl l
" A ý-
" CO
" " M M M (h N [I' (0
M C0 'CD
r-i 0\ W1 0 in N a
?. z (31 O d' N W N H O0 U1 '. D '. O co
6% U) H '. 0 0 01 N in CO tD 01 Q' 0) 01 H O) O r--1 dl H N
H H ""i '"H r"1 H r-4 H H
4 fly U W 0 X H 4 Z O C7 C7 C7 d
100
rn
w
H
t: H
0 V)
U, ca
w 0
CIO
r4
uä
s- w
0 ce U w w 0 x H a z 0 r ö ,
" Cl Ch %D rl H P H N 01 e-1 d N co tD In _ A N N Ol CO O H M l0 SD O' lD l0 M r-1 l0 O In M N Iq Cl) m c(1 lD M N M N
4 N M C4
+
C5 '. D Q .Z N (-) Ql CO N (') ri C) N M N O M Lt-) N N
2. (IN tN a) 0) %D O' N kD SD r-i QA (V Ol in c') ,.. -4 N ýD uA (3 O N .' o m O . -+ rn tz rn C' p
.ý r-I 1--1 1-4 . -i H H r-i -a
" ri i-i O C'') M CV CD CO N H 0) a) C" SD O\ "T A to r1 In Ql rl N M N M tD lD !I w t0 H r-1 d' r-1 U) M N S M V' N In ('M (V M N M C M M L
Wl r-1 t
O U1 il z Z -f CO N m [N (V 0) (V to 0 U'1 M co iN p) M v " " " " " " " " " " " " " " " "
CO tD ºn H O M M Ul 00 O H rn N O) N p
" rn 0 In rn 0 co r 00 un Vn C) M . -i to Ln Ln .. q tN ON Orn OD --i CO N r-$ Ln in 0 In N . -4 M r- en SD in
" U;
" C')
" r4
" F)
" M
" to
" CV
" to
" M
. N
. M
" M
. SD
. N
" N1 c" '1
+H ( ý
p Q 'V' N .O l0 (\J t+1 d'
' O1 O O N M CC) l0 C-)
Z co C"1 e-i SD '. D C` ) H t0 01 cV Q N QJ 00 N CO N N U') N N M
S " co
" 0: )
" O
" r-1
" m
" t-
" tý
" N
" Z 4 , H . -i H H
" '. D SP ß1 '-i N 0 LA (A to tN (V II) d' 0 0 . -I "" A d' ao O co Sr M '0 O to co 0) M rn in H N N C/) M H 4 M V' N LP M CV M CV S; N M M N
4m C") (V to (') l0 d' 0 CO M CC) u) 9 i--1 C'1 N M
2 Z 0 Ln tD O t0 N rn M 0 (NJ Nt' C in a% N ' v " " " " W Ip N U) l'41 M N M pp p: O O
. F. rl r-i H H H . -1 r-1 rI H
" ri Ul 10 H N tD SD PD C") LA Ol M lD u -f M p) q q O O a'. N ro q N co (7, N Ln M qr M 'p t-
d' N "
U) "
M "
N "
W "
M "
d' "
N "
tf1 "
M "
N "
M "
N "
V "
N "
M "
M "
.' V2 CV) N
1 fl 4 N
v, N CO H 00 d' M r-I Orn N d' M O 0) (l ýp c7 N % d' t0 tD M ' r-1 N '. O O M N N N O
" co
" N
" to
" c')
" d'
" H
" C") W" Cho O" O"
" Ol
" O\ Q" ý
" r-1
" N
'-4 r-J H ri H ri H r-I r-i
" C") . -1 tD CA S to CO r-i t0 (V C') CV N r-1 H O\ ^ Q t0 01 M Q1 In H to c) t0 O 01 Cp M N pý 00 1/1 SD
" VI
" M
" r-1
" V'
" C')
" SD
" M
" In
" C")
" N
" d'
" N
" SD
" M
" m
" en
" '
N
(i Qt' d' %D H U) to 0' cý Q' c"7 Cl co C') %D O ul N
A z D CO N Ul SD H H O U) M r-I M al to N
v 0 pp " N" vi c'') " " t0 " H SD " " W " O'. " H a1 " " Q1 " Ol " W " O " cM r-I e-1 r-I H ri H H ri
" .i t0 to tf1 N .Q
[T (V N l0 d' Ln tf1 N M 01 Q IN 'D O O M O h M m S 01 01 O) O' N r-4
ON M M r.. 1 ; el! d' M U-) M N m (\i «r (V N M N F-4 00
4 r-4 2 v SD N SD H co N Q rq en to z z SD % O co U) ttO M
v 0 Cl)
N LA N C' O d' N 0) O m C O' CO O "-i H H H H H { --i
7 x T 4 cA U w r4 C H 4 Z O C7
101
to In WOm
WÖ w
"
El a 0 U O- (0 x H 0 C U T
w
N a O O 6 N Ct' CV 0 r-! cl' lD r j e N I- O C U l7
O ip (n tD r r LA O LI) !n 0 N un C 0)
vl + r-; En r1 in W
t ý N V- M 0 C- N in 10 D N C C
Z 2 ý- { H In M r--! CO (. n ri N I
, r (\I 0) : M N N H H N M (V r ri
N N ý--! N N r-1 f-1 N N N
0 M r N CO -I ý '_1 N N n " I-- t0 (1) GO M UI t! '' r l: 7 CO N lo M r-:
,...! " +n ch r r r cr O ýn 4 in t uý rn En l0 r-1 U) r"! E4 I C9 to If
LO ý; r C\j ON r- ID 0) CO Lo cm H 0 (\I co Ln C) "o Z z '. o u H m '4 cv N C ' c-, 0 O H to r 17 O . -i O r tD CO r N N Cl t. 0 CV CN ri H r-{ (0 lV CV (V CV H H N N H r-1 CV CV (\i
co CY) N ko (n Lr) .. 1 0 '-f Lo .. " rn u) 1.0 d' O C v co O n u O % N O O N in
0 a
" to
" cn
" t9
" t0
" co
" LO
" 0
" in
. '3.
" to
" qr
" CO
" en
" to in
" ß1
tf1 N
k 0 11) p CO (ý
' Z Z t! 1 to CV , -H l0 d lD C' N N ri q H M v sý
S+] N d O d' M lp r h 'p N N O r Cl CO ý'
r{ H r) N N N N r1 H CV N N H H N N
M O " '' C H O) y t d' h t0 u1 M en N ko
IV N
A " i)
" cn h"
" c0
" r "
In . O ý. . .
ºn " 4' r " d' . in " in " C) ý
cn c! " rs' I Qcü
O M O d' to N h 0 0 0 C\j N Z Z,. N O H M CV W in h N H CO l O ri ." "
r ýn "
. -i in "
Ln "
ui "
r" r" r cM H O (0 O w M ""! <n N N N lV H H (y N H A N N (14
O C O CO ":: V 00 O op r O O C)" 0'% cV d'
r d' N
A "
" to
" en
. tD
" tD
" CO
" [0
" O
" u)
" in
" In
" C1' r"
" to lt) t! )
" co
MN Xl
p Q ccm 00 N CO r h Ö r O in rn 0 CO -o u_ rn '. o
Z Z tO C') (V l0 co I) N ri 0 Cn Ql In CO V) r N v h in H lD r M tO N CO M o Oý C;
M N N fV N H H N H H H CV (V N
ýp N pp rn d'
O to
r O
M M
cý i
N -%
V IT
" 1"
l0
Ln 1.4 h r CO ý O W r in
O
to
to d;
M
oo
H
in
(. D
u)
u1
'j rn VINr-4 Co 133 Q)
dn
N r N ('4 CO [N CO r'1 0 M Q'ý 00 c0 (D E-, C, Z z in 4` N N ff , v r7 Cl cc) tp [Y: tD tD Q .. "
h " " O r H H 06 h ý" e "
CAD "
O "
ri "
ON "
rq "
to { (ý N r) cV N ri H N r-i N N r-1 N N
in tD d' 0) O N M tV to " l0 Ul O W C11 d' d' h in N in N CO Ol M N
0% 0% 0 LA N [! h CU ºn O In N t/j co U)
r 5 p
H 0 to O d' H m !n r N cy, 'co z z co M O) 01 CO M h in CO M l0 d' to h m v
ao lq: r O in M O CO 11; Co N m O H-! r H Ln H "I en N M N N ri H N H N N rl N
.ý c4 U W W t9 x H .a O c Or d
102
C' w
a
54 H
C) u)
z H
V) r.
W
4'
a
v:
N'
c°.. '.
O L-4 s 0 + N cl) u cý a C' c>
N N c+ C- ca ,n ca (ý o O iC u, CC in In ý. cý N In O N O M c+ in (Ti c-4 q cD n N N " " " . " " " " " " " " " " " " " C Gý M N (n V' Q' c^ u1 (") M M N M N %Z
(1 iN v7
Q N v /C 0 C-' X) c, N O as r-I mot` to O al C-) :L F.. UN U a, C'- CO CO "-,
0 (N (Ti cn d' in W o`. [-
n 0 lD Q' a Cl M c-{ 0 (1* c+ N O N N M c-i ý. c-4 ci c-i " c-+ c-i r-I c-- , -+
" o in c+ m CO (Ti r, c- (Ti O C- w (q r) w ^ G1 O c4 (Ti O al (\J M in r-4 lf) a\ N c-I N Q) Q
v- Cn M N M 4
M M In M M Cl N fit' M c. 4
N ul IT) 0 N
(D Ln 11 Q In CO 0) C. ) co lU N N [! ' N N Ul M 0 Gl cI a.
Z 2 cf Cl) Q Cl CO 0A LD c", d" CO O H M 6 O IN " " . . " O ýo v C T (ý v c+ °Q c N co
stir' r+ c-1 ci* c-+ Cr C-1 -4 ` TA T-4 TA
" N co 0, co (n C) L' Ql [f' cI Crl in in (V o c, 4 Cl N N O ct N ct' Id) cI M Cp M al CO
Cm in M (V M M in M M M 4
d' (V M (N .4 wMM
n kn 0 Iq N N N 00 d N O 00 a Ci' to aN V' N vi (Ti
,Z z V4 0 M c') l- M M al to o) N in ID to N C
O N O O Cl N O N c+ c+ c-i pp c-I M ('n cI T"1 c-I c-I c{ c- (1 cI c-I rl c'? c-I C"1
" c-I l0 Ca c'r 00 0 fM 0 O N C1' N N c7' ID c-1 Q Ci O O in in cn in N (r N N in (Ti to 0 N
d' ) OD (V N 4 4
<Y' m to (4 4
M M C' N M M (n (n vM 7
(n 11 f^ to Cl L7 N N l0 N C) M [N al M 0 40 (V 0
Z Z c-I N N lC in to w ; i' a) d' (V v7' l0 N f- 0 ý e o [a? O N cf' O c-i N N c-( N N c-I ri w ri (NJ
t'4 c- cI cI c'H c? c-4 ci t"? cI r4 tI
" V' ci' u) c0 w 10 1T IH cO ON N cM t- r4 M G\ A N o 0 in N
" N. M o v' V O O d' 't' 0 in
C' Ov
to 1"1 (N In C' V ' M
ýri r) 4
ri 01 v
'ri ri r7 v
OD M U)
Q to 11 N CYN ýt Ln r- co J. 'r a\ Z Cl N co O O w IT M " c-I (N c-n C)
. eý W O [: C' e-4 C-) c4 N 1 CO N Q c-1 O) cqQ cr crl
rl Cr c+ c+ c-+ c'' " c1 r+ c'I t1 c+ c-+
" v in V' N a
t- lO N
c+ in
tD CD
N 't'
[N ýT
N I
" at 0 N CD
A "
0) "
Ol "
Cý "
c "
tý " " " " " "
c " f
M "
tp fM c-I ý' t. 0 U] N c-4 M fi't' i! ' M In m c"l M M CN "4
" M
" cal
" Lf
V] (v N
t' c i!
v " " " s o " a W Ol Iz
4 N t11 O f')
0 co N--4
a X-4 0 c en Z 'r-t r-4 c-4 ci H T-4 c-1 4 V cc-I
O c-i r-I
" (n N in 0 '7' 0 v1 M cfl C) v' t0 C) C' d' M ^ Q 01 O a) (V V" u1 UY) cu N to ci N (V 0 N S tf1 " " " " " " " " " " " " " " " "
C) 'C' in N N (n 4' M In M f+) M M 4
C-) M M [T V; ci N W1
n < T-4 M M co co t', Q0 p N to N c-I N Cl c1' cq d
Z z (, vi (Ti N ci CO 5 1.0 " V' N d' v to M
o ° n c , c c °ý 14 c ý TA IH ° T4 o 0 0
x 4 m u W 54 O H a : E: z O Ö Cf Ö C
6,
103
co ON
H
ei a r4
U, n
w w 4
Gr
cr)
äý
U ý7 «t, U x H :' z, ti O + N co '- Q a CY
"ý ;ý CO " O 01 00 N 0 N CO M N Q LA) N
O Co r- :R Co N V' Co M N CO Co N Co N ZIP CV M M C'
(ýT + c-4 f+I In
,t ul N 0 N U1 NN c1 co Co co UD N Co cc) co U) z Z W tt) H T-4 CC) 0 Co W Co Co Co co [- M !V . r)
of UD 0 O CO Co O Co Co 4 O N- m O fti
Co
" Ol O (V 'Y' 0 c' If) Ol C' U) t0 Co C) c-i '-D Q q7' c+ V' o? O 0) cP Co LI LI CO LI) u-, Co cr1 0
"Y O l7 Co N 4
Co mal' N CO CO N M N '4 N M Cl) tr c11 L^. (v W1 (D Ln U Q U) 0 to u') co O 0) V 0 CO (V O (V Cl) cM al O z Z rA U) CO (Y) 0 C' c1 CO H UD 0) 0) O U)
0) ko d' O C' M O co O c-I O 6+ ci c-I cý c-I %-I c-i c-I V-1 c-i
" Co p) M to co C4' cI 0 kD to N (C) CO V- C' to "" A u) Co c-I O CO O) M Ln '-D LD Cl) Cl) CO c-4 Co O M " " " " " " " " " " " " " " " " "
'P co U) Cl) c-i d' 'f' d' N l! 1 M N cM CV 'c2' (V CO C') ü1 W Ln
O tt) u 1y V' Co u) c-1 CO d' W d' LI LD tD O (A d' c+ ci 2 z ON N Co CN to it) O c cat c-+ v-4 N O O m
CO N 4'' O ('4 c0 1 O of O r+ c-I W N C\l c ( c 1 -1 Q c1 c-{ r+ c -1 i c-
t. 0 :Y 0 iI (V CO a cq M a) U-) I'D N c-I N . ý. A U) 0.? c-I ci to N co cM [b N QO l0 00 CO V' c-I to " " " " " " " " " " " " " " " " "
NP tn V) c4 c-i 'd' c}' 4
N CO Co CV Co N eM N f4 Co U) N'M
ý 4c ý1 N W cV ca 0 O" %D N O O O ai q; r rn O CO z z Co 'd' (n M CO Co r c-I t4' N CO 0) N Ln N (l
0) N d' 4
CO (V CO O 0) O O O CO 0) N C) c-A c-I Irl rl c-1 r-i c- ri %-i c-4 IH
" 0 O C) LI) O) 1S d' d' A ON CO cn CO ON l0 l0 N co t0 (X) 0) N M N
'Y O U) Co c-: Co d' '3' N Ul CO N M N .4 N M M 4
2MN
OA u a `\' O 0 ri W (n ' 0 ` ° z z w °n ) 0 (0 (0 N N öM CO
C') n
°' r o m o c m w TI c-i rq 1-4 c-i I V4
C\j N r-j C\J 0 u to Co A CO a) N CO O l O t0 13, 0 N Q IV
, 01 Ln L/! .
Co c-, -T V. '1 m LIl Co N M fV d' (1 M fM fit' u! N (V
U-: )
u < CV 0% v-I N ri '4' U'1 Co lV O Co ic, 'P r) t+7 C') G 2 c*) co w rn GN M Q: cri c+ 19 c1 cv c tCi cý .. Q " " " " " " " . " " " " . " " " W 01 lý ý ý ý Oý Oý O Oý Oý CD O
c ý ý c c -I
a' C° `-I CO N -% Q to co O O (0 O 0) 1N co N rn ON N En Co c0 N 14 N d' N CO CO d'
M Q r-4 Co N M c-I d' N N In 1qP c-i 0 rI m a\ O'. [-
v O d' M N N (0 0) U) (0 : Cl : CD U')
14 N m rn (N a) o 0; Co r4 -ý
O
4 0o u W w O x H a z z O N M ýt
a (Dr cr cr
104
rn rn
d Eý
ýi H
0 U) a
En G
w
U)
w U) x
cn T
Ü
U O" c O C'
t
. -. . Q w t.! ) O S Co in e lD ' N N lt) () l
L() Q ci tom- LA N ci N I() C) ý)' ri c-{ tv1 ri M OD ýfi +
i r N 4 O in C in d Co d' ýn 't o; W`
ý7tý 1f S N vi cU () s-I Co if) C) t- ) ci c'i ti)
a) N 0 G) O ýD fir) c LD c+ C O ý7 N - N s 2: vi c-i M cI N CV N N ri N N N vi N (\J N
.. " N 1- ra d' !: ) ' c-1 c-i N C) 61 Ul N Co CU G'+ rn
- Co in to C' co C' O
. c-4 c+ ul u'i Co c-i Co c-+ O
O N LC) (V
. ti)
" LI)
" M
" N
. lD lG 111
. Q
" lD
a V'
" in
. d'
. IN
. ' in
" lf)
" Gl
W1 c-f (D U) Il 4 u) N 0 ýj
Z Z c' ci cm v" U 1 O N r-4 O d' N . " " . " " . " " " " " " " "
m cn cl c7 O S ter' c-I I-D N M c- I ü7 c-+ to v' Fý c'1 c) N cV cV N N tV c-+ N C) N ca N N N
n " ßý O 01 in N t- M d' 1'' lD C) N fT O S C) N Q N Un N CV Co C) Ln M Q M in N to (r) O G " " " " " " " " " " " " " " " " " cv (n Ln Cl) Ln Co CN t- S O iD to in d ' oý d' in in a1
W j ýn u 'l CD «e C'ý fn W OD v-, CO Ln t-
Z Z c-1 M i ri In C) ill r) O N Co CO si N O .... " " " "
Co (
in N N N lp Q\ LA l0 CV N fV N !V c-i N C) N H vi (V N
.. " M tf1 ß1 N CV 'V' mot' Cr) N l0 N Co CA lq N A l0 M rI to c-t cý S V Co O In M in Co lD to
" f a " i " " " " " " " " " " M C U) to M r N CO in O li
4 in
i co in in W
' W to II Q (7 0) t0 l0 M l- N Q' *7' C- e cm co lQ N C) l-
2 d r r- O cv in CO to . -4 t- CO N s cn rn aý ¢ W d' 0ý N ýn in ýn ci tD N c{ N In O d in
V-i T-4 (V (V N N N N x-' N N N vi cu N N
.. " CO �0 to O N M
M N
Co N
o 0't Q O C") N Ol to
d p
" Ch 'M in
" 0
" " CD 0 0 O 0 CV V] If)
O
uý cn "
u-) "
M o0 CO 0 tQ LA in Qv CO in LA in Co Wi (9 Ln !i <N 9 c-I d' c-I if) c-' lV n ýp
Z Z z ßl CD d` N O Ol 7' to ON d' N c-) (V Co .W O
" " rn
" N
" N
" "'
" d'
" c
" ýo
" N
" O
" cN
" " O
"
cn "
I-o N cI N N N N N lV c-I N N N
ý N N N
c! ' A CO 0 Ln C' O 0) tN N cý in O S 2. U') " . . " " . . . . . " . " . . " W CV C, 1 Vl I) CA r! N CO N O 1D LO to d' to lt lU OD
a ., te 2 Z fry c- f Cr (TI N N O in [ý cc O 5..., it ;
{tj G7 tt) Oý% �z3' r, H l0 O g c Z ri [-) N N cA N cm (Ni c9 c -I N N Iri N N
n " c-I 0% d' 0 ln M s oo it M d' c-f c'+ lf) M LI)
ýo A to ri rn to O W N N e c) 5 c-+ O in N ýn ! v) c-+ N v) u) to in r co to O ýO er) to ý4 OD in to Oý Cw7 8 d n c 3 2 , E e N in
Z z to iD c- tO O M 0 ) 0 n O v cý
WWW " s " " " " s " O
a N r. '
N N M N N N Irl N %-4 N vl c'4 N
N
p'.
x Q m b W w H a X Z Q Cr) c ä
105
oz o
r Ww x
HH
CO
W4 c49
a
0 z
w A z
H
a En m
H
ö
H Q m u w w 0 x H a Z z O 1 O N 0
cM O
' O
tp N vi e In '. 0 In Lo N M w In Ol N 0 Ol I/l Vi l0 v1 Ci' '. 0 d' N d' O V-4 In O N de M 0 W+ . . . . . . " . . . . " . . . t7 In O O O O O O O O O O O O O O (D O Q ý0 I 1 I I 1 1 1
d' 0 M ýp d' Ol d cý N 0 M N In c, 01 In l0 t/j lp N d' tý di -ýZv In Ln 0 VA Ul cm In O N N
ON 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 QI "1 I 1 1 I 1 1 1
le M N dp lS d' 0 lO N CO N d' [N N t-4 00 N U] Ul N O N dý cY) N L2 N O d' CV r O ri cý
" " " " " " " " " " U1 O O O O O O O O O O O O O O O O
4 d' 1 I I I 1 1 I 1
d' 0. cf N O1 01 d' M N c-, N S-4 Co c1 e c-I O Ul de cm O v M v 0 v w N O N N c' 0 vq N
" " " " " " " " " " " " " " " "
(D Lt) O O O O O O O O O O O O O O (D O
d' t0 N 0) t0 d' In Lý W 0 Ol l0 (N to to In d' V! M M O O (1 c-) vi 0 Ln N O O N N O n N O tin c) O O O O O O O O O O O O O O O 4N I I I I I 1 I 1
Cr In M M N '. D d' d' cý N N N In CO N G1 N ü2 N N O O 0 ' M O _l O N H N 0 N d' Co
(D O O O O O O O O O O O O O O O O' Q t 1
m M In 00 () O W 01 CO OQ C% e u) M O N O (0 S1 (-i c-r N i
" " " " " . . c) Ö
O O O O O 0 0
O O O O I I I I I I
Q 4 w u w w C7 x H a z O
106
ýi 0 H
H
F4 a z 0 U)
U)
H
C9
w
H H Ha
CQ
m
z
w a z a
w
a a 0
FA Q W v w w (7 x H a O
Ol N co ul In d d M O O 0 W S l0 N M U1 vi in c- Ln In N c-{ l- c-, cm in ri cý cI d' N
U') O ö o O O O O O O O O O o O O O <0 1 1 1 1 1 1 1 1 1
c7 ip in 'O M O O to c-{ O O d' a) N O c" Uý lD N d' (V e I ci Lri M O d' (Y) c-i cI N O
OLa O O O O O O O O O O O O O O O O Ln 1 I 1 1 1 1 1 1
d' c cm N N In vi al In r -i d' O 01 O N c11 O C/1 Ln N c-ý M N V-f in (1) O N M cI c1 c-I CV
ü l O O O O O O O O O O O O O O O O
q d' 1 1 1 1 1 1 1 1 1
le ;e M N () N M to m 00 e to p1 pp G c-I to tq d' m c) vl co O N O ul N O O N c-I O O N
" " " " " " " " " " " " " " " " O
in O O O O O O O O O O O O O O O O
IK? Ln CD M c [- Ol in O O ri Co to to 0-, VI
d' 0 vi -1 N O -4 d' N ri O N O O M
ON O O O O O O O O O O O O O O O O
d' tO O Cý d m N O l0 l0 l0 (Y) N v! N cn N c-i M M M N N 0 vi cm Cl) O O e e
00 O O O O O O O O O O O O O O O O 4N 1 1 I 1 1 1 1 1 1
p ý ýp m M to U1 N N N O Cl) 01 t- e cm Irt ul m 0 N ri l Cl) Cl) N O O (1) e o
Ö to in
12 ll l' l' l' ' 9j O O O O O O O O O O O O O O O O
4 PO U W W C7 x H . -l Z 0 vi ri ; ei Fit
a cs
107
z 0
a
N 0 V: ci W
H
ä
ul
z
4 w a zd
H N z dH
4 ul W
ü
aO L7
H Q 4 w U W Ga 0 H r4 X 2 O N Q
CJ in M vq I d ri CO c-' 00 0 "O v CV lD 44 U') N M N c In d' O O O N d N M ýp W+ " " . " . " . . . . . . " . U in 0 O 0 O 0 O O O 0 O O 0 O O O O 4O 1 1 1 1 I I 1 1 1 i
0 N W v-ý 0 v cý N co c4 to l0 O OD O CO UI to CV ca 0 0 to N 0 In c-I 0 O M 0 IH i M
0 In O O O O O O O O O O O O O O O O Lf) 4 1 I 1 1 1
d' N 00 0 ßc'0 ý t- Ol u') 0 O N a1 N N CY) V) tll N O O r M C') 0 u1 O O O IC-4 O O M c
u ) CQ O O O O O O O O O O O O O O O O
4 d' 1 I I I 1 I 1
d' d' '
pp -1
pý 4
O -ý
ri CV
ri fV
t cv cn Cl) 0 M O) '
CO Ul C ) c v M c tf) 0 (NJ c-I CV ri 0 C ) N
U' En O O O O O O O O O O O O O O O O 4 C") I 1 t I 1 1 1
d' in O M to L0 CO 01 d' N tN to in un C N En m C'n CV CV to t1 0 O ul c-I M N d' O O r-
N O O O O O O O O O O O O O O O O
M 1p ap d' dI Cl) U0 N ri N co N rl CO (ON VI (V M CV 0 l0 t0 hfl C'') LI) M c-I
N ri ca
" " " " " " " " " " " " " " " " N o 0 0 0 0 0 0 o O o 0 0 0 o O o 4
p ul d' 00 O EI c-I o o CV in OD S c`') l C') 'r' r 00 Lo cy. ) %-I O LA O O O O O O O O O O O O O O O O 4 ri I I 1 1 ! I 1
4 CU U W P4 C7 x H º4 Z O N CY) ri
C), a
108
H H
ri) M
ca
H H,
4
ä
ci w MQ xw H£ H tx H
ÖQ
ö `ui w ö
Q Q 0 U w w C7 x H i-1 Z z O ON
Gl O 0) d' M M cI M M N O O Ol CO G\ V] M M N O X, 0 v' N T-1 O O d Cf' N N V in OD in O O O O O O O O O O O O O O O O Q 1 1 I I I I I I 1 I
vi d 'z in N CN d G" 0. ý M cº l0 M OD d' l0 in Ln N OD OD cm c-{ 0 O N c-I N 0 O M M N M ('') U) < in O O O O O O O O O O O O O O O O in I I 1 I 1 I 1
ci N O 0 in OD M N N c-1 ß N i O X-4 M N N N c-4 0 c -4 N N ( N Q in O O O O O O O O O O O O O O O O
d' CO C% N d' M OD CJ ýp lý
W d' CO O N d O O r cr c-I O O M O N M
OD in O O O O O O O O O O O O O O O O < cri 1 1 1 1 I 1 !
Co N Ul d' N dl Co r- CO cm D (n 0 UI M N
" N
" N
" in
" c
" N
" " If) N
- -i OD N 0 O cm
0' N
O O O O O O O "
O "
O "
O "
O "
O "
O "
O "
O "
c; 0
le l0 O x-' tý d' d' N [b in 0 N in CO N e M ü) N N (V v1 in in 'V' N in M M N ý( O CO O
0Ö U O O O O O O O O O O O O O O O < ýS I ! º i º I I
O Co t/1 IM IH O c-1 M l0 in N O p N r, m N c-º N 0
in O O O O O O O O O O O O O O O O Q 1 I I I
Ö
4 on u w w 0 x H a Z 0 ä ä öý
109
APPENDIX 10
ALTERNATEF0RMRELIABILI TIE S
0FTHEPRIMARYS0URCETRAITS
BYSEXAGEANDS0CIALC LAS S
110
1 ,)
C-.
CS
r.
.:.
c
W
ri) :y
! -i
Ca
1<
Q
c;. Ca
G+ 0
Q
cý
C
ii C\ e
if) if) '
L'') t
N '
CC M N II) Cu 1- t i 1 I .J i- l'; _
() t'7 M M M t'T C7 i) Cl1 C) (1)
i y ri N H Ui I IX7 N C: ) C-1 JI) C14
. L4 ko C) co Lr)
(n
s
C C) C? CO a1 Gl Cl Oý Cl Gl ON _ _. _ . _... _
p C1 -4
-I --t l0
{ -o v. M c. 1) . --I l^ C: ý /) CI . U r- -i N C c C-) ri O
rý G7 N M C) C+: Cl) M f y) N M cr M
(y) M
CO N
c. . -H
Cl W LI) Cý Ol M N
N N
G) N
O O 0 Ol CD C) C3 m 6
j Olt M O LO L') L.. ) y N O. C7 if) CC) to 'C I N Si . -1 M ýS N CA cV lO l0 {
H CO M 1-1 O\ lO N O in in co O Gl . -t O . -i N C7% N N CO Cl to
. c
. ul
. . tt1
. tll
. d'
. V'
. un
. to
. .
O, l7 ON r-q .O t Q N . -1 G1 O ri C' Ö
O Ö
H
M ý
N (`) CO T (N N N C) N (X) Al
i " aQ N . -ý N lO N f- ). O k. 0 I-T i
H r-1 r-1 'A ri r-4 r-C s-"1 -1
r-"f . -i
CO CO (0 OD N U) H Ol cf) Q CO N N 0) H C) CO llý t' CO N
.6 N N CJ N CO N N (N N N N
?ýr, K- in CO N N M ý) O C) N CO M rl N N Ol O N G1 ri N r-i r-ý -1 te -4 p r-q H 0 4 -q r
r 4 H r -i -q a
H r -i UT
(N '
CV N to O vý -'ZT to 0) ti In it) N N N . -I if) tt)
( CV "-i N
C N. IT Cd r-"I N If) N . --I (I) N t
S4 N
-4 M tN
-
"
lD N l)
O -i tf) N OO H 1 ý'
CO Cl
[ V
C C v
(0 'Ci"
N '7"
co M
X74 ý"
() G'
"" C'
6 t0 ýD N O c
co l0 O u, O r"1 to to to l0 o CO 0) %D -i' lo lJ N
t! ) 14 M
C4 M M
M c'1 cM M M M q
g . -i N co to . -i N
N C-1 Ln C) CC) C) -. LO N 01
: 0) co (Y) O1 r"-1 (N
r- GO Oý [- Cp tz C6 Ci7
C: ) 4 4 r1 m -4 %D C, 4 co m LO 0) "i H r-7 ) 0) 0) OCR N 17%
U) N M M C") M N N N C"; N
if) O ". -1 N 01 Ol O (fl Olt C) C N lD H CT) N O V N ýT t'
t; 1 C^ O'l Gl (n C') C) C) CO ) Cl
Z '' -I '. D t") Cl) O rl 00 LO C' lD I
CO O t- C) lfl N co CO r) in c" t`4 C' (N H r-1 N C') N
i1 00 f") r"1 CO N
1 to (n N 0) (\ C")
p) [- O) O 0 O (`d r4 C- co to v.
. "i "
. cY'
. l0
. U
. Vý
. C
. C
. ý'
. to
. C
.
1 ý Ü
rU r-4 U
N U
[il Q
H O
CC) N C' ýO {
- I r1
ul tr) ºn to i F
(-A r
to .1 co
S N ýt' ýD Ui
-1n l
-- ti
v)
p4 P J
a a
ý L"7 L`+
O (Yi
E
j Q r; O OD C) O m H ra N { >w G^ Cý N N- co Gl C' 6' CD CO
.: L) M 61 c. ( w CO L') '-i t 'n CO
''-i U) C') ti' H
co I
N N .o Uc 16 N N '- N N N
`�yi O to N -4 L' N 01 C`'1 N O J-l a O O r-! CV O O ul O r-I r-1 ri
N (Nl CV N N N r-1 Ni N N cV
'
f
' C : i Lp Ln M v C ) r-i M CO H C; lo CO N Cl -i CO
iia C ? . N N lC lC N- N l0 l7 lG N- l7
z ý. M (n L7 I '
L) "-i N 61 O in co : [l Co N r) ) C) 1 Cl U) N- N
ýp c'ý r1( M <t h ý+ N V l0
in . -4 r-I O OD N CO co N 61 C) C' M N M
' %D co N co N- -4 N
M v' ý' v mot' r'1 M C) Cr
Ul l0 co l0 lý O0 lý N CC) ri O q OC) 01 ri CO l0 03 Ol CO O Ol
. -i 1 "-i N H .4 r-! H r1 N r1
C' O M rA . -1 N N C C) N "-1 H lt3 ýf' CO N N to ri N fit' O
to N . -! ý1ý N Ln
_.. Gl CO t-o l0 '-; m
O O H (") C) co 4 O O H O N N N C N; r--i r4 N N N N
rýr
ýy co ri tD in C) N Co 1444 (V C) co 'P 7 O ri In 00 C: 114, t- O l0 O W
ý w rV
[ý Cý l4 Ui Iý Iý l7 lD \O [ý lU
N 0) . `ý to t11 N N- %D in N r-4
ýy M M
ý 00 ýi -4 in ILO C) -, v C> Q0 P -i 'V" -T tO d' C" 61 CO U) .1 in
cM t0 V-1 CO co %0 C") ri m C N p co p) Ca N l0 CO O1 O G1 O O
f! ) r ri N N t-i 1-4 . -f N r- N N p
O N
in l0
-r r
N- N
'--1 ff
lD r-I
l0 e-i
r-ý l0
O U)
co ý'
N C')
V `. . [- 1- f- W 00 co r U l- 1- lam.
O N co O N N M O 00 M U) ON C1 O N H 0 "-1 O C1 (N H O
r1 N N N N N N H CV N N
- C) O N- Y C_) 4;
N M
tl tf)
in Cý c. CV N-
C C'
q C
. -i N
co N
( ) ri
co N ýo tý r- r
Z r r-1 l0 C"1 M CD '-i Co in C U I
cr1 O l0 C) 110 N 00 C (''') in IY' N %r N r-! H N M N M
O 00 O N- ul a1 v' co (N H v. }d %D O a1 l0 O1 Co N- co ' t0 (N 01
N .
ý" .
C .
Q' .
<N .
C .
N .
cr) .
q' .
-, v .
M .
`a n ' o o
w + rý jý "-i H H Lin in it-, th H . -4 N l0 . U)
117
ý1
cn
Ui f--t
Ql
-- ...
N0
a
r a
CIA
o H
H W
E
?, , ý.
Iý
a U H
8
U
ý
' U) H
`' C
M C')
rI C7
N N
O C)
I- O
L! 1 C1
M -!
7 t
1 1 t . 1-
L) LO N ') C) C
I7r ` In tI; Ill 4
(1 `: ' ý+ I H
i U) ` L. W ON O Lfl 1-4
Ol to ON
C4 C, %
ýý CO
cM Gl
rý C'.
rh { O 1
4-4 18
'O I I'D (N C7 410 ' "-4 CO N to c 1' C C) H C) cý ca: c- O
ef' !n t. ll ý(1 d' c`'S ri' V'
Z Gý r) CI LC) U) in '-f N Gl C) Ln CO t7 C I- c'_) t L'1 I-- N
r-4 C) c*+ N 'ý2` 1 C. ' to lCl
I i ri N r-1 N N O N N O CD N 0
N 1Si 'CF' lf)
al ý7'
l7 In
tD c7`
O Ul
0% C'
N 'cy'
N Ln
. -t Lrl
Q lD ON r-i if) N Nil N Cl Cl (N '3' cl lo r1 N C) In of lo CO rn m
vý cri C4 4 C4 Ci (, i C4 (n
lo OD
CO C'
M U1
W 6 -
V' Cl
00 lfl
N CO
N O r-S N N
l l N 61 Q7
n H -1 r-d H-{ r-i r-i H H '--I r-I
l0 C) N Cl N r-i v` O N Co H C) 0'. Ln N H CO C) H r o Q
M M C7 M M "IV (11 (+') mot' M ct'
l0 al H (V ()l lfl 0) Zi' 110 O M Ol CV OD M O '. 0 tN C 01, , -1
LI) (V mot' N N U) C) -zv l0 Cl un t. r3 N N N r-, l. o I. r) c'1 --t 1, - (N ýw N r-i N (M N H M N
00 O q I- t. l1 r-1 M OD O lfl y4 N co M O % CO N 910 co I- M
'
. . . . . . . . E
-
. . .
N O O r11 ('1 O Ä
p to O M
(') N
O N
CO Cl
O r-{
CO 00
to t-
O Q`
r-I l0
N ! f) M
Yý Cn to to
r--1 u1 rA
tp H H
U H H
Ln H r-i H
U) U. U)
[c], ti 7 {]
l0 CO
O al
d' O
to to
Ol O
G' iN
ü1 N
Mv O
-^- CO l7
d' 61
N ON
f C (.
5 C M M M c' M M (4
CO -i
CO V*
O O
C) 10
N U)
ül lD
( Ch
%D 1
l0 -
't' M
U) N
H Rý
r
tll tS) . f) L7 tf i Ul L()
9
Cý'
!
i) L') I f)
2 cý H ID cl cl O . -i CO in ;r 1-0 r`") N
C) Q'
'-0 C4
C) r-I
lD . -i
fV N
0p co
(-) N
C) H
if) M CV)
M CO LI) N 'V N -4 L0 rl iD 14 C H Ol 'd' ul V' v N (1) 'C'
Ln Ln r ul to M t. f) U)
H [ý ý1 Iiý
FH H
Vý. ý
--r "ý
r ýJ U U O
% 0 co z . i o 1
O w 0
F,, t t . -1
ui N
t mot'
L l0 CI1 co
r
113
GV
ýf=
W
:4 LA v
c..
üQ
b O
Q
N
T
II ýý
r-i rtn o NOOLn (w tnLn
N" a' l0 +
N M U)
N H %D
co (-I U)
m I-i in
UUU ý Ü Ü W ý L G
8 U)
I In I
C) 0 k. 0 CC)
Ho 400
IIA
}
00 0 N
c,
`4
ý_
ZS H
;f
U4
r
Ü w
H
U,
JJ F--t
r-1
4-a
Q
6
ýi
M rN to C) c, ", z;, O rn _t N i r . ýr r cý cý ýv to lzr e, ct,
. i
-: in r! C P-) LO O ýt O U1 N O C) cV cam: c, CO to ! r) r ri c
" -4
ö CI) c v' C') ' ' ci
rl cr M M If) N C) (') t' CJ Q rý - (N Cr) N N to c O C; t'
^r v ", ; 4
14 L`? 4
i' to L! V' I
M Ol CO tr) . --1 -A l0 LO c" t- N-
QJ to C'') U7 O : -i L r) H - N- O
(ý () to N O O' N ('1 N N N N N r -q r-4 H H H -1 r-l
H H H
m Cr1 O LI)
Ln Ln -i N 0) O Lr) z CO Ul a rn C"% -11 oll LO r- C-li
t" " CU H t`'l C N- ý' N lJ l0
N to CI) U) t co (N l0 co to O CV "--I LO H Co (V N -V r-i
i4 kD lfl Ln Ln l0 il- lO tO l0 lU lD
m c) C) . --l N N C) N O I tt) 6 (1) t. () Ol O N Ql ti' M N CN N
cn ýr vw m "I; r IV NT ýr c4 cV '44: r vl p - to -: 3' tO 00 r'1 . -i O LO CO to
r( N N O H En u) N N to O C ý rs tt7 d' N H M t'7
r"H r-1 f-i H ý--ý r-l ri r-i H r-i 1"1
CA to Co O LO U) ) En t- to (V W N O ý-1 .I M t[7 Co tO -I N
ýw 'gP "; Zr W to to 'r C' a' Lll tY'
. -t N O IV O N l0 al CO to ri C'1 V' CI) Ll in M v' CO H tD 01
In C-4 c) CY% CN Cl) IN -4 CY) r-1 H H 4 H r-i H H H r-i to N C' N N L)l O C ... __ _ _. _. ý.. _ ý.. _
z Ul LO CV N N- H t to CSI H f' N N H N cn N H C`) N
N O O n7 ('') d' N N to CO r-"f " 'ý
M (N >n Zr I O Q r-1 'Zr M C) W tD ýzS' l0 Zo [- In l0 to to to
, ca, d' O M tI? 'Zr CO al N CO f- 6 C t o oo . -i to r- In O 01 [-
'C ' cN C M tf) L11 vý cr Ul ý4' S1'
O C M H Ol 114V O C) '4 N LI) N- N 00 to N- En O to lfl H M
lO H O N 4 4 e+7 f et ý` i
N M M IS co to cn l7 [g O O Iý
oe. LA U; 4
1. n
ý! ý 1° C3
[ý Ch
C) In
U) r
lO Ul
co "Zt'
O (1
tf) lD
N q! '
t- [l-
lO co
ßl r-i
P tt1 N O 0) ! N N N H N N r"ý r--i rý r-'i -i r{ r" 1 1-4 --1 H
z -i to CD -i co in kD
M O tO O l0 N W c U) w N C N rl H N In N H rq m
N CO N N G- Ul tO -1 M M N- CO C- M m N H Cr CO U) C') to
. U)
. U)
. C
. tD
. N-
. lo
. to
. to
. to
. to
.
}U. }ý y.
ýý in
"ý
l
" Ct Coil
G C1
ý ý ý ý f 5+ ?ý >+
c1 Ü Ü Ä H
o tD CO
ü i o co t l t tn ri N c' vý CO
115
0 0 iI
tS. N
I
ä
LI
ti
U) ! -i
! S: C1
A
D
G
V
ýi
C7 a Gl lD CO -f N l^ Cl
cl
O .
b 03 t- N ri m m GZ .
N N N C) r N N
co 1.0 C\) CO : I'
f, 4 " N Lý (\ Lr) I) C i O (1) II) + ..
L r1 N ('1 t1 N N N CV
L^_ r-i r-4 H r-1 r-I H ri r-1 r--1
kD N l4 l0 t- O CO N, tý ' rq'} ri G co r- Ni i: 1 l11
l
N M r-i C) [- N NY' U V' co C lO Cl lD ri .ý U Gl 1
O N M %'l N N CV H ri r-i ri H A r-1 ri r-"! H -1
Z ßl m C) U) Ili Iß'1 H (14 C1 Q LI) a: ) in c3N N hl M ýYý dl LO iý N
CO C' ri m r- '3+ CV 7 lU
M CO N- N H cV cý . -f N al M
cri "-i to '- tip CV to CO co N. M $4 `V' C' M M to Cfl mot' M ýt V`
tý C) co to c, N r- - (j cc) 00 al Cl .-I CO CO Ol C t0 C)
I v Ql Co N N N N M N N N
rý mod' pl
Oh N
M Om
(n M
to L') co H_A
r4 (V ch M Cl M N N N N (") w .c f. _{ H H H H r-1 r-i H H . -t H
C) CO N N fit' O 1-1 l0 M v r-1 O ft lD 10 (N N LC) N
Q NT M M N M C4 M M
a4 N CO H to -4 -[r U') ei
r. -C' QO ', ' C') O CO O 1 4
co
ý to
4 ß r ý r"i - H C t'{ i -1 r{ r1 r{ e'{ l e{
- N cV N Ul O C 'a' lfl Ol L') In N N N i-"I lD i-fl
t ý'' N wow N H N Cl N ) N
to (N 00 N co m Cl lfl C") C) U)
C') C'r) In I11 T (`') Ln C
pi N r-1 N N C" N CA %D p M 00 C) M (N m N O r ""i QO
f7 C' N C1 M M N N M M N N C)
C U) C, O e-1 C ') ON r-i C> N cfl G
G H H H H [sa H H
" -1 H
r -1 ' -"1 T
.-! H
r -i r H
4') to Cl CO 00 C) M r-1 U) co (N Cý "l r N N O co to I, CO N. N
N M l'''1 M M -: r t4 M M M M
y r-1 N 6'. %D U) O N O CT f"') . -i O "q' N Cl N NI' O ül "--1 fý
M M N N N N N
H to M M O . -{ CO to lD M O l0 O ID N 00 CC) C") fl l
CZL- I- -4 M .a c ̂
O Ln . -i 111 ý1 f7 CO
M N O Lf 1 tf1 N
In "
M C") LI) LI 'd' - C In IM
N
-4 CN E4
--w Iýq t. 0 co N 'C I
i- p -4 -f d) Ln th En O %O Ott
.I N -%' O N
116
)
0 ci
51
H
Z-
G
t..
Ca
cn ý+7 E
El) H
all
CQ a
L=+
(U
x
M cl 00 Cý; -4 CD ^" ý _
` .
s L) 10 ifl .n to
-' M M m N tc ýt: H . CO le 10 t. f. C G7
ýn - N N [ý1 CD r) N H (N CD N
-4 n i 1
tYi . _. _... _....
L Ll-
ý ý` ` to to C) to } (N f'- -V n to " Cam' : _r-
t. n Q () ]1) ! sý c r . -ý cr c """ ý. fr in : tl Ifs lf'i Lf) Ul U) LO til 1ý1 Lý'T
ý-1 ul a CO c -i Co (N co c 1r Cý r7 N N . -? CO N N r-i N (N N
Cl C) CD i'1 2f :" 1 rt . -i (14 Cý C- II) i; ') lf) 61 N CD C1 c« { lf? h- N
M c". .ý ., w t r.
CO d' r-1 CO to -4 O to V al N 14 (N U) C' ct' l0 W N N CO CD -1
In a1 U^. 10 CD 0) G} 10 N- (D r- Q CO N ýD N O r-i N Ol N
(n In !n Ln Ln lD U) U) cri lei Ln En p 110 co N 0 V' 0 -I N O H M ri 61 ßl O tý r- 1-4 lqr Ol C1 10
O ß r4 Cý M N H r -1 . --1 H --1 i -i "-i ri H ý--1
"" .. .. N m Q (1) H co 10 117 1n N (N
-" M r l0 lD N ri CO C' to CO Q
Ln ul to Ln in ul Lo W Y) Ln to W Fý n
CO In . -H C) 1-1 CO co fl- . --t --1 uy N rl O r1 M 0! to N
r-I r-4 r-q ri M N r-I -4 N r-1 ', ý r-1 . --1 r-i "-i ri H H H r-i "-I 1"1
u7 (N -zr (d N 1t) CU if) LO N N N. . -t 10 if) N ý-1 [-- N cC t .I .t N c` 1 N 1 <h N
.i to co to [- O kD "V co Q) I-W t CO l0 N '-0 1- to N l7 CD N (11
[v) d' N 10 ll) N C) N to N lz: r l0 tl t. [) rd C) lD U) M rý 7
V) 1; UA t1l ID In to 1. n to Ul Ln Ln
CO N to l.: v' -: CO co O ON co N LO O N mot` Ul p) p)
-4 -1 rI -i
co mot' c" C d' lfl C% M N u1 O un to ul U) t[) tr) to C4 L'
CO - 6- CO (' 1 1ý DO 00 `d H
cr M N N CD D C N ce t1 cy -4 'A
rte --- Yd'
. -i lD M ý
C-*) V-1 '
to Cm C) W C) to N oD c. ) M to N C' (N r-i . --{ (N CO N -4 (0 t-)
(0 "$ GA N 03 G1 N H i-I "'7` C. ) c" fV to afl CO to M N N I` ýO N- N to to [ý- N lG
1 cil C'ý1 ý
ý v
ý "ý
G . -t u N u
W Q O O
00 11 ( (ý
N I
a ' ü
%D o 8 `ý
tD -4
co t E-4
4 ID
117
\ý
r! /ý v
`-4 T",
H
S H C d
ý f" i
V
ýi
C'
iW
r
c, I E-ý
G
Q
:u 0
H
W
i LI) C)
: r, VD c
co cl t ý; M C
ii ri ýý rJ rJ cý c) c -t r; rý r"
ý" n C) 6ý } p () In O C'7 D) C"
(n H r-1 HI r-4 O r-A
Al C; H
C H
Gl
.
Qý 1
1
_ c; 1 C3 O N Cl O fý r-t i` C"1
-.. F-. !
rT in M C [') t. 'M r) ("ý f. 7 fß 7
}""r. I
c= '"---+ t-ý
h N
r-I
h ý7
r-i . -i
O O
ri -4
. -! Ol
O
r-i
«J Co
r--! r-1
O in
-! H
"'D co
O
. -1
M Cv
r-i H
C Cl
H
C)
H
Cl Co
M in
O G1
U) }tt
N} LI) M
to ()
N t1
c` O N
U) (N
ON
cJ
I-
V`
C)
r1 (
r-
M
lfl
v
1-1
t
. -i
V' f
in
Cti
CO
in
M If) Co N tN N . --! M N d' If) () in in in (ft in C, Ul U) (ft L'l
(N r-
01 co
N in
L") M
O O
O' N
(1) I- I, -
cU O
CO N co 6
N N N N M N (N N M N N
0.1 C:
I- CO
N N
O O
O in
c' Cc
N O
%D H
H H
N CO
(Y) LI)
(N CO
Q 0. O r1 N N N -{ . -i y
Q1 . -1 0) M N C` M co Ln co in tý fý CO Co 61 O1 N lJ CO I- Q
N N N N N N N N N N N
t- N
in to
m N
H O
H O
m C) , -1 O
In O
v to
h 11
m O
w1 M M .r.
m r
m i- i
f r1 rI
w. . -i
M '1 'A
M r-i
M r-i
M '--0
M rH
`. t I! ) in
N to
C ' (V
N t-
(N I-
In -4
O l0
-tv to fV
ko ' -i
Cl N
(13 c' N H N in (N . -1 in N
1{ M (V
N- ý4i
N . --1
ý4 N
'V in
t- al
u1 CO
o a'
t. I M
M C)
CO O
07 C' r) N M t'') cV . -i N M M
.4 m (7% r-i
M (V
N Co
't ul
. -4 in r -A r-I
O O
c" U1
r) N N
CA M f7 M M M M M M (l7 ('7 (l7
V. Ri V' Ol
N N
CO q
H t-
l0 O
in in
t- ßl
N CO
C)
. Ol
CO M
W :,
A t- cz CO 01 r-: Cp O
Q C IN
c' tf)
oll ri
O O
CO tip : f)
. -i t"')
O 'i
-q M
M "
< M t'7 c"7 m m m ('7 c7 M M M
co q: v 61 cN N Q1 CO N r- O r-i O km N tom- N N o
'H
in C'. in C1 Cl 61 co* 1l 0l Cl 0)
C' M . -ý O
l0 %D
C, O
M l0
CD N .H Co
CO r1
ul (C)
v` U)
ko ;: v
N (V r-1 N M ('4 "--!
N r- Lo
co -4 -4 t
C4 `
co r-
t- C)
N co
I- oll
00 rl
. -1 CV
C' C C') L7 LI) U') '
C C' 'C
F] >+ ä >+ >-I a
Ü Ü Q o o k
C! ) N v' ý !
to to N
in C
t+ l7
C) LI)
Ei in co rl
}. 0
6 H 1-4
x
1111
2 Ca
in
z r
ý5
Ii
0 ý` M r-i CV c' O . -i CO u; c- n tn ,n co r r- cr. f rn i ' N :' N N cam' N (mot iJ iv cam; r.
Gi CO N N . -t cý CO . -i f O CO N a C CO N Cý t9 CO
to C1 CO CO CO W Cý C0 q CC os q
p q V7 ýv rA Lrl M LO Cam) %, O M c:
r' FCC M M M M' M M M t^1 t; M ti
N CS. r4 N CO C tD (N N CO O 61 Cl ri lý CO C7 N -N M CN
Cl t`- tý CO Iý r- c CO O) CC CO
n tom) O CO CO t11 "-4 N cl O CO
CO CO Cý N M M V' Gl If) V- N
N CO -i 61 CO N to r-i 4 CO CO 14 CO O CJ O c-, lf) )`N to CO O OJ
th M .
N .
r1 .
M .
M .
N .
N .
N .
M .
N .
c r- "
r- r- -P M N r- '
CO to CD to U ) to CO N to l0 ýY lJ l0 l9
N N N N N CR N N N
t9 N N CO C) "; zr O O 7 M r-(
_
to (71 N N )- to N to N . -1 3
-1 CO N N CO N N O C) CD 0 CO
*
r-4 lqr O lO CO CO co CO N to N
Q CO M M C. <. ' CO tU fit' 3 r- L f)
a, m (q m M M N M C'1 M M M
CO to
CO C
N )n
O (
to O
NY' r-1
M O
tO ). f)
N V7
M C
O Ol
Q
. 2 00 r r- r- N N C C) N c- r- : U1 N ' N N CO C)
1 IV d to m
Ui Lfi . z"
N N N ri 10 CO LO CIA . -i CO N
CO lqr O " a) CO OD (Y) ý 'IQ, CN
14 -q , N ' Vi l I
N 0) t- -4 C) U) CO M V M N N 0) M N
N ä tN
ue co n co O co In s -
N N N C tV N N N N N
Ly ýj ct' co CO CJ . -t [V N r"i cl CO
O "Cr O N (N N ON a% (V D
ß1 Oý týU 40 Ol. . 01 00 c G'; G1 C6
rl N N -_-
'd' -: Is
co O
N U)
01 U)
a% C
co N
O t- r-1
rd) N try
Q M C') eß'1 M CO CO CO r') M M CO
co U') -f CO to r-q m F )m
to d'
c, CO CO Cl
. t+')
O )t) )-
N Q1 N N Ln
, co '00 .3 0. ý 0ý Q' c as 00
r-i XD C') CO O . -4 co U) C+ tO CO C) lO O l0 N CO c') CO CO -rr r4 `; r N r-) ri (N M (N CO M
t "
C) . -i 0 O 'i' to CO (r) CO O if ) M N N
14 (11 .
N .
N .
9-4 .
N . .
N
. N
. N
. . .
is W ru
U U CQ cl ý d N2, C' m ýq %D co
n n o L ) t to to . co
N eT %D (n
119
)
M r4
90
x v)
co p
q
rz R
ö
U H
8
z a
w
CO N .O
M C) C' en
ee
M M M M r r') M (") i'1 ii rte;
' Ul N
- c-
ri Li
. -i W''
C7 (n
' to -V
. fl r-t
- L:
" C"
b'+ . ý1 O C) CU CS r-; O ß1 QS rJ O CO
L) r3 l0 lC co co o r c, 3 - Lii Q Lil lJ C" t-O U) 'tS` C`. ýn *f)
«-3 ti r' Iý M O O to co Ln N
M r'l
Iý N
"c' CO
3++ Q (N M N rt 14 N N .-1 ý r-4
Z CO n Cl G N
t! 1 to -t N CT) C) 11) N
r^
fNýl rr C)
SO " N N NNE U-) O M M r-1 r-t M M r-1 .
(N N
N r''3 N
O 110 ý
to N m r14 M om '
M 1 to ON r- V l0 lý
q U? M M cr) M m ei ri Cr) c- M Cr;
N O
$0 Cl
N- ri
Co º.
N- 1'
lO Cl., . -1
L1 . -t ll
co N ,
CJ 2 C> Ch cl CN r-4 C) CS O O 01
ý-+ r-i ý"-1 ri
9 O W
ßl r-
(n N
t") U7
m Oa
%. 0 W
C) Cl)
. -i M
[-- i- G'l N Ln C-
W Q VJ M tý M t7 M ýýT M f`) fy; (4 M
y C14 -W m 1-4 rý co 114V to C%3 f- % o sL' L P" dl a1 ti) N N O1 00 co , d'
4 r4 r4 r4 M N r4 1- 4
r4
Z ý
Zn N N N LO
r-i c- "o
N N . N N '' N r--i N Cr) N . -t Cr) N
v' M .i "d' l0 tý . -1 Cý M lfl }4 r 00 CO .N O LO Cl) M CO m O Cr) .
Cr) .
N'-O .. .
Cl) S
N N M M M
Li co ý-l
(A CO in
-4 U) C4 N rOi in
m C) N 110 Ics, r
* r- CO t y-
uj C4 M "M c4 .4 C' C v) ch M M
", ak 'm N -4 O tf1 N CO CO . -d 151 4 -V L ý N N tfý Ln N t- N
h e-t r-I rt r4 rf 4-1 ra O p
i .'
cC O to
M M
CS M M
N tn (In l : " "
ci' m . (`1 M Cr) Cr) Cr) Cr) Cr) C7 M Cr (Ml
co CD lqr . 00
N rt r-j N N N -i r4 ' . -1 r -i
C) O LD O tO N co co
Cr) l-r Ln l0
N d' N . r-1 . -i N CI) N Ln M C)
, -4 cq r-i r-4 h S ß ý C) 00 cn CO Cr N r-4 c i r -1 t N lD "--l $4 N N r4 N N r-i O . -r N N
` U U Ä ° r ` c w L o ca c
H
120
U)
0
Q
i:. Q
W it 3, 'sý' R. a
w
W
S
Q CO Ln 00 Ln ý j - CO ` y
N Cý: C3 (N M M tV N Cr) r3 LV
n ri . -1 N N O H v'1 CO Cr) M Gl v M ýr
. Cr) ) 0)
u) `' rn ö ri r{
ö c ö ö c cd
J) C) r-I -4 O CO CN (N CV Ol C rH N (ý N ch M M N M fr3 r1 cat rZ rý M
N r-t %D "-t v0 Cr CO w co L! ) 2 G; N Ol co N O 'SO r--i Kr 00 01 (N
"; Q O O r-! N - O rd r{ O O ra
e-i r-i ri º-ý rl r1 . -I r-! r-i rý ri
G1 m O Lr) LA ul . -! cV m Cl BIS 0.7 (A 0% C- P1 Cr) -" m Lr) Cr N -P CO V' -i M N C' N to
-" N O 00 N CO M tO lO N ON $4 N
N .
r-4 N
.
M N
.
N O
.
0 r-1
ON P`1
CV C'1
.
ON H
.
H N
.
C> M
.
.4 N
.
M- to . -1 O ß1 co G1 00 N tff CO CO to to t1 0 C LA tO O r- l0
N N (N N N N N M N N
(j jLý to O CO H C) 'q' r-i N M tO
a 1' O rn N to u1 ra to o
H 0 H c) C; H H r -1 r -ß H H H H "-1-1
N ýa aý +n c in m ON . -i rn Ä ri C) O CO H N H at H r^{ r-1 M (4 M N M M 14 N M M Cry
O r- sý
CO in
rl CO
r! LA
M N
ýi' 10
M O
N N
C' to
00 (1
rI ri CV N r-4 r-i r-i r-1 r-i r-i r-H r-i H r-i H H H r-i
to N --" N N (A C 4" cP to O'% U,
_
N H (V
M N
(N M N
O co N M to (%A N to . --1 N O M N H N
N C) W N to N
M Gl
-I N N .
N .
rl . .
() .
N .
r-i .
N . . .
sT d' M N N Q ON co C31 O 01 O\ N M O CO
U) N N N N M N N <'t C4 N
U) O N M C) co C) U) C- N ( eA N
. N rn C " N H co
C) C) C) 0 M 81 lO V7 N 4
" 01 C3 CO r-I r i CO OD Ö
Ö
N " M " N N M M M C'4 nj M . th.
( CO %D ON CC) l
in 00 '
(7% O C) lx) Q (V N N L O -G CO
. ü1 C tQ
Ol O r4
ri ri
r4 ri
O ri
O ri
G) ri
O ri
_^ - ..
O . -i
O t-
C rý
. 7. C H to M (A O . -1 co ._
U) %r lG (A O tO O %D N CO M. M to *4 Al C' N ri '"'1 N M (NJ H C) M
O to ri U) (A O U) ON oD )f) u) . -0' N
tC r-4
O N
LO O H
O ',; r
CO . -i
H N
0' H
c7 N
th N
dl ý V V
Cl 9
a sr sr yý ' co
u j om tn l l H tD co ri ' l D U]
121-
x
) w
Z-
uk
p D
Q
W P4
ý LO) H a
Hý 8 w
" co GJ Ln r- cc r- in N- Cv +^J ri
LF r--
p N
. r-i
gyn. " C)
Z., f
(N CV C3 (4 M M N Cat Ci (1! N
r-i H N N Co ri . ̂"1 CO -4 vs M 1 (, y In v" C" a% G1 t; D re) Q' M M M
LA a Co ra r-H Co C' Ca Co O O CD
.. ... ... ._ ._ F f7 Cl C. (N co O {` r C C) v) [! ri ri Co CO CV C14 N Cl C ri N
rs V? ( C"1 (Ii M N M rn M N M rr) M
[-" rt vý -4 -. O C) Iýw CO to co LA .j
Z;; c N ON co N O w r-i "zr co V7 (N L fU O O r! N H 6
-4 H O O H . c. rH r-1 H H r-4 H H r-i r-i . -i
% c m Q N 10 in N at r
Co Hfl CO LA O N- CO M a\ L ) CN N Cr CO --S' H CO C t' v" (N to to
"
(. N O CO N CO CO to to N ON
I{I{ N N
C-4 N
C) N
N O
01 r-i
ß\ CO
N N
o H
H N
O M
H CN
M in e-i 0 c71 Co Gl Co N t! 1 co Cfl tD to If) rn G1 Ul l0 O 1ý tD
N N (N IN N N N N M N N
Ln Co co 4 O "Cf' r-i (N e2' CO to
H H H r0-i r-4 H r-i . O4
r0-4 rH-i
A lam- ri
l0 O
.
Ol O
.
1n co
.
H C-4
"
O N
U) r{
t31 (n
.
rlti rl
. -4 r-{
O\ rl
U3 M M M N CO M r`'n N M M M
(D ßr1
H CO
H in
M N
C' CO
fM C)
N- N
C' tO
Co ü1
A .4 CV N A 14 A N r-1 ri ri
Z tn Ln
N 111
C' (V
N N-
N r-
Ln s-S
C %J
Ilzr to C, 4
to H
Ch - r
.r r1 ( cn N . -i C) (V
O co N M LO lD 'N c, to H (v
Cpl .
N .
N
.
H
.
N
. (n
.
N
. rr- i
.
En
. N
M
. r-4
.
Q Ol Co DO 0) XT ) 0) N C''1 00) CN- CO
Cl C)
%D co
CA N N N N M M N N M M N
N O N- M O CO l1
(7% Ul N N- 1
M M N N CA (n w
; N lqp . - CO
; C C G= M
CA C7) O
to Cý
V co C3 rH1 "i Co
m Ö Ö
N M N N M M (r1 Cpl N M . C+').
cn 00 to a CO LA co '
C'. Co C) '
to
4; v,! t3! N N N- O C CO
. U) mot W
" m O 14 4 O O G; O O O
W M C O . -4 t0 t!? '4' '0
C) N
O d'
to N
O e-1
a ri
(N N
co f"1
m N
M . -i
to M
lqt' C`)
O to ri tf LA O tt C\ CA In Ln %T N
to CA .O N
in O
to H
O V
Co r-i
H CV
CSl rl
M N
C') N
Q 1-4 U
cv
V Ä ý' Cl *9 "4r ý co 8 co , -i N c ýo in El) H
121
/
)'. 0
ui V4
O
c2
N
I,
'
CO N N
C1 co q lýD z
,J -- I , V' N O it "Y' .c N. M N -,
roc 41o c . -i (1) CO C: ) 14 r-4 co H H H O O O H H H r4 O
Lo H H H H-i H r"t r-i r-1 H ri
Ii Co O C' N Lfl V' C N. (N to 'C' U) Q M Co M N N N M r -1 C> \ t''
r LJG 4 v c' V' d CY V d v V C
Q ti Ol Co
N Co
O Co Ol . -{
Ol N
CD Ln
Ci mot'
. -i r `
-c7' ()
O M N V ,
H H . -4 O O O H N H ri H
Q1 ("1 O Co ff) M C
N ON C N Lf)
d' co - H Co N. v N t! ) V
14 CO al
tO H
'ct' N
Co Co
to O
N O
L; l H
Co Co
Co CO
Ui N
00 N
U1 .
l0 .
U .
tD .
lD .
lJ .
l0 .
U) .
to .
lO .
Ln .
Q C') N
Lii to
' tf1
N O ! ll M
Co 'tom
d' M
to co
lfl M
O l7
d' mot' d' LO
N N
Lii If!
tO C1
M 11
M N
N Co Cl
O M r-1 N
tO H
CO ßt1
("') N N -r4 N N N Co N Co N rt r-1 . -4 H H 1-4 -1 H r-i
---- ------ --- C-) C1
(D O
m r-1
a% O ß1
CC) O
C) 01
00 CO
CC) co
M O
Lo O
(Y; ýy d. Vi
N O
N ..
to ýi'
CO l0
G% "
to al
co GO
N t11
Ol Co
H O
CO V'
M N N H H H N Co cV N N r-H ri H ri H H -/ r-1
Ln Co
N Co
-T' N
N N
N N-
Co
r-i c,
t. 0 cv
in c' (
lC H
a%
N M N -4 N
Lii O
N O
C1 Co
('4 N
M Lii
N b
O N-
Cil ýS'
to Co
to W
M M
Ul tD Ul kil Lo \O 111 U") l0 to tf?
Q lO co
Co d'
O Co
T to
N [`N
Q' to
Co M
C) to
N t.! )
rn to
N d'
ýr ýr a dý
L=. N CO
d' m
N Ul
lD r-1
H cl'
O rf
to M
Co . -I
N l0
C ý2'
Co dý
w %" 431 Cl C1 C1 ON c c1 O d1 C% CA
M O " C-i n CO c) r- i r- Ln G1 ( 4 ( O Co H G l Co H
" "' co V" sip d' qw Cr d" f c
y O t0
N H
C H
Wo Co
Co "-i
U1 O
O N
M Co
O Co
d` a1
LO c')
Q ý O' - 0' C7 ý + o o C) C1 Ql O .4 O %
. ý+ C; r M . r-1
O to to
M Cl
Co tO
O N
f 00
oo (-)
ul M
d' U)
to e7'
N 'a' N H N Co N . -"i Co Co
CO 01
w to N
CO N
ON fV in
Q1 Ln
N rr co
r-! (I) Ln
14 LO .
u) . Ln .
in . Ln
. . . Lr) . .
Ln
. t
.
}ý Gý . -t N H s7" 'd' C
9 ko CO
ý ül t ' Lr, 8 H li ,O O) U r-4 N n st' t0 N
122
-3
T-4 1110 101
ti Q
i-I
r
r
w
C7
ä
UJ
H
r_ p
b A
+ý
Q
r
b
Cý
ä ý3
-4 Cl r- tC) C O CO in
Cý N rl 14 N N M cýl
Q ýý lJ ý1 t0 ()
-. Co co r 01 co
r yý Q r-4 in G1 r M Ol '-i C) r! ko
N I 'ý G1 co t- r- Gl w CO CO CO co 00
<-'l fV in N- in CN in r--1 in r-i w M o c t7 LJ7 lß"1 '-t ýI' C N
w tt- N C) in N CO C' ' c-. M Lr) L-. O
O Q' r in O C'i () i N -1 QL:
O M CO r oO 41 C) ON CT; C r-i
r N co in CO
Ql r L ? -!. in
H G1
' Lr)
O N-
+n : ý:
"" c-J t 1 r)
I, - N- N N -r c, G1 in in ri O to to co to H CA ON C) r d' lii N N O H M M r-1 N H N N
H -. f' H "W CO H CN ;W m C) <D CO in C). co Ol
M N N N M N tV N in N N
gy N . -1 1-4 tý) 1-4 M C O H r-4 CO a sý
to W r-1 d' i in C) Ol O :
14' N
ºý+ " co co Cp. al co of r- r- .0; U)
-W O' in H r in CD VW O N N r . -1 6A r to M b O 00. M r-1 O
M N N N M C+ M C' M C4
y H O M in O U) N H t0 N r-1 i^=. ý
in 01 tD tO 01 0p N Li) O M O }ý
+ a% r N- tO N- t- OD co CC) co CO
in N N N in C) 'C. cr ýO 61 T ul in cri r r H ko U) N . -i N N in (N . -t t r5 N
in in in CV M o0 in t- r-S in 0% CO -4 M M N N
. . º-i
. . . . . rA
.
Ö
. . N
N
.
N co co '9' ) lo " (V al tO in d' N co in rl O O
in N N (V M M C'V N M M M
-rr c: > cl, M 00 C, 4 -4 in th lD t0 N ON co Ol M ' t' M
QQ a1 00 t- t- 01 a1 QO Op 00 co co
A ri N Ol N N r-4 O to M (VI N M M C4 M ch
M , I* N N- H in
O M
0' 1
O M in (n Co N C 4 N in . -i 5%
i ö cri rn ö ö ö ý- 1-4 -4
ON 0 -4 r-i
rn
i 9' -tzr H Cl) C U H co U) M O t O to N co, C'-) M ºn
N H H N M N H M M
. - M O M in in W -t N m r4
- Co N co H ý N
.
N
. O
. r-
. . . r -I
.
N
. -t . .
N
.
N U tU UUU W
ý W '
Oj q N (r]
C) C> N
o ) U) in t in r-i N l2' W U)
123
)
r-
ix.
94
A
Cý Q
A
rz-:
F r . -t ýý
ä x
w
Cf 1 {M M tr M M t"a f'} M
i o co gyn. co N o o LO C) o cv
ý! 4 ý"t 'C Lri M Ln LI; in l7 LI)
UN . z. CO CC) uö c, CO cc m to Ci co LZ
N ttl CO (V 01% O N O 'd' -a' Gl ^ to <f' N M iý l0 ST C Sý Ui c")
x N Cl CD M M M t-) M M CD CD CD
^i N r-A r- N ^- N N (4 N llýr "+ Q4 - in U) [`N N LA CM in O CD
C3 . -i -i N r4 rt r_4 r^f N r--1 r-i r-i 1 -4 r-1 H H H H r-t H
Gil t'7 C Lrl ; J) If) . --1 N .T d 1-1) c0 in Gl ['- " rI) C mot' m LO tN (N
ý, _ (") t' (N w N l0 \D
! ") H M O co -' U'1 N co CO al $t Cl Lo 0% co \U VA N ON 0 -4 -4
ý'1 .
C .
r*1 .
N .
ýf` .
V .
ý7' .
N . . . .
0) N in CO M CO OJ lid r-1 C7 to A 0% N O O '. ' : 'I 0) 0) H CV O
N M M M M (11 N N r'7 l) f")
Co N CN [I) r
In N
N O
m N
r-# 0
CO GU
U) M
t0 N 1
CO CO C0 Q) CD oQ o? 0) CJ CO q
(tý -W O O CO N N r-1 Ot l0 C) N U) L') CD v'' d1 tt) M to CO "C *`
O M (M D cD) M to M M M M CD
Q (I) Ov O N (7 00 to ON %D N l0 lU
'C in CO . -1 N . -i in l) r1 d Ö
(ý O ri r N H H ri H .-! i _i 4
. e. Zr) N Q' N N LO C) l0 C% LID LID N N l'- H %D In N H N N C' tV r-1 (N M 'N H M N
M 00 co [' "Cr l0 1-1 '4t: r rýr O CD N co CO O ui M H o". co C' %D M cf' I! CM LA ti' C' N LA C M
r-I %D 0) C) UO N 0% N 'O 110 lO
. -i M C' -w Q7 V' H e-1 N N (14
Cý f; C14 114 C4 C4 C4 C' H
M Rir 0X
CO lD
if) 11)
C) 00
N 0%
01 lC
1-4 W
01 H
N 00
C) r
CD co a7 . 01 W OJ CO 00 01 06 C0
Q Co to U1 C co 01 to . -4 q r 4n N M Ln ko Ln H in %JD m M M M M C4 r'; (VI C4 C4 Cý C4 CN in
h ao
N CO
N t-
) CO u"1 to
03 u1
a1 c7
u) N
W to
P, .9 C, Cý -4 4 r4
d' H lQ M co O H CC) '
LID d' lD (ID O %0 4 to (N CO f 7 M to `cr N l7 N i-I H N M N H M M
c l0 M Z CO co ý L .0 C-4 0, 0 O al 0) L f)
'i Cý
ý}a6i ý }nýjj
ý ý ý"-Y y}
rý""
` r-t U cV t a ý` 10 C' ýr >4 U o
#, o CO g
i I `_ Li)
vt un L n
8 to CO
r-I N "G' %D U)
124
ýr,, ý: ýJ
N
l
. -1
ti x :n
ate.
fl
64
ti
cv)
G1 UC) F C)
ýT' M rl
(N C
M tý tý
c ýT M ýJ ,
H) P--3 LI) r) M rl c) r
C) O N
N LI
Ol (N
C) N
C\ GJ
O t0
lD -1
N LI)
fam. l0
M co
O t ri
. -. N -4
M N -4
r; -1
-4 -1
ri H H H .4
. --i r -t
L^. q M .I Ol co lý O r-{ (N [) fv
Ef) M M N N M M: M M M C) M
^ lD tD N
ßl 'IV
C' O co q
O N
M ri
M (
r-- N
J 1 4 N
W O N M N N N N ' r-i N N
1 Z ON co
M LA
O Gl
1-11 N-
_4
f) M
fn Li) M
-4 d"
N C
Ol to
C) I-
tf) N
.4 N- v
ß1 N
M -4
N kD
M d'
61 Ul
d1 01
C to N CO
r- N C v ý' M l0 Hfl M V' U' to
Q to '' Cl CD M' 00 O N r1 r 1 0 0 M
q U) M m co (V m M M M
m rr) M
G1 to [- .
N r- N N CD CN co ti l0 N N r-I N Li O CC) r-1 O
G+ (n .4 N M H H . -i O -I ri 4
,A .4 . -i .4 H H r-I r-i . -i r-I r-i
t0 I- co a'. to N CO l0 V` (V C) " Q CV .
0) .
IN .
C N 0'. O M N O ýy I
Q ch N N N .
M M N M. C. j
+ c""1 ý
r! 00 N
ul CO ý n oll ý
li'
w O . -i
-4 r-1
M M N . -{ r-1 r-i N
Z LO U7 C-4 N N
.4 t. 0 U) ('4 -O CN
N ýi' N H N M (V .4
M (N
34 N 00
N N
N N
N ,
N ßl
lo C)
O C)
Cr) V'
N N
a1 a%
03 Cl
cn .
-i' .
M .
d .
r .
In .
M .
v .
C .
M .
ýS' .
6 tLi c(N
N rr-4 9 to
M M M C
6 C4 M th M M M M M M ; M
fü L~ti Ln to (7, r4 co
N co "0 N ýi L) C) O
W N
C) rt
. O 1 N C4 N N .4 C4 C4 Cý W . I r .4 . -4 r-1 -1 r -1 , .1 H V-4 -4
A M C14 N
(14 O
rn a% n
El- '. O
co O
m N ýMj
-4 M
U) r7 t_; (4 N M M M Cß'1 C4 M
N O
co tD
LA N
(0 U1
(") r-S
O f,
I- fý
N U) ()
N C'.
r. ) , ( -: Cý h N th N
cn N C) "W to N .4 1.0 -A " N CO M cy) CV
M
-4 n M (1) ('ý1
:J M N N M O O u1 (- M r-
tf) to V' M N ko w mot' to t! 1 ill
Uýl ý 'ý ý ?+ flT CQ r-t N W
ý ý
Ä ýr ý ýý-1
ý a U U G7 w co (%j ICV to 0
1ý + Q lý rt rt
,n v Ell H co
)
.,
H
1ý + ( ý
ä" ý
cr
r
%-4 X N 14
£+ 8
Cý ~ r"ý
U r3
ti s-ý
w i
ul
Ott E0
2
1 $4 1ý
a
w
[i AN
q
iý m 2, -9
z
$-i
(D -
1-0 t Co , --; 1- t: . -i ci
C') 00 C ^' C'S LV (*) CJ
MNO
. -i -4 -4 '--!
-4 O [- ON. Cl c`1 C) CO
O ON w cr. G1 _j CO
c') MN
ON o 'LO (--; in Qi i c3' CO '1' --A
l! 7 CQ 01 E*- C\
lD lfl .Oi. '
O V' r1 N
rt tn to o0 4 4 4 4
M O . -i ;: f' Li 1O 03 Ol
U) C M -i F-4 r-i
to C N i M ttl lfl Iý
O to M CO . --4 fV
S C Ifl
. If)
. ýP M N
. --I
t11 CV ý"' N Lf1 U) N N N C N
I- O C") CO co O c') '9 mot' LO to li 1
r-I . Cw) O C O Cl c!
L! 4
hfl fit'
M O O IN O 00 N O
C) r-t to to r-{ O OJ G) .
. -i -co 1fß M 0) Ol N
ri rý r-1 ri
e1ý r4 l7 M Cl) Cl to O N C N r-1
tn r o co 0
w 1- to to
2 2 2 2 a cr "; r N V ýO
I # u'1 ul N
r-i N "; r to
3 lÖ C' ) ý' C C) X31
Mv cý n N f--i CC)
CV C\: rl M 1 . -i . --1 -4
Lr) M co N 0
Ln 4c4
if Cl rý N
NNM -4 r-1 r-t . -4
LO 1. -1 -4 " fý) MC Ol
N C\ to N G1 Q' to M
tT OM al N r- 'c1' CYl 4
'4 fit' 4
[") Ur lJ .
M cr ýt) H r-I H r-i
N Cd aN C *% UN
CD tl- 1: 31 03 DO d' Cil
CV Ui OC N . -i to U'1 . -i N (1) lei
LO 07 [N c7' Nm two t! 1 '0 Ul Ui ! f)
M . -+ "r o 'O o CO r1 ºn v. ui
t- rn O cr1 ý-1 MN
CO .-M fV O 0) a1
CO NN CV W CO 41 Cp
r-! 4
C> OD t0 N 00 M r-i C'S (Y) N
I ß1 mot' N
Iý l0 lD \D
Ü Ü Q
8 U)
12E
ý? G ri
N C} Cv . --1 , --f . --i
Cl Cpl lý CD G : -
Ll v; ": z
ON C", r- o co rri ýýi cv
c Co in in r- N CV lD .D
C Co C' al lD C lfl tD o
co LO ýT ko
Ln r1--
110
444 - CC)
ýr ýr cri
C--, en U) ra ýý Un aý ýr ri N O' Ol
ri 1 r1
ý: ' lD Ol
OGN cn ILO lJ Ul U')
O r-i O
U7 II) Lf)
O co O U,
e-1 N +-1
co Mq
dý to dý
MO U) ill C) O
N ri C, r-; r4 . -1
Ln 'KV t. 0 M LO d' r-I MM
U) CN
C> C: ) ri -, A r-li -A Hw co
lb
APPENDIX 12
ITEMANALYSISDATA0NTHEHSPQ
127
'0 2I Z
NN J
CD
ä
0 U-
J
a 6 22 w a J
O N Tq
W
H
* 7_ <
WS Q 2 1- < - üc of (1) -<H W- O - M O - I cV 00 - - In In 'CT ON In N - O ý- 00
-M of z ci L. 1- - OX to
2W
1ý v ao p r O r N in r ýr ýt - ko o) p-, co "t 00 0'
I0' IO N
I M r
I r M M
I IN 1°' M %0 N r
ý J Xt
I 0
Cy O O M N K) 0 (N - O cV - 0 M O O O -
. qr co co C, 4 0 14'% I I N 1 I- - I I 1 I O
I I I
d - O N N N - N O O (N N - N 0 O
U) (14 It'l EK) Iin '
M rv
«) %0
20 I ao o
I p r
Iý O I j q% Orn Nr
r jr- I- Ißt W GO. N 0 O O O
N _ ö O
J () M M til ýt M ON r- r- in %b %0 /p O% O ýD cf N in CO U) O M
Itfl I Iv IN r %0
I I 00 Ln
'rrl in
IO Orn
Ir It11 N
IO In r
I O O - N N O O O O O N -
p O Z
M r Oý p N r- M to " CO u"a M - %0 (V - O Cc Z -ý
I I I p I J I
N I Icv M O O O - O O O O O O 0O O M -
0 -
W lo ýp IIn -
tin in I Oý N
O I r Iý lc r O\ 'ct N1 Q% I M - 'D O I '. O - 0D O'.
o O O M - O O O - O O
IO O
M N in pN '. O - º^ M 0 M ON u1 N M Cp t+1 %0 0 3 N Irr R- I IN I IN I I I
- O - N O - O - - O M O Ö
N - L J O CC) N ul N N In n %0 O' M t0 M 0 O CO CO a O
Iý N -gr
I 00 M r-
I O n Ln
I - 01)
IN %D
I- N) ý - 00
I- I
O - O N O M O O N - O O - U. j V)
W ON pr) - rn 0
I- - 00 I N N CO WN
I I- "ýt
I 9: r U'%
I C%
I (7%
I trN I IN 04
ca Ö
O O N 1- O O O O O - O N
O O - w
w N
' v I 1ý Cl) in fit r - 1t 0D IÖ 0
00 00 %0
I %0 (N
I r 00
N Ul
i(V -
tV N IN N I O O N - O 10 O O O O O - O IO
o N 1
O O M I I I
r- Nr I
r- (N 1 Iý CO
r- I
ko n I
N 0 Ü O N - N N N 0 N N O
N -
p !M
_ O V
N Lrl I ý Iý O
I I I r
I : zr 04
I O N O 0 M N - - ct O - IN IN O
0% wi CO Ö
I0 N # o 0
co 0 to
I CD 0
V) I 2 I l O 0 0 O O i0
0 O
0 f-
N Ö o
M
C14 I I I 0 I0 t- M
I r i jLj-j I IN I0 I 0 fn I O O O O O O M O N 0
O W
a J W Q
ONJ Gý W-Q c/)
Öa Z '0 O O "D O '. O - M '. O v C)N (yý
LY L M
N - N - 0
N O N N
00 N
In Ö
O cn Of x-
p M N - - - - - S U) S U
OQ .i Ou Q
Ü Ui-Ui N
5 Z N ;t Cc v' to 0 N ON a% (0 N '. O 0 N N 0 c% M %D M ul Öm g L. r O O in 'i 4 r, a 00 r, 0 N U-% O N - M N O N O N - - - - - -
Üýö -
CL 3
W OD OD tt in (X) 'Kt in cT ON 0 CO u1 0 0) N Il In CO 0 0 O in in Kl d C in In In 'd' M 't in 143. Ill vt In cY ct V . Zt %0
x W6 1-
N Nl 'q7 if) %0 IN 0C) ON 0 -
- N M IT In %D IN OD ON O - tV - - - - N N N
128
Z ...
CD Q
w
Z- 27 -1 oQ
1) w
r F- J 4
iQ
W 0.
J
'r 0 6
W J Co 4 H
*Z cr QJ
:uMQ U' I- -Q 1- W
N - N O - - d N N O O O O - N N - - V U-) O
ZýZtn L - }- -
OXm zw
D N N N N n 00 M CO CO t- cr V 00 M ' 1- co 0' I I I IN I %D N
I I I 0
C>1 O O O O O O - O - N O O O N
111) n q 0 N N 0' N O M -t ü1 "O t[l Iý M 0 rn
I o D 'O
r UD
ýin co N i-
tO ý Iv I'
III I"
d Io O O O N N 10 O N O O 0 N I p Q
ýf tN - N Ul 4 O M 'O Nl N N ;; t N ON C) 10 CO CO Cl) N V
I N CO il Kl 1.0 in O'
I 0 M 0\ \O ül
I 'O
1 CO y- -
J d - 0 0 0 10 0 I- 0 - O O - 0 0 O - - O O
C) 1- N "D s O M 00 oo r n O L(1 ct co - N rr i N r, O N N M
I kO
I O t0
I l N O CO ct
I I` - N N
I %D (N
j f- -
O O - - M O - - - N O - - - N - N O 0 Z
r- 0 f` M in - - d to _ N ýf w) N n (-4 M z M rn 8 -
I Ö p C14 M Q 00 4- K)
I 1 Ict LA
- O 0 Q - O O O w
- i N co N zO C) M p' in M Kl O' ýt 0 - N
I ( I
U'N I O 0
Iý IM N teý %a V)
I ý p
I
n N O O N O J - O - O O O O
;; ý S (7%
I CO
I 1 I IOS I I ýO
IM IN rn 0
- O N - O O O N p ý
L J 5 0 in ON f-- N - N CO CO CC) OD N - N 0 O n Orn rt < M N U) o0
I M ýr
( O O M O O
I rn ' d %T
1 N U1
I -
N -ý O O O O M - O N O M N 0 cr- w N Q N in N M in CO _ in Q (7\ o\ U1 . i- M co oo in
U- IN C N ýt N ý1'N N M t[, \ tT (
N 1 -
I N q: r
I 00
I CO N
l M
co i0 O O O - O - - - - O O (14 v O O N O O 0 w ( "t W OD CD O - . '. T M Q GAD M CO N - M -zt ýt I
M N J
0 I I"ý
00 NT I I Iý I I I
ö - - O - N O O O O O O
O N (N i r- N IN IN r- in 1! -\ ap O (: )N O c* in ý2
O j C7, CY\ N CA
j I-*- (N til
I - -IT rn
I I 1 -
I 1 = O O O O N (O O O O O O O N N -,. T N O V O "It
U I Iý L
%D C) - O I I0 I I ! I (i
N O O O O - O fle) O O - N i O
co n (X) in i- T d - in ýo M M P- M
CD M l I I I I- I I
0 I- IM I
O 8 O O O O O O O O O O
0\ co rn - N CO in -; t n ýt in 1 - in in O rn
Q l I 0 O I
C, 4 I
0 N I
- 0 ä ö 0 0 i0 - ý - - ! 1* - - 0
w Q J Ü
0N -J w - 4 u g
w is
O Z ' O N tom- 0' N of - in m in '0 0' O in p' N `t O V O' M N N M N CN r- M3 Cr 0N 0 O M O M d 0 C `N r-4 N O N
_ ÖÜ ' m x
Cam- 1-- O
pQ J
w- 4 0U)EX v w-w U) G_)
N Z in M CO N Hfl Co in N N ON - CO N in N, - in - [t O 00
L. M M M 0 r ý 0O 0 O
_ O N O N N - M Q NN N N N - - N O
ax i- " OQZ z 0-- F- 00 a 3
F- lZJ Oý i W N +7 N ýh CO
N CO ( CO q: r mot' in in IN M U) in
N-
W Kl et ßc1 % r 0) 0" 0 N Nl 10 un 0 t- 00 0' 0 - N M
O N N N N N (V N M M M M M M M in IN in -Z
129
.. w M
a
z
J
0
0 U-
Z cn
cn W
rv,, t- zo
z !J
0 a W
J w
04 J
mw I--
H-
Q J _ w
S F- <- (D oý O N
Ui - N M d- Oý N O - N - Ln M PC) 0 - -ýf (N 'Z- pr) 0 O 11 ZV) S L 1 - OX u]
Zw
I00 6+ d IN in ul i- a' a0 N N CO N c, 4 CO %. o /O I I- I0 I I 0 0
1 in
1Ö
O1 O IÖ N
- O 0 M O O O O Ö Ö
N n (N CC) n tf) I I IN I I- - I I 1 I- r
I 1
4 O M N O N - O O - N - O O O
N rn d O O 0 0 rn O' C- N O i t' r r ýl V) N 'D %0
I I a.,
I in
( co
I O - Lf'l 0 0 N OA - of O N
fi
J O" O O O O - N M O 0 O - O M - N N O O N - 0
U r (, 4 O n O /o d v r /o IN O (n ýr oo O
Irn 1 O 0
Iv ao
I(N Ir Iw ýr
( N a' O r
I N co
Its Ip f K) O (N O O O N 0 0 0 O M O p
O Z
V a' O in 0 O' M N N o c0 M a) (V rýl tp u\ N a) --, j co
I Pr) CN
I 0
I- I I ,
O !O - O 1 O O O O O 0 - N 0 O 0 - 0 I -
LU O %0 'ch - p\ in t_. O\ N- M 1 O u1 CO O N tO W
- I - l O t 0 IO I- I - - o o o - o to o - - g o in _ - 1ý tfý - O N r Oý V N d (N M d \O O\ O O' d
3 1 - - I -- I I - - f I 1 N In %. 0 l - 0 0 0 O N O O O d 0 0 L
\p co O IN M r M If 'O - O M ý %o a' N 0 w V 00
I I IN I
O I IN I
0 W) in N r - Ln oN O I
d d - M O M - - O - O O N O O N N - - O - - O --
w M i1 1 M 0 ý1 M O D O --t r (V in
U- O to I - d O Ico lqt I Lr, 0' iI , ( ö c I IN ý ý 1 tr% CC) 0 I co I CO O C'4 N N N O 0 N
- N M O
O O Ui cV ! fý Ip o (V "o O ýO c:, m r '. p N r O N try a'
V) w I
N I IÖ IN I
Ö 0 I
Ö I IO IÖ
Ö ý O O - -
Ö - N
Ö
Cl) C) 0
IN Iý r - a,.
IN I 0 CO 'rr
I w'T - (N
I IO C"i O
is
0 p O O - - M N N O N U _ O C11.1 . zr -; r ct
U , ` I
O N r f I I
0 f I I
O N - N O re) O O N (N 0 (N N - O O O
LCI Eft r M M . Ml 1r) r M O\ a CO cr 0 p\ CO , a' 1 I I l I I ý * O O - _ m O O 0 0 N O O O O O O
CO r. GO U'\ I; r
I m I N r-
I ý N
In I_ I ` I_ I I_ qT UN
0 O O N N N O O NJ N C) 0
w O J 0 'j, J
Ü
W-< t/) - V c r Z rn r 0 r 0 v co in r CA CO in OD 0 ao '. p ao tO O r - L to N d - C\ v 0 IT 0 N %0 -t - rn in co o N WO O - N - M N N N N t'1 N N M N N
z 0 co U- F-
O -
J w O< J W-
OtoM U W W (1) R ÜO-
z N in in to M d \O N O d' to to in in in - to in - M WHmL. 3 N- - kO M co O r '. D in OA 00 - to u1 . - (]\ in Ll - 0 M
_ 0 O -- - - O N N - N - O - - N N - N - - (N -- OUz S
_ Vo0
a 3
W N- N N in - N O r in M %p M - 0c) N O\ O r O 'ý F- 1 CD in to tt to in -3' M ýt in to 't u1 in qt In i in 0 OD co
W" d in ý N 0) O' 0 M . in to r op a' O - N M ýt 1- O - v J v v Ln LA Ln in in Ln in in in in tO to w .0
130
M
Z
J
(D
X7 0 U-
FS
ui w M
J QZ
W
N
1- _
v. J Q
*Z <J
W 7Z < 2Q- 0Wa U)
Q F-- W - - 0 - (N - O O 0 N C' 0% ýt O O O M: Of Zcn i- - H- -- OX co ZW
0 O MO - O - O N N N ,; t rf1 , ýp +ý 10 rn N In In of
ýO Kl ct et 0' 0
Is (N
1 I- IO w t-
Ico
O r1 N O O I! M O O O N N N O O 0
- 0 [- M Q% -
I IO i0 N
I I0 I Co In - (O M
I In
I 1 N
1 U-N Ct
It In
I ýp N O N N 0 O O 0 Ill O - N
co -gr M U1 n t0 ON ;; t '7 N I O O ko N N V N 1tß N N N I- S -
I - I- I -zt ON - O% N d N d O 00
(
- N I J O- - O N N N O O O - - O - - - O N O IO
0
U tko CO N - C-4 ON --; I- 00 O %0 co - O ON N c. n et M V) O (I'`l -, t %, o
I d n
1 M
( N
I ,t In N
( co
I
- O\ I i
co 0 ( M N N N O O N O O O N O O O O N U
Lrl (, q 10 o) '*', C)
z U"N I
0 - ýo rn 1
M l
Ö I I
- I I - I
O O O Ö O O O O O O O N - O
W N O II UD ý N d - '° O I C\ C)N (N n . I O
I I I I I I I I I
O 0 - y- - N O 0 O 0 N 0 O O O O S
to o oo n N 1N C' O - %D - Wl ýD In IN In O r- In ()N 3 M: I l IC14 I I I J j
N O - N 0 0 - O - O 0 O - 0 O - O O -
M M K) N In U\ O 1N U*) N In 0 O - C) N IC' I
_ I IO
CO I to it
I u 1 _
I O O
IO 1 O co 0 0 N
I Ißt OD N In N 0 N C) N N O 0 N O O - - N
W V) N C4 %; t In n p n - lL
IN In N
jUlt j 2 ý:
IN i°
I IO N
I IO Ia) p- ff)
IN
CLJ - C14 - - -
O Lij co n C, 4 r) In
W 0 I I I (rl
%D (1 I I I
- (
C14 I
N 0
- N O O O - N N 0 O O N - 0 0 1
, - u*, M '0 N 'O Qrn 00 ' 0 In N O
,O I r-
I
I l I
It p I I I FN
CN I
N r' - - O O O O N O N fn N O 0 O O
O U
I I1
( I I- Lr)
I Ö
Id IO I N
1 ( 10
- I O
O - N O - - N ý
O
co O I
O I O FN I I
- - I- I ln
O O N - O O - O O O O O
n le) cl ; - 0 I O
I I M N
I 1 0
I I N
I I I I 1
O 0 N O 0 O N In - N 0 -
O N
-'
W
W Q 0uJ C) w-< cn UO cr Z ' cT to - -, Zr N co d kO -2' N I 0 In CO 0 'O In p' co
WL 3 'D (T - N N- N CO O M In n C*- r- Pei Cj\ n ON N NJ -tt N O O M N M 0 N Ul N ^1 N 0 0 N C' N 0 M M
e: U- F-
O
-1 W Cl < J W
Ov Ü
W- W V) g ÜON
Kº - t i Z - N N In '0 'O ý 00 0 In % In n ON co t4) 'O 00 kO W 1- n L. '
3 C' M In - - Q% - M In N - hol in N In O' N O\ d' '0 I d O 0 N N N O N M N N M - - O O - - - O N N N - OZ I U
2 F- 0Ö
Q.. 3
W c " F- LU
T N - IT 0' 0 T N
0 In
00 O\ In co (V N In In In ; -t %t N In Nr u1 N I 1.0 'O N In ýY N 'O I -r 'O N '0
W Z In %0
\0 '0
f- '0
CO %
ON 0
0 r
- N- N N
M N
I*' N
I N
% N
N- N
CO o 0 cc)
- 00
N ca
M In o N N co cc) c
131
'0 M
Z
U, J
0
Q
w c
z z 0 F- N w
Q
W N
1- Z
d
J
O NQ
N J
mW
*z a4J WI< I F- Q- 0 ol: of (n F- W O N rn O O O N 0 O M N C' M - to O N rý 0 0 -
OX Q] ZW
fl tý M Oý M M tý Iý - Iý M 'D to r ýt kO (- p I I f t I ` 0
I I I I- ` I
d M O O 10 0 0 INN O 0 N O 1 O O O N rn
N it1 t ßr1 O ýO d ýO ýD n N- r- N to CO ct co M O
I 0'
I N %0
I I - d' N- In
I N d - CD I ct N 0 N n
ý
O" M N M - M N O - - O - - O O O O - - -
CO co co c"j (, 4
1 I- l rn
l I I I I l
W O. c) M - O H i- - 0 0 JO - O N M O 0 J
U In CO in co N O CO M LC\ 00 1ý CO I- 0 -c* 0 O N to U) O in N-1 v
Ißt I O
IN o0
IM (N IQ 9
In IM O' O
I r-
1 IC: )
II I r,
N - O O N 0 N O C O O (N O O N O 0 Z
CO In N I co tp I N M N O N in in I ! kp to
M tpý I 0
I
- O N M O O O O O O N - O O O O J- O O O
I - I I I I I
%. 0 !
O I Io (
O 0 - N - 0 O 0 0 O - O N - N O O 0 I ~
CN --T N- 00 00 N tl Ul Q' M ul 0 O - 00 _
I ( I I- N
IO N 0%
I 1 ko %. D CC) CN O OD
I
N - IN N O O I- N O - to O N - L
Iý v ` ýO M
°ý O
0 U) I.; r
ö tO
° '' ao
8 'D
ö 'O N-
ö 0 lö r
Pr) in
0 O
- }in N) ; 04 - ö O --
cr -
lL M Ö IÖ IN Iý
M C% Iý
O 1 IN
O N Iý IO
Co in U) - tri i o
Ö ö
w co -4T r- r- qzr 0 Cc) co w
I I l I I I IM I-- IO I O O O O O O N O O N O N - O O O
CO r4 M u\ %, 0 M N- d' Kl to - in %o d' to " O
IM I I -
I I I ICO = U
O M O O O - N O O - O - O
O U
ý I IO
0 O I I I I IN I
N N O O N O - O O O O O
r- r- Ln u-1 O g ' N in %D
- N
IO O
I 0 0 N
I 0 a
I I I N
I
, -t l
j 0 - - - 0
I _ N
I I0 I IN I O
IO co CC)
I 0
O - N iO O N - - N O O O O
W O -J W Q
O cn J U W- Q l/)
sý ÜOa Z O' W o' O CO U'% M - ^ M kO in - to O In O' in in N- CO
}-- W L. In qt O\ M
Ot M
N - t
00 N
r M
- "T
0 -
N N
ct O
O O
d O - M -
M - in ý7
cr- = Cl) O M N O - 0 0 M M N
OU cn m
O
J W O Q J W- Q
Ocncr U W-W U) N ÜO
L Z N-
ý ýn pý N- in CO
M - -
- co N
N 00
N- in M - N-
M CO co
N - M 0' It
to -q
in - W F- m 3 t O N1 o - %0 O :f ýp p\ O N N O M M (V N M O O O N O N N N -
U OO
~ d 3
W %0 O\ N O in N- CO O to i - T 00 d- N - in N N
t-- Lei In -t qr qt 4 C: r N in tO Vt N in in 't In 00 in -T \O N tO
Ui O to
c N O 0 - N
+ M '. 4r In to N- co N
O M O
in to o co C D 00 at 0 O rn at at at at at rn O O O O O -Z - - - - - - -
1142
., ýo
Z
N J
C7
ä
0 LL.
W
N
J Qz
N
W IL
in N
S ýp
., -
*Z QJ
WI S l- Q L^ ýý to -Q F- W - CO - qzl- N V u1 0 U -N N u N r - - - N N S 2' Z v) L
F- - OX co ZW
oo - Nil tn
I cr,
l (14
I r- V)
I 0
I V) 1,0
- OD OD (14
I Iý tel
I I
O" i O O - O 0 0 - - O O 0 O N
Lr) l I- ý N
I 1 IN IN O
I0 r
O" O o M - N t 0 Jo - 0 i
V, N r ul C O M N o a' (N - a' M M O O r V)
I I N
, ö v
, co NJ- Ö
I M O
I Id -ý ß
- - O - M J o
J
L) U)
co I r- I 10%%0 ( I I I IO I I I
t-- I
O 104
O - -
ý'`, - O O - O N N
N O O 0 Kl (D Z
O %0 r N O N co rn v . am .,: T Q CT
Z 1N
0 I J
- I I I I I
0 O - O f 4N v 0 - O O - N r'n 0 O
C-4 0 , zr . 0 ON q cc) r re) co I 1 1 (N
I I J I- I J
N N 0 O O 0 0 O - O O - 0 0 O =
- N Uff co N R a' M oo M N n r O r QO t') r "-% N
I O
I Iý I l \O N In N
j ý - In O O
M 0 O 11 º^ o o M O O O M L
co N r cV in "t 00 O t tzr UO U I r Iý
N - r 0 N rt- O O l0 V U)
r IN IN
- J
a' i (N E
- O O - - O M - - O {O M
co
o 1 i !ö 1 \ l jr- 0 - - M m U- - ö l ý j S 0 \. O O iQ (N O ý t! l I d M r r tD ' r r M ýO t0 U) O U) U) N Cl) w N
Iö ö
Iö 1 1ö Iö o
I ö
I k ° 0 o
I H J - - 1 -
o o M Ul 'D M t m rt M O r - u-, U-\
p Ln O p I I Ir
r- i o-. CO
ý T h co co 0\ N Ist in '0 co -r
I rn in II N N O O O O O f M O O D
U _ p M O 00 00 (7, O , 4) ý r- (N r o o V M M t
U ý
CD CO 1 ý
r I
O 1 ý
N ýt N I C+0
t U r
l l M ö
- C14 C14 M O O O C, O d (N O
ý ` 43 a 0 0 N 0 N
- I p p
0 O - O O Q
0 O
t11 r r a rn r V) c, U) - r pp 0
I f M
I U
I - CD r- CD
1 cc)
O N O O O N O O O 0 0 0 N
W Q J LLJ
CU) -J Ü
W-< U) Z , C14 M \O d' UO M d r In - d co U ýY N W
F- WL 3 . OD N
- t0 co d' M O U) 00 - O N - cc) c) o 'C) O cl) O - - - N o
OL`m S U- F- , Cý;
J (L pQ J W- <
Du U W-W N ÜN
'o \0 . 4t in In M M - Oý N N W I-- U) L
2 3
M -
- N O
N In
O\ N N - UO CO V' - in 01 -
'D a' r N - CO pý N
O O - - - N - N - N O - - O - O - N i O - OÜ
Z 2
_ _ V OO - a 3
ý H t M 'ct in V
in M ' '
M I 10
U u s n Q t t0 V l
WO 0 N M "I* In U) r U) rn O -" Z
O -
O -
O - -
- - - - - - - -
N N N N -
N -
N -
133
%0 2
Z
N J
CD
ö
z z 0
J
F5 ä- w a
w N T-1
I-
*Z a<J U1 M< SH<-
Q I-
WN O - U O 1,0 N N d- N e S2 Zcn L.
OX 0D ZW
in - - 10 0 ö l- IN X 0 0 i - lö l G. 0 1 ttI O ýC M '2' O N Il> M 0 N M Co N O fO
laD Ir I - %0 M 00 -
I M M
Ißt N t0
ý M
O> O O O N O O N O N N 1110
to N 0 00 _ 1
M J
p l'r, 8
Ul n O I M
M N %D w O" O O 0 O 0
S O O - N N O - O
J Ü
00 %. O e '. O a% ýO 'o M ý rn MD '. O
- (f) 0 1 l0 0 I- N I 1g ä lö iN - 0 0 0 - Co CO C'4 N O M 00 QD M M N O O N z -
1 I 0 r- r- 0 0 Ö ý
- - w
M V u-% - N U1 �O N O '. o o0 N %0 I I I l N
i 1 - Ö
N - N O O O O 0 - - -
(V N
lo OD
N OD
O IM lu-ý -
- -t
M (
p', r-
l t0 I -, 0 t
I
O O - O O c" i O O O M
J r- ` N Q ° N
Iý Oý
lw i ý
Iý M
1 Q M
ý -- -
IM M m
' O
O 0 0 O N M O w N %0 %0 in N co Co M -e . Ki- OD N - O !O aO
LL. O l
- m %D 100
- O 0
1 N
- pý
1Pr)
I p O Co 0 - - 0 O O N
LLJ U1 00 in 1.0 M ch O% a% 00 %0 00 W W qt O r-
l n iT M O% - O O N
O O N N d ml M N '-O rn ýt O M u1
O p in 0 N IK I
I\ M iý I
Ö O O I N - O N N U
O l n
U i l IN I0
M N I
U i I
Kt !
0 p O N M1 - O N O - M
CD i lö ý - _ 0 - 0 0 0 2 8 0 l0 0 , I 0 Im 1 N
jriÖ M
1 ( l i i lÖ
0 8 O O 0 O N 0 0 N O O
w p J
_ O ccc i Ü w-¢ cn
_ ÜO2 Z rn %0 N N - OD N O v d rn in in '. O in cm
w F- W L3 e N d - ý ý - M N (V - N N Co O - cn O N N O - N O N N - - - N M O V N
On 2 C> - I--
O
J w
w Oc ii0 Ü
gN OO Z
L CD 1N fN M U, 1 N O O' CO N- 2 o Co
O m p O+ r M
O N
O - - - 0
O N N O M
- z 8U
O
v c2 O C
w 0% N tV - 0 N 0 N CO m - O N CO M O'
I- ct M UA st M N (A n M Ln Ul ýt '. O
- CO N M frý M r, CO u 1- N (V N in M M M M
M in K l Z
134
ýö ý0
Z ...
J Q
C9
Q
0
w
z
I
r J
W CL
_J 81
N vi
W J co
Q J W d` Q. ' O 0
aQs wTO =ý- Q- OWi Lr) -< F- f- -- - - N O - in N o' N ct kD n 7 M O tý C I tY ZUL - F- - OW Z0
C
ýp d tý V N- N o' M M1 00 r- Co M O o' V OD N
iO' CO
I 1-
I Nt Co N
I N O
{ co
I %D
I dT Co ý' O tf\
J O" 0 0 - N - N O N - O - - - O M O 0 O
p\ N1 - C, % M M n N O in - O\ M ff-I d M E p Kl C
I Co - N ON 'O N r r p\
{ 0'% ýo
I Ul N
I in
{ ON
O- - O O N O O - O - - O N O O N - - - O
N 0 O0 -4- 0' N N et n
- t0 MO ON O ul h "0
ý 1 a - o o ö - I - I V - - 0 - t O - 0 1 0 F O Oý m N n N N o - ý0 V o En O c itl co t+1 Z O d N-
1 N
1 I I UO N M d N1
1 N
1 i
1 c7 f`r) 197
1 N N
I 0 ";: r
w O 0 - C - - O - O 0 - 0 - O 0 - O N - :E V) ui z J too N N -- O \ ý' O\ 'mot O\ ION h 'D E M M
N i O I
(O Ö
{ O (
0 0 0 0 IO
IÖ 0
Ö t4rt {O
M Ü (j) O U' i- (Co fi { IN IýO (00 Iip IO\
0 ON ýfi 1 (Co N IN N m- 1 Co M I
O O O O 0 N O - 0 O O - 0 0 O 0 aZ 0Q W S
til O
N- N
N -
(14 -
0 d
00 %0
Iff) O
N O
O\ -
Co N-
co O
CO 0
I f- N
., t O
O' '-
Iz II
I ul 00 { d' tN
cn W O N - - O - O O -
1 M O - -- W
Ö= -, r - N IO GQ o' ' N ,t C' d1 N- O ' M o M r- u-
H-- c9 1 rn O
I0 I S N 0
I 0
( "CT (IN
I 0
- J f - - - i v - - - - I ' 3 Ü
-L M En r N 1` M in iz Co M N o' O\ C7% -,: I- N t- w
{ O
l I I -
I V I I 1 O 1 H 0
- O - Ö
- O O Ö Ö 0 Ö
N O
1O
Co N` co -, t N N- N n -,: r -. -t L)\ in W O
{ 0
I { I { CO t Ö `0 - - O 0
O O O O Ö
- O O Ö
CO %0 lzr CF\ t. 0 r Co O
( I ( I { I- I (N
{ IN Co
0 - N - N - N O 04 0 N O 0
04 V) t %. 0 U
t H {
O Ißt
:T t-rN i I Un
IN { (N
1 0
Ia) Im±'t O 0 O 0 N N
O
11 11) r C
0 o
o " o r- o Q O Q 10 ISO
0 N O
In
O N
- V 0 0 O
I n
Iý {Co IM N
I ICA (En 0 00
l
O V O N O - N O 0
I- w OZ -j
LLI - w 0 vý g ÜO$
Z M in CO - N- %0 rn NT C; \ O ko r- M M - \0 Co N- a) W F- F- L3 qzt N. M rn O\ OD co to in in 0 ul v O\ 'T %0 -- ON N
Oj 0 N - - tv - O - O N - - N - - - 04 0 0
OUQ U-0 h
LY_ - }- :L 3
Z J D
Li ü Ü
O I- - cn E 7- I- t) N M - N in Co in r- CO Co n M Co
a I- L3 O N-
Co N in M 4 N CO in tO in 0 N -' N ch tO in r- 0V 0 -gr M M in -; r M %r M ýt M M :r zt in -t M M ct M N
p Ü0 U _ D ä ß
E-- a. 3 2 0 ul d u1 Co CO oD ref Co O\ O\ - - - - M N- 1ý S
LU < N s- in N- CO - IN 00 CO - 00 0' "O to 0 00 Co CO 8 O -
F N in J in to N- CO ON 0 N M -i' in to N ý 0\ 0 - o N
-Z
135
Co N
w J
H
0 %0 tu)
v
N J
C7
d
2
wu ff ý_
Z
NC
J
T5
W O.
I -- z z u, Ci c WS0 SQ
n < F- F- - - N O M N - O - - O O v' O O - N rn Of zU ý- - -
O Z U-) 0
a-
00 N r- d 0 N ýh I-- 0 M oo 0 O to /D N 00 N c? co (V
J O'
1 U1 U"% 0-, (, 4
I N ýt
( l co V O ON -: t N
I CO
I
O" O O O 0 - M O - O O - O - - O O O
M N C-4 - V O' LC) o c\l , - M ON IN O M to u-N QO a) M O
I O
ý (IN co
I N
Ir l UD
1 Lr) "D M M ,t
Icy IV l 't M
IN
O " O 0 O O 0 O O N 0 O O O N
N 0 - 00 N M ' ý ; O N "4- ýt r- C14
I I ( I (M I I I I
d O O O O O O O O O O - O O 0 -
z 7 - N 0 0 N N N %D n N N (14 N
I 1 IN N I O I I I IO N I O 1 M 0 O O 0 0 - ý - 0 N J
l
o I I0 I I I
co ÜN o O O O O O C) NN O O O - O N O O
C) 0 - N N CC) 0 00 C'4 0 (-4
0 ON 0 00 I I
r- I C'4
IN I - %0
I- 1 m ON %10
-
az - O O O O O O O O O
Q Z: w M: "t r-- co l %D r- N o-, N U)
r
UN 8 I-
N p 0 I I I I- I l
N I
O O - ý ý
N O N
I I J I I L ( I 110 co
3 O O - O N O O O O O N O O N (-
o ~ W
I IN M
J O
f- I I I 1
- O 0 0 0 - ,p O O N rn ---; 0
-7 M - 0O CO to o-, N r- .t C o' N N N W a0
I V N CO /O N '. O
l M '. O
j M N
I 0 co
1 N
1 ON
1 O
1 00
1
O O 0 0 0 0 0 - 0 N - O - 0 0 - 0
-zr r- rn O a-% vt ko d c0 M O ko O d n %0 O 0 0 M 0
IM IO' N N M
( O
IN t'"%
III (V ct
I N
I '
III
- O O O O O N O O O O O O N
00 (A n r- U
I I IN 18 r- 0 r-
J - :t
I OD
I 00
( 0 fr)
Iý
O O O O O - - O O N O O N ý0 0 0 D M
CE) ö
tN ö O o 1
N 1
M u1 o N o ct t - I OD 00 I ýO I N I o o , o O 0 0 0 0 - O O O CO %0 N 0 C)
0D a\ Uý
ý - Q %0 IM %C 04 - CO 2 1 l0' I. ý I I O 0 0 0 0 0 0 - O O 2 O -- O O M M O O
Z M
J
N Ü
0 O$ z ul M M - ON N O' %7 - 0 0' N N (V CO - Kl to (V i U L - 11, tf1 ýt - M ! fl M kO N M ýi M U-\ N N N
OU ciý O - O 0 ' C7 ff) N M p n; CV N N Lr OUO =
F-a. 3
F- W Q Z -i 0W Q
ui V) Ü N WOt Z 00 ul r' (V - N - n O N LL N a) N C-4 M N
ý- ý- L 3 '. O N 0 - O M N N It M C\ M N N U1 O\ O\ fý ci
=j O M - - M M ul M M M n ;t Nl Ln N '; r 'Ch Ct t- ýt 000 0
--
F- 0- 3
< N N "t %0 co M kO VJ 7 m O1 O\ r - W C' O' C' 0' O V - 0\ ir1 d -t. d DO kO O ý E- lei _
. .
N N Lr\ to F- Z t N N N N N (V OM M M M M
M M
M K l M
'0 M
ON N ri
Z z il l Of ¢ w= 0 T 1- Q li. C7 lY I N
-< f- E-- - M O - W' - Lf) OD N O O d' O O - M O - SIrZU L - -
f- -ý OX0 ZW0
S
ul r O' 0 N- ZT 0 M cy% V N N M ü\ 0 N I- I 0
I I I 1 O 1
O, O IO O N - O O O N 0 O O O O O
o N N r1 M N N- O d' O -- O Ln N
týl l o 1C I
ýo I- I - rn { VD l KN . - OD = K) I - C) In co
ö. ) N o - - o - - , N Oý 00 , 41) "; ', N O ko to - o vo Co I- K 1 1 I r I 8 tn
O- O O O O 0 O - (4 04 0 0 0 O -
M "') O rn ý % 2ý O O N) 'O O
I I (- - M
1 0
( I0 (N
' S 0 0
O O 0
O -- - O Z g
cn r1 un p\ M ' O rn N N Ln O jO O ko co N- OW M 1 M 1 M ul M I - N
r L11 M M t0 Cp n
1 ct ul N
Q 0 0 0 o o - o I o I- o Io 0 0 o - F-U UN r ýp u1 M n O N t` N O! t') LC% 0 - -- I I IN 1 ( I
D C9 ý O O 0 - O 0 - O - O O o LZ C
11) I Ipr) ( I 1 IN
D¢ O M
O O O O O N O O O 1- O W
I- r r- LrN O O
IO Co
I I I j i 1 I(N Ir 0S
N O 0 O - - O O 0 N 0 0 H 1- L3
0 0 Co N - - 00 N- o' - O N M r) a\ U U- N - N 0 1
I 00 - n n M L[ 1 1: T I I N M
I
L N O O O N - N -
N - - - N
~ N 'V DO 0 ý O r- ct N - oý - 0 -4- M
l ý I o j 0 O 0 IN o { O 1 I 0 0
10 0 J - - J 0
0 0 ö M 1 N N N- O M ko Ln c0 - 0\ O O M
I: I Iý 1 -
I I 1 N N
1 N
N O O 0 O O N 0 O 0
o\ cc) Ln N Q o - N r- rn C4 O
IO N 0
`? (I l O
( Io I
0 O - - O - O O O O
m 0 0 1 0 - ö ° 0 0 0 N - 0 0 0
0 , 0 0 0 N M N N N d' CEO O\ Ln N a0 00 C' N N O o' I - -
I I IO O
O O O - N O O - - N O N
az J 0 (I)
0
i± 9 L' 001 Z 1- _ 00 d N N O - O' O\ 0 M 1- O Co N M N ui ý- L - - to
O0 O N O O N - - - N N N O - N N J 0 =
U-C I- - Of Q F- U 3
DZ -j aWW < WCXM U ÜR
W01 Z U OD to N co ul N ýr 'o 8 - O\ U Co - - L1 M L - ON M ' 0 00 Co ul ul IN O' N -- co Ü
O N N
M ýt M M un n Z N M ýt l11
U Cl m Z
Otr2 1-- i- CL 3
0' 1- U1 1N N O, N CO p N tn co W< - U 1- CO N 0' - N N
O '- N- O' M O C) 0 CO CO -.
W - N M V Ln to N aD 0 HO v q- - . v v wt ý- - U U-1 U U-\ ir ,n Lc) u1 U. z -
137 I
% M
zw wSO SýCZ
<H j_ M N M M 0 0 N O 0 N - O N - 0 N ,: t co co 22ZU ý- -
0 ü 0 ZZ
0
in t0 N ý- a) ý- a' t- M M M u1 O O N I I IN I
C), - O O O N N O O 0 O N O O O - -
r O N N N N M N CO N 0 N CO N O L, -, I I I O 0
I0 1
d i- O IO O 10 - N - v - N N - - 00
o C o a. M u\ - N ON n l- O' 6\ m0 ul a) M N N in _
I M
I 00 ON M - ül
i o O' in - N 0 in N Nl
(> N 0 O - 0 O O 0 0 - - - O 0 0 - - O
%0 N N t- ' O N N C)% .* N 't Oh ,J O O' 1 I- I I ( Ir, 1
Z O N O O O N N N 0 I- O O 0 O W OW N { N M CD aO n O "I' ill N Oo M a0 N - Lr%
-' J 0 0 ö 0 0 0 I 0 0 I0
lö 0 - -- 1 -- 0 1-- 0 0 N
ul CO - N ON d N I\ CO CN p I- co M C c I I I I I0 1 1 I
OZ - O N O O - O O O O - N 0 0 O O
dZ d ("4 O f- w M N r- - ýO 1' ) ,, O V l
2: I J I I J ILn IN I J
N I
W O M
N 0 0 O - O O - 0 O O 0 U) W w
o' in 0 ON in N Cs /O r- a0 in (7% ý Ul N CO O - CO (D I: r
- 0 0 - l - J 0 0 0 l ß 0 0 - - 0 0 0 1 1 1 - , 0 0 O-
kc) - 10 L I I IM I I {aD % O 104 I N o% to I 1ON I 0 N O O O o I- - ö I-
ýO V N in O' N ul t11 CO L[l O L" N I- N W
IN I I 1 ( I 1 I
0 0 0 0 - - 0 O O o - - - 0 0 O
(7, (N 0" N (CO 0 O
1 S
I lo Iý 1 I
N N O - N - ,0 O O O N
f'f CO CC)
0 'D IN IN
co N I- IO ( ( 1
N O (- I0 I- IN
O O 0 N 0 O C) 0 0
n Ul O N Ln - r- V) ` r- m )
i ! (
I N -
n CO Kl I r-- 'O ýf N OD
J co - ;t ý ý
O O O - O O O O 10 O O O -
rn in in N N ---t M ül N ý Q, 0. ') a)
< Ir
rn C: ) tfN I
0 1
00 "t O\ I
r- I
r- I
V) I 1
M O O O O - N - 0 O O N N -
J OZ
O u) Z C) Q N
- M U-1 N et Q\ M N ; Zt -4t O N N O L- r N N O Nl aD CO N O - N O in N N _ u1 N
OU O N
N - N N N O N V N N M O O r_ T ZD I 0-, C) _ v-o Cl:
I- a s
OZ J W w W <
u '1 ý F- - cn WOl
Z 0 M N - CO in ýY' N M C - in N O\ M CO in CO in (N (N v in rn - in N rn in co O 'o O -
Ix OU O ýt v in Pr1 V' N t ul M V ýY in in N OÜ0
_ Z-0 F-
w F- a 3
i ýt N K: r ct O M X11 N N N CO N O U-\ N N ;t O' W< 0' M - in CO CO N ul - Nl N 0 ýO in in ýh ý* t0 N
-W F
-i - - - - - - - - - W O\ 0 - N M "It in 0 N OD O\ 0 - N M V in ýO N
O in %0 CO %0 %0 \0 '0 "O %o tO v0 N N N tN N N N N Z -
138
''o M
Z
N J
0
O- z
ýS 21 Z }- OL
r J
F5 U, c w a J
N1
W J
H
z w
w20 Uc1 cn -Q I- H- - - v v N O O O U 0 N G' O O O O O O M N S Of zU L
oxO Zw0
(if d
p OQ N 1N M OD N IN o N M CO C, M C M N tp a%
I CD
I CC)
' 0 %D Ln 4 N
ý0 ýO N 0 o
' O
d O - - 10 0 0 - J - O 0 N
0 %D v co O %, o N n OO %0 O 'r M 0 co N Wl 0 V
' N
I M ON
I I ýY N - 00 N - I-) Lf
I p'
J
d - O - 0 O N N N N O M - 0 - - O O
V) rn -qr N ff) C, (14 ON CYN 14)
I IN S Cýj
IO C14 -, zzr
I I
N r- 00
IO r- -
I0 IN 0 0
d 0 O O O O 1 -
O 0
I- 1(ON I ( IN 5 I J
Z 0
0 10 IO 0 N N - 0 O 0 0 N O 0 LLJ
U) N ,.: o - OD 0
(Y% NO ko
- N p M L
o' V)
co OD " r
0 I o p o
-- 0 0 0 n
I0 0 0 0 Q
- o o 0 ,0 - J0 O
U to ' ,O O N ii ul 0
O' 0
C' 0
IO CO Ian co OO 1kD OD r-
co 0 pýO
I 0
M 0
M 0 v
C7 O o 0 i0 0
- O I oz
_ a- Z - 0 N N N o-, v D %, u1 co OD (7% N ßt1 a: 0% q N
1 r- 0
I- I M
ýI ý
0< O O O O 10 O - N O O 0
Lu M: LLJ "'
- (X O v o r- cc) N M v 0o oo ( rn OD v N m (0 N to M
I O t[1
ý M O' N n OD
I 1`
l n OD %D Ln i
I
O N O O - O N - O - - O N O - m
U Li. 100 I
M to O ýT it11
O N I IN
Ln CO l
r- 1(14
10 M r M N
I- ~
OO OD d' ()N ý OD try kp M %0 o 00 C O aý w - ko
j - U-% fn %. 0 n r-
I rn r- co
1 C-4
I- C-4
1 p
( I
N O O O O O O O O O O O JO N O O
Q ill %0 o' 00 't CO N M ul 00 N N O 0
k4 I I I I I N
O - 10 O O - I - O O - o
00 -
U O I '
- J J
- - N
J O O 10 N - O O N O - O O N
(7% % N "o NT tn M M m N
f 1 O \
1 d' - p\ O OD If\ lsl
O O O O O O - O O - - O 0 O O
Cq 10 r- V) tl'\ r- r V) qr kl) -ac %0
I cc)
I i N
I 0
1 0 N
I0 ISO
- O 10 0 0 - O O 0 0 JO
w az .j - Nw < D
w- v) Ü1 Z N (V \O co U-1 OD CO M tD O - M - N t`N u1 OO
w 1- F- L" N \o O O 0D v tO to rn - ao ao M 00 rn co n N 0 of O= 0 - - N - 04 N N - N N 0 N C N1 N N K) 0 N (Y M U L0 a
F- a 3
t- ui Z
Cl W
W wN U WOI
Z %D N ul N O 0 n Nl M W% %D O co 0' U'\ rýl %1' ix F- F- -3 '0 - tll 'D CO CC) N O 0 - N 00 N N \D C' tO N 00 Ö20 O M st M 'q7 ct Ul 'rY U, \ (V ' n w 'Cl- t Q' X11 Pr) M
L) L) a m
i-- a 3 M M h M N 00 N n %D OD r i
ULJ O% OD 0\ N - N
0 N N 0 O N l W
w CO 0' 0 - N M d" 1[1 tO N co D 1-- Z 1ý N- CO OD 00 CO 00 CO CO CC) CO O D O\ O\ O\ 0'
0 O\ O'
139
'o M
a
z
N J
0- Li- F 2 W4
-C Z1
L
i-- c
t- J
W CL
J
S U
F-
~ i
w
cr-, c Li M:
oQ1N -< I- 1-- - ON M (14 CO 0 tO M 0 0 - 0 - 0 vt ýY 0 in SZU L - - - ox o Zw0
Of
00 Ln V-1 CO I ( I (M I I I I ( I (
O" 0 I- O 0 O O - N 0 O - - ( O O 0
- - t1- co IO co
I r-
I 0 00
_ 1 re) r-
I co _
{
O" O O O O - o lO _ - O
IÖ 0
I IC I IÖ 1 I Ö
1-. t (
IÖ (Ö in
I , Zr C).
0 C) N Ö 0 - -
p -
v co v co mit ý u- 'r N O O cc) to Ito un - in O I 1 Ip I(
pý I
C14 K) (
1: T I
00 I
N I- iO - 0 0 0 N 0 N O - - - O
0 0
Z LLJ
(J) M 0 N N -t to O N to - in N N O O
r ý
O
IO
O
I -
O
I IO I IO I
Q O O - O N O
o 0\ 0 0 CO r- ON N 0% - 0
I- I
US r- r- I I-- I 1 1
- I
U' 0 - O O O O i- c'1 0 O O O 0 OZ -
a- n co 1.0 04 r- a\ Lf. ý cc) N N 1 1 I- - In
{ fl-
I 0 rn
I \0
I
OQ 0 N 0 M O N O O W
0 0
I I0 1 I I I M
O2 p O O Ö
10 O O O O O M
10 O O 0-
cc) rn cc) c\ fn W
I0 I I 1 lCA I I IQ I1
L O N - O - N O 0 O - O O O N ý
~ N a\ 00 O h
g 00 M h o\ ¢ CC) N
w ko GO
ö IM I N M ö co pel 1 I (14 fn I N I U-\ I I I f O O O O - N 0
0\ "0 c) 0 tO - 0 ö ) IM I o Iý IN Iý 0 0 IQ 1 - N 1 n ý0 co 0 N V
i 1 I- 0
I I N
I 1
- 0 O p O O O J- N N M O 0 O
tf1 - N cy% in Q M 0 r-- M 'O N m
ý N CO jr- O
I I - co
( I j r- U-1 N O O 0 10 'O - 0 O O 0
Ö O
Ö O O O O
o m pn 00 I (7% 0
I 00 %0 Ln
{ r-
I r-
1 M
IN rn ('4
I
JO o O 0 O O - 0 N 0 0 M
. 10 N
C) Z J
Eii N O
Ü
_ U) 001
L Z - - N ao
M O+ h
O' N 14-
to T N
0 0 to
N CO
M - CO dr 0 O O N Ü Ö p M - N C N N tp co Q WO0 0 O O p N O N N N O N N N N 0 7
p0O C)
Of Qý I- a- 3
WZ O
Q
w N L) Ü U, WO1 Z il 0% M p' M 'i' %0 00 - p' N h tp p\ %0 1-- in -. Co
3 F- 1- L M OD N co - 'IT to 0 'I' in to I- N O\ h - h 0U p M N N (V ý7 M "T in ý' M M M1 qt ct M d M
M p ÜÖ U Z
F- 3
co I- in in in L1 CO M M p N tp N (V co 0 zr - ýt u W o% 0' r-- O' 1- a0 0 to -3- in to in 00 O h N cV M p . . . . . . . . . . . . . .
W6 I-
1- CO aý S - O N M
0 ýt 0 Lr\ 0 to r, co N M T in
m at c7% O O O O - - - - -ý
140
0 M
zI Z
WX0 =Qs
N < I--- I-- - M u1 N 'D M M - N -ýZt O - C M 0 M M d M N
YCz c) L OXQ ZW0
tY
.;: r O ul N jo' O CO ' 1,0 C) 0 ct -
CO -k ON t ICO r- M jN 'IT co rr-, co r ý M N b 1 'ct ý I is N Jr ' V N I
\p -CI- zr j co tt1
0 0 - - o {N 0 O O N - o 0 - {o o - - (N iO u O N rn c o N pr) u N IN a' IIl co o N N
T 41 I I I I
O J I
% I
ff) I t
ON I; zr l(N
N O' !O - O (, -1 N O O - O O O - 0 - 0 - N
0 rn K) 0 N N 0 o co o O }O - (N
(O tr% d N I
I d
I o
1 vý)
I K) OD
J ý pll to d
l 1(`J N )t0 N
O O O N M 0 0 0 O lO 0 0 O O `O N
r - V) N <1 n M d %0 M d O "') Im ICO N IOD O d' 0
j 1;; CO M - n
ý N 0
1 I- p CO
ý CO CV
I O 10 M to
- O N N - O 0 O M - O O O O O - I N - 0
p' ill S? N O N O %0 M CO u1 n - O co -zt
-' I I N . Zt o I o 0 O I 0 I
o o - jo ö o N o -- J iO o o 0 I- W M CO N Ul CO 0 N M N d N N CO M - - N N
V) - n O N N 1 I I
d' N I
N I
kO I
M I
to N N OD OW O O O O O O - O O O O CN - - O - O N 0
1Q 10 M o% - N m CO - I- \0 CO 00 d- M M t- %o N- U
2 ý l
N O N K1 N I
'14 d- O a0 M ,
O m I
V) M 1 N
ý
-T Q N O O N - O O O N O - O O O
O z
Ix - V, % CO 0\ 04 lzT CO oo aZ cD do qt
I I d ao
1 N v
` %D N
I CO O
l IN OD f
M I
co J
r- rn O O 0 N - N IO O N O 0 O - N O ) 0Q cW
W 1'' O N co M ýt 0 N 0 00 N N co O C" N N N
-a W ý 1ull I
~ - (
r- ro
O O O O - O O O - O O N O . 0=
I-- ý- O- N V) CO t0 - d a' . 't a0 r- M N
:L 3-1 W 1 0 J
o l - - N - - I- 0 0 0 10 10 0 0 -0 - 00 N 00 of co tn - m CO 'CO oo 111) Ln N - r- 0
J- I l I I I
O O N I- M O O D - O O - O - D -
LCN CY, (N (D Lil 0 U
IN IK) I(N I 1 ,. T (CO I I I I
00 I
N I
0 M N O O O O O O - 0 - N -
CO M M M QO Oý d r C, CO N 00 N 0' n N N N ý I I I I
Q O O O O - O O O O O O - O O - O O O
ao N ý O i M r Ö "f) N i t l = ý
8 _ I
r %. o -
uý O I r 1 It) Ln
I ý
c) 0 0 O 0 O - O - - O O -
h- W oz --j Ü
O(1) W N ÜO aI
Z CO UN - kO a M n M N tN 0 QD ýD M - N - %0 N F- I- L3 M M CO N O\ 00 O CO 00 N O CO 'r -t 0'% - - CO CO O W O0 O - - O - O - N N O - N - O - - - N -
O C-) - IU-c. Cc - H a-
I s
I- W z -j OWW <
WNM 0
t5 iý O U) i
Wol z N - 0' u M rn V) ul N Lr\ ct '0 co N oD cr r` M N N F- F- L3 - a' O% O N lfl O
U- N M N N 'O '0 N - 0
- ui I- ÖZ0 Q - "Y ýt M d t1 ýt M M tit M d' M M
000 S
(L 3
iZ - %D - CO N 00 O O CO Jt - OD - CO v 8 N m ýcl to W¢ -- M - N1 If) Qý - d lrý - OD If) OD N O D Oý CO d' %D I- W
W ýO N co o' O - N M d' If) %o TN co a' O - N M d' n 1- Q - - - - N N N N N N N N N N M M M M M M -Z - - - - -- - - - - - -- - - - - - - -
141
0 M
z v
c9
dý O w
w
Z Z O_
H N
} F-
3 N
W a_
141 cn ri w J
z ZW
w 0 = S F- Qi
Of ý - I- t- < L rý - to O SQZU L
x O O
ZW0 Q' 0-
ts ", C' 04
lo -
d N .
0 UD pr) 0
- O
0 I
O ' 0
co T Nl N M
I M tý
d - O - O
0 CC) CD N O O O
H
- - - ý-; r (j) 0 0 w J
0 1-) 0 C) U) O p - - - 0 (. D
N a- Z N N - 0 0
O C9 N 0
( N
W 0 O O 0=
co oo ýzr u, \ 7 0- U- O lPf)
0 N ; N
Ö - O
-L w o fn to F- W - N " s. 0 N SO O O
p n N - M O N - M - O N N
co 1 U 1-
l - N N C
1 C)
O M O N
OD M - 'zt -
a) co 1 I
(N 0 0 0 0 - I Ic l° (ö lö lö
o o
aZ J <
ON W-O U) 1 T 50 0, r, co ") r- co
L - OU 0 N
N O M - 00
0 7- C) -0 F-
m cy- F- O_ 3
U-i az -J W CW)
Ü F- -O cn ý WO 1 z - - M Wa N U)
it C) C. ) C) in
000 2
F- CL
Z 1Sl N W¢ \0 - U) C' - Ul
w
w r- co M F- O Ma M
142