virginia tech...the perceived dimensions of jobs: a multidimensional scaling approach by kathryn a....
Post on 05-Aug-2020
0 Views
Preview:
TRANSCRIPT
The Perceived Dimensions of Jobs:
A Multidimensional Scaling Approach °
bv
Kathryn A. Welch
Dissertation submitted to the Faculty of the
Virginia Polytechnic Institute and State University
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in
Psychology
August, 1989
Blacksburg, Virginia
The Perceived Dimensions of Jobs:
A Multidimensional Scaling Approach
by
Kathryn A. Welch
Roseanne J. Foti, Chairperson
Psychology
(ABSTRACT)
Recent research has revealed ambiguous evidence for the validity
of the cognitive complexity (CC) construct. Some authors (Bieri, Atkins,
Briar, Leaman, Miller, & Tripodi, 1966; Scott, Osgood, & Peterson, 1979)
have suggested that a potentially useful method for examining CC is
multidimensional scaling. The present study examined such an operational
definition. The present study also examined the perceptual dimensions
that underlie individuals' perceptions of jobs.
Three hundred and five subjects rated the similarity of pairs of
job titles, completed the Role Construct Repertory Test (REP), and later
rated videotaped vignettes in a performance appraisal simulation.
Multidimensional scaling extracted the subjects‘ dimensionality. Due to
an unstable solution, the study°s first three hypotheses (that
dimensionality would predict rater accuracy, that dimensionality would
predict rater accuracy better than the traditional Role Construct
Repertory test, and that dimensionality and the REP would be correlated)
were untestable. Multidimensional scaling was not a useful approach in
this context.
The fourth hypothesis stated that the present data would replicate
the three-dimensional job characteristics model of Stone and Gueutal
(1985). Results indicated that the Stone and Gueutal configuration was
not supported. Thus, job design efforts predicated on their model appear
premature.
ACKNOHLEDGEMENTS
The author wishes to thank her advisor, Dr. Roseanne J. Foti, for
her expertise and guidance in the preparation of this dissertation. Her
enthusiasm for research and dedication to her teaching are contagious.
Moreover, her patience in the face of “inheriting" new advisees made a
difficult situation better.
In addition, Doctors Robert J. Harvey, Neil Hauenstein, Robert
Madigan, Christopher Peterson, and Albert Prestrude have all provided
considerable guidance and support. Their efforts are sincerely
appreciated.
iv
V
TABLE OF CONTENTS
A Page
ACKNOWLEDGEMENTS ..................... iv
TABLE OF CONTENTS..................... vi
LIST OF TABLES ......................viii
LIST OF FIGURES ..................... x
INTRODUCTION ....................... 1Overview ........................ 1Defining Cognitive Complexity.............. 2Operationalizing Cognitive Complexity....,..... 8Construct Validity Issues................ 14Importance of Cognitive Complexity in Work Settings. . . 18Summary and Hypotheses ................. 33
Performance Ratings.................. 34' Frequency Ratings................... 35
Job Characteristics.................. 36
METHOD .......................... 37Overview ........................ 37Pilot Study....................... 38
Pilot Study Sample .................. 38Pilot Study Measures ..............,.. 38Pilot Study Procedure................. 40
Main Data Collection Phase ............... 41Main Data Collection Sample.............. 41Main Data Collection Phase Measures.......... 42Main Data Collection Phase Procedures......... 44
RESULTS.......................... 47Reliability Estimates.................. 47Deriving Dimensionality................. 47Summary of Results ................... 72
DISCUSSION ...................,.... 100
REFERENCES ........................ 110
APPENDIX A: FAMILIARITY QUESTIONNAIRE .......... 118
APPENDIX B: BIOGRAPHICAL QUESTIONNAIRE.......... 121
APPENDIX C: JOB SIMILARITY QUESTIONNAIRE......... 124
vi
APPENDIX D: REP GRID MEASURE............... 141
APPENDIX E: BIOGRAPHICAL QUESTIONNAIRE.......... 146
APPENDIX F: PERFORMANCE RATING QUESTIONNAIRE....... 148
APPENDIX G: PLOTS OF GROUP STIMULUS SPACE ........ 152
APPENDIX H: PLOTS OF GROUP STIMULUS SPACE FORFOUR SUBSETS ................ 160
APPENDIX I: SUBJECT WEIGHTS MEASURING THE IMPORTANCE OFEACH DIMENSION ............... 189
APPENDIX J: PLOTS OF INDIVIDUAL SUBJECT WEIGHTS . . .. . 196
APPENDIX K: CORRELATIONS FOR DEMOGRAPHIC VARIABLES.... 203
APPENDIX L: COMPUTER PROGRAM FOR DETERMININGACCURACY SCORES ............... 205
VITA ........................... 219
vii
LIST OF TABLES
Table Page
1 Stress and R2 for Dimensionalities Two to Six(INDSCAL)..................... 50
2 Stimulus Coordinates for Four-Dimensional Solution . 52
3 Stimulus Coordinates for Four-Dimensional Solution:Subset A (INDSCAL) ................ 57
4 Stimulus Coordinates for Four-Dimensional Solution:Subset B (INDSCAL) ................ 58
5 Stimulus Coordinates for Four-Dimensional Solution:Subset C (INDSCAL)..... . .......... S9
6 Stimulus Coordinates for Four-Dimensional Solution:Subset D (INDSCAL)........ . ....... 60
7 Stress and R2 by Subset (INDSCAL) ......... 61
8 Coefficients of Congruence: Two-DimensionalSolution (INDSCAL) ................ 64
9 Coefficients of Congruence: Three—Dimensiona1Solution (INDSCAL) ................ 65
10 Coefficients of Congruence: Four-DimensionalSolution (INDSCAL) ................ 66
ll Coefficients of Congruence: Five-DimensionalSolution (INDSCAL) ................ 67
12 Coefficients of Congruence: Six-DimensionalSolution (INDSCAL) ................ 69
13 Stress and R2 for Dimensionality One to Six(ALSCAL) ..................... 73
14 Stimulus Coordinates: Two Dimensional Solution(ALSCAL) ..................... 75
viii
15 Stress and R2 for Dimensionality by Subset(ALSCAL) ..................... 76
16 Stimulus Coordinates for Two-Dimensional Solution:Subset A (ALSCAL) ................ 79
17 Stimulus Coordinates for Two-Dimensional Solution:Subset B (ALSCAL)................. 80
18 Stimulus Coordinates for Two—Dimensional Solution:Subset C (ALSCAL)................. 81
19 Stimulus Coordinates for Two-Dimensional Solution:Subset D (ALSCAL)................. 82
20 Coefficients of Congruence: One-Dimensional Solution(ALSCAL)...................... 83
21 Coefficients of Congruence: Two-Dimensional Solution(ALSCAL)...................... 84
22 Coefficients of Congruence: Three-Dimensional Solution(ALSCAL)...................... 85
23 Coefficients of Congruence: Four-Dimensional Solution(ALSCAL)...................... 86
24 Coefficients of Congruence: Five-Dimensional Solution(ALSCAL)...................... 87
25 Coefficients of Congruence: Six-Dimensional Solution(ALSCAL)...................... 89
26 Coefficients of Congruence for Stone and Gueutal°sFinal Solution and Six ALSCAL Solutions ...... 91
27 Stimulus Coordinates: Stone and Gueutal (1985).... 92
28 Descriptive Statistics................ 96
29 Correlations for Hypothesized Relationships ..... 98
30 Analysis of Variance on the Effect of Work Experience
on Differential Elevation ............. 99
ix
LIST OF FIGURES
Figure: Page
1 Stress and Proportion of Variance Accounted for as aFunction of the Number of Dimensions (INDSCAL) . . . 51
2 Proportion of Variance Accounted for as aFunction of the Number of Dimensions: Subsets A & B. 62
3 Proportion of Variance Accounted for as aFunction of the Number of Dimensions: Subsets C & D. 63
4 R2 and Stress as a Function of Dimensionality(ALSCAL)....................... 74
5 Stress as a Function of Dimensionalityz SubsetsA & B (ALSCAL).................... 77
6 Stress as a Function of Dimensionality: SubsetsC & D (ALSCAL).................... 78
x
INTRODUCTION
OYERYIEK
During the 1960°s and l970°s, cognitive complexity (CC) was
considered a variable of importance in the study of stimulus perception
and social relations. More recently, the construct has fallen victim to
benign neglect at best or to disrepute at worst. In short, it seems that
Foa and Turner°s (1970) prophecy of an increased focus on cognitive
structures by the year 2000 may not be accurate in the case of CC. The
literature reveals both an explanation for this turn of events and yet a
possible justification for the reexamination of the cognitive complexity
construct.
A reexamination of CC is potentially important for research in work
settings. For example, if raters' cognitive capacities constrain the
manner in which they perceive stimuli, the potential importance of CC in
work settings cannot be overstated. Such workplace endeavors as job
design, job analysis, job evaluation, and performance appraisal may all
be susceptible to the effects of individuals° cognitive complexity.
Accordingly, another look at cognitive complexity is undertaken here.
First, this dissertation defines the CC construct. Second, several
methods of operationalizing CC are described and evaluated. Third, the
present paper suggests a number of contexts in which a new look at CC may
make a contribution in work settings. Fourth, one potentially useful
method for operationalizing CC is proposed. This methodology permits the
measurement of not only cognitive complexity but also its convergent and
1
predictive validity. Fifth, the present study attempts to replicate the
underlying job dimensions of Stone and Gueutal (1985). Finally, the
implications and limitations of the present study are discussed.
DEF[N1§G COG§ITI!§ COMPLEXLLX
Human information processing involves two distinct classes of
information: content and structure (Schroder, Driver, & Streufert, 1967).
Content variables provide information concerning the nature, magnitude,
and duration of to—be-remembered stimuli, whereas structural variables
concern the manner in which people process stimuli. The existence of
structural variables has numerous implications for the study of
individual differences in stimulus perception. Cognitive complexity
(Bieri, 1955; Kelly, 1955, 1970) has been proposed as one such structuring
variable.
Derived from Lewin°s (1936) notion of differentiation and from
Kel1y°s (1955) personal construct theory, cognitive complexity was
originally conceived in terms of an individua1's capacity to
differentiate stimuli along various dimensions. Kelly viewed people as
scientists who use their own personal systems of constructs or mental
representations to impose meaning on the world. According to this view,
people use constructs to understand, predict, and control their
environments. These constructs are bipolar dimensions (such as shy versus
outgoing, energetic versus lazy, etc.). A more cognitively complex person
would perceive greater dimensionality (i.e., use more constructs) in
determining the attributes of stimuli than would a less cognitively
complex person. According to Bieri (1955, 1971) and Bieri, Atkins, Briar,
2
Leaman, Miller, and Tripodi (1966), the term gggnitiye comglexity refers
to the manner in which individuals structure their social worlds.
Specifically, Bieri et al. defined CC as the capacity to "construe social
behavior in a multidimensional way" (p. 185). The present paper extends
Bieri et al.'s definition to encompass the perceiving of any stimulus in
a multidimensional way.
Bieri et al.'s definition suggests that CC is a unitary construct.
But, there is considerable evidence that CC is a multidimensional instead
of a unitary construct. For example, Vannoy (1965) discovered evidence
that CC is multidimensional. In a factor—analytic study of a battery of
CC measures, he extracted eight factors, none of which included all of
the measures. Vannoy's investigation revealed evidence for several
important dimensions of cognitive complexity: (a) the degree of
sensitivity to judgment variables, (b) the degree of discriminability on
judgment variables, and (c) the degree to which the world is viewed in a
narrow and ordered way.
Seaman and Koenig (1974) also provided experimental evidence for
the multidimensionality of CC. In a study of 146 undergraduates, they
employed seven measures of CC derived from the REP test. Results
indicated that three factors accounted for 78% of the total variance in
ratings. These factors were "positive affect cognitive
complexity-simplicity," "negative affect cognitive
complexity-simplicity," and assumed dissimilarity (p.387).
In subsequent research, Jones and Butler (1980) suggested that CC
involves not only dimensionality ge; se but also the fineness of the
distinctions made along the various dimensions. They employed subgroup
3
analysis to examine the factor structures of environmental
characteristics ratings for high, medium, and low-complexity raters. The
operational definition of CC involved a technique adapted by Jones and
Butler (1980) from Bem and Allen (1974) and Solomon (1977) which assesses
the variability of responses to a work-environments perception
questionnaire. A standard deviation was computed for each individual°s
responses to each questionnaire composite. Finally, the authors summed
these deviation scores to obtain an overall complexity score. Although
high-complexity raters provided six factors (vs. four for low-complexity
raters), some dimensions were consistent across subgroups. These stable
dimensions included leadership patterns, work-group interactions, and
role characteristics. However, there were differences between subgroups
"in the degree to which organizational level characteristics were
distinguished (a) from task and other domains and (b) among themselves"
(p. 8). For example, the authors found that low-complexity judges
produced a component which reflected a more global description of the
organization (such as role conflict), while high-complexity judges
produced components which made finer the distinction between role
conflict and organizationally-based conflict.
Another proposed aspect of CC is integration. Crockett (1965),
Schroder et al. (1967), and Scott, Osgood, and Peterson (1979) argued that
abstract thought involves both dimensionality and integration. Indeed,
Schroder et al. asserted that there is no necessary correspondence between
the number of dimensions one perceives and the level of one°s information
processing. What is relevant are both the manner in which an individual
combines attributes (dimensions) and the distinctions made along the
4
various dimensions. Streufert and Streufert (1978) have revealed that
cognitive integration is related to moderate levels of differentiation
along various dimensions. Unfortunately, the usual methods for measuring
integration (Schroder et al., 1967; Zajonc, 1960) have been more
problematic than have efforts to tap dimensionality. Scott et al. (1979)
have summarized these shortcomings by suggesting that there is no
consensus, at either the conceptual or operational levels, for measuring
integration.
Because cognitive complexity is multidimensional, Scott et al.
asserted that it would be more accurate to use the term dimensionality
when speaking of differentiative functions. The term gomglexity, they
argued, is more accurately applied to both differentiative and
integrative functions. Thus, the argument for examining integration in
any study of CC is compelling. But, any examination of a multidimensional
construct must begin somewhere. It makes sense that a reasonable first
attempt at reoperationalizing CC should focus on dimensional capacity.
The present study, therefore, takes this focus. More importantly,
however, the present methodology lends itself to an extrapolation not only
of subjects° cognitive dimensionality but also their integration
(Schroder et al. 1967). Thus, the present study also lays the groundwork
for an examination of integrative complexity. This issue will be
explained more fully in a later section. (It should be noted here that
henceforth, except when otherwise indicated, reference to the term
comglexity means dimensionality.)
Another important issue pertaining to CC is that of domain
specificity. Although Bieri and Blacker (1956) provided evidence for the
5
generalizability of CC across stimulus domains (people and inkblots),
Sechrest and Jackson (1961) did not confirm such generalizability.
Similarly, Crockett (1965) contended that an individual°s dimensionality
is specific to a particular stimulus domain. To test further the domain
specificity hypothesis, Durand and Lambert (1976) administered three
versions of the Bieri Role Construct Repertory test (described in a later
section): one for automobiles, one for toothpaste, and another for
interpersonal relations. The results revealed that although CC was
significantly correlated for automobiles and toothpaste (r=.33, p <.05)
and for automobiles and interpersonal relations (r=.22, p <.O5), the
correlations tended to be low enough to be of doubtful practical
importance. Thus, as a practical matter there was little generalizability
across domains.
Consistent with the domain specificity hypothesis, Schroder et al.
(1967), Streufert (1982), and Streufert and Streufert (1978) suggested
an interactive view of complexity, positing that (a) the level of
information processing varies within and between subjects and (b) there
is some optimal level of dimensionality/integration. Streufert (1982)
asserted that complexity involves a U-shaped function in which behavioral
complexity is plotted as a function of environmental complexity.
Different curves occur for cognitively complex versus cognitively simple
individuals. Along these lines, Schneier (1977) suggested a cognitive
compatibility theory that posits an ideal match between task/environment
characteristics and cognitive complexity of the rater in a performance
appraisal task. But, subsequent research has not supported this view
(Bernardin, Cardy, & Carlyle, 1982; Lahey & Saal, 1981). In response to
6
such null results, Streufert (1982) argued that a behaviorally anchored
rating task, such as the one used by Bernardin et al., does not
necessitate a high degree of differentiation and integration.
Peterson and Scott (1975) have taken a compromise position in this
controversy. They asserted that cognitive style is “to some degree topic
specific and to some degree generally manifested" over several stimulus
domains (p. 372). Peterson and Scott concluded that cognitive style
depends not only upon the stimulus domain but also upon the interaction
of the person and the domain.
The issue of interactive complexity is important for several
reasons. If complexity is both person and situation specific (i.e.,
interactive), then generalizability across persons and situations may be
difficult. From the discussion above, it is evident that stimulus domains
such as job titles must be examined individually. In addition, the
interactive nature of cognitive complexity must be taken into account in
any measurement strategy. This matter will be discussed further in the
methods section.
The present section considered some definitional issues pertaining
to the cognitive complexity construct. In particular, evidence was
presented which suggested that CC involves both differentiative and
integrative capacities. In addition, it was suggested that CC involves
the fineness of distinctions along the various dimensions. As it is
conceived here, CC is not a trait which is stable across various content
domains, but rather a process variable which may interact with features
of the stimulus. Finally, because research has suggested substantial
stimulus domain specificity, the importance of examining
7
CC-dimensionality within a particular stimulus domain, such as jobs, was
discussed. These definitional considerations have important implications
for the operationalizing of CC. A discussion of operational issues is
presented below.
OPERATIO§ALIZl§G COGEITIYE COMPLEXITY
As Bieri et al. (1966) and Scott (1963) indicated, the measurement
of cognitive structures, including cognitive complexity, has proven
problematic. A variety of operations for measuring the construct have
been employed. Indeed, Bonarius (1965) has enumerated at least ten
different methods for operationalizing CC. Because of the plethora of
measures, the comparability of CC across studies is difficult. Some of
the approaches to operationalizing CC include the Role Construct
Repertory Test (REP), Scott et al.°s (1979) Listing and Comparing Measure,
and factor analyzed semantic differential scales. Scott et al. (1979)
have also proposed a geometric solution (to be described later in this
section). Additionally, Crockett (1965) developed a Role Construct
Questionnaire (RCQ). Finally, there is the multidimensional scaling
approach suggested by Bieri et al. (1966) and Scott et al. (1979) and
adapted for the present study.
Among the measures of dimensionality, the most commonly used is the
REP test. Actually, the REP test is somewhat of a misnomer (Neimeyer &
Neimeyer, 1981). These authors asserted that "the REP is not a single
test, or a questionnaire with fixed content, but a broad variety of
techniques sharing certain methodological features" (p. 194). One
version of this instrument consists of a grid with columns (for persons
l 8
to be compared) and rows (for the constructs underlying the rated
similarity of the comparison persons). Comparison persons hold such role
titles as parents, friends, teachers, etc. Persons are compared three at
a time. Subjects identify a construct along which two of the persons are
alike yet different from the third person. Then, using a six-point rating
scale, subjects rate all persons on all characteristics. An index of
complexity is calculated by comparing the grid ratings in each row to
every other row. It is important to note that in this approach subjects
generate their own constructs.
There are several reasons for questioning the usefulness of the REP.
First, it lacks the requisite isomorphism between the construct and the
measure (Scott, 1963). As Scott (1963) noted, "the procedure for
analyzing data should be structurally appropriate to the property being
assessed" (p. 269). What the REP seems to provide is an indication of
the constructs that are used and whether or not they are used in the same
way across the ratees. Thus, the REP is not a direct measure of the
dimensionality of the rater. Second, the REP assesses cognitive
complexity less directly than other methods such as sorting tasks and MDS
(Scott, 1963). Third, support for the REP°s predictive validity,
particularly with respect to the prediction of rating accuracy, has been
weak (Bernardin et al. 1982; Sechrest & Jackson, 1961). (This issue is
discussed more fully in a later section.) Fourth, the use of the REP is
constrained by potential bias in subjects' sampling of constructs
(Bonarius, 1965). Fifth, the traditional REP task is very tedious for
subjects. Finally, the traditional person-roles REP may be confounded
by familiarity. That is, subjects' ratings may be differentially affected
9
by knowledge of the ratee. One benefit of the REP, however, is that
(except in its modified version) it does not assume the constructs a
priori.
Bieri et al.‘s (1966) modification of the REP test provides subjects
with constructs that they use to perform the rating task. This approach
offers the advantage of saving time and reducing subject tedium. Although
Bieri et al. (1966) and Tripodi and Bieri (1963) assert that provided
constructs and subject—generated constructs produce similar results, the
evidence on this issue is equivocal (Crocket, 1965). Thus, caution is
required concerning the use of experimenter-provided constructs.
Another measure of dimensionality, Scott et al.°s Listing and
Comparing Measure (LCM), requires that subjects
generate a set of objects as well as successive subsetsthat are alike for some characteristic.Objects that clearly differ on that attribute arealso clearly noted, but a classification of everyobject is not required (p.104).
LCM dimensionality is computed from the measure of dispersion, "H"
(Attneave, 1959), which is an index of the number of dimensions in the
subject°s grouping system. Logarithms are required for the calculation
of this index. By their own admission, Scott et al.°s LCM is constrained
by the representativeness of the domain. Because the domain of objects
has been generated separately by each subject, it may not be a
representative sample of objects.
Another method for estimating the dimensional capacity of
individuals is the application of Scott et al.°s geometric model to
ratings or checklists. This approach requires that subjects rate a
(standard) set of objects according to a (standard) set of attributes.
10
It is then possible to estimate what Scott et al. term "redundancy"
between a pair of attributes by computing the squared correlations for
the set of objects. Then dimensionality (D) may be computed according
to the following formula:
D=—Ä‘*j‘ [llK+-2Er
"where k is the number of attributes employed by the respondent and Erz
is the sum of all the squared intercorrelations among the k attributes“
(p. 108). This geometric approach has not been widely adopted in studies
employing CC-dimensionality as a predictor variable.
The semantic differential scale (Osgood, 1962) is another means of
estimating dimensionality. This approach imposes constructs by
requiring subjects to rate persons or objects according to pairs of
bipolar adjectives such as: (1) ineffective-effective, (2) shy-outgoing,
(3) weak·strong, etc. Ratings are then factor analyzed. Foremost among
the criticisms of this approach to recovering subjects° dimensionality
is that semantic differential scales do not in fact recover subjects°
dimensionality at all, but rather the dimensionality of the scale
developer (Goldstein & Blackman, 1978; Pervin, 1973).
As a means of extracting dimensionality, Crockett (1965) developed
the Role Construct Questionniare (RCQ). The measure consists of short
descriptions written by subjects within a three-minute time limit.
Vignettes describe eight individuals who fit particular roles. These
vignettes are scored according to the number of interpersonal constructs
ll
employed by each subject. One major limitation of this procedure is that
subjects° performance may be constrained by verbal fluency.
Although multidimensional scaling (Kruskal, 1964; Kruskal & Wish,
1978) has been employed most frequently to describe results for groups
of subjects, Tucker and Messick (1963) paved the way for the use of MDS
to tap individual differences in dimensionality. Indeed, Schroder et al.
(1967) and Scott et al. (1979) suggested that such an
individual-differences approach would be useful in the study of cognitive
dimensionality. Another important development was Carroll and Chang°s
(1970) algorithm (INDSCAL), which facilitates the retrieval of individual
differences in dimensionality. More recently, Peterson and Scott (1983)
have employed MDS to reveal the number of dimensions that individuals
employ when perceiving lists of nations. In so doing, these authors have
obtained considerable convergent validity between MDS-dimensionality and
other measures (checklist description and rating scale approaches) of
dimensionality.
Because of its usefulness, multidimensional scaling warrants
further discussion here. MDS assumes that the similarities (or
dissimilarities) between stimuli may be portrayed in a multidimensional
space. The associated spatial representation is derived by having
subjects rate the similarity of pairs of objects. In doing so, raters
impose an implicit set of attributes on their ratings. It is the task
of MDS to recover the implicit dimensions by estimating the number of
independent dimensions defining such a space. To do this, the INDSCAL
algorithm produces a geometric representation of the psychological
distance between pairs of stimuli. It is assumed that for each subject
12
there is a unique configuration of stimulus coordinates. From this
configuration, the procedure extracts the best n-dimensional space for
the group. Dimensionality is defined as the scaling solution (number of
retained dimensions) beyond which additional dimensions do not
significantly increase the proportion of variance accounted for (Kruskal
& Wish, 1978). Unlike traditional MDS techniques, INDSCAL accommodates
individual Variation while the traditional approaches treat such
Variation as error.
Though beyond the scope of the present paper, integration may also
be inferred from MDS analyses. Schroder et al. (1967, pp. 180-1)
summarized MDS°s usefulness in this regard, indicating that: (1) The
presence of multiple dimensions provides suggestive evidence for a
schema. MDS dimensions, they argue, are by definition integrated. A
dimension will not be extracted unless it contributes to each judgment
of similarity. (2) The number of dimensions uncovered by MDS analysis
reflects the number of sources of information. One may also determine
how important each source of information is to the rating of similarities.
This information is found in the individual subjects° weights. (3) "Often
the meaning of MDS-dimensions may be revealed by the arrangement of
stimuli on the dimension" (p. 180). The upshot of this summary of the
usefulness of MDS is that MDS may provide more information than merely
the number of dimensions which contribute significantly to average
similarity judgments.
This section discussed the relative advantages and disadvantages
of several methods of operationalizing CC-dimensionality. It was argued
that, although a variety of operations have been applied to the
p 13
measurement of CC-dimensionality, one potentially useful approach (MS)
has received insufficient attention. This operational definition may be
a useful means for examining CC-dimensionality in work settings.
COMSTRUCT YALIDITY ISSUES
Construct validity (the extent to which a measure assesses what it
purports to measure) is assessed not by a quick yes-or—no assessment but
by a painstaking accumulation of evidence. This evidence includes: (1)
the reliability (e.g., test-retest) of the measure; (2) the internal
consistency reliability of the instrument (does the measure assess one
construct or a number of constructs?); (3) the content validity of the
measure (i.e., is the adequacy of the sampling of the content domain);
(4) prediction (does the construct "behave" as it "should" -- that is,
does it correlate with the constructs the theory predicts it should and
remain uncorrelated with other constructs?). Por this latter line of
investigation, questions of convergent and discriminant validity
(Campbell & Piske, 1959) must be assessed. In other words, the measure
should correlate highly with other measures of the same construct
(convergent validity) and correlate little with similar-method measures
of different constructs (discriminant validity).
A fundamental consideration is the reliability of any measure
because reliability sets the limits for validity. Schneier (1979) reports
a test—retest reliability of .82 (p <.00l) for the REP. However,
Goldstein and Blackman (1978) have reviewed the reliabilities of various
modifications of the REP. These authors found test-retest reliabilities
ranging from .46 to .86. Their conclusion is that although the
14
test-retest reliabilities are statistically significant, they are
"somewhat below the level that is generally expected" (p.11l). In their
literature review, O°Keefe and Sypher (1981) found test-retest
reliabilities below .70 for most measures of complexity. Furthermore,
0°Keefe and Sypher (1981) report that the REP is unusually sensitive to
Variations in its administration and procedures.
Regarding internal consistency, as indicated in this introduction,
research has shown that cognitive complexity is a multidimensional
construct. As O°Keefe and Sypher have concluded, the Various measures
of cognitive complexity may not all be assessing the same thing. This
fact would argue against overall cognitive complexity being internally
consistent. However, such a multidimensional feature does not preclude
the possibility that any one component of CC, such as dimensionality, may
be important. Indeed, in other areas of industrial-organizational
psychology (job satisfaction, for example), an examination of the facets
of a multidimensional construct has been somewhat more fruitful than
relying on the global measure.
The internal consistency issue is relevant to content Validation
in the sense that any effort to tap only one aspect or dimension of CC
does not sample from the entire content domain. It would appear that
content validity in such a case would be necessarily low. Again, however,
this does not preclude the possibility that one component of CC may be
worth examining empirically.
As indicated earlier in this introduction, the convergent validity
of CC-dimensionality has been explored by Peterson and Scott (1983). In
their attempt to arrive at fundamental measurement, these authors
15
examined the intercorrelations between several methods of measuring
cognitive dimensionality: (1) pair-wise comparison, (2) rating scale,
and (3) checklist description. The intercorrelations ranged from .53 to
.70 and were all significant (g<.O0l). In another study (Schneier,l979),
convergent validity was demonstrated between Scott°s (1962) measure of
complexity and the Bieri (1966) grid form of the REP (; = -.19, g<.05).
However, the correlation, while statistically significant, was
unimpressive. Despite the apparent relatedness of various CC-measures
revealed by Peterson and Scott (1983), other studies show that the
convergent validity issue is less certain. For example, O°Keefe and
Sypher (1981) report intercorrelations between Bieri's REP and Crockett°s
(1965) measure, ranging from -.19 to +.14.
Evidence related to the discriminant validity of CC suggests that
CC does not appear to be significantly related to such variables as
intelligence and verbal ability. (Bieri & Blacker, 1956; Crocket, 1965;
Leventhal, 1957; Scott, Osgood, & Peterson, 1979; Wicker, 1969).
Streufert and Streufert (1978) have reviewed the literature concerning
this issue and found that any potential relationship between cognitive
complexity and intelligence appears to be at the lower end of
intelligence. That is, persons of very low intelligence are not likely
to be multidimensional. More recently, O‘Keefe and Sypher (1981) reviewed
the literature and again found that correlations between between
cognitive complexity and intelligence tend to be low and nonsignificant
(-.20 to .20 for Bieri°s REP measure).
Cognitive complexity also appears to differ from field
dependence/independence (FD/I). Field dependence/independence (Witkin,
16
Dyk, Faterson, Goodenough, & Karp, 1962) reflects the tendency of people
to rely either on the field or themselves as referents. In other words,
field-dependent individuals rely on some aspects of the environmental
field when they view stimuli. Field-independent individuals rely on some
internal orientation or standard. This is a different notion from
differentiation along the various dimensions or constructs used in one°s
conceptualization of the world (Witkin & Goodenough, 1977).
Indeed, Streufert and Streufert (1978) have outlined the nature of
the distinction between these two different usages of the term
"differentiation." They cite Witkin, Goodenough, and Karp (1967), who
viewed psychological differentiation as a kind of differentiation
relevant to perceptual-motor tasks. In contrast, according to Streufert
and Streufert, "differentiation" for research in complexity theory
concerns "cognitions involving verbal °concepts,° 'dimensions,° etc." (p.
14). Bieri (1971) suggests that cognitive complexity is a more specific
notion than FD/I. CC, he asserts, pertains to differences in the way
people process information about their social worlds.
Regarding spatial ability, Thompson, Mann, and Harris (1981)
conclude that cognitive complexity is related to spatial task performance
for males but not for females. On the other hand, FD/I correlates highly
with a number of spatial ability tests (Vernon, 1972).
Research reported in this section illustrates some of the
construct-validity issues which have proven problematic for cognitive
complexity researchers. It was shown that (1) reliability, which
circumscribes the limits for validity, was somewhat weak; (2) the content
validity for measures of cognitive complexity is generally low; (3)
17
internal consistency is low; and (4) there is mixed evidence for the
convergent and discriminant validity of the construct.
POTEELIAL IMPORTAECE OF COGEITIYE COMPLEXITY IE EORK SETTINGS
This section describes the potential importance of
reoperationalized cognitive complexity. If MDS-dimensionality can
actually be demonstrated in this context, the implications for behavior
in work settings are enormous. As Jones and Butler (1980) pointed out,
it is likely that cognitive style influences the manner in which
behavioral outcomes and causal inferences about those outcomes are
derived. This may be so because raters who are constrained by their
ability to dimensionalize stimuli may err in the judgments that underlie
performance appraisal systems and wage structures. Thus, job analysis,
job evaluation, and performance appraisal could be particularly
vulnerable to the effects of CC. Furthermore, job design efforts, which
are predicated on rating judgments, would be highly susceptible to the
influence of such cognitive structuring variables. A brief discussion
of the literature in the areas of job design and performance appraisal
follows.
As indicated above, the job—design literature is relevant for two
reasons: (1) job design is based on the perceptual dimensions underlying
jobs and (2) job design efforts, which are based on rating judgments, may
be susceptible to rating errors which in turn may be influenced by
cognitive dimensionality. The perceptual dimensions underlying jobs have
both theoretical and practical importance in the area of job design.
q 18
From a theoretical perspective, one may have an interest in the
manner in which people perceive their jobs. As a practical matter,
practitioners cannot redesign jobs unless they know the dimensions along
which jobs should be redesigned. The task attributes approach (Hackman
& Lawler, 1971; Hackman & Oldham, 1975, 1976; Stone & Gueutal, 1985;
Turner & Lawrence, 1965) emerged to provide such dimensional information.
As noted by Stone and Gueutal (1985), most investigations of the
attributes underlying jobs have employed measures such as the Job
Diagnostic Survey (JDS) of Hackman and Oldham (1975), the Yale Job
Inventory (YJI) of Hackman and Lawler (1971), the Requisite Task Attribute
(RTA) of Turner and Lawrence (1965), and other similar questionnaires.
Stone and Gueutal (1985) pointed out that all of the existing measures
of job characteristics are rooted in the work of Turner and Lawrence
(1965). However, since by their own admission, Turner and Lawrence
identified their underlying dimensions a prrgri, it is unclear how well
the RTA and other similar questionniares parallel employees° perceptual
dimensions.
In a similar manner, Hackman and Lawler (1971) and later Hackman
and Oldham (1975, 1976) refined and supplemented the concepts explored
earlier by Turner and Lawrence. The result was the Hackman and Oldham
(1975) Job Characteristics Model. Its five core dimensions are described
as fol lows:
1. Skill Variety (SV): reflects the degree to which the jobrequires incumbents to perform a variety of activities.
2. Task Identity (TI): concerns the extent to which the jobinvolves the completion of a whole unit of work.
19
3. Task Significance (TS): is the extent to which the jobmay significantly affect others.
4. Autonomy (A): is the extent to which the incumbentsare free to schedule their work activities andto determine procedures.
5. Feedback (F): is the extent to which the incumbentsreceive clear information about the effectivenessof their performance.
According to their model, the five core dimensions (task identity,
skill variety, task significance, autonomy, and feedback) relate to the
critical psychological states (experienced meaningfulness of work,
experienced responsibility, and knowledge of results). These critical
psychological states lead to personal and organizational outcomes such
as enhanced performance, increased satisfaction, reduced turnover, etc.
It should be noted that Growth Need Strength or GNS (an index of the
extent to which a person desires an enriched job) acts as a moderator
variable for both the relationship between core dimensions and critical
psychological states and the relationship between critical psychological
states and personal/organizational outcomes. In addition to describing
these five critical psychological states, Hackman and Oldham (1975)
arrived at a combinatory rule for the establishment of a unidimensional
index of job scope:
Motivating Potential Score = (SV + TI + TS)/3 x A x F
At first, the above research appeared to offer a framework for
future job-design applications and job—design investigations. Such
research also held the promise of recognizing the basis of individual
20
reactions to jobs. However, this early research was also vulnerable to
a number of criticisms.
For example, Dunham (1976), Harvey, Billings, and Nilan (1985), and
Roberts and Glick (1981) have pointed out that there is a lack of
independence among the dimensions. This finding contradicts Hackman and
Oldham (1975), who argue that the core dimensions are independent.
Roberts and Glick (1981) argue further that a five-factor model, such as
Hackman and 0ldhams°s, is not consistently demonstrated in the
literature. Indeed, studies by Aldag, Barr, and Brief (1981), Dunham
(1976, 1977), Dunham, Aldag, and Brief (1977), Lee and Klein (1982),
Pierce and Dunham (1978), and Pokorney, Gilmore, and Beehr (1980) have
failed to capture the factor structure of the traditional Hackman and
Oldham (1975) job characteristics model. Roberts and Glick (1981) also
criticized such traditional job-characteristics research for, among other
things, inadvertently capitalizing on common method variance and ignoring
the question of perceptual vs. objective task assessments. The
method-variance problem in the use of the JDS has also been documented
by Harvey et al. (1985). The issue of perceptions and their role in task
assessment is one of the major criticisms by Stone and Gueutal (1985) and
will be discussed more fully later in this section.
Other work casts further doubt on the earlier Job Characteristics
Model of Hackman and Oldham (1975). Having meta—analyzed a number of
factor-analytic studies, Fried and Ferris (1987) revealed that while
there appears to be more than one perceptual dimension underlying jobs
there may not be the five posited by Hackman and Oldham (1975). In a
recent attempt to shed additional light on the factor structure of jobs,
21
Kulik, Oldham, and Langer (1988) contrasted the factor structure of the
Job Diagnostic Survey (JDS) with a revised version by Idaszak and Drasgow
(1987). They found that the revised version conformed more closely to a
five-factor structure than the original JDS. Idaszak, Bottom, and Drasgow
(1988) reported further that -- with the exception of a new, somewhat
weak sixth factor -- the revised JDS has an invariant factor structure
across the five job characteristics of Hackman and Oldham (1975).
However, this finding does not negate the criticism that the Turner and
Lawrence (1965), the Hackman and Oldham (1975), and the Idaszak et al.
(1988) dimensions were not empirically derived.
In another study, Stone and Gueutal (1985) attempted to empirically
derive job dimensions. They reported that three dimensions underly jobs.
Those dimensions are: (1) the complexity of the job, (2) the extent to
which the employee works with the public, and (3) the physical demands
of the job. According to Stone and Gueutal, multidimensional scaling
stimulus coordinates for the first dimension correlate highly with levels
of "skill variety," "complexity of job," "job requires a great deal of
reasoning," etc. (pp.387-388). The second dimension indicates the extent
to which incumbents interact with people on the job and provide a service.
Finally, the last dimension is a reflection of the extent to which the
job involves working in an environment that is clean and safe as opposed
to dirty and unsafe. The 1985 Stone and Gueutal study provides the
advantage of determining subjects° own perceptual dimensions of jobs
rather than deriving them from a priori assumptions as did Turner and
Lawrence and, by extension, Hackman and Lawler (1971) and Hackman and
Oldham (1975, 1976). Stone and Ruddy (1987) have found evidence for the
22
generalizability of the Stone and Gueutal (1985) results to samples
outside the laboratory. More recently, Zaccaro and Stone (1988) have
reaffirmed the Stone and Gueutal model. These authors claim that the
dimensions of jobs are probably best protrayed by a model different in
kind from the traditional five-dimensional job characteristics model. In
addition, they assert that such early approaches unnecessarily "restrict
the capacity to account for variance in such criteria as job satisfaction
and intent to leave an organization" (p. 249). They report greater
success at accounting for variance in job satisfaction with the use of
such empirically derived job dimensions as the Stone and Gueutal
dimensions. For example, their results suggest that two of three Stone
and Gueutal dimensions (those approximating complexity and physical
demands/danger) significantly predict overall satisfaction over and above
motivating potential score (MPS).
It would seem that the direction of current research may be shifting
toward an embracing of the Stone and Gueutal dimensionality. This is
despite the fact that research rooted in the job-analysis tradition
appears to bear a somewhat different message concerning the factor
structure of jobs (Harvey, Friedman, Hakel, 8 Cornelius, 1988). Using
the Position Analysis Questionnaire (PAQ) and the Job Element Inventory
(JEI), Harvey et al. (1988) found that the five-factor structures of the
PAQ and JEI are similar. The factors revealed by Harvey et al. are: (1)
decision-making/communication/general responsibility, (2) skilled job
activities, (3)information-processing activities, (4) physical
activities/related environmental conditions, and (5) using equipment/
providing service (p. 641). Although apparent similarities between the
l 23
Harvey et al. and Stone and Gueutal dimensions exist (complexity vs.
skilled job activities; physical activities/environmental conditions vs.
physical demands; using equipment/providing a service vs. works with the
public), the number of dimensions retrieved by Harvey et al. differs from
the number captured by Stone and Gueutal. Moreover, the use of the job
analysis technology may be preferable to requiring subjects to report on
their perceived similarity of job titles.
In view of the brewing controversy, additional work documenting the
replicability of Stone and Gueutal's (1985) three-dimensional solution
seems warranted. The present study, with its large sample size, provides
an excellent opportunity for such an investigation.
Job—design research (Cherrington & England, 1980; Stone, 1976;
Stone, Mowday, & Porter, 1977) has suggested also that a number of
individual—difference variables (for example, desire for an enriched job,
need for achievement, and work values) may interact with job
characteristics to influence incumbents‘job perceptions. These
perceptions may in turn influence job outcome measures such as
performance, general satisfaction, and satisfaction with the job itself.
A logical extension of such individual—difference studies is the
hypothesis that dimensionality also may moderate the relationship between
job characteristics and job characteristics ratings. According to this
line of reasoning, individual differences in ability to conceptualize
to—be-rated stimuli ought to affect ratings of the stimuli. But when
Stone and Gueutal (1985) used multidimensional scaling to examine job
similarity ratings for trichotomized cognitive complexity subgroups, no
effects for REP-CC were noted. Specifically, Stone and Gueutal found that
24
multidimensional scaling solutions did not differ significantly for high,
medium, or low cognitive-complexity subgroups. This null result
underscores the controversy surrounding the cognitive complexity
(dimensionality) construct. A clarification of cognitive
dimensionality°s role is one important aspect of the present research.
Another area which may be influenced by cognitive dimensionality
is performance appraisal accuracy. Most prior research concerning the
REP test and rater accuracy pertains to the accuracy of predicting
behavior. In other words, researchers have typically asked the question,
"How well does cognitive complexity (operationalized by the REP)
correlate with peoples° accuracy of predicting others° social behaviors?"
Bieri (1955), for example, found that subjects who were high in REP—CC
were better predictors of others° behaviors than were their low-CC
counterparts. This finding has been supported by subsequent research
(Addams—Webber, 1969). However, it should be noted that in the Bieri
study, the author examined the relationship between REP scores and the
efficacy of predicting a classmate°s responses on a social-situations
test. As Goldstein and Blackman (1978) point out, the results were
significant only when a one-tailed test for significance was used. In a
variation of this line of investigation, Leitner, Landfield, and Barr
(1974) summarized the literature by concluding that cognitive complexity
does correlate with accuracy of predicting differences but not
similarities in 0thers° behavior. Their conclusions have been refuted
in a recent study by Lal and Singh (1984), who found that one°s level of
cognitive complexity did not predict one°s judgment accuracy. It would
seem that CC°s ability to predict interpersonal behavior is uncertain.
25
Addams-Webber (1969) explored the hypothesis that one°s level of
cognitive complexity can predict the accuracy with which one predicts
another°s constructs. The results indicated that high-CC subjects were
better at identifying the constructs of their partners in the study than
were low-CC subjects. On the other hand, Sechrest and Jackson (1961)
investigated whether or not high-CC raters would better predict the
construct systems of others. Results indicated an insignificant
relationship between complexity and accuracy. However, the ambiguous
evidence regarding CC-REP°s ability to predict both interpersonal
behavior and others‘ constructs does not address the issue of REP°s
potential ability to predict performance appraisal accuracy. It should
be noted that neither the Addams—Webber nor the Sechrest and Jackson
studies investigated performance appraisal accuracy. Unfortunately,
early research in the area of CC has given little attention to performance
appraisal accuracy.
To discuss properly the issue of performance appraisal (rater)
accuracy, it is first necessary to identify pertinent definitional
issues. Separate from the matter of cognitive complexity
(dimensionality), rater accuracy in performance appraisal situations has
received considerable attention in the literature (Becker & Cardy, 1986;
Borman, 1977; Cronbach, 1955; Murphy, Garcia, Kerkar, Martin, and Balzer
(1982). Cronbach°s landmark (1955) paper established that rater accuracy
may be broken down into four components: elevation (E), differential
elevation (DE), stereotypic accuracy (SA), and differential accuracy
(DA).
26
Thus, ratings will depend on the overall level of a rater's ratings
(E), the average rating for the dimension upon which a ratee is judged
(SA), the average rating for each ratee across dimensions (DE), and the
residual or (DA). According to Cronbach, DA indicates the rater°s
ability to differentiate between ratees on any particular item.
Differential accuracy, as it is defined, is really more properly
rater INACCURACY. This is so because accuracy is conceived as the overall
difference between ratings and true scores, hence inaccuracy. Similarly,
the components of accuracy (of which differential accuracy is one) reflect
aspects of the overall difference between ratings and true scores. In
other words, a score of zero indicates perfect accuracy for that
component.
A few additional comments concerning DA are relevant. Differential
accuracy (or rather inaccuracy) is described by Murphy et al. (1982) and
Cardy and Dobbins (1986) as the "finest grain measure of accuracy" (p.
674). This is because differential accuracy is a measure of the
difference between ratings and true scores (when controlling for overall
rating level as well as rating levels for both dimensions and particular
ratees). In other words, Cardy and Dobbins suggest that DA is a measure
of non-systematic noise. Cronbach's equation for differential accuracy
is as follows:
[2]Il ij
27
Note that rij and tij are rating and true score for ratee i on dimension
j; ?i_ and Ei_ are mean rating and true score for ratee i; F_j and F_j
are mean rating and true score for dimension j; and F__ and F__ are mean
rating and true score over all ratees and dimensions (Becker 8 Cardy,
1986). (This equation will be used in the analyses described later.)
Additionally, other components of inaccuracy (elevation,
differential elevation, and stereotypic) may be calculated according to
the following formulaez
a2=(?__ —E__)2 [3]
DEZ = 1/n;[(?i_ - ?_ _) — (EL - E_ _)]2 [4+}1 .
SA2=1,’kX[(?_j—F__)—(Z-j_ —E__)]2‘ [5]1
Cronbach (1955) also identified subcomponents of differential accuracy,
differential elevation, and sterotypic accuracy -- DACOR, DECOR, and
SACOR respectively. The equation for these are:
DAr = r ' '‘ [6]
r,t
DACOR
where r indicates ratings, t indicates true scores and (°) indicates that
these values are deviation scores.
DEr [7]1. 1.
DECOR
28
where ?i_ is the mean rating for each ratee across all dimensions and
Ei_ is the mean true score for each ratee across all dimensions.
SAr = r;·j E~j [8]
SACOR
where F_j is the mean rating for each dimension across all ratees and
?_j is the mean true score for each dimension across all ratees.
Cronbach (1955) and later Becker and Cardy (1986) have pointed out
that the above subcomponents may be important aspects of ratings, which
should be examined. Cronbach (1955) and Becker and Cardy (1986) also
have contended that one‘s choice of an accuracy measure ought to be guided
by the purpose of the research.
In the area of performance appraisal, research has provided some
evidence that individual difference variables such as cognitive
complexity may affect raters° perceptions. For example, Leventhal (1957)
determined that cognitive complexity (CC) influences one°s perceptions.
Campbell (1960) revealed that people scoring low on CC were more likely
to categorize ratees as either good or bad than were their high-CC
counterparts. Some investigators of performance appraisal have suggested
that such cognitive individual-difference variables may influence rating
behavior (Bieri et al., 1966; Feldman, 1981; Schneier, 1977). Feldman
(1981) adds that "since cognitive complexity is based on the number of
categories and relationships among the categories used by an individual,
it is not surprising that those with a more differentiated category system
would be able to make the relevant judgments with less error" (p. 141).
29
This, of course, assumes that highly differentiated raters do not, gg
ayegage, attend to irrelevant dimensions.
Schneier (1977) examined the effects of REP—cognitive complexity
(high versus low) and rating format (Behavioral Expectation Scales versus
graphic rating scales) on response-set errors in a rating task. Results
indicated that cognitively complex raters gave lower average ratings
(i.e., were less lenient) when using the BES than when using the alternate
format. Conversely, cognitively simple raters gave lower average ratings
when using the alternate format than when using the BES format.
Cognitively complex raters had less halo than did cognitively simple
raters for both the BES and alternate formats. Regarding restriction of
range, cognitively complex raters had less restriction of range than
cognitively simple raters for the BES format. For the alternate format,
restriction of range results were not significant. Schneier°s efforts
were directed at supporting his theory of cognitive compatibility. This
theory suggests that complex raters do better with complex rating formats
such as BES whereas cognitively simple raters do better with simple
formats such as graphic rating scales. In addition, his findings suggest
enhanced accuracy for cognitively complex raters.
Contrary to Schneier's findings, Lahey and Saal (1981) assert that
the combined effects of rater complexity and rating formats do not predict
rater accuracy. These authors examined cognitive complexity, using REP,
factor analytic data from REP, and a sorting task. In each of its
operational definitions, cognitive complexity failed to significantly
predict rater accuracy and failed to interact with rating format as
Schneier°s cognitive compatibility model would suggest.
30
Other evidence poses difficulties for the validity of CC as
previously operationalized. For example, Weiss and Adler (1981) have
questioned the validity of cognitive complexity as an individual
difference variable in rating tasks. They found that subjects differing
in CC did not necessarily differ in their leadership prototype ratings.
Bernardin and Cardy (1981) examined the effect of REP-cognitive
complexity on rating accuracy. No significant effects were revealed by
that study. One other study, Sauser and Pond (1981), involved a subgroup
analysis (high versus low cognitive complexity subjects) to determine
whether CC moderated the "effectiveness with which ratees are evaluated"
(p. 571). The results suggest that REP-CC did not moderate rating
effectiveness.
As indicated earlier, Bernardin et al. (1982) also failed to reveal
effects for REP-CC. In a series of studies, Bernardin et al. concluded
that (1) CC was unrelated to rating accuracy, (2) CC was unrelated to halo
error and rating formats when police sergeants rated patrol officers, and
(3) the construct validity of CC was not supported in a study of student
ratings of managerial performance and instructor effectiveness. These
results tend to confirm Sechrest and Jackson's (1961) finding that the
REP is not significantly correlated with differential accuracy in
ratings. Thus, although a logical case may be made for the existence of
a cognitive structuring variable (such as CC-dimensionality) in rating
tasks, experimental evidence supporting the viability of this construct
is ambiguous at best. For every study purporting to demonstrate the
effects of CC, there are studies arguing the opposite. Unclear at this
point is whether or not the difficulties encountered by investigators
31
of cognitive complexity are the result of: (a) the lack of evidence for
the validity of the construct itself, or (b) the operationalizations of
the construct. Accordingly, additional effort at resolving the
operational issues seems warranted.
The present study makes additional contributions. Although MDS has
been employed in previous studies of the dimensions underlying jobs,
several differences between those studies and the present one are notable.
When Burton (1972) used MDS to reveal the perceptual dimensions underlying
job titles, he did not constrain raters to base their ratings upon the
work activities for the job titles. Although Stone and Gueutal (1985)
prescribed that the ratings concern work activities, these authors
examined group data; they did not assess individual differences
MDS-solutions. Dubin, Porter, Stone, and Champoux (1974) also examined
the dimensions underlying job titles, but were concerned with dimensions
more as a function of organizational role than as a function of cognitive
individual differences variables such as cognitive complexity
(dimensionality). Peterson and Scott (1983) considered the convergent
validity of MDS dimensionality and other measures of dimensionality
within the context of their effort to arrive at dimensiona1ity°s
fundamental measurement. However, they concerned themselves neither with
job activities nor with an external criterion such as rating accuracy.
In contrast, this study presents a methodology for examining
individual differences in reoperationalized CC within a performance
appraisal context. Moreover, the present study requires the raters to
dimensionalize job activities. The convergent and predictive validities
of MDS-dimensionality also are considered. Finally, the present study,
32
with its large sample size and methodology parallel to Stone and Gueutal
(1985}, provides an excellent vehicle for assessing the generalizability
of those authors° three-dimensional model of job characteristics.
SQMMARY AND HYPOTHESES
In sum, it has been suggested that efforts to investigate cognitive
complexity have been constrained by the operationalization of the
construct. Various methods of operationalizing CC (REP, semantic
differential scales, etc.) have been discussed. The approach suggested
by Scott et al. (1979) —- that multidimensional scaling be used-- appears
to be a fruitful method for freeing CC from its operational constraints.
Indeed, it has been argued that MDS dimensionality is closer
conceptually to notions of dimensionality than is REP-cognitive
complexity. Consistent then with the recommendations by Bieri et al.
(1966) and Scott et al. (1979), the present study employs the MDS
operational definition of cognitive complexity (dimensionality) in order
to identify individual dimensionality. In addition, the apparent domain
specificity of cognitive dimensionality appears to require the use of jobs
as a stimulus set.
Accordingly, a methodology detailing the selection of such a
stimulus set is outlined in the following pages. As indicated above, the
present paper offers several extensions of the literature on cognitive
complexity. This paper suggests that job duties are a focus in a
job-related activity and that individual differences in MDS
dimensionality are at issue. The paper also employs an external criterion
33
and assesses both convergent validity and differential correlation (r
REP,DA·vs. r MDS,DA). The following hypotheses are examined:
Performance Ratings. Although past efforts to operationalize
cognitive complexity (dimensionality) have yielded disappointing results,
the literature suggests that an MDS operational definition may be
predictive in performance appraisal tasks. This leads to Hypothesis 1.
Hypothesis l states that MDS dimensionality will besignificantly and negatively related to rater differentialinaccuracy (noise). Thus, raters who have available manydimensions with which to perceive their worlds will be lessinaccurate (will produce ratings with less non-systematicnoise) than will raters who have few dimensions available.
But a test of Hypothesis 1 leads only to the conclusion that
dimensionality and accuracy are related. Although this issue is
important, an issue of greater importance is whether or not MDS
dimensionality better predicts rater accuracy/inaccuracy than does the
REP test. Accordingly, an additional hypothesis must be tested.
Hypothesis 2 states that MDS dimensionality will predict raterinaccuracy in performance ratings better than will theREP test. In other words, MDS dimensionality will bemore highly related to rater inaccuracy than will REPcomplexity.
Another matter of importance is the issue of convergent validity (Campbell
& Fiske, 1959). Any effort to tap a construct ought to be correlated with
other measures of that same construct. But, as explained earlier, the
REP measure of CC is not isomorphic to the concept being measured. For
example, although Scott et al. (1979) have found moderate
intercorrelations between measures of dimensionality, Goldstein and
Blackman (1978) have argued that -- unless derived from a common
instrument format -- measures of dimensionality are not intercorrelated.
34
Despite this apparent controversy, one would expect that the two
operational definitions of CC would be related but not isomorphic.
Indeed, evidence for convergent validity is expected. But because REP
is a less direct measure of CC than is MDS, convergence is expected to
be moderate. Hypothesis 3 follows from this point.
Hypothesis 3 states that MDS dimensionality and REP-CC will besignificantly related, but not isomorphic.
Frequency Ratings. In addition to these hypotheses, some
additional questions arise. For example, is dimensionality related to
the accuracy of frequency judgments? As indicated earlier, one criticism
of the Bernardin et al. (1982) behavioral observation task was that it
did not require complex judgments and that therefore MDS dimensionality
and/or the REP may not be related to such ratings. However, others
(Murphy et al., 1982) present a differing view. These authors have noted
that, even though behavioral observation and performance evaluation are
not totally independent of each other, they do require different
processes. Behavioral observation involves such judgments as the
frequency of a behavior°s occurrence. Performance ratings, on the other
hand, require evaluative judgment concerning the quality of performance.
Indeed, Murphy et al. suggest that accurate behavioral observation is a
requirement of accurate performance rating. It is therefore of interest
to examine the question of whether or not the present data reveal any
relationship between dimensionality and frequency ratings.
l) Does MDS dimensionality predict frequency ratings?
2) lf MDS dimensionality does predict frequency ratings,is it a better predictor than the REP?
35
Joh Qhgracteristics. In light of the continuing controversy
concerning the Hackman and Oldham (1975; 1976) Job Characteristics Model
and the more recent Stone and Gueutal (1985) three-dimensional model of
job characteristics, the present study will also examine the issue of the
replicability of Stone and Gueutal°s (1985) findings. The present data
will permit an examination of the overall (group) MDS solution for its
agreement with Stone and Gueutal°s three-dimensional solution.
Hypothesis 4 states that there will be congruence between thecaptured dimensions of Stone and Gueutal (1985) and the retrieveddimensions of the present study.
The remainder of this paper outlines the methodology and results
of the study. A discussion of its implications and limitations is then
presented. Finally, possible directions for future research are
discussed.
36
MEIQQD
OYERYIEK
A pilot study was conducted to develop materials for the main data
collection phase. The purpose of this preliminary investigation was to
assure adequate familiarity of subjects with the stimulus materials that
were used later. Additional pre-testing assured that subjects understood
the directions for the measures.
In the main study individuals° cognitive complexity was to be
estimated, using the number of independent dimensions along which
subjects perceived stimuli to vary. This estimation involved tapping the
underlying dimensionality of subjects° perceptions of job titles. Then,
dimensionality was to be correlated with rating accuracy to determine the
predictive validity of cognitive dimensionality. Also in the main data
collection phase, subjects completed a version of the REP, modified to
include job titles. This REP was included so that a determination could
be made whether or not dimensionality via MDS better predicted rating
accuracy than did the REP. In addition, the hypothesized convergent
validity between a traditional measure of cognitive complexity
(dimensionality) and MDS dimensionality could be established. Because
the job-titles REP was used (i.e., the content domain of the two
predictors was held constant), significant differences in predictiveness
between the REP (jobs) and dimensionality would suggest that the procedure
rather than the content of the measure is important.
37
The dependent measure was differential accuracy (DA) in performance
appraisal ratings. DA was determined as follows: Subjects were asked
to rate the teaching effectiveness of four instructors presented on a
videotape. The videotape consisted of four vignettes employed by Murphy
et al. (1982). (The videotaped vignettes are described more fully in the
procedures section.) Differential accuracy was assessed according to a
procedure developed by Cronbach (1955) and later employed by Murphy et
al. (1982). Differential accuracy scores for each subject were used in
subsequent analyses.
Finally, correlations between dimensionality and the REP, between
dimensionality and DA, and between the REP and DA were standardized,
using Fisher°s r to Z transformation. A determination whether or not
dimensionality was more predictive than the REP was then made.
PILOT STUD!
Pilot Study Sample
The pilot study sample consisted of 39 (20 male and 19 female)
undergraduates in psychology at a large southeastern university. A
laboratory sample was employed since findings by Stone and Ruddy (1987)
indicated that the dimensionality which underlies college student ratings
is similar to the dimensionality of subjects drawn from neighboring
communities. Extra course credit was offered to these participants.
Pilot Study Measures
The materials for the pilot study consisted of: (1) a set of 20
job titles (to be described below); (2) the job titles familiarity
38
questionnaire; and (3) biographical questionnaires which ask respondents
for demographic information concerning themselves.
Based upon prior research (Schroder et al., 1967; Scott et al.,
1975; Sechrest & Jackson, 1961; Streufert & Streufert, 1978; Streufert,
1982), cognitive dimensionality is best examined within a particular
domain. Job titles and/or performance dimensions would appear to offer
job-relevant content domains. However, the use of performance dimensions
would impose dimensionality upon subjects. Thus, the use of job titles
seems preferable.
Earlier, it was noted that Stone and Gueutal (1985) found no
significant effects for trichotomized CC subgroups on job-dimensionality.
One matter of interest, then, is whether the operationalization of CC
accounted for Stone and Gueutal°s null results for CC. To facilitate an
extension of the Stone and Gueutal (1985) study, the present project will
retain job titles as a stimulus set.
Job titles (n=20) that were used for this study were those selected
by Stone and Gueutal (1985) in a procedure described as follows: Titles
from a larger stimulus set (n=52) were originally drawn from the
Dictionary of Occupational Titles (1965). To reduce their original set
of titles, these researchers first determined the subjects° (88
undergraduate management students°) familiarity with the target jobs.
The familiarity ratings ranged from 1 ("not at all familiar") to 5
("totally familiar"). Job titles eliciting a rating of greater than or
equal to 3.0 were retained in the stimulus set. A final reduction in the
number of stimulus job titles was accomplished by selecting those jobs
which were "(a) both heterogeneous with respect to such criteria as job
39
incumbent°s "collar color," employment sector, job status, and other
variables, and (b) generally representative of civilian jobs" (p. 381).
Retained job titles included cashier, janitor, fire fighter, telephone
operator, bank teller, mail carrier, taxi driver, file clerk, accountant,
short—order cook, actor, assembly—line foreman, machine operator,
elementary school teacher, flight attendant, waiter/waitress,
construction worker, tailor, lawyer, and retail store manager. The job
titles were randomly ordered on the pilot study questionnaire.
Directions for the pilot study required each subject to rate his/her
familiarity with the characteristics of the work activities of
incumbents. The purpose of this phase of the project was to assure that
subjects were sufficiently familiar with all of the items used in the
questionnaires. This measure is included in Appendix A. The ratings
scales ranged from l ("not at all familiar") to 5 ("totally familiar").
A demographic questionnaire (Appendix B) also was administered to
subjects. This biographical questionnaire permitted the examination of
the potential effects on familiarity ratings of work history and other
demographic variables.
Pilot Study Procedure
The research was conducted in accordance with the principles for
the treatment of human subjects prescribed by the American Psychological
Association. Informed consent forms were completed by all participants.
Subjects were reminded that their participation was voluntary and that
they were free to withdraw from the study at any time. Names of
participants wishing further information about the project were recorded.
40
Subjects completed a biographical questionnaire and the familiarity
rating questionnaire. Subjects for the pilot study were debriefed and
thanked for their participation.
Job title familiarity ratings were analyzed to estimate whether or
not the stimulus words for the similarity measure would be sufficiently
familiar to subjects. As in Stone and Gueutal (1985), all job titles met
the criteria of a 3.0 mean familiarity rating and were retained in the
stimulus set. descriptive statistics for this measure are contained in
Appendix A.
MAI§ DATA COLLECTIO§ PHASE
Main Data Collection Sample
Three-hundred-and-five undergraduates who were enrolled in
introductory psychology completed the main data collection phase.
Ninety-five percent of the subjects have held paying jobs. Fifty—three
different job titles were represented by subjects' occupations. Subjects
were on average 18.893 years of age (range=17 to 31). Most participants
had worked at paying jobs (M=2.164 years, range= 0 to 14 years). Of the
original subject pool (n=344), 25 were dropped from the study because they
did not return for the second session. Thirteen were dropped because they
failed to complete one of the hypothesized predictor measures. One
subject was dropped from the study because that subject°s data prompted
a fortran warning during multidimensional scaling. Extra course credit
was provided to participants.
41
Main Data Collection Phase Measures
—The predictor measures for this phase of the study included the Job
Similarity Measure and the Modified Job REP. These appear in Appendices
C and D and are described more fully below. A biographical questionnaire
(Appendix E) also was included to permit the assessment of effects of
demographic variables such as work history, sex, etc. The criterion
measures were performance measures identical to those used by Murphy et
al. (1982). These questionnaires appear in Appendix F and are also
described more fully below.
The Job Similarity Measure was identical to that employed by Stone
and Gueutal (1985). Instructions for the questionnaire directed subjects
to rate the similarity of all possible pairs of listed job titles.
Specifically, when making similarity ratings, subjects were asked to
consider the job duties for each job title. Subjects were not given a
dimension on which to rate the comparability of each pair of words.
Instead, they imposed their own underlying dimensions. Each pair of words
was accompanied by a 6-point similarity scale. Anchors ranged from 1
("not at all similar") to 6 ("almost completely similar"). Word pairs
were randomly ordered within the similarity questionnaire. Although the
200 pairs that made up the list included only 190 unique combinations,
all 200 word-pairs were used in each set. The inclusion of duplicate
pairs allowed for the assessment of the reliability of the measures. This
strategy is identical to that employed by Stone and Gueutal (1985).
To determine the relative contribution of the dimensionality
approach, a REP-type guestionnaige was developed for use with job titles.
For this measure, subjects generated constructs that underlie their
42
perceptions of jobs. They were presented with lists of triads drawn
randomly from the list of twenty job titles in the pilot study (used for
the similarity measure). Subjects decided what made any two of the words
in a particular triad alike and yet different from the third word.
Subjects indicated this similarity by selecting a descriptive word.
Subjects also indicated the polar opposite of each descriptive word. For
example, one subject may have perceived the jobs, telephone operator/bank
teller/construction worker, using the pair of words "automated" versus
"manual". Such a bipolar pair could then be used in the second part of
the grid task, which involved assigning a numerical value to each job
title for each pair of bipolar opposites. Thus, for example, the
above-mentioned hypothetical subject may have rated the job, CASHIER, on
a scale of -3 to +3 for a total of ten bipolar opposites that were unique
to each subject.
Grids were scored according to the procedure outlined by Bieri
(1966) and described below. The rating in each row was compared to the
rating directly below it (for the same job) in the other rows on the
matrix. A score of one was given each time there was agreement. This
matching procedure was carried out for all unique row comparisons (900
comparisons in a 20 x 10 matrix). A high REP score indicates low
complexity. In other words, a high score indicates that the subject used
the same numerical rating frequently. Such a rater would employ construct
dimensions in an identical manner.
Regarding the performance measures (Appendix F), the first set of
performance-related questions asked subjects to rate the frequency with
which each of twelve behaviors occurred on videotaped performances of
43
instructors. (For example, "Lecturer spoke fast," "Lecturer gave clear
answers to questions," "Lecturer spoke in a monotone for a sustained
period.") Responses were anchored on a seven-point Likert-type scale
(l="never," 2="almost never," 3="a few times," 4="about half the time,"
5="often," 6="most of the time," 7="all of the time"). The second set
of questions asked that subjects rate the lecturer whom they had just
observed for such dimensions as thoroughness of preparation, grasp of
material, organization and clarity, speaking ability, overall rating,
etc. Responses to these questions were anchored on a five-point Likert
scale (l="very bad," 2="poor," 3="average," 4="good," 5="very good").
Main Data Collection Phase Progedures
This phase consisted of two sessions, the first lasting about an
hour to an hour and one-half and the second lasting about 30-40 minutes.
At the first session, each subject completed an informed consent form,
the Job Similarity Measure, the Job REP questionnaire, and the
biographical questionnaire. All subjects answered the questionnaires in
the same order because the presentation of the REP Grid task first could
have primed the subjects for the similarity task. To minimize possible
fatigue effects, the experimenter told subjects that they could take a
quiet break from the task if they felt their attention diminishing.
Approximately two weeks later, subjects returned to view four
videotaped lectures. The lectures were selected from the eight vignettes
developed by Murphy et al. (1982). Scripts for the videotapes (one
concerning crowding and stress and the other self-fulfilling prophecy)
were developed by holding constant the content across the two same-topic
44
lectures. However, the thoroughness and organization of the lectures and
the clarity of answers to questions varied both between and within the
same-content videotapes. For example, one script concerning crowding and
stress contained a good, well·organized lecture, but poor answers to
questions, while the other contained a poorly organized lecture but good
answers (Murphy et al., l982). Lecturers were drama students from a
Houston-area university. Low, low-moderate, moderate, and moderate-high
levels of performance were represented. True scores were established for
the videotapes by Murphy et al. (1982), modifying a procedure developed
by Borman (1977) and described more fully by Murphy et al. In essence,
Murphy asked "expert raters" to generate true scores by having them rate
the videotaped performances. These true scores (mean performance rating
across all the experts) facilitated the calculation of accuracy scores
that were used in the analyses.
Rating instructions informed subjects that they were participating
in a teacher evaluation study and that their ratings would be used for
research and feedback purposes only. Subjects were told that their
ratings would not affect the employment status and salary of the ratees
and that individual responses would be confidential. Subjects were told
also that shortly they would view four instructors on the videotape, each
one at a time. After subjects viewed each instructor, the experimenter
would stop the videotape and the subjects would rate the person they had
just observed. This aspect of the procedure was intended to minimize
potential memory effects. Subjects were asked to read the cover page of
the rating booklet and to familiarize themselves with the rating
questionnaire.
45
Following the rating task, subjects were fully debriefed.
Instructions for the rating task had suggested to subjects that ratings
would be for research and feedback purposes only and would not affect
ratees° employment status or salaries. At the debriefing, subjects were
told that, although the instructions suggested the ratees were teachers,
the persons they had observed were actually actors. Subjects were asked
not to discuss the nature of the experiment with anyone. Names of those
subjects wishing further information about the research were recorded.
Subjects completed extra credit point sheets and were thanked for their
participation.
46
RESULTS
RELIABILITY ESTIMATES
Scorer reliability for the REP Grid measure was assessed by randomly
selecting ten grids for rescoring by an independent rater. The interrater
reliability estimate was .96.
A reliability estimate for judgments indexed by the Job Similarity
measure was derived by correlating ratings on the ten pairs of duplicate
items. These correlations were calculated separately for each subject.
Individual subjects' correlations were transformed by Fisher's r to Z
transformation. The average Z transformed value was .829 (;.d.=.52).
This index of subject consistency compares favorably with prior
consistency estimates for this measure (average Z=.76) by Stone and
Gueutal (1985).
DERlyI§G DIME§SIO§ALITY
In order to examine dimensionality°s predictive validity, it was
necessary to derive subjects° perceived dimensions of job titles.
Following is a description of the results of this preliminary analysis.
Dimensionality was assessed by submitting pair·wise similarity
ratings for the 190 unique similarity items to non-metric,
individual•differences multidimensional scaling (INDSCAL algorithm).
Program option specifications were as follows: level = ordinal, similar,
shape = symmetrical, condition = matrix, model = INDSCAL, maxdim = 6,
mindim = 2, iterations=10, converge = .001, untie. It should be noted
47
that maxdim and mindim are set at the maximum and minimum (respectively)
allowed by the INDSCAL program. In addition, the iterations required to
attain convergence at .001 were always less than ten. First, the overall
(group) MDS solution will be reported.
As indicated earlier, INDSCAL employs a least-squares criterion for
fitting the data to the model. In other words, the iterative process
attempts to keep S-STRESS (overall "badness of fit") at a minimum. The
program iterates until some specified level of S-STRESS reduction is
achieved. Program output indicates R2 as well as Kruskal°s STRESS1, which
is another "badness of fit" measure averaged across subjects° matrices.
According to Davison (1983), the major difference between S-STRESS and
STRESS1 is that S-STRESS uses squared distances and squared disparities
while STRESSl does not square the distances and disparities. The results
indicated that a two-dimensional model accounts for .64 of the variance
(STRESS1=.245), a three—dimensional model for .67 of the variance
(STRESS1=.184), a four-dimensional model for .70 of the variance
(STRESSl=.137), a five-dimensional model for .73 of the variance
(STRESS1=.113), and a six-dimensional model for .74 (STRESS1=.096).
These data are presented in tabular format in Table 1 and in graphic
format in Figure 1. The data reflect values (variance accounted for)
which compare favorably with those Young and Lewyckyj (1979) termed4
"moderately good" (p. 103), but not excellent. INDSCAL also identifies
the importance of each dimension averaged across subjects. These data
as well as the stimulus coordinates for the four-dimensional solution are
contained in Table 2.
48
One convention (Kruskal & Wish, 1978) which guides the selection
of the appropriate dimensionality per number of stimuli is that I (the
number of stimulus objects) should be greater than AR (where R is the
number of dimensions). In other words, a six—dimensional solution may
require approximately 24 stimulus objects, while a four-dimensional
solution may require approximately 16 stimulus objects. The present data
(with 20 stimulus objects) satisfies this criterion for the
four—dimensiona1 solution.
More importantly, a determination of the dimensionality reflected
by the data depends also on the satisfaction of a number of other
criteria. The first is whether or not the addition of a particular
dimension contributes to a substantial increment in R2. Note that r for
the group solution is calculated by the following equation:
2 1/2 [9]K?·(1,j,s)*iö1js · 6. . .>“Z(1,j,s)<61js · Ö. . JJ
where 6* is the grand mean of the actual scalar products and 6* is the
grand mean of the estimated scalar products, ij refers to the stimulus
pair, and s refers to the subject (Davison, 1983). In other words, öij
refers to the proximity (distance) measurement of stimulus pair (i,j).
A proximity measure is an index of the degree of similarity of pairs of
objects. But INDSCAL provides more than a group stimulus space. As
indicated in the introduction of this dissertation, INDSCAL also provides
a direct means of assessing the importance of various recovered dimensions
for individual subjects. These importance weights are computed by the
49
>~•4-3••-I
»-1TI!Q r-1O »o cxV} „CIOE
-r-1Q
U1 C'1 Q*4-a cx qx .O O O A
. . QCI OO *:*4
·•-4 U2#3 E H
A U O <1.>A r: ··-4 m oo oo ><¤ ¤ m Q Q N cwCJ "•-4 E
•···• •··• C) WU3 Q; . • • 4}Q CU E WZ ·¤-1 xx•-1 V1 Qxx Q bl
<l>< : ~o Q- r~ m UH O 01 0 r~ Q N cnvg ~•-4 N •—• A
•-•Q
gg . . . „ ZO C:
»-—•4-3 O
E s-4•-1 ·•-1 Q}***3 O C "O
3 r\ oo •—< N Q Gzu E- .:: N co -—« oo «~ »o zuQ U N N -4 «-• 4-1Q >, :5 . . „ •
-Q)
ff.} 4J {-in *-*:-•···• -¤•-• $.4 QQ O >Q '4-a O 6*1 U1 U1 M) (DO
•'< U1 Q O Q Q *:-4·•-4 N 6*1 N N :-4 •—• Hgg „ . . • „ ,:4
Q OE) H~·-• >«Q r~ ¤o Q »¤ N 4-»
N 01 Q O N Q '•·*
s-4 xx ~D ~¤ r\ r~ 1~•-•
Q . . . . „ :5Q CI
· ON*-4Z V1
Z/J C2’UU1 V1 Q N 6*1 Q O
c: cu xt oo m —-« cx E:5 $.4 N
•-4 --4 1-4 O—»-4
$:3 • „ • „ • "QU3ua EO Z3x-« E4-3
"'*W C>x '·"*u E-»-1r—• G.)Q „-E
ä 4J--1 N 01 Q I-/1 Q vz5;; ·•4
G
E O.„, EQ r-I
50
LO
>~„
S2Q)„„.§E
QM-J
°.
N*6E
OO O3·»«¤%<¤$~?%?98 EEYU
SSSMS g2
S3’U
<°äNm
wm;C.93
<rg;§¤wä-
mg
C)
N
O O OO
51
TABLE 2
Stimulus Coordinates: Overall Solution (INDSCAL)
Dim. 1 Dim. 2 Dim. 3 Dim. 4
Cashier -0.7107 -1.4787 -0.9338 0.1874Retail Mgr. 0.9179 -1.2090 -0.6739 -0.9854Bank Teller -0.3921 -1.1328 -1.4195 0.5982Waiter -0.4785 -1.0359 1.0723 0.9322Cook -1.4438 -0.9054 0.7681 -0.5960Tailor 0.2856 -0.9645 0.8882 -1.5965Actor 1.4734 -0.2234 1.6309 0.9317File Clerk 0.2103 0.5466 -1.5393 0.4589Lawyer 2.0704 0.0571 -0.5335 0.7375Teacher 1.6552 0.3144 0.7455 -0.1756Tel. Op. -0.4855 0.5923 -1.0409 0.9007Taxi Driv. -1.6196 0.3028 0.3778 0.7758Firefighter 0.3490 2.2694 0.5504 0.0331Const. Work -0.4173 1.4743 0.2744 -1.5891Mail Carrier -0.6023 1.3408 -0.3423 1.3453Machine Op. -1.2478 0.5309 -0.7065 -1.2603Accountant 0.9779 -0.9189 -1.4132 0.3367Assembly Frmn. 0.4091 0.2979 -0.4236 -1.9071Flight Attend. -0.1503 -0.6380 1.0704 1.3692Janitor -0.8009 0.7802 1.6484 -0.4967
Variance 0.2051 0.1812 0.1742 0.1431
52
following equation:
{[2 · ·«6*- -6* Fm; ·(3*--€* 2 1/2 *10]
(1,j)~1js ..s (i,j) 1JS ..s)]}
where 6* and 6* are the mean actual and estimated scalar products for the
particular subject, s, (Davison, 1983). lt should be noted that r, the
correlation between each subject°s actual and estimated scalar products,
indicates the strength of the relationship between the dimension and the
subject°s matrix.
For the group space, R2 is plotted against the number of dimensions.
An "elbow" in the ascent of the plotted line indicates the dimensionality
beyond which additional dimensions are not of much importance in
interpreting the group space. The plot for the present study is presented
in Figure 1. This determination is, by necessity, a matter of judgment
since there are no tables for INDSCAL (Davison, 1983; Kruskal & Wish,
1978; Schiffman, Reynolds, & Young, 1981). Although Monte Carlo tables
have been developed for statistical hypothesis testing of more
traditional approaches to MDS, there are no such tables for INDSCAL. Nor
is there a convention such as in factor analysis. An additional
limitation is that techniques such as policy capturing via stepwise
regression requires an external criterion. Thus, while policy capturing
may provide additional information about the relative importance of
particular dimensions in the prediction of rater accuracy, the technique
is not appropriate for the "internal" statistical hypothesis testing of
similarity dimensions (Davison, 1983). The plot of Rz against the number
53
of dimensions appears to favor a four dimensional solution. However, it
should be noted that at either three or four dimensions R2 and Stress are
somewhat disappointing.
In addition to the criterion of proportion of variance accounted
for, other criteria are important. For example, as Shepard (1972) has
maintained, parsimony, stability, and visualizability are important.
Generally, a trade·off arises because when dimensions are added fit
improves at the expense of parsimony. The parsimony criterion would
suggest that, since little additional variance is accounted for by the
addition of the fifth and sixth dimensions, the group data is best
described by a four-dimensional solution. Indeed, Shepard contends that
except when using INDSCAL (as the present study does) dimensionality
should rarely exceed two or three dimensions. The stimulus coordinates
and variance accounted for by each factor are indicated in Table 2.
Because they maximize fit, data reduction techniques such as MDS
and factor analysis are susceptible to shrinkage. Accordingly, an
assessment of the solution°s stability, the second of Shepard°s (1972)
criteria, is important. This stability may be assessed by randomly
dividing the data into four subsets and comparing the results of the four
analyses. Therefore, four subsets, each containing one—fourth of the
subjects° matrices, were extracted from the data set. These subsets were
then analyzed separately by INDSCAL, using the same program options that
were used for the full data set. Stimulus coordinates for these data are
reported in Tables 3-6. The results appear to suggest that the original
solution is stable. A four-dimensional solution again appears to describe
the data (R2 =.717, STRESSl=.134; R2 =.653, STRESSl=.148; R2 =.737,
54
STRESSl=.132; and R2 =.720, STRESS1=.l35 for the four subsets
respectively). These results are presented in Table 7. Figures 2 and 3
illustrate R2 plotted against dimensionality for this subset analysis.
However, when coefficients of congruence are examined for the two subsets,
stability seems less certain. The Job Title Stimulus Coordinates were
used to determine coefficients of congruence between subsets A, B, C, and
D for each dimension. These data are presented in Tables 8-12.
Coefficients of congruence (Burt, 19&8; Wrigley 8 Neuhaus, 1955)
work directly from the scores rather than from the deviations (Cattell,
1978). Cattell calls the procedure "joint mean and deviation factoring,"
which retains information concerning the "variation about the means and
the absolute means themselves" (p.468). ‘The equation for this statistic
is found in Gorsuch (1983):
C12 = XPV1PV2.._ ._. [ll]
@*31 @*31
"where C12 is the coefficient of congruence between factor one and factor
two, Pvl are the factor loadings for the first factor and Pv2 are the
factor loadings for the second factor" (p. 285). The formula may also
be applied to stimulus coordinates in MDS.
Coefficients of congruence index the extent to which a factor or
dimension is stable (i.e., is reproduced) by another. Consider the
dimensional solution illustrated in Table 8.
55
There would be evidence of stability for dimension one if the
coefficients in column 1A were high (in the high .90 range) for one of
the dimensions in each of subsets B, C, and D, but low for the other
dimension within each subset. In Table 8, for example, column 1A should
indicate numbers in the high .90 range rather than -.87, -.83, and .94.
Conversely, coefficients of congruence for 1A with 2B, 2C and 1D should
all be much lower. In the present study, changing the dimensionality did
not substantially improve stability.
The third of Shepard°s criteria is visualizability
(interpretability). A visual inspection of the three and four dimensional
solutions is not particularly enlightening. And, as indicated earlier,
since external criteria were unavailable, statistical relationships
between the captured dimensions and particular external constructs could
not be empirically demonstrated.
In light of the aforementioned instability of the
individual-differences scaling solution, the data were unsuitable for the
tests of Hypotheses 1, 2, and 3. Therefore, the analyses did not proceed
to the tests of these hypotheses. Instead, the data were submitted to
group multidimensional scaling (ALSCAL or Alternating Least Squares
Scaling). Program options were level=ordinal, similar,
shape=symmetrical, condition=matrix, model=Euclid,maxdim=6, mindim=1,
iterations=10, converge=.OO1, untie.
Results revealed that a one—dimensional solution accounted for .579
of the variance (STRESSl=.373), a two-dimensional solution for .607 of
the variance (STRESSl=.249), a three-dimensional solution
56
TABLE 3
Stimulus Coordinates: Subset A (INDSCAL)
Dim. 1 Dim. 2 Dim. 3 Dim. 4
Cashier -0.6245 1.2302 -1.2014 0.5032Retail Mgr. -0.2179 -0.6645 -1.8402 -0.0607Bank Teller -1.1455 0.7552 -0.8372 1.1625Waiter -0.8364 0.9189 0.2477 -1.1448Cook 0.4267 1.5233 -0.4688 -1.0516Tailor 0.6493 -0.1298 -1.4118 -1.3439Actor -0.9841 -1.2069 0.5639 -1.6701File Clerk -0.2365 -0.4549 -0.0526 1.7253Lawyer -1.3280 -1.9705 -0.0520 0.2995Teacher -0.1718 -1.6497 0.2263 -0.7973Tel. Op. -0.0471 0.3123 0.4319 1.3956Taxi Driv. 0.1077 1.6425 1.0568 0.1011Firefighter 0.8102 -1.0343 1.8279 0.2197Const. Work. 2.1461 -0.3064 0.3794 -0.2798Mail Carrier 0.0147 0.3513 1.6669 1.1372Machine Op. 1.6059 0.8934 -0.2243 0.7314Accountant -1.3479 -0.4436 -1.1454 0.9814Assembly Frmn. 1.2847 -0.7055 -1.0450 0.3441Flight Attend. -1.2365 0.5373 0.8046 -0.8555Janitor 1.1307 0.4017 1.0731 -1.3971
57
TABLE 4
Stimulus Coordinates: Subset B (INDSCAL)
Dim. 1 Dim. 2 Dim. 3 Dim. 4
Cashier 0.3401 1.3399 0.6699 1.2448Retail Mgr. -1.1773 0.9337 -0.1445 1.0986Bank Teller 0.3054 1.6295 1.0394 0.2887Waiter 0.4152 -0.7986 1.0605 1.2019Cook 1.4921 -0.2337 -0.3312 1.3557Tailor -0.5230 -0.3084 -1.1178 1.6715Actor -1.3451 -1.7526 0.8217 0.0876File Clerk -0.0671 1.0849 0.7011 -1.2201Lawyer -1.9896 0.2041 0.4678 -0.9782Teacher -1.5084 -1.0689 -0.1278 -0.2319Tel. Op. 0.7941 0.5806 0.8990 -1.0068Taxi Driv. 1.5627 -0.7359 0.4760 -0.0816Firefighter 0.0066 -0.9835 -0.8053 -1.9818Const. Work. 0.4275 -0.1225 -1.9900 -0.7668Mail Carrier 1.0529 -0.1683 0.6934 -1.5154Machine Op. 1.0470 0.9659 -1.3613 -0.1931Accountant -1.2240 1.3922 0.5493 -0.0878Assembly Frmn. -0.4560 0.5823 -1.7798 0.2475Flight Attend. 0.1403 -1.0157 1.3814 0.7956Janitor 0.7065 -1.5251 -1.1007 0.0716
58
TABLE 5
Stimulus Coordinates: Subset C (INDSCAL)
Dim. 1 Dim. 2 Dim. 3 Dim. 4
Cashier 0.5131 1.6877 0.5497 0.1336Retail Mgr. 0.9807 0.5581 -0.6234 -1.5093Bank Teller 1.0851 1.3815 0.6202 0.2593Waiter -1.0951 1.2502 -0.4441 0.6526Cook -1.2412 1.1559 0.8089 -0.5761Tailor -0.8180 0.6324 -0.4036 -1.8103Actor -0.5468 -0.2756 -2.3876 0.6111File Clerk 1.3246 -0.0064 0.9559 0.3631Lawyer 1.6105 -0.1857 -1.5142 0.2670Teacher 0.0957 -0.7601 -1.6498 -0.4945Tel.0p. 0.9010 -0.5724 0.5997 0.9874Taxi Driv. -0.8931 0.0998 0.6307 1.5057Firefighter -0.5093 -2.2716 -0.2778 0.4578Const. Work. -0.7921 -1.5376 0.8181 -1.1141Mail Carrier 0.1803 -0.7879 0.6939 1.8077Machine Op. -0.1493 -0.4390 1.6575 -0.7178Accountant 1.6743 0.8743 -0.2449 -0.2149Assembly Frmn. 0.4135 -0.9377 -0.0029 -1.7396Flight Attend. -0.9994 0.6321 -0.7493 1.1100Janitor -1.7344 -0.5520 0.9629 0.0216
59
TABLE 6
Stimulus Coordinates: Subset D (INDSCAL)
Dim. 1 Dim. 2 Dim. 3 Dim. 4
Cashier 0.5140 -1.4909 -0.5167 -0.8782Retail Mgr. -0.7631 -0.7341 0.3722 -1.6712Bank Teller 0.1284 -1.5394 -0.9508 -0.5289Waiter 0.7025 0.7147 -1.3615 -0.5609Cook 1.4970 0.3097 0.1858 -1.1682Tailor -0.9615 0.3475 1.5558 0.0818Actor -0.9964 1.8103 -1.1437 -0.3810File Clerk -0.7520 -1.1046 -0.0733 1.1492Lawyer -2.1488 0.0708 -0.7694 0.3149Teacher -1.2778 1.1441 0.0246 -0.8875Tel.0p. 0.2485 -1.0539 -0.2868 1.0704Taxi Driv. 1.6510 -0.0153 -0.4546 0.6725Firefighter -0.1030 1.3711
l0.6160 1.9339
Const. Work 0.5640 0.4568 1.9065 0.8115Mail Carrier 0.4062 -0.1243 -0.5380 1.9754Machine Op. 1.0295 -0.9062 1.4775 0.3462Accountant -1.2828 -1.2289 -0.7672 -0.3958Assembly Frmn. -0.1443 -0.2555 1.5654 -1.3306Flight Attend. 0.5003 0.7707 -1.4949 -0.3735Janitor 1.1883 1.4575 0.6533 -0.1801
60
— TABLE 7
Stress and R2 for Dimensionality by Subset
(INDSCAL)
Subset Dimension Stress R2
A 2 .236 .6633 .178 .6854 .134 .7175 .111 .7336 .094 .754
B 2 .265 .5823 .199 .6094 .148 .6535 .121 .6816 .104 .695
C 2 .238 .6523 .175 .6974 .132 .7375 .107 .7606 .091 .778
D 2 .244 .6533 .178 .6804 .135 .7205 .109 .7536 .092 .761
61
LO
2m
‘°.Q__ (0
98 Sg ¢.€2
näE
cu.?I1fü
Eo00¤o0¤oo°§°Q'%¤Q¤¤<t"?^!·tQg
·v-4
ZH Q“6
CI
.3=¤‘2LE
@2< wm
2 gwU) wwmi-4
..€.§¤ga.N
OOOOOOO
62
LO
E¤ "°@é-9%
ESie .22Q2“°EÜ
NZQC
.3
@'¥“?LO’<ü"?"!fQ .3¤ZH"5
noäE
ID;uCmvv <T.9--.¤ mw
= c:N) 'JE39
Er;
cx:
OOOOOOOOre ¤¤ M2 <r ~2 N < @
63
OG ON „
•-4
¢‘¤ON•-J
¤ o•-<Q ,-4 „„
U FIO
I-1v 4=
O•·*NI: U omcQ N . . .
--4 •-IQQ4-* I3I-|O
U2 ·¥·I¢O<fP'*NQ ·-< U C>or~r~Q Q
•·-• ...•s: I: •-•oooQ} O IIQ
-•-4
H V)¤0 3E Q) { JF0 E ommcomU ·•-• ¢¤ Q¤O4*¤>N
Q Q N•••••
*4-I I •-IOOOO
EQ Q III
GU14uE-< : 14 4:4: 4:Q) O OIhf\€'7I~N
E *4-I E OOI-hGOU\·r·I V) PIOOOOOI4-1 Q II*+4 4-*4* ¢UO CI
'U OIh¤\<!'Ih<„°®E
:.1] om¤¤mr~¤o·-4
O FIOOOOOOL) I II
V}D
--43 -k ·k·!< ·¥=E omnr-nmO<t~1'
TJ E oo¤o<r¤o¤.nmc~I/I FIOOOOOOO
IIIII
IhO
E V'PI ¢¢GGUUGG B4O
•—*N•·¢Nr*N•·•N {
64
Q(*1.
· vd
O4Q OON
„•
PIC
Q1-HNQ OOQ¤-1L7
ONFWÜ(*1 ~•••
Q •'•OC)C)21 1 1UU1 "I¢Q Qc¤•-•<¤¤'~Z CJ OOIÜWFU1-« N
.„••«
-« MOOOQI
I1O~•-I
^-* ·¥¢: Q1~c~11nOO•—· U OO-•c¤NOQ
•-1 .....1
U3 ·-<<:>cQccII
O -1:-1:+:-1:•··< Ou'11~c«N¤OOU1
(*1MOCQOOQ
E 1 1 1U ·•~
Y ‘FÖ :.1 -1: -1: -1:ZL-•
knQgg_:N ..„.....
5 E- -«c>c>oc:ooc>I II
Ch UI-•
'~•« *9O ·I<-1< ·I=
anUI<€V} G) Q OSCI-h!\C<'7®~T
[-1 ,.1 ,.1 -1 .„.......: Q •··OOOQOOQOq,1 : 111 III
U °U·•-1 L1*-1 O -I<-k ··I=-k ·I=‘•-4OU
1/II
¤—*OC!OOOOOOQQ III II I
ZE·•—
-l=-¢= -!=·1< ·§=-R
JcnMOOQOOQQOQOII I
·k·k-KII
I
1hO
E V·1·I <<C<¤t¤¤1UUUQQQ ¤«Q -•N<·‘1·-<1~1c*1•·—<:~1¤·1•-•¤1c•‘7 -l:
65
I I I IIgI S. j
A I —·I
I IOn2 QQI I
I
IQ ccc I¤·<¢OI II I I
I I I II I ¤ -I¤I II=I cä-:>I II•-¤I .... I
I II
I·-<cQOc· I
A I I•-Z< IU I eg I I{0 ¢•-·•¢
U¤O*I¤ u cQr~Ine•s~1 Iz II-1
I I-QQQQ¤I
I II: II
Iva Ce I I I¤ UI ¤cOc—1InIn. IO I QCUGOGG
NUI I I III, I
U •-1I IE I
U E ICD O I
"* UB —• •-•„„„„.
2 E Q "'°‘?????°IUI
Ihl ~« I
. ..1 ·« c I IE Q I I
5 "' I I ”'?°??°‘???I I„„ H I I I
Iu c ·‘ I I
CI—
1 • e e IQ
IuaU
IIII I Ivz Z I .I- I I
I I
I
I I I I••
I c er ¢ I IuI
II I I 1
III
; I ZI I C I C I C CI
I
~ I I
IIII I I
I • q er I II I II
I -·ccccccccccccccIII I I
II e I III(I
I
II I IIIII I QI I '.‘„I I VII••I,¤
eI
F O~r~M~¤•~ FF2 F
F .. FFF
FF FF F . F FF oF·»«·~~c~»¤¤¤~1·F FFSF FF F -•oc¢?ogccF F
F F FF IC FFF
«c¤ccc¤ccF | I III
..F F F F< F F¤ F • • F FVI F QBIENQINSÖIV-an-F
"'·*OOQO¢¢OGG IV
FF Illu F
2 F F . . .F•-IFQ
OOCOO GQOG3„. F ··. .= F FF: „-F FU QG G _ e c ¢ FN G F F
F? 1; :2F FQ E ·—•OOQG¢O¢O¢OG F·« U ll I I I
F·· „. .. F ·hl c Q F..1 F F c anE I) N F F
m FF II F FFF F F F
u« I Cav =¤ F F F
Fu F Vccccccccccccc FE F F III lll·¤ F
FI- F FO F IC ilGFUF¤F
F —¤¤c¤cccccocc¤cF3 F F F F F F F
F F:: F F2 F F . . FuW
FF
FF
F Fil { ¢ I
FF
F Ü “°?‘?°‘?°°°??‘?‘?°°??F FF F' F FF F
‘ FF F ec ¢ ee F F
FF F
F
FFF ee an • en er FF F
F F wggqcoqgcgcccgccggoF F
F I • ee ec
FFIhF~F
FQ
VV V V
V V VVV V V VV VV
V VVV V V
V IVV V2 V V
VV Vä V V V VV VV V V VE V V V
u V
ä 2 V V =„„ E ä V VU
V VJ *6 E V I: 2 2VG•• ' Vai 2 u' V VE 2 s V V V5-
••« •„V
VV··~ IE E VVV .
¤ [‘
VE V V'¤ V VE VV
U
2 V V V V.-V V VE V V V
V IV VV VV6V 2V VV V V VV V V'2V $5V VV V —=VV VV V V VV9,} 3:8
V VVVE, 6:22V V *éééV V SV V
LV
68
I I IO canonlg 6s6„„„g::I lI "?°‘?°?‘?°?I I
I II
MIVQN I
~cccocccco‘IIlII III
II
I II I Ucan Igl l”'?°°?I?°???°I II I
l II I c an I.Il
III I
_
I I I II ( I Il I lII I —cooccoccc¤coI I
A lI IIII I I
I° I I A * IV7 Oonca II"‘
I •-Ooooooooooooo lV I I II I III I IlF I I
S I 6„„„· ;· IIII —cccccccccoococ
3 l l II IIII I III,c „ Iw ¤ I I IIE Ü I I m I I
~? MQ I
N S = I —occcoccc:occccc |__S l l Il I IE: I
7 “ I I
Ö E E ISI I2 Il —occc¤c¤cccccoccc I
¤ I I III I ,E °
I I IIl
I I M BI M öl*5 Ohlhßüxfßl
·‘
II
l: I —cccccccoc¤cccococ'-GI I III II I I I Ill l
6 I ~ IQ l q e ce IU ll
2 I I¤ I ‘ I
E I I l•.• I ar ¢ a 4: I ,m l II I II I
I I I II 1 a ca IQQONISINN-j—»o~1~1oon~omr~-~l lII~occgogcc:oqccg?QqqcclI
I an ce e I_lI
I II I I II I ¢ q cc c II
lII I
lI I an ¢ ce a ¢ el II l
I IgI.
Ia V¤I
IC II
I 11 II
1 I 1II 1 1I 1 II1 1 I1 1 I 1II 1 I
: 1 I I6 I Iin 1 1 1¤ 1 1 1z I I IV IO¢:
I I IE 1 1
1 1 1 IIU1: 1-1 1 I- 2 2 II1: In c 1- E·•1u
C Q 1
2 2 = I1za E 1 IV 2 2 1¤ 1 2 I I IQm E E IS ¤1···1
°1 1
‘1
E 73 2 1 1 ” 1E-
-• we I‘“1
T1 3 I Q•¤•6 E %1 =z=2'J3 1 I ·-QI IE I 1 I¤ I IO 1 IQ QI c-v~I I1¤ ~v 1 ·-·ccI IE 1 1 1U1M QQQQ1
·-¤Q¤¤II 11 11 IISI I1 1 •-·¤¤¤cI III
I I I
I 1 1 I1 -1 1E I1 —2=2=21 II S 1
I21 1I
·-·cc¤cccI IIIIII
I1·
1 I121 11 1 ~2222=2¤ 1
I 1 SI1 1 I
70
for .623 of the variance (STRESS1=.l91), a four-dimensional solution for
.679 of the variance (STRESS1=.140), a five-dimensional solution for .694
of the variance (STRESS1=.1l2), and a six-dimensional solution for .710
(STRESS=.100). These results are presented in tabular format in Table
13 and in graphic format in Figure 4. Stimulus coordinates for the
two-dimensional solution appear in Table 14. ALSCAL subset analyses are
reported in Tables 15-19 and Figures 5 and 6.
Regarding dimensionality, only the two-dimensional solution
indicates an "elbow" in the plots of stress along the number of
dimensions. (For ALSCAL, stress rather than Rz should be used to
determine dimensionality, according to Davidson (1983) and Young and
Lewyckyj (1979)). A two-dimensional solution seems reasonable since
authors such as Shepard (1972) recognize that, except for INDSCAL,
dimensionality should rarely exceed two or three dimensions. An
additional criterion requires that stress be at a minimum. This would
appear to pose a problem for the two-dimensional solution, however, since
that dimensionality permits stress (or "badness of fit") to be .249.
However parsimonious a solution, this is a disappointing result. The
stability of the solution looked somewhat more promising. When
coefficients of congruence were computed for four subsets of the data,
the two-dimensional solution did appear relatively stable. Relatively
high congruence was indicated within dimension, but across the subsets.
And relatively low congruence was indicated across dimensions. These data
are provided in Tables 20-25.
Regarding Shepards°s (1972) visualizability criterion, at first
glance, Dimension II looks as if it approximates Stone and Gueutal°s first
71
dimension, which they named complexity. However, closer examination
reveals a number of inconsistencies in the ordering of job titles for the
present Dimension II. Dimension I in the present data does not
approximate either of Stone and Gueutal°s remaining dimensions. This is
evident from both visual inspection and the coefficients of congruence
presented in Table 26. Stone and Gueutal's job title stimulus coordinates
are reported in Table 27. The present data suggest that Stone and
Gueutal's (1985) three—dimensional representation of job dimensions was
not supported.
SUMMARY OF RESULTS
As indicated in the previous section, Hypotheses 1, 2, and 3 were
untestable because of the unstable INDSCAL solution. Hypothesis 1 stated
that the number of dimensions subjects had available to them would predict
their differential accuracy in performance appraisal ratings. Hypothesis
2 stated that the number of dimensions subjects had available to them
would better predict rater accuracy than would the more traditional REP
questionnaire. Again, no conclusions regarding this hypothesis were
possible. Hypothesis 3 stated that MDS dimensionality and REP-CC would
be significantly related, but not isomorphic. The outcome of such a
comparison must likewise await further study.
Hypothesis 4 stated that the present data were expected to replicate
Stone and Gueutal°s three—dimensional job characteristics model. Several
lines of evidence suggest that this hypothesis was not supported. First,
a two-dimensional solution best describes the present data, while a
three-dimensional solution was reported by Stone and Gueutal (1985).
72
TABLE 13
Stress and R2 for Dimensionality One to Six (ALSCAL)
Dimensionality Stress R2
1 .373 .579
2 .249 .607
3 .191 .623
4 .140 . 679
5 . 117 .694
6 .100 .710
73
LO
FQ3“°2
U')<r
5 §“ü'>E
C _3{Ö GJ cn
E §Q ··-<
~ é .
EE
Ö C) Ö Ü Ü Ü Ö Ü cu
$$6.143 ZC!}
QD Q)§"CJ
W ENm
v E L4.9 OJ
Q S*9 ev {Bf.§ “
N O
O O O O O O Q OC9 N *9 *9 '¢ *9 N -1
ZH
71+
TABLE 14
Stimulus Coordinates: Two-dimensional Solution
(ALSCAL)
Dim. 1 Dim. 2
Cashier 0.8545 -1.2078Retail Mgr. 1.3557 0.4914Bank Teller 1.1682 -1.0508Waiter -0.3517 -1.0063Cook -0.9727 -1.3249Tailor -0.4444 1.1695Actor 0.1057 1.8595File Clerk 1.0957 -0.5783Lawyer 1.5833 1.1592Teacher 0.6235 1.3674Tel.Op. 0.5063 -0.7793Taxi Driv. -0.8909 -0.9811Firefighter -1.3603 1.0969Const. Work. -1.4888 0.6293Mail Carrier -0.4379 -1.0334Machine Op. -1.0949 -0.3889Accountant 1.5645 -0.1018Assembly Frmn. 0.0023 0.9303Flight Attend. -0.2618 -0.1781Janitor -1.5560 -0.0730
75
TABLE 15
Stress and R2 for Dimensionality by Subset
(ALSCAL)
Subset Dimension Stress R2
A 1 .368 .5812 .246 .6173 .187 .6374 .137 .6885 .113 .7036 .097 .720
B 1 .399 .5222 .267 .5523 .205 .5724 .150 .6295 .124 .6476 .107 .659
C 1 .348 .6242 .246 .6283 .177 .6734 .135 .7125 .111 .7316 .095 .747
D 1 .361 .6002 .245 .6293 .190 .6344 .135 .7015 .112 .7216 .097 .730
76
RD
nn
07
cn vg;E EBSq)
CU?.¤ mg?:
\•/U) ...>\Of;
2”
E
2ssang 3
«¤ ä¤U)
in VJCJ)
ZGL4#2;
< .9..¢•E3
*°<v§2 · .2.%U') Ncjm
OOOOOOOON 9 LQ <'Z N N <
SSBJLS
77
LO
LO
2**.95•—E E5= OE?‘”
cu2_11C1
B-2
Q
06666606 °¤¤'%<¤¤’?<r¤‘?<‘4* S-v-4
E@53
CT}
nä2<r"’§
L) ExpE P°<vE£(D E:-0 am3
N
O OOOOOO“?g<¤¤¤<t¤’2<=!<
SSBJLS
78
TABLE 16
Stimulus Coordinates: Subset A (ALSCAL)
Dim. 1 Dim. 2
Cashier 0.9478 -1.1827Retail Mgr. 1.2427 0.6264Bank Teller 1.3901 -0.8854Waiter -0.4524 -1.0814Cook -0.8873 -1.3549Tailor -0.0242 1.0777Actor -0.5416 1.7420File Clerk 1.2565 -0.3582Lawyer 1.3482 1.4781Teacher 0.2745 1.3738Tel. Operator 0.5358 -0.6782Taxi Driver -0.7729 -1.1945Firefighter -1.3551 0.9337Const. Work. -1.5702 0.6047Mail Carrier -0.0986 -1.0386Machine Op. -0.9790 -0.3357Accountant 1.6595 -0.0201Assembly Frmn. 0.0381 0.8487Flight Attend. -0.4154 -0.4628Janitor -1.5965 -0.0927
79
TABLE 17
Stimulus Coordinates: Subset B (ALSCAL)
Dim. 1 Dim. 2
Cashier 0.9342 -1.1721Retail Mgr. 1.3436 0.4701Bank Teller 1.1483 -1.0333Waiter -1.1472 -1.1447Cook -1.0594 -1.3325Tailor -0.6788 1.1148Actor -0.6082 1.5944File Clerk 1.0897 -0.5493Lawyer 1.4599 1.2352Teacher 0.5471 1.3682Tel Op. 0.3603 -0.9376Taxi Driv. -0.9418 -0.8911Firefighter -1.2718 1.2067Const. Worker -1.4831 0.6871Mail Carrier -0.5433 -0.9756Machine Op. -1.1630 -0.1986Accouhtant 1.4844 0.1344Assembly Frmn. -0.3212 1.0355Flight Attend. 0.1419 -0.6418Janitor -1.5080 0.0301
80
TABLE 18
Stimulus Coordinates: Subset C (ALSCAL)
Dim. 1 Dim. 2
Cashier 0.8192 -1.0569Retail Mgr. 1.4394 0.6034Bank Teller 1.1422 -0.9000Waiter -0.3472 -1.2096Cook -1.0301 -1.2794Tailor -0.8885 1.1302Actor 0.7093 1.5284File Clerk 0.8788 -0.7673Lawyer 1.7167 0.7825Teacher 0.7516 1.1606Tel. Op. 0.4602 -0.6738Taxi Driv. -0.9488 -0.8625Firefighter -1.3322 1.1954Const. Work -1.3879 0.8963Mail Carrier -0.7704 -0.8205Machine Op. -0.9323 0.3592Accountant 1.4556 -0.2119Assembly Frmn. 0.1096 1.2400Flight Attend. -0.2235 -0.8017Janitor -1.6216 -0.3125
81
TABLE 19
Stimulus Coordinates: Subset D (ALSCAL)
Dim. 1 Dim. 2
Cashier -0.5216 -1.3718Retail Mgr. 1.3940 -0.1073Bank Teller 0.8866 -1.2086Waiter -0.2785 0.6995Cook -1.2015 -0.9674Tailor -0.8179 0.9811Actor 0.5475 2.0159File Clerk 0.8876 -0.8693Lawyer 1.7449 0.9907Teacher 0.8709 1.3151Tel. Op. 0.4107 -0.8136Taxi Driv. -0.9286 -0.7285Firefighter -1.2913 1.2051Const. Work -1.5999 0.0144Mail Carrier -0.5938 -0.8867Machine Op. -0.9602 -0.1111Accountant 1.5658 -0.3347Assembly Frmn. 0.3332 0.2477Flight Attend. -0.0450 0.7697Janitor -1.4459 0.1596
82
OG O
rx I-n .I-1 •-1
<UC/1•-"I<]U
EO
0 ··-•U +-3E 3 ·l<Q3 •·* OO5 O · U OOL4 U] v-Q .•
cb ¤-••-<:
«—~ IO <¤U Z
O O(N bu ·•-1
O I/ILS C-1 m QG2 •-3
G3 D-•-•
I -I:-I:U Q3 OCNON
***3 C OOIUI
B- r-1QQ0 L4
•
O OQ
'~•-•U}GJ
+-3GJE
·-4'UI- —I=·I<~I<O ONGNO0 <£ OccncuL; +-4 ....
«-•OOOq; I I'J
E·•-•+-3V1
khO
VE
·•·1 <}¤¤UQ Q-IQ I-••-••-I-• 4;
83
OQ ON .
-4
OQO OO1-4 ..•·*O
rxJ<1UU}>-Z<¤ -1:W ONC'7
N ..„O
-•co--4•-I5-«Ofl]
·I<•"‘ O<fLh~'fG U OOCNO
U : P1 ....:2 0 -4ccc:.1 -·-· IQ2 ChZ L"H OGC E: --4O Q -l< -1:L}I-4
OQcqcg .....
O ·-O:>c><:>Ls}1.: :1: L1Q .1 O4 :
~•-•Ev C)
U 0‘}¢ “I¢
·-I •J O•·*<"xO\®¢'7~« <¤ Q OO¤~C>¤¤NQ-4 Q
•-1 •„„.•„
Q) ‘•·I FIOOOOOQ
’¤I I I
(J HOO
CJ
va -1: -1: ·1:2 O¤*«¤¤O•-•O<t
·-4 <:>cc¤~c«c¤c¤
g N•...„„„
E ·-cocc>oc--4•JU}
-1: -1: -1:
4 c>c¤~cc«-«¤o~,-4 ......„„MOQQOOOO
I I I
LÜO
E V··4 <1<lQQUUQQ 3-
·k
84
OO
gn .•-I
CGI2 OON ..
·-O
CNNG CONI-1 .„•
•—1Q?
·1=
<2UU} ‘¥I-1 OcI~co·-•<: U O~·-•I~OQ N ....„
•'*OOOO: I IlO-•-1«-JI! *1*·— C>«?:•'IO•··••-•O U Occoc-•c
U] •—1 .„....·—c>cOC>c>•—1 I II
füEO
··-· 1< -?=·1=Tlan 0-~m¤I•-·~o~o
gg gg gn ...„„.„U E MOOOOOO; --1 IIC) 23 IZ- Q)ca 4) -1<~k ·1·=·1<:I-•C‘
.*.2N ........
:—I ·-OC>C>C>c0cN M-1 In I I I
O OLi M-1I-} </IIn 1-I In -1< -1<<£ CZ C) O<\1<‘J¤3G'I¢'7<'IJ'7€*IE-• C} H CD OOOC7\OOCUG•'*
·•-•zu
•-1 ......„„„U E •'*OOOOOOOO
--1 ~•-4 II II
$-1U OO O ·1= -1: -k-1=L} L)W
gn ..........3 •"*OOCCOOOOO
I-1 II I IlZ!E•-I -1: -1=1= -1:1:V}42
OOGc~O•—Inc~Or~r~N ..„•.......
•'<OOOOOOOOOOI I I
·1< -1= -1<
OGOUIONCNMONNCI-1 „.•.•.......
II I I
IhO
VE-—< <:<<¤¤¤¤¤¤uUU¤¤¤ ¤-Q ·K
85
I I IQQ„ QI„_I
I GQ IISI QQ II -¤c
I QFSFSQ Q-sc II .. .INI-«¤c I
I II I IGQQIEI cocEII |-«co¢?I
II II I 1I II GN-•G¤> I
„I3 QQQQT ·I ¤·¤GQQQA IJ
vl ¤¤~¤¢¤g~~oI1 2 ‘?°f?‘T°‘2°2‘~¤I -«c¤c¤¤ä I I1*.. I . ID I I G~DG~D¤¤¤¤~f*; ‘ä„
?‘?‘?"!""I°W I -•cooocc0 I IIU ve .E E IS o 1L J I_ Q-«-•«•u¤-•«e~·•
IQGQGG-•GG Io si I ·-·ooococc Igg {J —
I
I I I I I^‘I. n I
IJ O LI I 1 1 1| I¤ ¤ S I Q•—«un•-<r~G·-•^¢~I I: 2 2IäI QQQQQQQQQI Io ' -·ccc¤ccc>cI I•• L | I s u z u I3 ß I I
·· I I 2 2:1..-2% II
GI , •-<QQGGQQGQQI II III IE I I I II ISI I ,. .. I ,
U I„ E °2°Z°£?‘T"?°Z"I°%"‘? ·S MQGGGQGQQGQ II I
u I ¢ 1 Icn c~<r-•¤l'1¤¤'¤Q~¤-¤¤'*IIE GQGQO-•GG<7~G•-•GI “°?°°???°???II I I II
I 1 1 1 IG¤¤Q-¤Q¤v'¤~1¤~0-·•"•¤> IIf,I I
I I ”??°??°°°???°I II II 1 1 II II I ‘???°°?°°???°°I II I I
I II I 1 1 II I II I
I I III . . . I I
I:II
I III
III··*
<<<<¤Q¤¤UUUUQQQQ ¤«I
I II „„„„„I IE II II I (I IQ~1f'\~b~0QRI cc-Icuc:
I *?°??°I II Q-•I•'•§~r~<7r~I3 II -<QO??OOI
I
I Q-«~Inc:«-•InI II
I I III I II ' ¢ 'I ¤~^—1~In:>~vI¤••~I2
II I I I I
U 1 IW O~cI•eInc~•-«•-«NIn II*·’
III:I IE I IU I ' 1 1IB OI'\~7»ß0¤~0N¢•‘\¤\¤~I•\I
*5VJ ¤-IOQQGÖOOQOQI IU I Il ll I IU •-< ‘I: ¤ I I
IH•·• I I OQN•n-«r~¤n„1In¤¤9I‘é°E ISI ?‘?°E?°°2"‘°°‘?°:9I0 QOGGQQOOOU I
MI I_ Il I IÄ 1; 6 I I Ikl ¤I I I 1 II I‘
IA ua > I , c¤c~¤¤¤r~c¤¤n»¤¤0•¤nc II 2 E I': ISI =2<%=2·:=2"?9¢2=2‘2‘2=2·:(-· zu I I ·-¤c¤c¤<:cccQo¤~·«
A1 I I I lll
° ° I I I" "" I I II•« I Iu ¤I I I 1 cc Iua ua I I I8 ja I
EI I
I‘¤
E I · II II I
IB
IA ll I I II I
I IU 1 1 ¢I IUI II
I"‘°???°°‘?°?‘?°°?°?I III II ( 1 1 I .
I IOO-IOOGÄO-QG¢~-¤•-OQ¢*QI I
IT -·ccoococcco;>ooocc§ '
I II I II I I I III I I
1 1 11 II
IIII II
I I -1 1 II I III§I
II II I II I 1 1 1 I II
I II I I,,'I .qII IEI
lv.
IQ
I IIII II
III IIII IIII IIII IIII I§ I I I I
1 II IIIII IIÜ II II.3 II I
I 3 I I I2 : I‘
2 I II2 I IE Iä Ü a I II = I, II IN 2 z I I I2 2 ~ I I IE E E I I II3: II I,’·’ E I I II
E II II‘§ I I5I I II“ I II
II IIII III2I äI III I
SSI II°I MI II IIII 22SI
IEI **331IIIIIII $$35I II I —2?2I ·II IÜI
I;If„Zfä§.EE‘.§‘.I‘$;'$.’fI‘2I‘d%’~.“$äEI?.Ei'$äI2‘I
· 88
I I .I
•I IO~g~•'v¤0••-•¢~4-•ISII
I
IiI g•—•~a¤r~~~t¤~¢'•~~1SI
II I._
I I3 °29°2'1‘?‘?°Z?"t"!°2‘2 I¤-IQQQQQQQQOQQ I
e eO•n~c•n~~•n~¤In~c•~ III ”?°°???°°°???· I
^ I I IJ II I 1:
2II ¤-•OQQ¤QOQQQ¤OO¤I‘·* I I II II I I= I
3 I g —¢IuII
8 W I I I I II I Iä 2 I: I: I a ac IH OjaO
¤' I •-OQGOQQUQOQOQQGQ,,„ U 3 I lb I I I II°‘ —·~·· I C II I an ¢• I
¤ I SI
IEIIU In I I I I I I I I I0 c I
I,
I: In I I e ¢ I¤» ua I I8 2 ISI
II I-GQEFQUUQUQUQOUQOGQ
IE I I Ic I I —¤ 1 I
·•
··IU
—-OUQGQGOUOQQOQOQUGGI 1 I I I 1 I I I I I IE e ‘ an | Irn II; I
I I II I I II . . . I I
III
I‘
·g cc I I
I I
I I II I IGC '
I III
I I
IInnI?Iv
B II
1' 11 1 1
1 1 I1
1 111
11 1 11 11I 1 I11 1 11 1 1 1
1’
1Q 1ä 1-1‘ 1 1 1G 1 1Q 1
··· 1 12 1 1ä 8 1 1E 1 1 1Ä; SB g 1 1• G —«2 6 2 1 1G
E 75 ä1 1 1
111 E2 Ä 1:1 31 1zu E ua _~¤1 .1 1
Q 3 ~ 1 ”I 11- _: E 1 1 1 11 J 1IJ l 16 6 121 221 1
N 11:11 ..1E 1 1
”°1E 1 18 1 1 1U
1;151·
•—•GG19i 1 111•·1
1 11“
121 $$851 11 „ -:::1 11 I 1 11 1 1
Ä ::~1o·-·12r 91919 11 ·-ccc: 11 1 - II
1I11EI 999999 1
1 1 -·c:coo Q1 1I 1 1
1 L 6„6„„2-S1 99999991
·-ccgcog 11 _1
11 QQN—•1~1r~•—•1h 1S1 99999999
1 1 —o::c:¤:11 2
E V1••1G1•-«lVI'5~1'¤h~¤¤<lN9~1•l'\~O•-•t*4l'\~?•t\v¢¤¤•t~1¢'1~Y•t\~¤II
TABLE 26
Coefficients of Congruence for Stone and Gueutal°s
Final Solution and Six ALSCAL Solutions
Stone/ Stone/ Stone/Gueutal 1 Gueutal 2 Gueutal 3
S/G1 1.00000 0.73636 0.72182
S/G2 0.73636 1.00000 0.77666S/G3 0.72182 0.77666 1.00000
1ALSCAL1 -0.29827 0.10203 0.37681ZALSCAL1 0.26280 -0.13711 -0.40776ZALSCAL2 0.42502 -0.03385 0.214923ALSCAL1 0.24890 -0.16622 ‘ -0.415903ALSCAL2 0.42822 -0.00826 0.217273ALSCAL3 -0.05650 -0.33256 0.170754ALSCAL1 0.24266 -0.16610 -0.423134ALSCAL2 0.41535 -0.04075 0.165634ALSCAL3 -0.07128 -0.40897 0.130824ALSCAL4 -0.08087 0.01956 0.010275ALSCAL1 0.24603 -0.18473 -0.41649SALSCAL2 0.43042 -0.00185 0.21182SALSCAL3 -0.08136 -0.40317 0.14770SALSCAL4 0.07594 -0.01016 -0.01125SALSCAL5 -0.02561 0.06281 -0.064906ALSCALl 0.23349 -0.17851 -0.428376ALSCAL2 0.44024 -0.02192 0.202446ALSCAL3 0.07872 0.40372 -0.14730ÖALSCAL4 0.06092 -0.02050 -0.01655ÖALSCAL5 -0.06775 -0.00370 -0.035976ALSCAL6 -0.01630 0.07603 -0.06864
91
TABLE 27
Stimulus Coordinates: Stone & Gueutal (1985)
Dim. 1 Dim. 2 Dim. 3
Cashier 17.70 30.27 5.28Retail Mgr. 78.89 18.94 34.03Bank Teller 28.25 31.43 1.00Waiter 21.48 82.09 32.93Cook 19.93 61.43 44.09Tailor 53.33 33.42 59.32Actor 88.77 77.42 46.37File Clerk 31.10 22.16 16.80Lawyer 100.00 40.07 29.26Teacher 83.04 55.15 40.27Tel.Op. 19.52 43.88 26.07Taxi Driv. 22.22 70.85 50.21Firefighter 47.95 44.65
A81.09
Const. Work. 30.77 27.09 94.48Mail Carrier 20.70 54.55 41.50Machine Op. 16.17 14.38 73.34Accountant 70.74 9.51 8.41Assembly Frmn. 60.97 20.59 73.24Flight Attend. 36.95 85.38 38.36Janitor 26.94 52.54 79.07
92
Second, a visual inspection suggests that only on one dimension is there
even moderate convergence between any of the present dimensions and the
Stone and Gueutal dimensions. Third, coefficents of congruence computed
for Stone and Gueutal°s job-title stimulus coordinates revealed that
there was substantial overlap between Stone and Gueta1's first and second
dimensions, first and third dimensions, and second and third dimensions
(coefficients of congruence=.736, .722, and .776 respectively, p<.05).
This calls into question the independence of the Stone and Gueutal°s three
dimensions.
Although the study°s hypotheses were not supported, some additional
results are of interest. lnterestingly, although REP did not
significantly predict differential accuracy in performance ratings
(r=-0.027, g>.OS), it did predict differential accuracy in frequency
ratings (r=.148, p<.01). Furthermore, REP also significantly predicted
elevation in performance ratings (r=.143, p<.01). Descriptive statistics
appear in Table 28 and correlation tables are presented in Table 29.
Other results indicated that differential accuracy in performance
ratings (DAP) was significantly correlated with differential accuracy in
frequency ratings (DAF) (r=.20, p <.05). This overlap is not surprising
since the two rating processes were not expected to be totally independent
(Murphy et al., 1982). One unexpected result was that there were a number
of small but statistically significant intercorrelations among some of
the different accuracy components. For example, differential accuracy
in performance (DAP) ratings was significantly correlated with elevation
in performance (EP) ratings, and with stereotypic accuracy in performance
(SAP) ratings. Correlation coefficients are .18 and .17 respectively
93
(Q<.01). Differential accuracy in frequency (DAF) ratings was
significantly related to elevation in frequency (EF) ratings and
differential elevation in frequency (DEF) ratings. Correlations are -.31
and .21 respectively (Q<.0l). These results are despite the fact that
these components are presumed to be additive and orthogonal (Cronbach,
1955). DACOR, DECOR, and SACOR were calculated for frequency ratings.
Again, no significant relationships between these subcomponents and REP
were revealed.
The demographic variables were unrelated to either differential
accuracy in performance ratings or differential accuracy in frequency
ratings. However, some small but significant correlations were noted
between the demographic variables and other components of accuracy. In
particular, it should be noted that the number of years a person had
worked significantly predicted differential elevation in performance
ratings (r=.13, Q <.05). To examine this relationship further, a subset
analysis was performed on the years-worked data. The data were
trichotomized and submitted to a univariate analysis of variance (ANOVA)
procedure. The analysis indicated a significant effect for level of
experience, §(2,283) = 3.23, Q <.05). Means for these groups were .52,
.57 and .59 respectively. Results from this procedure appear in Table
30. Tukey°s studentized range test revealed that there were significant
differences in differential elevation when low and high experience groups
were compared (Q <.05).
Additionally, whether or not a person had been a manager was
slightly but significantly related to both elevation and differential
elevation in performance ratings (r=.13 and r=.12, respectively, Q<.05).
94
Although raters° gender had no significant bearing upon differential
accuracy in the present study, it was correlated with differential
elevation in frequency (DEF) ratings (r=.l3, Q < .05). It deserves noting
that although there are some statistically significant correlates among
the demographic variables, in every case the correlations are small and
are likely a function of high statistical power (large sample size).
Thus, the practical importance of these demographic correlates is
unclear. A discussion of the implications of these findings will follow.
95
Table 28
Descriptive Statistics:
Dimensions, REP, Accuracy Components
Variable n Mean SD Min. Max. Range
REP 305 221.413 56.730 129.000 478.000 349.000
Yrs Worked 286 2.164 1.948 0.000 14.000 14.000
Age 290 18.893 1.406 17.000 31.000 14.000
Diff. Accur.(Perf.) 305 0.487 0.101 0.245 0.809 0.564
Elevation(Perf.) 305 0.492 0.306 0.006 1.556 1.550
Diff. Elevation(Perf.) 305 0.556 0.197 0.094 1.357 1.263
StereotypicAccur. (Perf.) 305 0.308 0.114 0.086 0.972 0.886
Diff. Accur.Corr. (DAr) 305 0.160 0.233 -0.505 0.820 1.330
Diff. ElevationCorr. (DEr) 305 0.810 0.136 0.165 0.999 0.834
StereotypicAccur. Corr. 305 0.489 0.295 -0.619 0.929 1.547
(SAr)
Total Accuracy 305 1.040 0.517 0.244 3.606 3.362
(Perf)
96
Table 28 (Cont°d)
Descriptive Statistics:
Dimensions, REP, Accuracy Components
Variable n Mean SD Min. Max. Range
Diff. Accur.(Freq.) 305 1.817 0.195 1.278 2.476 1.198
Elevation(Freq.) 305 0.498 0.323 0.004 1.588 1.584
Diff. Elevation(Freq.) 305 1.144 0.297 0.356 1.915 1.559
StereotypicAccu:. (Freq.) 305 1.233 0.236 0.714 1.939 1.225
Diff. Accu:.Cor:. (DA: Freq.) 305 -0.198 0.135 -0.564 0.199 0.759
Diff. ElevationCor:. (DE: Freq.) 305 -0.702 0.264 -0.998 0.949 -1.937
StereotypicAccu:. Cor:. 305 -0.125 0.281 -0.736 0.571 1.311(SA: Freq.)
Total Accuracy 305 6.660 1.051 3.958 9.621 5.663(Freq)
97
OOO
Q •-;
OO
Q •-I
O‘~0
O O
AF'N
O
O Ih QO Ih N
Ö O O O~¤
•-•Q Q
NJ I
•• O In O däIn O O N *0Q. c ·-I •-I N
,.5] O•-1
Q QQ
S*"* ·I= ··kH ·|< ·¥¢G O Ih •·I O IO*'•
O dä N Q NO •-• Q N O
ä° I: ·-• c c o ogq O I I I
NISdä ·•*N mz I:
@0 ·!< ·I=IJ .¤U I: N O •-I Q•" HI- O dä O N N Ih
A „ . „ •E-· NE cw - - ¤ o c c
:0 O•-I Q I I I
II-I
Sr.:W ·K ·¥¢ ·I¢°‘ I: I: I:U) O •'* N O dä M FIC1 A O w Q N N ~0 ON O •·< O •-I Q O PI#-5 FI O O O O O OQ I I IAQ}I. -kL-I A O N M Ih Q ND ID OO •·* O N Q Ih Q •* N OU O O O P*I O O O O O
•- O O O Q O Q QI
A A A A A A A A•—• N In Q Ih Io N coO AO O NO O O O O
G G U·•-• ·•-• ·•-I \fI•-II-I I: I-I: Q. ccC.‘>x O GO >s>•
· ·Ivo--• <v~·-•
uuHG J-I HH OG VVQ)!-• G UG dl}-•WB > W> HB ¤«¤«
D-4 *4-4U 0 II-IO OU I-4 H HI-I ··-•·—• uu <: zu <: ·I¢-I:Q4 hl Qui I/14 Q Q V) ·l<
98
TABLE 30
Analysis of Variance (ANOVA) for the Effect of
Work Experience on Differential Elevation
Source df Sum of Squares Mean Square F
Experience 2 0.2490 0.1245 3.25*
Error 283 10.9150 0.0386
Total 285 11.1641
99
DISCUSSION
The present study was expected to (a) reoperationalize cognitive
complexity, (b) compare the predictiveness of CC-dimensionality with that
of CC-REP, (c) assess the relationship between CC-dimensionality and
CC-REP, and (d) replicate the three-dimensional job characteristics model
of Stone and Gueutal (1985). The attempt to reoperationalize CC was
consistent with a number of recommendations (Bieri, 1966; Scott, Osgood,
& Peterson, 1979) that MDS is a potentially useful approach to examining
the manner in which subjects dimensionalize stimuli. However, the initial
(INDSCAL) results reveal that this may not be the case.
To extract dimensionality, the study submitted subjects° similarity
ratings of job titles to multidimensional scaling. The
individual-differences scaling results revealed that a four-dimensional
solution appeared to best account for the rated similarity of job titles.
This was determined by the conventional "scree test." In other words,
four perceptual dimensions seemed to underlie subjects° perceptions of
jobs. As indicated in the results section, the proportion of variance
in similarity ratings, which was accounted for by such a four-dimensional
solution, approaches levels considered moderately good (but not
excellent) by authors Young and Lewyckyj (1979), authors of the software
application of the INDSCAL algorithm. However, to determine the adequacy
of the solution, stability was also a requirement.
The stability of the Job Similiarity Measure was first assessed by
computing the average correlations of repeated items for each subject,
100
converting these correlations (using Fisher°s r to Z transformation), and
then arriving at an overall or average Z. The results of this assessment
suggested reasonably high reliability. In light of the moderately high
stability (suggested by the consistency of subjects' ratings of repeated
items), it seems unlikely that the instability of dimensionality is the
result of subjects‘ rating inconsistencies.
However, one particularly problematic aspect of these results was
the relatively low stability of the group multidimensional scaling
solution when a second method was used to determine stability. One method
to assess solution (dimensional) stability is the comparison of Job Title
Stimulus Coordinates (JTSCs) across four subsets of the data. However,
coefficients of congruence for the four random subsets' JTSCs on each
dimension suggest that dimensionality was unstable. This instability
calls into question the usefulness of the group INDSCAL solution in the
present study.
As indicated earlier, one alternative for enhancing stability is
to examine coefficients of congruence for alternate dimensional
solutions. However, the results revealed that changing the number of
dimensions did not result in stable dimensionality.
One usual explanation for solution instability is the use of a
higher convergence option (e.g. converge = .01 to .05). As indicated in
the results section, the MDS (INDSCAL) algorithm iterates until
convergence meets the researcher-specified option (or the default value,
.001). This means that S-STRESS (overall "badness of fit") must decrease
at least to convergence for the program to continue iterating. The
present study employed a convergence option of .001. Although greater
101
stability can be expected at this level, such stability did not result
in the present case. Similarly, the number of iterations that were
requested in the program options (iter=l0) could not have accounted for
the lack of stability because in no case were more than nine iterations
required for convergence at .001. So, although increasing the permissible
iterations may often result in a better solution, this strategy would not
have been helpful in the present case.
A related issue is the inferential testing of particular
n-dimensional MDS solutions. Although Weinberg, Carroll, and Cohen
(1984) have investigated methods to establish confidence regions for
jacknifing and bootstrapping techniques in multidimensional scaling, this
work is in its infancy. Thus, the statistically inelegant strategy of
determining dimensionality by locating an "elbow" in the plots of R2
against the number of dimensions will have to suffice.
The disappointing individual-differences scaling results led to an
inability to test Hypotheses 1 and 2. Thus, no test for the relationship
between individual dimensionality and rater accuracy was possible.
Consequently, no comparison could be made between the predictiveness of
individual dimensionality and the predictiveness of REP.
However, regarding Hypothesis 3, REP did predict elevation in
performance ratings. The less the REP-complexity, the greater the
elevation in ratings. Elevation, according to Cronbach (1955), reflects
the rater°s "way of using a response scale" (p.l78). In other words,
elevation, the average rating over all ratees and items upon which the
ratee is judged, indicates the overall level of the ratings. This means
that the higher the REP score (i.e., the greater the tendency to use
102
dimensions in the same way on the REP), the greater the elevation
(in)accuracy. In other words, raters who tend assign the same REP numbers
regardless of the dimension also tend to produce ratings which are
generally discrepant from true scores. However, it should be noted that
statistical power may account for this result and that the practical
importance of this finding is unclear.
REP also significantly predicted differential accuracy in frequency
ratings. This result contradicts a number of previous studies which
revealed no significant relationship between REP and differential
accuracy (Bernardin & Cardy, 1981; Bernardin, Cardy, & Carlyle, 1982;
Lahey & Saal, 1981; Sauser & Pond, 1981; Sechrest & Jackson, 1961).
Though of interest because of its contrast with the above-mentioned
studies, this finding may be less than compelling. The correlation
coefficient, though significant, was small and of doubtful practical
importance. Again, this result may be explained by the statistical power
resulting from a large sample size.
Another possible explanation for the small, but significant,
correlations between REP and some of the accuracy measures may be common
method variance. There may be some construct similarity between the
measures. Perhaps both predictor and criterion measure halo.
The present study revealed that the work experience of raters
contributed to their differential elevation in performance scores.
Specifically, the greater the experience the greater the departure from
true scores were the average ratings for each ratee across dimensions.
Although not breaking new ground, this is an intriguing finding because,
it provides further evidence that performance appraisals for subordinates
103
may be influenced by the number of years a person has worked. It is
possible that experienced workers may be more aware than less experienced
workers of the potential impact of performance appraisal on their
subordinates and therefore be less willing to shoulder the risks of
alienating certain subordinates. The result may be less accurate
appraisals.
In a literature review, Landy and Farr (1980) have documented the
mixed effects of experience on rating quality. Although some studies
(Jurgensen, 1950) reveal that one‘s job experience affects the quality
of ratings, others report that job experience may lead to more lenient
ratings (Cascio & Valenzi, 1977; Mandell, 1956) or to no effects on
ratings (Klores, 1966). The present study does not resolve the confusion
over this issue. Future research may determine the nature of the
relationship between job experience and rating quality.
Another hypothesis concerned the predicted relationship between
individual differences in dimensionality and REP. Again, this hypothesis
was untestable.
The disappointing individual-differences scaling results also
prompted a re-analysis of the present data, using ALSCAL's Euclidian model
option. When the data were re-analyzed, a two-dimensional solution seemed
to best describe the data. This is not unusual, because as indicated
earlier, two to three dimensions are usually sufficient to describe the
data, except in individual-differences models when sometimes more than
three dimensions are necessary (Shepard, 1972). The coefficients of
congruence for the ALSCAL scaling solutions revealed that the Stone and
Gueutal (1985) dimensions were not retieived.
104
This latter result provides probably the most interesting yet
unexpected finding of the present study. Although one objective of the
present research was the replication of Stone and Gueuta1°s
three-dimensional job characteristics model, the present study revealed
that their research may not be as generalizable as once thought (Stone,
1986; Stone and Gueutal, 1985; Stone and Ruddy, 1987). The sample size
for the present study should have provided ample opportunity for such a
test of the generalizability of their model. Moreover, the present
research employed the Stone and Gueutal methodology in the pretest portion
of the study. Still, little support for the Stone and Gueutal model
emerged.
Possible reasons for the failure to replicate Stone and Gueutal
include possible inadequacies of the Job Similarities Questionnaire.
Perhaps asking subjects to rate their perceived similarity of job titles
is an inadequate means of capturing what truly distinguishes job
activities. As Scott (1963) argued, requiring subjects to analyze objects
in a phenomenal space may burden their capacities of introspection. Scott
goes so far as to question whether subjects can easily "distinguish
between such common objects as a rock and a tree" (p. 270). In addition,
he cautions that inordinate demands upon the subject may produce artifacts
of the method. Perhaps the use of job-analysis techniques may be a more
fruitful approach.
Additionally, Stone and Gueutal used correlational evidence to
support their claim that the subset analyses converged. The appropriate
statistic is the coefficient of congruence (Cattell, 1978; Gorsuch,
1983). However, this statistic may provide vastly different results than
105
the correlation coefficient. Thus, the initial reporting of the Stone
and Gueutal findings may have been premature.
Another possible explanation for the lack of convergence between
the Stone and Gueutal dimensions and the present study is potential
subject differences. The present study employed introductory psychology
students (primarily freshmen), whereas Stone and Gueutal employed
undergraduate juniors and seniors. However, any differences between the
two samples° familiarity with jobs should not be large. It is therefore
doubtful that any such differences could account for the lack of
convergence between the dimensions in the two studies.
The failure to replicate the three-dimensional model is of more than
academic interest. Job redesign interventions are necessarily based upon
the dimensions along which jobs are believed to Vary. A failure on the
part of scientific research to delineate stable and generalizable job
dimensions can lead to misguided efforts in the area of job design. How
can the practitioner expect to make sensible changes with respect to "job
enrichment," for example, when the foundation for such changes is built
on an "earthquake fault"? One of Stone and Gueutal's criticisms of prior
job-characteristics research was that such work failed to properly
describe the manner in which people actually perceive jobs. Accordingly,
these authors have directed much energy away from a priori assumptions
about what workers perceive toward an understanding of what workers
actually perceive. Yet, despite this considerable effort, these authors
may not have achieved the breakthrough that was once hoped.
As indicated earlier, cognitive complexity is conceived not as a
trait which is stable across content domains, but rather as a process
106
variable which may interact with features of the stimulus and is somewhat
domain specific. In other words, there are attributes or characteristics
of stimuli that combine with the manner in which we process information
to affect social judgment. The manner in which we process information
is a function of those features, thoughts, etc. which we perceive
concerning stimuli.
The domain specificity issue must be further examined empirically.
For example, research is required to determine not only whether subjects
vary in complexity across broad domains (jobs, roles, etc.), but also
whether complexity varies within domains (i.e., for specific jobs). ln
other words, would one°s level of complexity be different for the job of
teacher relative to other jobs, such as dentist, pipe-fitter, etc.?
Streufert and Streufert (1978) have suggested that such variable
complexity may be a complicating factor.
Future research on cognitive complexity must address the construct
validity issues outlined earlier. Any future examination of CC must first
deal with the issue of an adequate operational definition. Efforts must
focus not only on the number of dimensions or pieces of information people
use to make social judgments but also on the combinatory strategies people
use to integrate such social information. Results gleaned from the social
cognition literature may provide some useful guidance in this regard.
Schema theory has explained social cognition in terms of peop1es'
use of a schema or mental representation of a person, role, object, etc.
(Taylor & Crocker, 1981). For example, people have some mental
representation for the role of teacher, leader, employee, etc. People
also have implicit personality theories or mental representations of how
107
the attributes of people are correlated. Regarding leaders, Rush, Thomas,
and Lord (1977) found that one°s implicit leadership theories may be
reflected in one's ratings of leadership behavior. More recently, Lord,
Foti, and Philips (1982) have suggested that ratings of leader behavior
are the result of a complex information·processing sequence. Individuals
enter the workplace with well-developed prototypes of an effective leader
and an ineffective leader. Raters compare the leader°s behavior to the
characteristics of the prototype and then classify the leader as either
effective or ineffective. Such a categorization process may thus drive
the rater°s overall perception of the leader (Foti, Frazer, & Lord, 1982).
Redirecting CC research by examining more fully both the content and
process of cognitive schemata may provide additional insight concerning
how individuals perceive the attributes of others.
Future research is required also to provide a qualitative
interpretation of the dimensionality data. The present study attempted
but failed to capture the three-dimensional job characteristics model of
Stone and Gueutal (1985). Therefore, efforts were not directed at
exploring potential names for dimensions. So, the number but not enough
of the nature of the dimensions was at issue. Additional research is
needed to examine the nature of the dimensions underlying jobs.
Additional research might also be aimed at determining whether
there is a quantitative representation of cognitive complexity. Should
other exploratory efforts demonstrate correlational evidence of a
dimensionality-to·rating link, controlled laboratory experiments would
likewise be a necessary follow-up.
108
Finally, a major focus of future research should be the development
of more reliable measures. Because reliability sets the limits for
validity, the importance of this prescription cannot be overstated.
In sum, the present study attempted, but failed, to provide a new
operational definition of cognitive dimensionality. Similarly, a
comparison of cognitive dimensionality with the traditional REP-test
could not be made. Moreover, an expected replication of Stone and
Gueutal°s (1985) model of job characteristics was not revealed. The
results, while not supportive of the research hypotheses, nevertheless
provide some answers to persistent questions about the operationalizing
of cognitive complexity and the dimensions underlying jobs. Meanwhile,
it would appear that, as Bernardin, Cardy, and Carlyle (1982) have noted
for cognitive complexity and appraisal effectiveness,it‘s
"back to the
drawing board" (p. 151). It also may be "back to the drawing board" for
the job-characteristics approach to designing jobs.
109
REFERENCES
Addams-Webber, J. R. (1969). Cognitive complexity and sociality. BritishJournal or Social and Clinical Psychology, 8, 211-216.
Aldag, R. J., Barr, S. H., & Brief, A. P. (1981). Measurement ofperceived task characteristics. Psychological Bulletin, QQ,415-431.
Attneave, P. (1959). Applications of information theory to psychology.
NY: Holt-Dryden.
Becker, B. E. & Cardy, R. L. (1986). Influence of halo on appraisaleffectiveness: A conceptual and empirical reconsideration. Journalof Applied Psychology, ll, 662-671.
Bem, D. J. & Allen, A. (1974). On predicting some of the people some ofthe time: The search for cross-situational consistencies inbehavior. Psychological Reyieg, 81, 506-520.
Bernardin, H. J. 8 Cardy, R. L. (1981). Cognitive complexity inperformance appraisal: lt makes no nevermind. Proceedings gf the4lst Annual Meeting of the Academy of Management, lp, 306-310.
Bernardin, H. J., Cardy, R. L., & Carlyle, J. (1982). Cognitivecomplexity and appraisal effectiveness: Back to the drawing board?Journal of Applied Psychology, 61, 151-160.
Bieri, J. (1955). Cognitive complexity-simplicity and predictivebehavior. Journal of Abnormal Psychology, jl, 263-268.
Bieri, J. (1971). Cognitive structures in personality. In H. M.Schroder & Peter Suedfeld, Personality theory and informationprocessing (pp. 178-208). NY: Ronald Press.
Bieri, J., Atkins, A. 1., Briar, S., Leaman, R. L., Miller, H., & Tripodi,T. (1966). Clinical and social judgement: The discrimination of
behayioral information. NY: Wiley.
Bieri, J. & Blacker, E. (1956). The generality of cognitive complexityin the perception of people and inkblots. Journal of AbnormalPsychology, 5;, 112-117.
Bonarius, J. C. (1965). Research in the personal construct theory ofGeorge A. Kelly: Role construct repertory test and basic theory.In R. A. Maher (Ed.), Progress in experimental personalityresearch: Yol Z (PP- 1-46). NY: Academic Press.
110
Borman, W. C. (1977). Consistency of rating accuracy and rating errorsin the judgment of human performance. Organizational Behayior andPuman Performance, gg, 238-252.
Burt, C. L. (1948). The factorial study of temperamental traits. BritishJournal of Psychology, l, 178-203.
Burton, M.(1972). Semantic dimensions of occupational names. In A. K.Romney, R. N. Shepherd, & S. B. Nerlove (Eds.). üultidimensionalscaling: Yol II (PP. 55-71). NY: Seminar Press.
Campbell, V. N. (1960). In W. A. Scott, D. Wayne Osgood, & C. M.Peterson, Cognitiye strncture: Theory and measnrement or indiyidualdifferences. NY: John Wiley & Sons.
Campbell, D. T. & Piske, D. W. (1959). Convergent and discriminantValidation by multitrait-multimethod matrix. PsychologicalBulletin, gg, 81-105.
Cardy, R. L. & Dobbins, G. H. (1986). Affect and appraisal accuracy:Liking as an integral dimension in evaluating performance. Journalof Applied Psychology, gl, 672-678.
Carroll, J. D. & Chang, J. J. (1970). Analysis of individual differencesin multidmensional scaling via an N-way generalization of"Eckart-Young" decomposition. Psychometrika, gg, 283-319.
Cascio, W. P. & Valenzi, E. R. (1977). Behaviorally anchored ratingscales: Effects of education and job experience of raters andratees. Journal of Applied Psychology, gg, 278-282.
Cattell, R. B. (1978). The scientific use of factor analysis in behayioral
and life sciences. NY: Plenum.
Cherrington, D. J. & England, J. L. (1980). The desire for an enrichedjob as a moderator of the enrichment satisfaction relationship.Organizational Behayior and Human Performance, gg, 139-159.
Crockett, W. H. (1965). Cognitive complexity and impression formation.In R. A. Maher (Ed.), Progress in experimental personality
research: Yol. 2 (pp. 47-90). NY: Academic Press.
Cronbach, L. J. (1955). Processes affecting scores on "understanding ofothers" and "assumed similarity." Psychological Bulletin, gg,
177-193.
Davison, Mark L. (1983). Multidimensional scaling. NY: John Wiley.
Dictionary of Occupational Titles, Occupational Classification Vol.2(1965). Washington, DC: U.S. Government Printing Office.
lll
Dubin, R., Porter, L. W. Stone, E. F., & Champoux, J. E. (1974).lmplications of differential job perceptions. Industrialßelatlons, gl, 404-409.
Dunham, R. B. (1976). The measurement and dimensionality of jobcharacteristics. Journal of Annljed Psychology, gl, 404-409.
Dunham, R. B. (1977). Relationships of perceived job designcharacteristics to job ability requirements and job value. Journalof Agnlled Psychology, gg, 760-763.
Dunham, R. B., Aldag, R. J., & Brief, A. P. (1977). Dimensionality of
task design as measured by the Job Diagnostic Survey. Academy og
Management Journal, gg, 209-223.
Durand, R. M. & Lambert, Z. L. (1976). Generalizability of cognitivedifferentiation across product and social domains. PsychologicalRenorts, gg, 665-666.
Feldman, J. M. (1981). Beyond attribution theory: Cognitive processesin performance appraisal. Journal of Annlied Psychology, gg,
127-148.
Foa, U. G. & Turner, J. L. (1970). Psychology in the year 2000: Goingstructural? American Psychologist, gg, 244-247.
Foti, R. J., Frazer, S. L. & Lord, R. G. (1982). Effects of leadership
labels and prototypes on perceptions of political leaders. Journalof Anplied Psychology, gl, 326-333.
Fried, Y. & Ferris, G. R. (1987). The validity of the job characteristicsmodel: A review and meta-analysis. Pegsonnel Psychology, gg,
287-322.
Goldstein, K, M. & Blackman, S. (1978). Cognitiye style: Fiye annroachesand releyant research. NY: Wiley.
Gorsuch, R. L. (1983). Factor analysis. NY: Lawrence Erlbaum Associates.
Hackman, J. R. & Lawler, E. E. (1971). Employee reactions to jobcharacteristics. Journal of Annlied Psychology, gg, 259-286.
Hackman, J. R. & Oldham, G. R. (1975). Development of the Job DiagnosticSurvey. Journal of Annlied Psychology, gg, 159-170.
Hackman, J. R. & Oldham, G. R. (1976). Motivation through the design of
work: Test of a theory. Organizational Behayior and HumanPerformance, lg, 250-279.
112
Harvey, R. J., Billings, R. S., & Nilan, K. J. (1985). Confirmatoryfactor analysis of the Job Diagnostic Survey: Good news and badnews. gournal of Applied Psychology, Z2, 461-468.
Harvey, R. J., Friedman, L., Hakel, M. D., 8 Cornelius, E. T. (1988).Dimensionality of the Job Element Inventory, a simplifiedworker-oriented job analysis questionnaire. Journal of AppliedPsychology, 12, 639-646.
Idaszak, J., Bottom, W., & Drasgow, F. (1988). A test of the measurementequivalence of the revised Job Diagnostic Survey: Past problemsand current solutions. Journal of Applied Psychology, 12, 647-656.
Idaszak, J. & Drasgow, F. (1987). A review of the Job Diagnostic Survey:Elimination of a measurement artifact. Journal of AppliedPsychology, Z2, 69-74.
Jones, A. P. & Butler, M. C. (1980). Influences of cognitive complexityon the dimensions underlying perceptions of the work environment.Motiyation and Emotion, 4, 1-19.
Jurgensen, C. E. (1950). Intercorrelations in merit rating traits.Journal of Applied Psychology, 24, 240-243.
Kelly, G. A. (1955). The psychology of personal constructs. NY: Norton.
Kelly, G. A. (1970). A brief introduction to personal construct theory.In D. Bannister (Ed.), Perspectiyes in personal constructtheory. NY: Academic Press.
Klores, M. S. (1966). Rater bias in forced—distribution ratings.Personnel Psychology, 12, 411-421.
Kruskal, J. B. (1964). Multidimensional scaling by optimizing goodnessof fit to a nonmetric hypothesis. Psychometrika, 22, 1-27.
Kruskal, J. B. & Wish, M. (1978). Multidimensional scaling. Beverly
Hills, CA: Sage.
Kulik, C. T., 01dham, G. R. , & Langer, P. H. (1988). The measurement
of job characteristics: A comparison of the original and the revisedJob Diagnostic Survey. Journal of Applied Psychology, 12, 462-466.
Lahey, M. A. & Saal, F. (1981). Evidence incompatible with a cognitivecompatibility theory of rating behavior. Jpp;na1_p§_App11eQPsychology, pg, 706-715.
Lal, R. & Singh, R. S. (1984). Cognitive complexity, attitude similarityand interpersonal judgment. Perspectiyes in PsychologicalResearches, Z, 4-7.
113
Landy, F. J. & Farr, J. L. (1980). Performance rating. PsychologicalBulletin, 81, 72-107.
Lee, R. & Klein, A. R. (1982). Structure of the job diagnostic surveyfor public sector occupations. Journal of Applied Psychology , 81,515-519.
Leitner, L. M., Landfield, A. W. & Barr, M. A. (1974). Cognitivecomplexity: A review and elaboration within personal constructtheory. Unpublished manuscript, University of Nebraska.
Leventhal, M. (1957). Cognitive processes and interpersonal prediction.Journal of Abnormal and Social Psychology, 88, 176-180.
Lewin, K. (1936). Theories of topologlcal psychology NY: McGraw Hill.
Lord, R. G., Foti, R. J. & Philips, J. S. (1982). A theory of leadershipcategorization. In J. G. Hunt, V. Sekaran, & C. Schriescheim(Eds.), Leadership: Beyond establishment ylews. Carbondale, IL:Southern Illinois University Press.
Mandell, M. M. (1956). Supervisory characteristics and ratings: A summaryof recent research. Personnel Psychology, 88, 435-440.
Murphy, K. R., Garcia, M., Kerkar, S., Martin, C. & Balzer, W. K. (1982).Relationship between observational accuracy and accuracy inevaluating performance. Journal of Applied Psychology, 81, 320-325.
Neimeyer, G. J. & Neimeyer, R. A. (1981). Personal construct perspectiveson cognitive assessment. In T. V. Merluzzi, C. R. Gloss, & M.Genest (Eds.), Cognitiye assessment. (PP- 188-232). NY: GuilfordPress.
O°Keefe, D. J., & Sypher, H. E. (1981). Cognitive complexity measuresand the relationship of cognitive complexity to communication.Human Communication, 8, 72-92.
Osgood, C. E. (1962). Studies on the generality of affective meaningsystems. American Psychologist, ll , 10-28.
Pervin, L. A. (1973). On construing our constructs. QpprgmpprpryPsychology, _@, 110-112.
Peterson, C. P. & Scott, W. A. (1975). Generality and topic specificityof cognitive styles. Journal of Research in Personality , 8,366-374.
Peterson, C. P. & Scott, W. A. (1983). Toward fundamental measurement of
dimensionality. British Journal of Social Psychology , 82,197-202.
114
Pierce, J. L. 8 Dunham, R. B. (1978). The measurement of perceived jobcharacteristics: The Job Diagnostic Survey versus the JobCharacteristics Inventory. Academy of Mapagement Journal,gl, 123-128.
Pokorney, J. J., Gilmore, D. C., 8 Beehr, T. A. (1980). Job DiagnosticSurvey dimensions: Moderating effect of growth needs andcorrespondence with dimensions of job rating form. OrganizationalBehayior and Human Performance, gp, 222-237.
Roberts, K. H. 8 Glick, W. (1981). The job characteristics approach totask design: A critical review. Qourpal pf Applied Psychology, pp,193-217.
Rush, M. C., Thomas, J. C. 8 Lord, R. G. (1977). Implicit leadershiptheory: A potential threat to the internal validity of leaderbehavior questionniares. Organization Behayior and HumagPerformance, gQ, 93-110.
Sauser, W. I. 8 Pond, S. B. III (1981). Effects of rater training and
participation on cognitive complexityz An exploration of Schneier'scognitive reinterpretation. Pegsonnel Psychology, 3p, 563-577.
Schiffman, S. S.., Reynolds, L. M., 8 Young, F. W. (1981). Igtrogugtipgto mutidimensional scaling. NY: Academic Press.
Schneier, C. E. (1977). Operational utility and psychometriccharacteristics of behavioral expectation scales: A cognitivereinterpretation. Journal of Applied Psychology, pg , 541-548.
Schneier, C. E. (1979). Measuring cognitive complexity: Developingreliability, validity, and norm tables for a personalityinstrument. Educatiog and Psypholpgical Measuremegt, 33, 599-612.
Schroder, H. M., Driver, M. J., 8 Streufert, S. (1967). Human informationprocessing. NY: Holt, Rhinehart 8 Winston.
Scott, W. A. (1962). Cognitive complexity and cognitive balance.Sociometgy, gp, 66-74.
Scott, W. A. (1963). Conceptualizing and measuring structural propertiesof cognition. In O. J. Harvey, Motiyation and social interaction.
NY: The Ronald Press, 266-288.
Scott, W. A., Osgood, D. W., 8 Peterson, C. (1979). Cognitiye structure:Theory and measurement of indiyidual differences. Washington, DC:V. H. Winston 8 Sons.
Seaman, J. M. 8 Koenig, F. K. (1974). A comparison of measures ofcognitive complexity. Sociometgy, 3, 375-390.
115
Sechrest, L. B. & Jackson, D. N. (1961). Social intelligence and accuracyof interpersonal predictions. Journal of Personality, 1;, 167-181.
Shepard, R. N. (1972). Introduction to Volume 1. In R. N. Shepard, A.K. Romney, & S. B. Nerlove, Multidimensional Scalingz Vol 1 NY:Seminar Press.
Solomon, D. (1977). Perceptions of similarity between striving tasks andthe generality of task preferences. Motiyation and Emotion, 1,181-192.
Stone, E. F. (1976). The moderating effects of work-related values onthe job scope-job satisfaction relationship. QrganizationalBehayior and Human Eerformance, 1;, 147-167.
Stone, E. F. (1986). Job scope-job satisfaction and job-scope- jobperformance relationships. In E. A. Locke (Ed.), Qeneralizing fromlaboratory to field settings (pp. 189-206). Lexington, MA:Lexington Books.
Stone, E. F. & Gueutal, H. G. (1985). An empirical derivation of thedimensions along which characteristics of jobs are perceived.Academy of Management Journal, 1;, 376-396.
Stone, E. F., Mowday, R. T., & Porter, L. W. (1977). Higher order needstrengths as moderator of the job scope-job satisfactionrelationship. Journal of Applied Psychology, 6;, 466-471.
Stone, E. F. & Ruddy, T. R. (1987). Dimensionality of perceived jobcharacteristics. Paper presented to the Annual Meeting of theSociety of Industrial and Organizational Psychology.
Streufert, S. (1982, August). Cognitive complexity. Paper presented tothe 90th Annual Convention of the American PsychologicalAssociation, Washington, D.C.
Streufert, S. & Streufert, S. (1978). Behayior lg the complexenylronment. NY: Wiley.
Taylor, S. & Crocker, J. (1981). Schematic bases of social informationprocessing. In E.T. Higgins, C.P Herman, & M.P. Zanna (Eds.)Social cognition: The Ontario Symposium (Vol. 1). Hillsdale, NJ:
Erlbaum.
Thompson, E. G., Mann, I. T., & Harris, L. J. (1981). Relationships amongcognitive complexity, sex and spatial task performance in collegestudents. British Journal of Psychology, 1;, 249-256.
Tripodi, T. & Bieri, J. (1963). Cognitive complexity as a function ofown and provided constructs. Psychological Reports, 1;, 26.
116
Tucker, L. R. & Messick, S. (1963). An individual difference model formultimensional scaling. Psychometrica, 2g, 333-365.
Turner, A. N. & Lawrence, P. R. (1965). Industrial jobs and the yorker.Boston, MA: Harvard University Press.
Vannoy, J. S. (1965). The generality of cognitive complexity-simplicityas a personality construct. Journal of Personallty and SocialPsychology, P, 385-389.
Vernon, P. E. (1972). The distinctiveness of field dependence. Journalof Personality, gg, 366-391.
Weinberg, S. L., Carroll, J. D. & Cohen, H.S. (1984). Confidence regionsfor INDSCAL using the jackknife and bootstrap techniques.Psychometrika, gg, 475-491.
Weiss, H. M. & Adler, S. (1981). Cognitive complexity and the structureof implicit leadership theories. Journal of Applied Psychology,gg, 69-78.
Wicker, A. (1969). Cognitive complexity, school size, and participationin school behavior settings: A test of the frequency of interactionhypothesis. Jourhal of Educational Psychology, gg, 200-203.
Witkin, H. A., Dyk, A., Faterson, H. F., Goodenough, D. R., & Karp, S.A. (1974). Psychological differentiation. Potomac, MD: Erlbaum.(Originally published 1962.)
Witkin, H. A. & Goodenough, D. R. (1977). Field dependence andinterpersonal behavior. Psychological Bulletin, gg, 661-689.
Witkin, H. A., Goodenough, D. R., & Karp, S. A. (1967). Stability ofcognitive style from childhood to young adulthood. Journal ofPersonality and Social Psychology, Z, 291-300.
Wrigley, C. S. & Neuhaus, J. O. (1955). The matching of two sets offactors. American Psychologist, lg, 418·419.
Young, F. W. & Lewyckyj, R. (1979), ALSCAL-4 User°s Guide. Chapel Hill,NC: Data Analysis and Theory Associates.
Zaccaro, S. J. & Stone, E. F. (1988). lncremental validity of anempirically based measure of job characteristics. Journal ofApplied Psychology, lg, 245-252.
Zajonc, R. (1960). The process of cognitive turning in communications.Journal of Abnormal and Social Psychology, gl, 159-164.
117
APPENDIX A
FAMILIARITY QUESTIONNAIRE
118
Directions: Please rate your familiarity with the work activitiesof persons holding the jobs listed below. Please use thefollowing scale for your ratings.
1: 2: 3: 4: 5:
NOT AT ALL SLIGHTLY MODERATELY VERY TOTALLYFAMILIAR FAMILIAR FAMILIAR FAMILIAR FAMILIAR
1. CASHIER
2. JANITOR
l 3. FIRE FIGHTER
4. TELEPHONE OPERATOR
5. BANK TELLER
6. MAIL CARRIER
7. TAXI DRIVER
8. FILE CLERK
9. ACCOUNTANT
10. SHORT-ORDER COOK
11. ACTOR
12. ASSEMBLY-LINE FOREMAN
13. MACHINE OPERATOR
14. ELEMENTARY SCHOOL TEACHER
15. FLIGHT ATTENDANT
16. WAITER/WAITRESS
17. CONSTRUCTION WORKER
18. TAILOR
19. LAWYER
20. RETAIL STORE MANAGER
119
Appendix A (Continued)
Descriptive Statistics:
Familiarity Ratings
Job n Mean S.D. Minimun Maximum Range
Cashier 39 4.231 0.872 2.000 5.000 3.000Janitor 39 3.949 1.169 2.000 5.000 3.000Firefighter 39 3.821 1.144 2.000 5.000 3.000Telephone Op. 39 3.410 1.208 2.000 5.000 3.000Bank Teller 39 3.641 1.246 1.000 5.000 4.000Mail Carrier 39 3.744 1.141 1.000 5.000 4.000Taxi Driver 39 3.667 1.243 1.000 5.000 4.000File Clerk 39 3.333 1.475 1.000 5.000 4.000Accountant 39 3.282 1.395 1.000 5.000 4.000Short·0rder Cook 39 4.051 1.050 1.000 5.000 4.000Actor 39 3.795 1.151 1.000 5.000 4.000Assembly Line 39 2.974 1.442 1.000 5.000 4.000
ForemanMachine Operator 39 3.308 1.454 1.000 5.000 4.000Teacher 39 4.051 1.025 2.000 5.000 3.000Flight Attendant 39 3.385 1.290 1.000 5.000 4.000Waiter/Waitress 39 4.410 0.677 3.000 5.000 3.000
Construct. Work. 39 3.667 1.402 1.000 5.000 4.000Tailor 39 3.282 1.376 1.000 5.000 4.000Lawyer 39 3.590 1.371 1.000 5.000 4.000Retail Manager 39 3.615 1.310 1.000 5.000 4.000
120
APPENDIX B
BIOGRAPHICAL QUESTIONNAIRE
121
SubjectNumber:Job
Similarities Study
Interview, Interviewer, and Interviewee Information
InterviewInformation1.
Date: Month: Day: Year:
2. Time: Starting Ending: AM PM
InterviewInformation3.
Name: Last: First: Mlzl
InterviewInformation4.
Name: Last: First: MI:_
5. Street name and number:
City: State: l____ Zip___
6. Telephones: Home: Work: _l____
7. Occupational/jobtitle:8.
Jobduties:9.
Nature of present job: [ ] non—managerial [ ] managerial
10. Employing organization°s function (e.g., manufacturing,sales, education, government agency):
11. Tenure on current job: Years: __ Months:_
12. Combined tenure on alljobs, past and present: Years: Months:___
13. Years ef formal education: ___l___
122
Page 2
14. Diplomas and/or degrees (check highest level of attainment):
[ ] High school diploma or GED equivalent[ ] Associates degree[ ] Bachelors degree[ ] Masters degree[ ] Doctoral degree
15. Sex: [ ] Male [ ] Female
16. Age in years:
123
APPENDIX C
JOB SIMILARITY QUESTIONNIARE
124
General Instructions for Similarity Ratings
1. Please enter the last four digits of your social securitynumber in the boxes in the upper left corner of both attachedopscan forms.
2. Be SURE that you begin with the opscan marked "Form No. l"in red ink at the bottom of the page. On Form No. 1,please mark your responses to similarity ratings 1-100 inthe spaces designated 1-100.
3. When you reach line 101, you will notice that the "1"circle is coded. Do not erase this code. It tells thecomputer that you have marked Form 1. (It is also a reminderfor you to stop working on this op-scan form and turn to thenext op-scan form.)
4. Then, on Form No. 2 (so marked in Blue Ink at the Bottom of thepage), please mark your responses to items 101-200 in the spacesdesignated 1-100. Please be sure that you are coding your ·op-scans correctly.
5. Again, you will notice that on line 101 of yourop-scan sheet there is a circle darkened. This time the "2"circle will be darkened. This will indicate to the computerthat you have marked Form No. 2.
6. Please turn to the questionnaire, read the instructions at thetop of each page, and complete the questionnaire by darkeningthe appropriate circle on your op-scan sheets.
7. PLEASE MAKE NO MARKS ON THE SIMILARITY QUESTIONNAIRE ITSELF.
8. When you have completed your similarity ratings, please proceedto the remaining portions of the questionniare packet.
THANK YOU FOR YOUR ASSISTANCE.
125
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
1. CASHIER andRETAIL STORE MANAGER
2. BANK TELLER andWAITER/WAITRESS
3. SHORT-ORDER COOK andTAILOR
4. ACTOR andBANK TELLER
5. FILE CLERK andLAWYER
6. ELEMENTARY SCHOOL TEACHER andTELEPHONE OPERATOR
7. MACHINE OPERATOR andTAXI DRIVER
8. FIRE FIGHTER andTELEPHONE OPERATOR
9. CONSTRUCTION LABORER andMAIL CARRIER
10. LAWYER andWAITER/WAITRESS
11. BANK TELLER andFIRE FIGHTER
12. MACHINE OPERATOR andTAILOR
13. ACCOUNTANT andASSEMBLY-LINE FOREMAN
14. FILE CLERK andTAXI DRIVER
126
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
15. ELEMENTARY SCHOOL TEACHER andSHORT-ORDER COOK
16. MACHINE OPERATOR andFILE CLERK
17. ACCOUNTANT andTELEPHONE OPERATOR
18. CONSTRUCTION LABORER andLAWYER
19. FILE CLERK andJANITOR
20. ACCOUNTANT andFIRE FIGHTER
21. LAWYER andRETAIL STORE MANAGER
22. JANITOR andSHORT-ORDER COOK
23. CASHIER andTAILOR
24. FIRE FIGHTER andMAIL CARRIER
25. CASHIER andJANITOR
26. ACTOR andELEMENTARY SCHOOL TEACHER
27. ACCOUNTANT andSHORT-ORDER COOK
28. CONSTRUCTION LABORER andELEMENTARY SCHOOL TEACHER
127
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
29. LAWYER andTAXI DRIVER
30. ASSEMBLY-LINE FOREMAN andRETAIL STORE MANAGER
31. ACCOUNTANT andFLIGHT ATTENDANT
32. CASHIER andTAXI DRIVER
33. BANK TELLER andTAXI DRIVER
34. FILE CLERK andFLIGHT ATTENDANT
35. CASHIER andLAWYER
36. FILE CLERK andTAILOR
37. JANITOR andMAIL CLERK
38. ELEMENTARY SCHOOL TEACHER andTAXI DRIVER '
39. BANK TELLER andTELEPHONE OPERATOR
40. ACCOUNTANT andELEMENTARY SCHOOL TEACHER
41. ASSEMBLY-LINE FOREMAN andJANITOR
42. ACTOR andTAXI DRIVER I
128
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
43. ACCOUNTANT andMAIL CARRIER
44. SHORT ORDER COOK andTAXI DRIVER
45. LAWYER andMACHINE OPERATOR
46. TAILOR andWAITER/WAITRESS _
47. ELEMENTARY SCHOOL TEACHER andFIRE FIGHER
48. JANITOR andLAWYER
49. ACTOR andFIRE FIGHTER
50. ACCOUNTANT andLAWYER
51. FLIGHT ATTENDANT andWAITER/WAITRESS
52. ACCOUNTANT andCONSTRUCTION LABORER
53. ASSEMBLY—LINE FOREMAN andELEMENTARY SCHOOL TEACHER
54. MACHINE OPERATOR andRETAIL STORE MANAGER
55. FLIGHT ATTENDANT andTAXI DRIVER
56. ASSEMBLY-LINE FOREMAN andMAIL CARRIER
129
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILARA MUCH COMPLETELY
SIMILAR SIMILAR
57. CASHIER andELEMENTARY SCHOOL TEACHER
58. ACTOR andFLIGHT ATTENDANT
59. LAWYER andSHORT-ORDER COOK
60. FILE CLERK andTELEPHONE OPERATOR
61. ACTOR andSHORT-ORDER COOK
62. ASSEMBLY-LINE FOREMAN andTAILOR
63. ACTOR andCASHIER
64. LAWYER andTELEPHONE OPERATOR
65. ELEMENTARY SCHOOL TEACHER andRETAIL STORE MANAGER
66. CASHIER andFLIGHT ATTENDANT
67. ACTOR andMACHINE OPERATOR
68. FLIGHT ATTENDANT andMAIL CARRIER
69. TAILOR andTELEPHONE OPERATOR
70. BANK TELLER andCONSTRUCTION LABORER
130
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: S: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
71. FIRE FIGHTER andFLIGHT ATTENDANT
72. BANK TELLER andTAILOR
73. MACHINE OPERATOR andMAIL CARRIER
74. CONSTRUCTION LABORER andRETAIL STORE MANAGER
75. BANK TELLER andMACHINE OPERATOR
76. ACCOUNTANT andCASHIER
77. ELEMENTARY SCHOOL TEACHER andFLIGHT ATTENDANT
78. CONSTRUCTION LABORER andTAILOR
79. ACTOR andLAWYER
80. ACCOUNTANT andWAITER/WAITRESS
81. CASHIER andFILE CLERK
82. FLIGHT ATTENDANT andLAWYER
83. ACCOUNTANT andTAXI DRIVER
84. FLIGHT ATTENDANT andRETAIL STORE MANAGER
131
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
85. BANK TELLER andLAWYER
86. ACTOR andTELEPHONE OPERATOR
87. ACCOUNTANT andTAILOR
88. ELEMENTARY SCHOOL TEACHER andWAITER/WAITRESS
89. TAXI DRIVER andTELEPHONE OPERATOR
90. ACTOR andMAIL CARRIER
91. ASSEMBLY-LINE FOREMAN andLAWYER
92. FIRE FIGHTER andWAITER/WAITRESS
93. ACTOR andCONSTRUCTION LABORER
94. ASSEMBLY—LINE FOREMAN andMACHINE OPERATOR
95. CASHIER andSHORT-ORDER COOK
96. FLIGHT ATTENDANT andJANITOR
97. ACCOUNTANT andRETAIL STORE MANAGER
98. BANK TELLER andJANITOR
132
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
99. MAIL CARRIER andWAITER/WAITRESS
100. CONSTRUCTION LABORER andFLIGHT ATTENDANT
' 101. JANITOR andTAXI DRIVER
102. ACTOR andASSEMBLY-LINE FOREMAN
103. FILE CLERK andSHORT—ORDER COOK
104. RETAIL STORE MANAGER andTAXI DRIVER
105. CONSTRUCTION LABORER andFIRE FIGHTER
106. ASSEMBLY-LINE FOREMAN andTELEPHONE OPERATOR
107. FIRE FIGHTER andTAXI DRIVER
108. ACTOR andTAILOR
‘
109. CONSTRUCTION LABORER andWAITER/WAITRESS
110. MAIL CARRIER andTAXI DRIVER
111. MACHINE OPERATOR andTELEPHONE OPERATOR
112. CONSTRUCTION LABORER andFILE CLERK
133
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
113. ACTOR andWAITER/WAITRESS
114. MAIL CARRIER andRETAIL STORE MANAGER
115. ACCOUNTANT andMACHINE OPERATOR
116. FIRE FIGHTER andLAWYER
117. CONSTRUCTION LABORER andMACHINE OPERATOR
118. ACTOR andJANITOR
119. CASHIER andMACHINE OPERATOR
120. BANK TELLER andRETAIL STORE MANAGER
121. ACCOUNTANT andBANK TELLER
122. JANITOR andRETAIL STORE MANAGER
123. CASHIER andWAITER/WAITRESS
124. CONSTRUCTION LABORER andTELEPHONE OPERATOR
125. MAIL CARRIER andTAILOR
126. BANK TELLER andSHORT-ORDER COOK
134
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
127. RETAIL STORE MANAGER andTAILOR
128. ELEMENTARY SCHOOL TEACHER andMACHINE OPERATOR
129. FIRE FIGHTER andSHORT—ORDER COOK
130. JANITOR andMACHINE OPERATOR
131. FLIGHT ATTENDANT andTAILOR
132. ACCOUNTANT andACTOR
133. CONSTRUCTION LABORER andSHORT-ORDER COOK
134. ASSEMBLY-LINE FOREMAN andBANK TELLER
135. ELEMENTARY SCHOOL TEACHER andFILE CLERK
136. ASSEMBLY—LIINE FOREMand andCASHIER
137. RETAIL STORE MANAGER andSHORT-ORDER COOK
138. LAWYER andMAIL CARRIER
139. ASSEMBLY-LINE FOREMAN andFLIGHT ATTENDANT
140. FILE CLERK andFIRE FIGHTER
135
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
141. CASHIER andCONSTRUCTION LABORER
142. FLIGHT ATTENDANT andMACHINE OPERATOR
143. ASSEMBLY-LINE FOREMAN andSHORT-ORDER COOK
144. FIRE FIGHTER andTAILOR
145. ASSEMBLY-LINE FOREMAN andTAXI DRIVER
146. ACTOR and .FILE CLERK
147. MAIL CARRIER andTELEPHONE OPERATOR
148. MACHINE OPERATOR andWAITER/WAITRESS
149. ACTOR andRETAIL STORE MANAGER
150. ELEMENTARY SCHOOL TEACHER andJANITOR
151. ASSEMBLY-LINE FOREMAN andFIRE FIGHTER
152. FLIGHT ATTENDANT andTELEPHONE OPERATOR
153. CASHIER andFIRE FIGHTER
154. BANK TELLER andMAIL CARRIER
136
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELYl
SIMILAR SIMILAR
155. CONSTRUCTION LABORER andTAXI DRIVER
156. JANITOR andTELEPHONE OPERATOR
157. TAXI DRIVER andWAITER/WAITRESS
158. RETAIL STORE MANAGER andTELEPHONE OPERATOR
159. BANK TELLER andCASHIER
160. ELEMENTARY SCHOOL TEACHER andLAWYER
161. SHORT-ORDER COOK andTELEPHONE OPERATOR
162. FIRE FIGHTER andMACHINE OPERATOR
163. FILE CLERK andWAITER/WAITRESS
164. TAXI DRIVER andTAILOR
165. FIRE FIGHTER andJANITOR
166. RETAIL STORE MANAGER andWAITER/WAITRESS
167. ACCOUNTANT andJANITOR
168. BANK TELLER andFILE CLERK
137
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
169. FIRE FIGHTER andRETAIL STORE MANAGER
170. MACHINE OPERATOR andSHORT-ORDER COOK
171. BANK TELLER andELEMENTARY SCHOOL TEACHER
172. CASHIER andTELEPHONE OPERATOR
173. SHORT-ORDER COOK andWAITER/WAITRESS
174. FILE CLERK andMAIL CARRIER
175. BANK TELLER andFLIGHT ATTENDANT
176. ASSEMBLY-LINE FOREMAN andCONSTRUCTION LABORER
177. JANITOR andWAITER/WAITRESS
178. FLIGHT ATTENDANT andSHORT-ORDER COOK ~
179. CASHIER andMAIL CARRIER
180. LAWYER andTAILOR
181. ASSEMBLY-LINE FOREMAN andWAITER/WAITRESS
182. MAIL CARRIER andSHORT-ORDER COOK
138
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR183. ACCOUNTANT and
FILE CLERK
184. ELEMENTARY SCHOOL TEACHER andMAIL CARRIER
185. JANITOR andTAILOR
186. FILE CLERK andRETAIL STORE MANAGER
187. ELEMENTARY SCHOOL TEACHER andTAILOR
188. TELEPHONE OPERATOR andWAITER/WAITRESS
189. CONSTRUCTION LABORER andJANITOR
190. ASSEMBLY-LINE FOREMAN andFILE CLERK
191. SHORT—ORDER COOK andELEMENTARY SCHOOL TEACHER
192. RETAIL STORE MANAGER andASSEMBLY-LINE FOREMAN
193. MACHINE OPERATOR andLAWYER
194. TELEPHONE OPERATOR andFILE CLERK
195. MACHINE OPERATOR andBANK TELLER
196. MAIL CARRIER andACTOR
139
Use the following numeric values to indicate how similar the job dutiesare for the pairs of jobs shown below:
1: 2: 3: 4: 5: 6:
NOT AT ALL SLIGHTLY SOMEWHAT MODERATELY VERY ALMOSTSIMILAR SIMILAR SIMILAR SIMILAR MUCH COMPLETELY
SIMILAR SIMILAR
197. JANITOR andELEMENTARY SCHOOL TEACHER
198. RETAIL STORE MANAGER andBANK TELLER
' 199. FILE CLERK andELEMENTARY SCHOOL TEACHER
200. FIRE FIGHTER andCONSTRUCTION LABORER
140
APPENDIX D
REP GRID MEASURE
141
I.D.
JOB DESCRIPTOR QUESTIONNAIRE
Directions: Listed below are ten sets of job titles. For each set ofthree job titles, please think about what makes any two of the jobssimilar to one another, but different from the third. First, write adescriptive word that indicates what is similar about the two jobs in the"Descriptive Word" column. Now write down the word you believe to be theopposite of that descriptive word in the "Opposite" column. Continue thisprocess for each group of three jobs until you have written ten (10) wordsand their opposites. Please don't repeat any words; use a diffferent wordfor each of the ten sets of job titles. When you are finished, pleaseturn to the instructions for the grid task.
Job Titles Descriptive Word OppositeCa) Cb)
1. cashiertailorlawyer
2. fire fighteractorelementary school teacher
3. taxi driveractorassembly—line foreman
4. cashierbank tellerwaiter/waitress
5. actormachine operatortailor
6. short-order cookelementary school teacherflight attendant
7. file clerkassembly-line foremanretail store manager
8. telephone operatorbank tellerconstruction worker
142
9. telephone operatoractorretail store manager
10. fire fighterbank tellerfile clerk
l&3
JOB GRID TASK
Instructions:
l. Please turn to the GRID Form on the next page.
2. Carefully copy the ten descriptive words and their opposites from theprevious questionnaire onto the grid form. Copy words from column(a) on the Job Descriptor Questionniare to column (a) on the GridForm. Copy words from column (b) on the Job Descriptor Questionniareto column (b) on the Grid Form.
3. Make sure that you position the next page so that the words "GridForm" appear centered at the bottom of the page (and Code No. appearsin the upper right column).
A. Look at the job CASHIER listed in the first column (upper left handcorner of the grid). Using a scale of +3 (very high on thedescriptive word) to -3 (very high on its opposite), assign the jobof cashier one number for the pair of terms (descriptive word and itsopposite) in Row 1. In other words, your rating for the job CASHIERfor the pair of words in Row l will be be ONE of the following: +3,+2, +1, -l, -2, -3. Place your rating in the square on the Grid thatcorresponds to column one, row one (i.e. the first square in the upperleft corner).
5. Next, moving DOWN the first column, rate the job CASHIER for thesecond word/opposite terms (in Row No. 2) and so on until you haveentered ratings for the first job on all ten descriptors/opposites.Be careful to place your ratings in the grid square that correspondsto the appropriate job column and descriptor row.
6. Now rate the job JANITOR in the same manner that you did the firstjob, and so on until all twenty jobs are rated on all tendescriptor/opposite terms.
1A4
Ä .tf=7> *7O
O* L<>*
S *7ä8
'Z
Q N+2LLI2D-§ “" 3 E
*=‘**'*·=‘~·'*<·*·<==*=**.*?.éS‘.*Q.*i*.·é.*‘3.'é°.i‘2«‘?L E
*3******** 3****** *****3** FFIIIIIIIIIIIIIIIIIIII ä**3******* F*IIIIIIIIIIIIIIIIIIII
****3**** IIIIIIIIIIIIIIIIIIII*******°^* **°'*°********°° IIIIIIIIIIIIIIIIIIII
*’*3******^****3*****^ ** r * IIIIIIIIIII*'*******3**** ******** IIIIIIIIIIIIIIIIIIIIuaaovauoouos Auvmawawa Fvn
---“m------.-u**°*‘***3*° 3*****°‘*** IIIIIIIKIIIIIIIIIII*"*”**“°**"**‘^**‘*’*****' IIIIIIIIIIIIIIIIIIII
*****3** IIIIIKIIIIII A * * Illllllllllll 7 22~v2~¤<>¤¤* ‘¤ IIIIIIIIIIIIIIIII
‘ 2>•¤**¤ **2 2 llllllllllllllllllll¤¤^·¤¤ ·¤*2 2 IIIIIIIIIIIIIIIIIIII¤*·¤¤*¤ ***2** 9 llllllllllllllllllll**3**3* *******3 3 IIIIIIIIIIIIIIIIIIII
*0****3*** 3**°***3*3* F IIIIIIIIIIIIIIIIIIII**3******* 3**** F IIIIIIIIIIIIIIIIIIII
**°****‘** F IIIIIIIIIIIIIIIIIIII**3******F IIIIIIIIIIIIIIIIIIII
M5
APPENDIX E
BIOGRAPHI CAL QUESTIONNAIRE
146
I.D. No.
PARTICIPANT INFORMATION
1. Have you held a paying job?
2. If you have held a paying job, for how long?
Years: Months:
‘ 3. What is/was your most recent occupational/job title?
4. What are/were your job duties?
S. Nature of most recent job:
[ ] non—managerial [ ] managerial
6. If you have EVER been a manager, for how long have you been one?
Years: Months:
7. Employing organization's function (eg. manufacturing, sales,education, government agency):
8. Tenure (length of service) on current job:
Years: Months:
9. What is your age in years?
10. What is your year in college? (eg. Freshman, Sophomore, Junior,Senior).
ll. Are you a male or a female?
147
APPENDIX F
PERFORMANCE RATING QUESTIONNAIRE
148
PERFORMANCE RATING TASK
Instructions:
You are being asked to participate in a teacher evaluation study.
In a few minutes you will view videotapes of four instructors. Please
watch these carefully because you will be asked to respond to questions
concerning the performance of the instructors.
After viewing each instructor's presentation, you will complete a
two-page questionnaire. Please use the orange op—scan sheets for your
responses. There is a separate sheet for each lecturer.
When you have completed the ratings for the first instructor, please
wait quietly for the researcher to begin the next tape. This process will
be repeated until all four instructors have been rated.
Your ratings will be used for feedback and research purposes only.
They will not affect the 1ecturer's employment status or salary.
Individual responses will be strictly confidential. Only group data will
be reported. If you have any questions, please direct them to the
researcher. Thank you for your participation. Your assistance will
enable us to better understand the performance appraisal process.
149
I.D.
Teacher Evaluation Form
Directions: How often did the instructor you just observed on thetape demonstrate the behaviors listed below? Please use thefollowing scale to indicate the frequency. For each of the twelvebehaviors listed below, place your rating on the appropriateop-scan sheet.
1 2 3 h 5 6 7
Never Almost A Few About Often Most AllNever Times Half of of the of the
the Time Time Time
l. Examples were presented which were clearlyrelated to the central topic. -----··-----·
2. Lecturer used purposeful nonverbal behaviors(smiled, pointed) to emphasize points. ----··-······
3. Lecturer stopped in midsentence. —-----·-—-··—
4. Lecturer was hesitant (said "eh" or num"). --—-··---····
5. Lecturer lost eye contact with audience. ----··-······
6. Lecturer provided facts or evidence to supportbroad generalizations. ·-···········
7. Lecturer spoke fast. ·············
8. Lecturer acted nervous. ······—······
9, Lecturer spoke in a monotone for a Sustained period. ··•••········
10. Lecturer gave clear answers to questions. ·············
11. Lecturer varied his facial expressions. ·············
12. Lecturer appeared unsure of what he was saying. -—···········
150
Page two
I.D.
Teacher Evaluation Form
Directions: Using the following scale, please rate the instructoryou just observed on the tape. For each of the nine categorieslisted below, place your rating on on the appropriate line on youropscan sheet.1 2 3 4 5
Very Poor Average Good VeryBad Good
113. Thoroughness of preparation ------------
14. Grasp of material --------—---
15. Organization and clarity --·—------—--
16. Poise and demeanor ------~------
17. Responsiveness to questions ---------·-··
18. Educational value of lecture —---·········
19. Rapport with audience ··-·—·——·····
20. Speaking ability -·---········
21. Overall rating ·············
151
APPENDIX G
PLOTS OF GROUP STIMULUS SPACE: INDSCAL AND ALSCAL
152
L4•
M CU Cs x EU I-I I-I I-IU
TU fü O O ü'«CM CD I- 3 U zuU ld Q) fü
Q)’U
M Q. CI-IM GEZI-. ZIM-I CI-II-IOQJQI ··-Im<‘.
:11-- A: 1Iu~-•··-•::..e-I,44.»Z Zü··*T'J I- fU>4—-JUMOG UI-=I E-#3 :2 cz--I-:u¤ <¤>.<I:> M 1a\ I—I I-dOM3DD'?3Q)UI—-4 MC lJ·*E—·$·-4
¥-I UFÜSS-MMUCGLUQ··—•·—·~ 1) OM USG. I·I-IU -I·I:JE,,:4.a
J .Z?J.;¢#-*X—IO1)>¤U 1I·•-IUUII-4„'i'•C}f1.)C·lÄI·I-I<: u‘:•-I:·-3~—•I-I-I3¤I-<><I-II:·I-—LIc,Icr1~p·:T.} I1UZ-¤3Ur•<Z¤¤-1&•&-IE-¤—L>EZ<tl<!3¤—*IE ...........„...„„..„
-I{
7 RS-5 7 E=.· 7 ·22 E Egs- ¤~I IQ.-II I gp,>I I g‘{ I {~I I g
¤ 1 I I
S? E 75m2 : ~ 2*E.} I {E; I ;°{ I {III‘· I z *9>=
IIn-Ä
_= 7 N=
aa; 7 9I I
gg 7 imS; I
‘6zu I-IIoraI ;‘*. I " *.
E ° 7I·•-••-•r·•»•»n-••-«•-••-1•-••-a»«•-••-••-••-«•-•»«I»-e•-•»-••«•-•»•¤»-an-IE»-«¤•-••-••-«•-•¤•-•»-«»«•-••-«»-•¤¤•-••-••-•»·••-••-•»··••-1•·••«r·••-« -:
EI I = i'PEE;'“
. E ., '" .··s{
I IQ
I“' I
§‘
*I
"‘
äi E 2..E2 E *I@=. .. I I"{ IE2 g .. Ä.Q;
{Tge I ?MI I II-I I I
ms I I
¤*. I *.9ut
I IN
>II II
-«I II
5*. I *°E I EI I ImIAIIIIIIIIIIIIIIIIIIIIIIIIIIIÄIIIIIIIIIIIIIIIIIIIIIIIIIIIIÄÄ
I I I I I I I I I I
9 "? 9 "! 9 9 9 "! 9N pa pc Q Q Q •-a -·• N
I I | I I
153
$,1 •
1.1 JJ EO :& 1:H $1-1 I1-4 $44-J1 1 :2 .2
“ä1.1 71 M ·y 1; U <U JJ".'J
;, w 1.1 Q. :1-zu GC
ggy_gg-•:; $-1E-J3 CJ C~-ILIUI-4 ¢T5>~1<C
> 1,,1 ;;\ 1-1 1«1o>.1:o:1_r6¤J4-1·—+ M1.1·1-•·-1
QJ Of-1 USG- *4-4-3 ·H3......4-3_:,,:1:;;+-1.z·-1OU>~.U JJ--cJv1—1„C‘.OJJOD··-1
<·_ 7;,.4:--13-1-11-J·—I3:U—1><$-E--UUUl·r<CL} 7.1 U TU 7.1 O _TU U ;-I Q _J.J _<1J TU :1-1 _C> TU CU U U1 1:-I KUU; ;;y_x13L;;—·<c-=-.-r•r-‘r~:-„>ZZ<*t2<·'1-=-·">
E . . „ . „ . . . . . . 1 . 1 „,•—: . . 1-1 -—1~:·w\1·m.¤1~::o¤=<:z¤<.;:‘.:.1:x..'J..»-4-:>1¢
I I I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIII¥IIIIIIIIIIIIIIIIIIIIIIIIIIIIMN{ : ~-{ { ä
ät : g'UI 1
·-«I 1 IEI.
: 0 IQÄ'! I {N‘*.
5 äMI 1 ;I < 1 g
{111-I n 1 In-Ä
1 II I§‘·I *.Ei E E
* I IQgr I I I'.-SI: N Iaf = 1I II 1 g
: I1MI 1 ggZt 1 >-I IOI Q 1 3EI. G l II 1 ¤ {~i E Lgg ‘·I.
1 .E IIII : M
In. Q:
g ..
ä · * FE 1-n—·x g15{
I {MI I IG1-- 1 ..gi I *7Q1 Isi : z°¥
"‘I E"!S1 1.1 I {T
ää I {MI- I I'MI 1 I(/1I I II I ICWI 1 IN>l I IIMI I IZI I IDI I II 1 II I I*1 I *,9IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIINI I I I I I I I I 1
9 "! 9 "! 9 I'! 9"‘
°N •-1 1-« Q G Q -1 N1 1 1 1 1
154
5..I •$-4 $1) Ec 4c E..I LI LI $44.4U') I1 G LS
L-4 7: L4 4.4 TQ
~ ?II ° 2 -I.2S 1)····
At.4.-73-*75 $-· ’U>H•-J$-«IOC'« I-I:1 2-3
L. _;4\-
L4OL4t>£:J€':QIu-4 L4C CJ—·r·!-4
*-4 C.§$-4’lJ.Z'.'2--•I·-4LJCZS.¤4-I€3
-·-•·—• UOS-IL)if)A
E ...............„„.. .—II
I-I·I ,
ES E Sag ° I $2sg s ;~-1IE
E · Sz IE:.
I-IQ:L ·gg : S.
2: E Sz I
‘.q
>: 2 .4 ·-·-; · SI 1
§S E SI2*: E . *2Ei = „ S'S ·=E-4
I·;
SI = EI 4LES „ I I. §
°· 0IL'*
Is: 5 SZT I fc!ég I g?
I
ä' I °' ·6; I S0_ { m 1m
Sl I.I I {Tjl I {
ES E S1 I2: : ZZ
E: : ··sr : S°§ I {
U I
1 I im
1 1 1 1 1 1 1 1 1 IQ Q Ih Q Ih Q Ih QN -4 -4 c o c -4 -4 AI
I I I I I
155
L-1 •I-• TJ i0 .~z E4-J I-1 L1 In-4.1
71 3 O O LuiL1 U) L1 3 1.1 gg1) IJ U 3 CU".?
J SC $-1 LL EI-JI-1 *:1:J 3I--I CI-·L1OQ)Q.1 -1-IQJ<'.
’-T-1J····At 1111--I··-I 5.1.1-IJ1.1
CZ fü-7: II- 11>•-I1-II-10 i 1.1ZI 2-3 U Ii-.iUI-4 3>¤<.‘> I-1 U\ -1 I-1OL1:D::111-1-1 uC U--M I- LSI-1I.112·-1 L•LJii.::1..1Q
·-1-- ll OH USC:. I1-11.1 -·-1:$E,:«.1J .i7:.:¢•-I.:t1·+3CD>~1UGJ·-•1JU1•—4.iO¤JOD-1-4< U1¢-I:·-¤·-<·-I-•3=IJ·—•><L1t:·—·U;1;r>·—•:LJ _7: U C': 3 _Q _€':
U ;- _5U _€U _3 :1-1 O 3 CU U U1 1-4 :1:ggE
.....1.....•...„....
1-1{888:88888§I8I88{88888§8l888;8I888:88888:8II8I=8|888:8888I1h
{ I §"‘1 1
*8 1.1- Q 1 --:1 1 1U1 1 1-1 1 1•-1 I ID
1111 1 IN>1 1 1U1 1 1
{ I {M1 1 1
1 1 11 1 1$‘.
I {**3-41 111 1 1-«EI 1 1
1 1 11. 11111 1 1
E{ I {
°* I {I : 29gg- 1
1••
I I **1 x I
=.1- 1 -SI I
: U G Q ·ll'1
n-11 1 86
g· : Iz: < U : M 8
°‘{ I e inE{•-••·A¤M1«¤1-••-•¤•-••-•>·••«•-••<•·4•-••-••-1•-•»-•»••-1¤•<•-•1-1:•«»-11-1»••·••<1-•1-••-•>«¤•-•1-•n•1-•r-11-1•-••·••-•r•»-••-••4r·••-••« ·ö1 1
§* ·~EIIII EE : .9
lI O„ 11
. 1•
..g g5; .„ . „ §H8 ·0 I
°1..1<lQ : ‘
äi ··‘
I‘*‘
-u-16· = §vg { ,., .**3
9* E ETä: 1 1
. 1 ·
== E EVI} 1 IO
„ 1 *'§: = EI·-•1 I t5*- I ·1:: I ä
I I 2-1111111111111111111111111•11111111•11111111111111t1IIl1111111~
1 1 1 1 1 1 1 1 1 1
9 III 9 ·I1 9 I1 9*·’! 9
6; 1-1 1-1 Q 0 Q•-• •-•
N1 1 1 I I
156
L4•
*-4 CU CIO .Z La {-Au
I-
O O CLII-4 71 S-I 3 4..1 :3_ 3) C) 1) cn OU-• SU S-I Cl- C2}-4X-I GC·-7 73:*-* Fc CTI-4$·•OOCD EUE
Z 42-2 E·—·3 CJ Z·—•.:u2-I C1>»<!‘.> $-4 lJ\
**‘$40%-ASODSQJA-J2-I In
O *4 LJ!-«QJ..Z2~—•S4L)ZZ„’J·4O~.«—-• U OH 11.:::..CJGO•I-4
<c ;n-»:·I-:>··-•+-•--I3:w;s-—I><»«::-I-Ic;:1v7·-•.·:LJ I'J'U¢U7‘3O_€UU;•·*Q_C1)Q)€U·r4OC\]<T$UU]··#<UE U1-’¤3Ur•<a-~r~&4E-3-U‘£.Z<<Za'¤E . ..............„....I-2 —•:~Ir·¤<r·¤^.-¤r~¤¤c¤<:¤¤LJ::2JL¤LD:—<*7:.¢
{IIlälilllzlllllzlllllglIiII;l*lIlQlllllgillllglllliglilllh
I =‘“
:1 I E1:ut I {.'I{ I {G.::1. I {QZ2 I 'aj 2 Q.
: 2 ES} 2 -.4*‘ • su
:•-•
äi — E *1ExlI°{I {
Q! : IQ
I V‘* I x z
ZI: I {gg Q
•
=N'. ° I‘·"
E"!:2: : ;·=Ol I {fl. Q :1
•·•{
~I -‘
E·•-n»•>-••-4•-••-••-••-••-••-••-••-••·•>•><>-••-•¤•-e•-n-••-••-4•-«»-«•-••-•Enun-••-an-«»<•-••-•»<¤•-••-••-«•-1»••-••-••·«•-••·«•-«•-•»••1•-•>·<>·•EQ
: Q?9.
'“I .I{I = I¤_: In
Q I Q?" I
E '§: = 2 “ . Ä.;... I ..
{1* " I TEQ '° 2 ~ Qnl ¤
I
s· · QIgt'.11
't
Q', I *.
EE 2 E1
·1c
Iuu• •I~
>:• :«
52{ I {°{ I {
1• 1‘. I *.**7
: : 1 1 1 : 1 1 1 I
Q un O Il! ¤ ll! 0 Ih Q
A .-2 .-2 A A
157
$.1 •
$-I GI) E0 x E•-J LI LI L44.»
V} G O O LCI5 U1 LI 3 u cg
U 1) G QJTJ..: :0 LI CL :I-ILI cz:IJ 713-Ia-ICS-ILIOGJQ)..
I!—•:! LI m>u•.1¤Q: LaLL] Z-43 Z) C···•£L)LII G>s<> LI 1)\ »—I LIC
U—'{r•!—I II-I LJ!-Iü)£.2··-III-ILDGCISJ-JO·I-•·—• 1) O3-I USB-..:G„£¢#;·£I·1O1)>xL)1)·-<CJU3I—I„:O1)OD—I-I
LJ TUUCUGO}! U···*¢'J1)1)€’,1--•O€¥J<'JUU7v-It'.}CQ Ux-‘I>3U;—•<CIl¤—7é·&·&·¤—UEE<<Z¤'1
E „.„..„.„............
1 1 1 1 { { 1 1 11.11111111x11111111I111111111{11111 11111111It1111111111111IIn
E{ nd
zwE<1I g
ur I 1L11 I 1E'. I *.9IL.1 I ;N
i{ I {'I I *.
"{ L I ,„ {1 I ;
S'. I {I'?§{ I {"2:. E {I-I1 I {
°{ I {{ °’
" I *.9g8 I 1-I
{ I ° {§{ I 1 In {48 I 1>-1 I 1gl I 1„,*.
‘? {"E•-I8 I IQcz: U I 1Ol I 1‘{
I {
-§ I·»••-•¤»«»••·«•-•»-••-••·«¤•-••-••-«•-•»•»•»••-••a>·«>•>-1»-••«»•»1{•-«•-er-neu»-<¤•·«•-••-«»-«¤•-••-•»••·••«•-1•-••-1•-a•·••-••-••-••·••-•·é
g{I•I .IE; “ I ·· {
9. I ."2
i ° I g?I N{ L')
··i : 'ä: „ 5 {„48 D I -|
II = §‘,.,. E .si .„ „ : §‘·*{ I ."IS!
°’I {T
.I1 I 1§{ I {:{ I {I/>• I 1c{ I *.9sul I IN>1 I III-I1 I 1"‘*. I {ä: I :
1 I 11 I 1I.n8-88I8888l8I8I8888888888l88888888888888888888888888I8l888l.¤;
1 1 1 1 1 1 1 1 1 I
Q Ih O QIhNI-1 I-1 Q Q Q I-1 In N
I I I I I
158
$-1•
:-1 1: ::C: Q: E
. 1-* =-· 1- pa-:I_ 37 =¤ 0 c :z.:U 1* >· 2 —· 1:SE Id 1: vs DU
, ___ ,1;:. ::1-Is-I :1:-1 Qaßdkcwg) --11:E ZJÄÜZ
Z--3 J: C--I"'U$-1CE3>~1;kl ij Q1 -1 $-11: ;;I-«1:..i‘.:‘.—»-1:-«L;::.:1:.:;*,
·;·-- .,1., 1:;:.1-1
„~• ,.,--1;)* U)- :$5 _~y_;1_;'2'I}_'3_T'JL:;·*€'J_’l)_’LJ€II··•C?J€’3LJJU}•;"fU
• ed; • • • • • • • •,• • • • • • • • •
QlllglllllälllllgltllIgillII¥ItIII:IlIII=IlII!zIIlllällllm
E =;·‘
_ I2,:. Ä .uu Ä *-1; I {;‘. Q ro
Z; ; Q", '
u
N: :.‘
: · *S
5. Q: am
Q' * Q"Q?
”E Q
HI 1 ·'
°: 1 Q ·^ Qva: Ä {Q>E I
:,4
-
‘ ·• u_,9 Q x(I 1
.
Ei · E iN* ¤ ¤
‘."Ä
ES 1•¤
ot 1 *§·. I S
‘ '
rl Q 1 Q•;Q? I ;·9 1E-
:1 •
„.§ 1 · QQ. 0 Q IL ·IlI
E = 7..= . Ä ·gl ’ Ä Q
Ei Ä"‘ .9
si = “ FI
E' Ä ·E; = 1 Q°i Q Qx .··1‘§: 1 Q7E'. 1 QIII I ··
GS 1 gQ'. 1 r¤wi 1
·,Q
>! 1
:S 1 VQ; ·
‘·
= 2 =*_ Q
:1:1Illlälllllglltil}!IIlK:IlIl!:IIItl{IIl!I:llIII:I!IIl=l!II—<\I
Q Ih Q III O Ih Q Ib Q
APPENDIX H
ADDITIONAL PLOTS OF GROUP STIMULUS SPACE FOR FOUR SUBSETS
160
L.; •
Q H GJ Go .z E
V7“
E *5*‘“
C'.! 7-I-EIII an III 3 u gg
es Q__ II: cu*¤2. -X-ILII CC
F iU$-I4-4 CHHOUU ·I·IGJEI] ’.I'13J·r·¢ .x ';Ji1J·-·4·I·I2I4—I-34-IU1 <1§I—I<¤ III ;;>2..1ul•-IQ c: I.:¤ ZZ-I3 za
H U\ I- {IIU1 1)*%*-4 $-4 LJ$-I U-C2-MHUGGQUO
·—I·I-I LU Q2-II 11:; II-III.: ··-I¤E.:4.IJ<CI
.·—I2
=°·‘
' {SY {S{ { .„ {III; I xSY Ö *.9III: I ,„>IIN:
:-”
x QI5*.{ *."!
EE E E"‘§Y {"‘
YE;.2’{
{ {"x I ;
·x I ¤ x< I ;22 I · 2.Üi In Ö xlémxIotIzx Q x
..§ „ I·;
gi EI Q I‘£‘{Q SE; Q InI °éI-I Q änI
SE ° I .:. I ß Iq
<x I •·<
I22 —~ = 2·pqI ') I I
gg Q zII9E{ ."!
S{ { {T.I•IQQ'.
I .»-Ix I x>-x I tmx Q x¤*. I Y9mx I x~>t I inIII: I“*. { .QS: I s
x I xx I x* {
*.“!xnxxtxxxxxxxxxxxxxxxxxxxxnxxxxxxxxxxxxxxxxxxxxxxxxxxxxnxxxw
x x x • x x x x x I
°‘^ 9 "‘ 9 "! 9 "3 92 -‘ 2 2 2 2 2 2 2
löl
Lt•
$4 1) E3 X E-,
, I-I I-I I-I III.:cu 0 0 :„:
<1 Ü LI 3 I-) ggJ II Q: TU Q.:#7 I-4 1 :I-III ~E
P ¤I—·-4 cs-III-«0:.IQ.: SQ;Ed
:I)‘r'*X QjU•,.4„,.q¤1‘_;__:‘_I
J1 :15--431 I-I u:.***3 G.) C.-·-I.T.UI~«I IU>~<!}
;-I *9* ’U\ I- I··4OI—Ic¤¤:¤:L:»-II-I ;.«JJ l:—•%•I-I I—« UI-IQ.1.:;=.·»-II-«;>.·:.c:..:¤I-::>
1 ·I-I·I·4 I) OH 1);¤, U-J.) ~•—4¤E__':;,j
·-4 £-'I$.:t#-*.:¤—·IO3J>~UQ.J~-•CL)<DI—I.ZOQ)0¤.„Ijfl U:"‘:°'-IO""*$-J?'3c¤*"I><$·l¤'•'!L)(J(J°]••-{¤
Q ZUTSUUII-ICQ
• • • • • • • • • • • • • • • • • • • •·—· ·—·^4<*~I'I¤~¤I\¤cc~<::z¤;;:.::.Ic¤.;¤:I:IA·—,;,2
I I I I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIII
;I
IN
· i-
‘
_;. I ·<I I IQI In I I»«I I I
;*. : *.9WI
I IN
>I I I-I I I
I. III
MI I I
g I I• I Q I
Zi I *.*IIA: I
I-•
QI I I
IIJII I>·« I'
E3 I I
GI X I {Q I IQ
gI‘ II•·<
3I I
3I I
‘I Q I I
Ei E . ·5} E imEI. I In IQ
ää I {III~§
·· I L
gi *„'I I
§§ I gQ
IIn
‘ I‘6
I •-1 inI .
g: e I{
,.: . tcI,. r~ I ·~‘§ I *1*ä, :
·‘ iH. 1 I -‘*} Ie z {gs I *ux I *_••'I
gv I lv-4
ä: I ä'
2* I‘
AI'I I
,.g I I*’§ I Iowg I
IN
Z{ I {'
zu I¤_
gi I :I E ?··II.III:IIIII:IIIIIIIIIII{IIIII:IIIII:IIIIIIIIIIIIIIIIIIIIIIN
9 “" 9 *‘9 *! 9 *3 9
Ä A •-Ä d o o A A N
162
}.1 •
111 11 gAJ A.J° zi 1. E
U1 IG O G 2-C1111 :1: 1-1 3 +-1 ru
" E1 E E :1.2 EEEM CBS-11-1 CHHOUG) •1-IG)::1 CZQ1··—• 1x 1111-1-«·1-4 ¤.1-11-.11.1U1 31-13 $-1 Q.1>;JA.15-1CG U:1 E--3 21
$-1 0\ -• I-ACI-1D¤3<UCU1-11-4 $-1vz 11-1E—·1-1 1-1 Q1-111.::*.--1111;):31:.::1-1C>
~-¢·-4 UChLJVJ
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1—-11
zäEESIWEtI =¤
ät I1I1
·er1gi‘= E im»-1:1I
1E: .. I EEI'I°tI ·=
·I/1·· I gc>I 1 {J-§ I :.1. 9 I ·.
II 1 InE; I I1 gg:z•1 gOl 1 gff I *.I»-« gu
1>1 119'I6 I" .E§ 1 I°1 I ."!I I II
I ‘gl I e1-1I
I Q:1 „" *1;g: .1 1
Igg2 v~IL11 gn1 1 ..S: : :1·gt I :Ct I ·”vs: 1
I *.9wu 1 uuZt I {‘KI I gct 1 g
t I tI_I Im
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII.¤1I I I I I I I I I 1
9 "I 9 "! 9 "T 9"‘
°ev ·-• ·-1 0 0 0 1-1 1-1 6}1 1 1 1 1
163
L4•
¥-I 1) CO A4 E
7}I-I L-1 1-1 11.11.1
LI In E éO LC4-)
< 1) U Q) »¤ In
. ¢¤ I- :. :1.11-1
·-‘ --1-1-1:*..1.1
Q.„ ·=·'j,“ 1)·
.::11-4 .
-1UZ $-11).:2. U-U ·r·—1D="'4:— Ars ‘ —1:>11>»;1 1)·•-I€U€fJ•··4.C ""“'.
S VI*-*¤'-··*¤··-1«-•-«3:·s·-1><s-«::—·-•;»8ä-EI-1IJE
äF1”l)2:°J;Ö_T'3L);-äil1)_1)_€V-1—O¢U€UUU]1—1I$
Z• • • • • • • • • • • • • • • • • • • •
—·Q
1-{I ,
::7; II.
2% 1 =E; I W?E: E Eu
n 1 ;
”E~ E I
g*_ 5 m um¤• 1
·”,,Q
Qt 1 ;wl 1
·Z- 1 ,.4 .¤• 1
‘
nt 1‘
Q 1 ·9; 1 {Q
2 ·1 ¤ E"
EE E ·^‘·
*2 : I§¤_ ¤ 5 {111Z2{ 1; I {°Qt u 1
·Il Q 1 (
I 1 (
N; I g5£•<•·•¤>-•••>-••-avarertv-11-11-«•-••·•um-av-11-er-an-•»«>••-•»••-•»-1E>••-a><•<•••·1»<•·«>-an-an-er-er-nnur-11-••-•r·•»••-••-1>·••-•><•·<v·«~§
„Q 1 . ·Q. W I .IE = E¤· II! I In
{‘
°Q E N {I
·1 H {
I 11 =¤<¤ ¤ 1 {1-;ä‘ I '„¥ " I iur 1 °
II
=§2 52„„
II la
ä: . 1 2·E'. 1 ·.
EE E I„·. 1* IQwc 1 ;N
>I1 (1
•-«•1F"_5 *
gl III
u 1·t 1 ;
{ 5um
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIllIIIIlIIIIIlIIIII~f~It n • • t t n t • 1
c IQ Q Q III c an o
N ¤·~ •·< Q Q 0 -4 A N
1Öl+
$-1 ":0 .x E
A 1; 1.. L-.1 .11.1'
·C O
”' "‘
1-1 vz 1.1 3 1..1 *5’_ CU €U’U
___ 1-·6 11-1-1 0..5;; E5:1 C'.·1J•v·1 x 1.10.1--1-»-1:.+-1,.1.1.1Z/1 73-15 La 1)>&-bu!-1C: AJ=¤ 2-13 11 1:~-1.::J1-..1 <¤>-<c:1 1.. 11\ 1-1 1-eos-1:-O:.'c¤<1J1-11-1 1-1CA U-1%$-1 $-1 L};-¢Ü;:'!'*$«I;J:¤.¤J¤J¤~-1--•
IJ 011-e ’JJ.:& 14....1 ·1-•‘.Z'JE.:4-1—I .25.:4*-*-At-101J>~U1J·¤4Q)U}r—1£OGJbD~H<CLJ _1’J IJ 5 ?3_3 _¤¤ U·-1 71} SJ _¤J <¤~·-· 3 <U Q U vl1—1 CU
E ........„...........—<I
I I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIItIIII:II1IIIIIIII1t1{ ' l'NI I E·I 1
..4- 1 .
EE E E;‘.1ät1 :~*t 1 ;
°{ =·E E§Ä· E 11 ZÜQ2 I - {6:- „ 1 ·..- 1 I;§ I io
.1 {Q
Q; .1 11 I
I 5 imN- 1 ..
E{ I „ {Q2{ ¤
‘“I {
~{ E {E;>«1-n-•¤»-••-1•-«>-•¤»-«•-«»1»•>«•«»1»••-••-«•-••-••-•»-«•-«»«1•-an1•-•>-«»·«•-••·•»•¤>«»-«»•¤»«¤¤¤1-•»«•·1•-•»••-••-••·1•-••-•»••-•·;
52 5 · §·E
“‘{
°. ° * I .*9I
‘?
ä ,. 1~ E-—n 1 I1
I 1 -ät 1 {6:· Q I 'J-Ä
6* = §*,..i I
'“.
11I 16: : E‘·". I {*9
I 1 I-n€22 « 2*
E: 1 :°W: v I Iowr 1 IN>I 1 ll-11 1 I“¥ I '.gl 1 I
{ I {I 1 InnIIIIIIIIIIIIIIIIIIIxIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII-N
I I I I I I I I I 1Q U1 Q sn Q _ Q an QQ „-Ä -• c c Q •-Ä 1-Ä N
1 1 1 1 1
165
)-1 •
$-1 1) CÜ .1 E
.1 *1 E"‘ "‘“
·O B-
S- U1 5- 34.1<Pkw_
1 L-A--153
-=l ¤C.)·—• x 1111-1-1-1-1:;.1.1.11.1vl €¤·—·<': s- ;1>4.14.11..1Q;; 1,1
E 21-3 :1 c::~—•.::.11.1 :¤>„<:..1 L4vl 11-1:—•>« $-1
Q) OX-1 1).SC- Q-11.1 -1-4Z)E,£.."J-I·—· .-':C1!.1%-*.1-<O1)>~.UQ)·1-4”J.)U'J1—4..‘JOC!.)Cl1°)•v-|ff'- ¤¤¢-•:·1·•C~-•u-—·3<!J·-><k-CZ~1-1;1L1:f1·—1:
Z• • • • • • • • • • • • • • • • • • • •
1- 1-1:~Jr*‘1<r·1.*1„91\:x3·;1¤<5‘_;·:1;;;_;g;;,;;:,..1.j>é
UlllälllllglllllgllliU=UIllU;IlUUI=IlUUI§IIlll=IIUlU:iUIUIh
2 =é*‘
:·. E xät : :-x 1
‘
EZ; I r=z-1 1 umEQ ¤ {•
*2 . E ;°x 1 ;n 1 ;m§.’ I ,5;
*£{ I {
EZ ä‘·
ot 1=t 1 ;
g:.“’I
" *.9{ I {
Z"~ I“'
E(U 1
t 1 {
Et : .. :.„En? m I icize 1 N ;CI 1 gzt 1 ;—··· : <~t1
'^¥ I 1.,E{Mr-an-11-•»•>·a=¤v·•>1-1-1-1>1•-••-•»1»«•·-1•-1•·a•·•>-;‘«1>••-••-•:>-1-•>1»-1-1»••-••·1•-1>1•-••-1•-•»1•-a-·••-••-••-a-1>-••-1•-••-«•·1>1•1 -6
Q2 ., g§‘
1E; : E°. I
"‘.9
| Q
E = §‘‘„ I .
st ··~ I EL"€ I E94 »« 1
1-•
ää = . { §'xi : 1Eu
| 0=‘·". I E"!
U 1 :-1ä: : ;·• 1E; : 1.-g 1~ g1/111 1 x• 1 tc
ut 1•~
ä{ I {'s
S:E : E.Äntxtntxsxxxnsnzxxtxrstxxxxx;•x•rxx•xx•¤txx•t••xttxttxttÄ¢~)
U t n • • t • I U 1
Q th O NQ Q Q Ih 0
~-1 -1 G G G -1 -1 N
‘1 1 1 1 1
166
:4.4 •I·• CJ C0 ;¢ E
_ I-I H $-4 $-44-1
4. is E é °“‘°
“’S4 E éi 4:.2 E?
FIII-44-I2-IGl)~r-· X 41JJ)·v·1~4-• 2.l-I,.]4-J
in :*.:-4:*: S-4 cL1>uu¤-IQ: 4.1:¤ 2-3 11 :••·•.•:U$-4 <¤>4<:E I- 0\ —* I-4OI-4C>DS<¤1JI-*¤··* I-aC/1 1J»—~&·4I-· $-4 U3-41).12.--4S-LJC’;'.4’J4-IC)
-4--4-4 3 3444 11.::14.Z'4€'J.;¢¢-#.:d—IO1)>«UQJ·4-43J1)»—4.ZCDQ)fJI}·»-•
< CfJ4-•4’I'4···•C·-·•4—4•—43C\1•—-I>·<$-4Z·»-•LJL)F4fJ·-¢’.1'U i741I?J€UC_C’.} U··-·!\‘1_1J1Jf’J·—4OIU!UUL¤r-ITUrg ;:2:=¤3:.2r·<:.‘¢=-.:-<e•+·:z4;.>z2‘.<t<c¤¤—>E 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
:-4Qxxtäxxxnrvättxtsgxxxxx=•xxx•$xtxx¤:•x••¤¤=x:xx¤•=•¤x:••x=xx•x•n
E E E"EE E E:I{ I *
4 x<41
-¤ 4 {u' 4 ;
NI4{I {
SIA 4 m 4.: arm:-4: u ·,Qmn : ,gn 4 ,
$*4 I *.EI I {
=O In äg‘§x^
I "
: : V_ •.:¥ " I *.‘i
E " IE'. : Im·-·¤
¤w•€é
ät . : * ·xx nn 4 Ä
.4Q E = Q~•-••-4:-4»-4•-«•-n«•-4:-•»••-•:-••-4•-•=•-«:-«>4-«»«•-«>-«:-«»4•-«•-«»4«E•-«»«•-4»-4»«:-4•-4:-4:-4»-«:«¤¤»•:-«•-«•-4•-4»«»·-4»-4»4•-4•-«:-«»•»« -q
EQ : Q?9; I .E; : Q°. "' I 4 .*3
{ I {?Q : Q
gs ¤ u• ¤•¤
’;.". I'i‘?
‘Q I Q'?ä
I.4.
‘ "‘I .
u.! I Iu I V- xgs • •
°*. I <€‘“
9{ I {T;‘{ I {E{ I {mx
• •
c? v I {9we ¤ «~>x ¤ x:h-tl I l
zu ¤ x¤• ¤
•• ¤
•I I I
'. I *.*3
Q an Q st} Q un c un Qex; 4-4 -4 Q Q Q -4 T er
167
X-4 •6-6 a6 :O .& E
Zn *:3 $5 $-6 $-48-)·
O " CL6 VJ $-6 3 6.6
Ag;
33 Q) U Q) CB Q)'¤
2-] CQ)‘6'* Ad 61) 1)***-6-+ C-$-*6-14-Jvu ::1-6:*:: 66-6 1)>A-66.6x.6Q: 6.62 :6 :·—·.:;6s-6 m>6<:ID $-6 _U\ ·—6 66-6Ol6-63DDNQJ*-*6** HIn U6-—+r•)6·« $-· L;$-|ÄU.;°:‘.·-4L«L):¤.C·UÜ
Q~:~-• 1) jh 1).:1 YH-) ·—•DE,.:A-I
.„ .-6C'J„;¢6-*.££* ÖU>~6U G)·6-•1)U'}—1_:OQ)GD·v·¢ff', C/}•·JC„'·6·*O·—•¢-¤6—3€B6*'*>€!~4G·•-¢Q)CJCf}·6·¢'.T.'JZ5')
Ux:Q31.Jr<2.-l%·&·&·k.L}"ZZ<¥.‘<z}Lr.":
Z• • • • • • • • • • • • 6 • • • • • • •
-6{III:IlIIt=IIIlI:IIlIIIIIIIIällIII=IlIII:IIIII=IIIII=IIII•h
2 = 5*6
1: I ;A.: 6 g
;·. ; 11‘;Q I
{"‘
'- I {Ӥ E EEl Q 6 um-; I
"‘ {.46.1e 6 g
Zt 6 x
V- : 1-.6:ci6~ • L': eo
fx I ,4;{ 6 :
'I. Q QI
x 6e 6 N
Es. I N {6zu q 6 g
ox -6 6·ze rn 6 g
,6Q ‘= E Q{unna•-••-•»••-••-••-••-••«•-•»•»••-6•-•>-••-••&§6-6v·•»•»•>1r·6•-6Q•-66-es-es-6¤»-«¤¤•·6¤>-•>-«•·•6-•¤•·••••>-••-«•-••-••-••-••-66-•>-«•n ·;§2 6 §·$2 6 ·..5 6 QQ 6 r-1 nn
6ä = §·-
6 .
äi E E><I
¤ IQ
Zi I '„:5 622 . .. E
Ei I iEQ " I Q°. I. .*0£{ : {1*.6e 6 x
i". I *.Zi I {mx 6 e
e 6 to
¤· 0 6-6
we 6 IN
>e 6 :6
&'{ I {
°{ I {
1 E•x e
• n e e x e 6
Q sn Q nn Q IQ Q sn Q
N 6-6 6-6 Q c o 6-; 6-6 N6 6 6 6 6
168
>—« S éü I-I Ed
VI <¤ 0 0 1::4:I1 E 3 " 3 “ II“ w :11 -0
dl I-I 0. ::1.11-1 gc:
.1-U}1*1-153 I-1 U>1-I1—J¥«ICC 4.1
=¤ 2-30EI-I 11\ -~
1-IG!-ecozcucua.11-I 1.4
V1 lJ··*%·I- I·-I UI-1¤.1.::··-I 1.01:::.01-1:_ ··:·-I U 9%-1USC-<C
V¥I-IZ···O·-~I-I-—13=<1·-I><I-«I:I··-«a1;1¤11-1-·I:LJZ
• • • • • • • • • • • • • • • • • • • •
·· —·N¢’7<1'L¤~¤I\¤3C!'I<€-’2LJ.’?„1;1l·Z.2Z!L'I—-‘1>d
I I I I I I I II.IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII1
*I IN
I 1
—« I:: I {Q1- 1 ..
-r 1 [N
>r 1 [
In 1 [
er I ,'I 1 [
5: I1.: :1:1~ In 1 ..
EE E I"Er 1 :
Ei I {I
I1; II Io
ä; I ;.{ I {
"I 1 Q g
ii gI igi I *NI I
Im
Ei 1 : II 2öEi " VI I {·; I {~§ ° E IE-1-1>«>41-11-••-ev-sun-1»·1><><¤¤»••-1•-«>••·<•<•·1•-«•-<>-••-<•-•»<I•-11-11-eau>-«»•1-«»••·•1-1»•>-11-«=»4>-•»-••-«•-4•-«•-••-«¤»•»«•-•zg
I I1
uiIQI-
I 4
Ei IEi .4 I ,„
. „g 1 ..I Q
1I = EI..; I ;
ä'‘ *
I I I
Z'. I *.9<I
IIv-•
ZI Igl
Bi N II•-•-Iäi~I :“I· E < TS§s I z·•-•I I I
Gi I {I
I IQ
Q. 1 0 --wr 1 IN
>r I II
I 2I
I I
II Inh
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIGII I I I I I I I I I
Q In Q In Q Q
61 Q Q
•·1169
kd •
$-1 GJ C0 .:4 E4.1 I1-1 In I-euU1 fü O O 2:-IZ1 an 1-1 3 1-1 ccCQ GJ GJ GJ fü GJ’O30 I1-1 S- E5-AH CC-1"5LLI :1)··-•:4 1)1)‘·*·1-IG-H1J•-1*CD :7·—1¢U$-12
Z1—43 1)I1-1 1)\
1-1U:IJ '.:I·-1 1).::.. 1+-11.1 —1«:¤E,.:a-1.1 :.'fü1z+-).ä1—·3GJ>«UG.J·—<G.Jv)—4.:3GJ¤1'J·—·( uu;-1:--1 Q·1—·•-11-13::-•><I-1:~-1:4;:1;)--1.*:L} TJ 1) fü fü C _fü U ;- Q _lI _GJ _<ü :-1 O ii fü U U1 :-1 füci: ;;:<=3LJr•<-=--1:-—•r~:—-r-1LJ„:.Z.<1'Z<t‘.-r-1*:
E . . „ • „ • . 1 „ • „ . „ „ 1 1 . . . .·—•1
1 1 1 1 1 1IIIIIIIIIIIIIIIIIIIIIIIIIIII;IIIIIIIIIIIIIIIIIQIIIIIIIIIILHg : 9:{ E {1111
;¤•
11
»«11
1gf
I'.?
$‘{I §"
-11
11
1gM1
1 1.1 16 1;
11
;ä'.
I $*3§{ z I · {"1 -1 11
¥$I $S{I {
1 1: 11
$I $?
gl1
~ ug la{I {
*II
Iext
11
11
;gt: Q 1 •-1 1N'.
I $*3-11
IIQ
KI1 {2:I 1{· ·
“I {¤{•-••-••·«•-«•-«¤•-•>-1>«•·«•-1•«»-«r·•»•»-••-•>-«•-1•·«»-«1-11-11-1>«•-«»-41•-«>«>-•»~«1-1»·1>-«»-1»-•>-4>-1¤1-4»•¤•-•r·••-1•-•r·«•-«r·e•-«>·••-••-••-1ää— = “ §·9{
E mwb ·
Y 1ä
8{E .*11
1I · E1u 1 . 1Q
I 1EI I {:*, 5 I 1, '.°3(I
I(laI"
äi 1 ::1 : °6: 1 {°$I $*3g'I ‘
{T.111
:E1 15 1
«·"{E E
V7;
: *98*: :1::11
*
¤1{I
f8:
: :{F {*2•‘11111•11111111111111•1111111111111111111111111111111111111~1 1 1 1 1 1 1 1 1 1Q Ih Q ID Q th Q Ih Q11] 1-Ä 1-1 6 o 6l7O
II.; •
M U CI0 .~¢ E6-J 1-I LII x-I4
' 4I. L2 I3 é °“‘ä
2 CJ CJ (U CUQ)’U
:0 H D. CLAN CZOi-4l·-IOUZIJ -I-ICD
ll C'„1)·•··,:6an:11--Im 1-4 r;)>44I-IO: 4
2 E-I3 0 I:-—I„:0I.I :¤>„<6::9 s.I ;1\ -« 5-IO!-«:„0:§:¤Q;4I—·• MC!} M
OH 1)..:C.. WMA-J ·I-4¤E,_:4-JI-Y .Z€'J.;£4-J.:d·IG1)>~U QJ·r-IZ1JU’J-4..:.."C}<1)OI}~I-•<;
I.......• „.........-·I
I I I I I 3 311111111111111111111111111111111111111111111111111I1:1111In1 I ig{ I
I .5{ I :I-; 6 1
sl 5 EZ>• I:.
‘°{ „I I :z: I :,5E1,
I {,.§Qt .. := :äi E E°{ I II,
:oQ1
~I g_·; j,.]
{ I {-1 I ;2*; I ‘
11 I
E?”
I°°>«16 IoEI I .
Ol I ;**. I E
3 1 O„; 5 E~z 6 “ — §·Q. In I .IIII1 0 :sg . i°. I !"!
'QE 5 {·‘. II, E
g: ° I Z„{ I {6I.. 1 I q ..§{ I
‘T
8*IE:
I 1°¥ I
*.“'3¤ ZI 1»•gl 6 II
.I•I-I1
6 1-1 6 1I/11 6 1¤*. I {9III.: 6 IN>¤ 6 ll41 6 1ZI 6 1¤• 6 1
{ I {*. I V'?111111111111111111111111111111111111111111111111111111111~
1 1 1 1 1 1 1 1 1 Io In o an o un o an o6 I-I ..2 6 6 6 A A IJ
171
Ln •I—· ca I:3 Q: E
l 4-J In L-• I-AAJJ? TU O O 2.G
n va I-I 3 I.: Q:2 U U GJ TU
’U'G
_ cc I. Q. :>-III :1*:rf ZI-I4-I CI-II-IGGJILI -I-IQ.2. CU·-*.:41/}
5}-153 5-I U>-I4-III-IO :1 IJ=¤ z..·—•3 cu :--+.:¤>.«
:¤>«<‘.: :.. lJ\ -« I—•OI-¤:0:I='.I1I·-I--I I-I7: :2-9;-. I.
;; 3;.1),::..-.::I:x6-I..x-•Q1J>»cJ~1I-—<1I;0«;:>:1JcD-««
C.} _€T'„U€Y}ZUO_1LJ···IT’§_'UTJT’J-¤<CJT’JVJLJZ/}r—17JLQ .Jz¤33LJr·<‘.3...Ir•&·&•L:.LJZE<!2<Lr.I"7E „ „ . • . „ . . . • • . . . • • • . . .—· ·-—':\J€*Wq*L¤»£>I\CI>C7\<:‘,1ll)2‘.C.LI2.C.1—-·'-;:¢
s z I I I I I I {IIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIQ
I I~
I · I.a· I I(I I Iut I I
"{ I *.9ix I Im>I I I-I I I
*. I {~:I I I
I I I
I I I
5*. I {**7•-•I I :»-<ät '7 I I
I I Iu.II I I
>-II I IDI I I
I I I
{ ‘* ””I *.9
QI I I-<
I I Q II I I
‘I "I I
(I I Q n II-I rd I I
SI I II II III!
rq. eg ..
HI I GI IQ
act I IOI I I‘*. I {_I
II
MI I I.·>1•·1•·••·•¤»-•>4»-1•-•r·•¤•-«r·•»·4•-•»•>·•r·<•-<•-•>—••-•I·-1»<•«>-••-•I¤ur-II-an•-•»-•»<»4•-•r·<»-<•-•I·•I-•r••-••·••-«»(•·1•·••·•r·••·1•·«•·•£ö
Ef I. g . {·. I -IIi “‘ · E_
I
° I ."!I {9I {
„ I•
„g I 5
5: : I•-•I I IQ
(I I IdI 0 KI in
ä: IZI I 9 {¤I : rw§{ I {7*§I E I.:‘{ "‘ {"’{
·{CI
I5.: I IN
?.{ I{‘
zx I IIS;. I
I'
g I I
: IIm
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIINI I I I I I I I I I
Q Ih Q III O Ih Q Ill Q
9 9 9 9 9 9 9 9 9
L4•
9 :1 .·:0 x E
_ I-I I4 L-: I.I:.:II ra 0 0 2:
_ I- JJ L-: 3 4.1 I1;J —.1 GJ G1 ::5 qu-I;_ Q_¤ 9 2. cs-In ct:
CI-II-0 um--•q,I
-2- ...,1-I .z :1„y·-I--:2.,:.1,;:.:./I 1'J-IM II- ;U>:.::.1;.1Q: gl
E Z73 ll :~·-«.:;1>-I ¤>»<1.-I I-I _J\ -4 I-I OX-:-O:·:¤Q1+.1I—·I I-IV} 11-IrI-I I-I ;,JI411;:•—;.;gg;;_g„IQ
_··-··-·
11 CI-I 3.1.:2. I-I-...1 -435;;;
-.Zf¤.z·-*.¥·-’311>„;111~—I;1;I:I-I;·;g;;;~.L)-„.:
fiW-JC--• O---I-I31'J~·><;1::·-I;);;yg.,.•r;_·
ri '<§1J€11_Q_?3 U;-I !QPLIGJfT§·-IOC'JI'„3;1Zf}—I<1§Q -:.'..:23J·r·<!‘.2..r-%·%-:..L;Z'.E<f‘.<.C:-·1
Z• • • • • • • • • • • • • • • • • • • •
·—· ,-—·€\1$"7<TI!\~9I‘GO9<¥'.f$\LJ2.kJCs-Q.T.I-•-;;4
I I I I I I I I IIIIIIIIIII IIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIUI
I g~
E : z*I
I I
.1- I .
<II I
21I I IMI I I·-I
I IO
wi I IN>I I I-I I I‘. I *."‘{ z I
‘
I I 0=
SY I *.**1MI I IM
Q{ I {wI Q I
,-I I I
°*„„ I {
I I IQgg. N I ~->I
I I-I
I I Q I
II I
"I I I
.:· I .
<II I
EI I ‘I
I IIIN. ,4 I ..
MII IQ
KI I M IQI I I
"*. IIE
Ä; :( ät)
-MMI-1MMMMMMMMMMMMMMMHMMMMMMMMIMMMMMMMMMMMMMMMHMMHMHMMHMMM§§II,Q.I .
IuII
sn I §¤‘I „.··e
° I v~ gI IE · ä
0 I QEI
UI I
MI I IQ
g I M
gl Q : guI
N. li I•
äi I §¤· E
I IM
II II
;{ I {..; I " ;>—I I I
I/II I II
I IQ
wII IN
>; I II,.,; I I
ZII I
GII I
II I
g I I
I I Ill?IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIN
I I I I I I I I I I
Q In Q III QIn·¢
* ·* = s· v 1· ·r
173
H •$-4 S'.) E:2 x EI.; LI In Leu
20 I1 O O CZ-ZI-I xr: I-4 3 u T1
LI O C) C1) 51 CJ’¤:13 II. Q. Q;-In ::3
E- ZI-Is-J cnncuaa ·-IQJI;.} S’i)*I'I Ai U<U·I-1-·-IQ.4-II-TI-I$/1 3-*51 $-4 1)>4·-Ib-II-4OC •—I:¤ Z-3 aa ¢:——I.:uI-I ¤¤>„<IZ2 $-4 7.)\ •—-I II-4;'}L«C‘„I}D€1I1,J¢.AI-4 I-•
uu 11-45-Is-I I-I LJI—•;J.::.·—I«L:x:':.g«-IO~—4·- C1) G$-• GJIL I4-·4-I -•D::.2.'I-•JI .2C¤:..’•-I.X•—·Q'l)>~IU $*-*97}-I.L.“O1JCO·v*<f_'Q
L;x¤3U°:"<-•;&‘?‘T‘;LIUzZ<<rLF1
E . . „„...„„I
I I { I I I I IIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIg 1 III
-, ; §GII.IL12
I 2Iz·- I {Q$*2 I 2"‘-l I II I IIFIÄ I
[I
I I z0 I
SI 1 IsnE2 I 2**
ID 1 II-III¤ I
CI I I{ "‘I {C,
*2 . : 2--I I I
2,:. I E·
I[ IE.; . I LI·-«I I IQ¤:I I-I I22 I *I--IIM 1-2 : - 2{un-er-4»4•-4•·4•-1»-«>••-«•·«»-4•-•¤•-1»n·4»«»·«•-••-••-••«»·4•·«>•>«1•-«•-4In-4>«»I»-I•-«»«»-«¤•-«•-I¤»«•«•-·I•-«»«»<•-••-4•·••-••-••-••-4·§
*2 ·=· *2‘
Q. III ‘EI4 1
¤ I lll. I ¤ ..I QI IE · E. I .nu 1 Ioi..2
"’I IQ1.. I -1 ..<I I I•-4
§* ’' §'
¤·· In I I‘ä2 I gI2 Q _IIIää E ETEI I II-I:
·I
52 I 2¤*. I *.9I.: I INZ2 I 2*RIA 1 ISI
~ II ¤ II·
I* I **9I‘I:I::II:II:IIIIIIIIIIIIIIIIIIIII:IIIIIIIIIIIIIIIIIIIIIIINI I I I I I I I I I
o III Q IQ Q IH o In Q
~ -— ·= 1 é 1 1* 1
174
$.1
IJ C'.O .:4 E1.1 L1 1-1
O L!-EZ A-J Q
31 2 ä ~“
EE" ..1:1:i- 1;:..1 $L«L.1’31J<lJ --«1J.;.1 29*-* M IJüJ·1·-1--1 C‘..1.J,..Iu:/1 :1J—·¢'J 1-1 1J>¤«:;.©: IJ23 E-3 IJ CI•·-1.:'.LJ&-1 €1>~.<CE $-1 -1 L1 QHCÜSQQJU- $-1ill
JJ " SSL *4-<•-J --1"'E;H—:
FJ1-I
I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIM{ I ä"g 1
*I 1..1- 1 .dl 1 I*:2 I {..• 1 I:1..I 1 IN>I1_:
1 I’{I I {
g ¤— 1 I:.1 1 Inn9.;2•/1I1 IZI 1 II.1I Q 1 I.EI 1 IQt 1 Ig m 1 II 1 IO§; :
“‘..1
I nä 1 Ig 1 I
ZI 1I «ä: _1 ..„
:‘.{ 1 IOä; I
"‘ {zu 1 {
"; 1 < gc
gg*1E;Q {
E . : he° v~ 1 o1 1{ Q {
é" „ 5“ ° E
:‘äi · 1 §*1. th •EI : Q
·
15* 1 Iux 1 *_“'§gz : {1*
1 I
2· .2 ‘—1
*»-I 1=
“¥: ·.¤§;" : :1•-I 1 ‘
II 1*_
F1'; 21 I
, 1 ImIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIN
I I I I I I I I I 19 us 0 sn 0 un c an 0·< ·* ¤ =? 1 v r 1
’S° ä.:d .,-
S-I L4 2-• L8-IU'! T'5 O O LC
I. L I-I 3 I-I 3LJ U LJ 2) T1!
€J’U
I-I Q- CI-I!-I CCE-•3%-*-*
CI-4$-42 U 1J ··-•1ILI Z ZJ"" X GJ U·-I-- 3..äJI-J-I1/1 :1-· T'.! I-4 U I-O C ·-I2 Z--3 12 C--·:14L- ¤>»<2 5-I $-4 C $-4 SDI! T'.! JJ4-II·—· Hvg gg-I5-:-I s-I UI-I€IJ.:.';.".~-I-.L;'.I.':.';•-I'.?
·•-•--4 ZUGI-IJ.,: t'J,.:,,/6-I.x·—I O 1)>~U U*··•
U Ir—•..’ZO TJ $-0--4<:'_]}•-I§··-•O·-•;J-·I3€U•">‘Ä$-I:"*L)LJU°}·-I'T-‘«
LJ T'3 1) I! T'.! O ,3 U ;-I ,U TJ ,3 ;- ,Q U L'} f I':zu L;:<.23LJr·<!‘.-=--rE-·r-L..„4.¢.<<2L7EI
I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIA
E‘
i"*‘II g
„• I {:: · ·txt- I *,9‘$‘«
i{"‘
-1 I gI I g
MI:”
E ES: I ·-··MI IIn I g
ZI I gI.uI I gZ- I ,7 „»-II I gQt U1 I g
{ I‘
9*-I:I , V"I I ä-1- I .<I I g>-I I 0 7S: I .. 5,Et- I ..
I IoZI I {
I„; „ I i•>1>1P·¢|¤|PI§-1I·(|·|I·I|·l!·1P·Ih-|PlP(|-|°I-|!l|-l|-\l•1D-lMPlI-I!-4{MMMMMMMMNHMHHNNNHHHHMMHHPCHHK;
ÄI9{I °IH IZ;IHIQ 0 I Ih
ä x In I °6II. I
‘
gt ‘* I :I I I
MI ¤ IQP-I I •«<I I A
I I I
Iäz · III.! I II I
°'. I {“'!mi “'
2 {TIn I xQ{ I 'iI-I I II~I I II/II I I
I I gcQ- I ..InI I IN>I I IIMI I IZI I IuJ· I Q „Qt I I
I I II I I{ 1 '."‘IIIIIIIIIIIIIIIIIIIIIIIIIIIIgttIII;IIIII:IIIII:IIIII:IIIIN
Io III c Ih c Ih o In c
~ ~ ~ = =a é :· I -9
5-I •
*3 3 äI-I II-II I-I I.I4.I
Y] Ü Ü O r-
.. I- 2 6-I ndä—· III PU 1I·¤I:X‘^
¤*"*'~I I—• ¤I>~I-II-Io: I.I=¤ Z·-I3 11I—• 1I\ I-I I·«I0I~I¤D::J<¤<uI-II-I I-IVI U-•%·I—· I-I LJI-IaI::.·-Is.IIu::.¤I-Io_ '··*'I" G) OH USC. M-Is-I --IjE__·;I.I
·—ggI-I
I I I I I II·IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIII IIIIIIIIIIIIQ
E ' VII
2; I I'„.„I I I
·-II I
»-II IQ
IIII I~
i{ I {‘. I *.
„yI•-I I I;
IIZ:I Ikll
QII
..
"‘{ „·I I{"‘
Qt II
§*‘. Q II I
Q: I In I
*. I *.9gz I z~
II z I
‘I I I
.a·Iä:') IInII
II N Ihn
I-III IC
BII I
2{ I {
~I =· I io·é
I I
0 I·•¥ ' E9-
I ~r~ I
!§ I §°
I 9 ,**2° ¤ =I I
äE..gIQI
s: : EI-II
I•·• IG
"' I“-I
*1I {I
EI I 2III I
E: 'UI IIn
ei I ~ iégz I z¤I~ I
I'
I-I II IQ
w·I
I~
iz I {'
§*. Q I *.
Q.I I
1 I {
gI Im
I I I I I I I I I I
o In Q Ih o ll! Q III QIJ IJ Ä <:> Q o .« I-I
~
l77
U]
c.>99 *-4 1::.$1-1E"@*-44-*
„& Q)Q)¥-4-•·1B.4-*L‘&-1UW
@·—•@ *-4 1:>e-*o1—«0: u¤¤ E-+3 1: ::~1-1.:31-1 ¤¤:>.<¤':> L4 c:\ »—1 1-1O1-«¤¤ <¤¤.:u1-1 1-1V} <U•‘*[—*$-* H LJ}-•G)iG-4 •L)¢::_¤;.:Q··¢·•·•
cJ OL-1 chic. 46..1.1 ·•-¢DEi4-*1-I<CU
@1:=¤<¤O<¤u·•-4c¤¤:c:c¤-1o<¤¤¤um·-1¢¤cg uacmzuz-•<:¥-=-..1e—s-&«¤=-L>zZ<t<L=.·1Z
••••¤ ••« •••• • •••• • ••
*-4 »—1t‘\A<"3<!'1.r'1»¤r\wc~<¤:;)Q;;.1;;„;_g;1..1»7>¢
I I I I I { { I QIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIII IIIIIIIIIII IIII‘\l1
3 1~1 = ä
Z. I .‘*{ I {:.1 1 1
-11 1o
1,1 1 nu
>1 ¤ 1..1 1 1
'. I '.,71 1 1
1 * L'! I
gt E:111
--1. 11A-Ä
${ I {E11. l
1-
Eä I {1 ¤ 1
1,,% I *.8>1 Q ac
1-1
1 ¤ 1g 1 1
-1 1 11*1 1.1.1u. 1
1g{ ° 111 M 1 {N'. "' I
'.“?
EE E E°"‘S I Y
{ I äM 1E{r11-11-nun•-••-1•-••-«1-•¤>1•-•¤»<»-«»-«v·1•-«»-11-1•·«»-1>·•»-•»·¢•-•E•-•¤>-11-11-«1-••·«»•>-11-«»•>-•¤¤¤•-«•-••«•-«1-¤1-••-•»<•-•>·1•·«»-•·;
1 1
ai 1 , I¥
¤ 1„§ 1 1. §Q 0 :
_¤q
g I M Z?1
. -7 1 ·..3 1
N 1E: 1 {1."' I i-Zgg {
E1
B: *>-•~ :.
ää = ¥°'
0 IYIII
; 1 1-«
ä; : . :·E*. 7 Y-.41
• I
>-1 : :‘”{ 1 -.9E1 1 :11-1 *
I
5*. I‘~
DI * I, 1 1
I {.-1IIIIIIIIIIIIIIIII:IIIII:IIIIIIIIIII:IIIII:IIIII=IIIII:IIII?:l
9 **1 9 **1 9 **1 9 **1 9e1; ,-1 J c Q Q .-1 •-•
~
178
L-4 •
H GJ CQ .z E
A I-I L-I I-I uu' IT} O O KLIC___Y1 s-I 3 u :¤* ä; E E -„.§3 29
2 II-I Q>·I-I·-II—C.'I a-ID -·I-..ZLJI-•
€U>~„<C
Z S-I _'D\ —• I-IOI-ISOZITJCUI-»·—I LIZ/1·-·I Q :.... II-I•-I ··-I :4-I
I-T .2TJ.>£I-I.¥I'IOQ)>~LJD·~DßI—‘.:OD3C·-I§ C/)·:*S·-I’§~·-„..J·—-13’?IK-*>'¢¥-I’I'«···IZJL)U'J·—dE
Q _TJ J TJ TJ Q _TJ U;—· ,D _D _¤ ~— 3 TJ TJ U Tl I-—I TJ
E .................. . .-· —·:~I¢··I\vm„cI~0ca<:=:.;:;:¤.;;:—«·1:6
IIIIIIIIIIIIIIIIIIIIII\=IIIII=RIIII=IIIII:IIIII=IIIII=IIIII|\
{ :•~I
-{ I {:1- I {Q1 I {
Z} II ..¤~ I IN
2} I {I. I I°‘{ I; {{
‘ ‘ä_. .„ : ·.~¤gl
:”‘
„: 2 :I· I .
E}“"
I {1 I g
g'. I ° {91 I ;,A
-2 E I..I~ I
Q I=S: .„ I :.„
N~ 7 II-•
-·&'{ I.- {°2: x g
Mpa I
Q; Eq·•-••-••-«•-••-•r·1•«•-«>·<•-••-•r·•¤•·1»••-••·«•«>•I-••-«>•r•I-•I-••-1»•E•—•»~•·<><•-<»•»·•>-«»-<I-•>-«I-•>-•¤•-•¤-•I-•»•r·4»«•-••-«•-«I-<»••-1•-«{QDI»•
I nQ. I .E I•·I I¤ I
InI cI = §·
. I N Q .I g5: I {EZ ~
‘°I EEI I
5* · 5A' II; 4 IzIgl I R‘-’{
I {"lgl I 1»<„{ I
{‘E1. : RI-«1 I 1»1 I 1I/*1 I 1
1 I IO¤· I ~·ISI: : 1NI-1 VI :°Z{ I {DI I 1
{ I {R I IM
I~IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIINI1 1 1 1 1 1 1 1 1 I
Q hh Q IH Q Ih Q Il! Q
IIS A A 6 A Cs;
p •II-I CD CO EI.) L. I-I I-I4.:
U'} K! O O I:-•C1I-I m LI 3 I-I TU9 ih 3 9 I-I.§E 99E-I ISI-IA-I OKI-II-•31)«1) ·—<1)I;} Cl)·•-4 .:4 C.)GJ·—<••-•C—#—·I.-]#·-I:13
:;—•c¤II-I :J>+-I#·4I-ISE #4
:1 Z-43 1) C--·..ZLJI·4 €'»Z>~<C:) II-I c;)\ ·—•IJOLSODQQIJQI x-
3;} ;)•—¢&•L-I LI ZJI-I1J;.".’Z·-II-•L)Z.'Z"-..’.¥#·4 O·-6--• Q) QS-I U-CC. wu ~·-IIIELL.)I-'I .;‘1‘T5.:4&-I.ZI··IOQJ>~IL)C1)·-—‘U7}r**."J.C¤'l.)3!l·*·4 ;,·;.I.«:-—Io·I-«4.I—I[3:·„;I-I><I-«*:·—·LI;)w~—I:CJ f¤1)1!!¤30_1‘JU;•-·I¤¤1),’l)_<U;~O€¤¤YJUUJr73an :.;::.:¤3LJr·<:=·-I&<r•r--¤-:JZZ<c<:‘.-¤·*=Q
E . . . . . . . . „ . . . . . . „ . • „ ...I
I I I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIQI
:EIN
-5 I .gl- III
V4 I 4:: : {Qig'
I 4-AI>I: I_:
I Imu.I I.
I : L.)Ä:Z: I I-AO, In I ..><I I I¤<SI I · {
2;** I‘.
EI ‘6 5{ U : sogl,I II-<; I {j‘.I *.
= ° 5I In I InN. I •-I*3 ..;: I .. {9
cu }=ff I In -i I “ I·•-•HH>4r·4>·•»-1H•-1•-1r·~r·•h<••1•-•I··4»4»••1»•»«>-<><»·1I·<»·4»-1E¤M><•-•¤»1r<»<>·4¤•-<•·•¤»•»1>4•·••-•r4•·••-••1r·«»1»4•-«MZg
5: Q I IMQ •-•: E9- 5 'E5 I 59. I .**3I OI glI. N :Q
·ge I *><I IO I IO>_,I P~
··<‘
I Anz: g I§·_ I .II.: fZEx*~·•tI 9*3@= I 5*ä; I r¤= E 5‘^{I no
W· I INZ: I v
I ImIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIN4 I I I t • I x I I
°‘^
9 **3 9 **3 ‘ 9 **3 9~¤ ·· 1 ·= 5 9 1 1 1
S-I •
I-I U EO Ad E
U} 4; $5 II-I nuO lr-:§_-I TI I-« 3 s.I :52 SIE E gg cu ut:
... CI-«$-I *;&· :*:I—•e-I OI—«I—Ic11¤) E1)E
C.‘Q)··-IAd ;)Q)··-I-·-1:.1-I..Ia-I
U1 f‘3—I<¤ I-I Q)>uuI-I3: LJ¤¤ E-I3aaD
I-I ¤)\ ·-I I—«OI«:¤:Im<uI-II-I InZ/J Q)-IPIS-• In ;)$-IQ).Z.'.’1‘.·-·IS-IL)IT.’F«..IGa-J6)_ ··-•·I- Q) Oi-I QJJIQ- I4-I4-I ···-I¤E.C.‘4-II—« :I¤.x4-1.:dr-·<O1)>»U I1)·-I'1)U1-·...:O<I):¤·—~<C I/J¥-*C'··IO··¢4-J•—•3€'I}-I>¢I-«:«·¤—~UL)U’}·-<CL}2
L1Z¤¤3LJ%-·<12·-I%·%•%·$¥-LJ«'EE<!.‘<:.—7
Z• • • • • • • • • • • • • • • • • • • •
I—·* «—~Nf"*<'I¤.OI\305J'~<!1€$I;J2£:-lk-;$Z.‘-·1>&
I I I I I I I II'IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII:IIIIIIIIIIu‘I
i I"‘
—{ I1; I ;Q: I g:I I {G
es I {
I I g
5*. I .„ ·.—I•-I I g,-1wt I IIl I I“£*. I '.I-II I g gGl I I
I 0 I Io
£; I .‘.:I I II I I
"I I I
<I I : I>-I I i5: In I InN. I 5 «-4 ..MI I IQ
Si‘“ I $
f¥ II I‘-
"‘ II * I IE£•-1•-•¤•~••-«»«>•>«r—••-•»1¤•-••-••-••-•»••~•·<>·e»··I•-«»••-<r·<v-«•-II¤»«¤>-«»-<>~e>-<¤»-•><»·«¤•·1»·<>-¢»••<•·4>••·<•-<>-•I»«•-•r••••·•
Q E M äl• I
•
•-I III ’I
°. I ."!I Q
I = I·I
-Q : N ;5I I ·Hl ~aI {Q
I,. I r~ ..‘$ I *7II I IäI E
‘ ‘I
J I rwE2 I $7ä: I {ul I I
*5} I 2I :
8-Q
EI I IN
ZI I I $'Il. : I,gz I I
g I I
=I Im
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIINI I I I I I I I I I
Q Q Q O Q
JI « Ä o 5 c •-Ä •-• AI
181
Le •I-I O EC x E
T Ä Le La GQ,‘- O G '— ···,„ ':· ß I- 2 I-I "5
”' gg H ZI <¤ a.•·::"‘E- CL-ILe CSC„1I•-II-IGTUCJ ··-I1)
;" gß"" J C11)-·—•·«-·IC‘..I-I...}I.IJ} ;4‘I*'¢ I-I CJ>uI-II.«;;: ,.;
Q 4.«—I3 lb ':·»-«.:IOI-e cv5;>„«;;~
‘—· _¤\ —~ IeGI.«:¤::I~;q;;.I„.I I,J3 lI·—*I—·=- I-3 ;;I—«:».::·-•s.«;;·:.·:..·.:.»«;,
·j·-· CD CI- 11.:; ¤-In --435;;,;·-I -:-I'5_„:E•-I;E···’.}1J>~•;;£I.)~»-•';I;r;·--;f_;;;m-,-egt JII:I...-—~..·--I—«3:;·-•><:I.I:--•;;;;g...«·,:·_
_T'J,\«3?'JC'J_':f_7JZ„:—f'J’?)1)<¤·-«Q:3:g;;y)„--ggg J.--¢¤3.«r<;-:-r·%·2¥-LJZZ<<!‘.Z=-TZ
• • • • • • • • • • • • « • • • • • • ,"‘I
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII ev
§ : E(I I [
:1 I '—-I I
gg
G- II.
.nII
[N
>II I
•I I[
II A [
ms I {
II [
2: Ä},,„
Q· I..
*'<I X I Id
II
[
EI G I I
WII I
I·I
„
>-<II I
GI( I
II I
II IC
I/I·I
„.
>I I¤•-«
II [
II [
‘*I I I
(I :4 I
gl ·ä
I I IIJIN. I
I.
~—<t In I QOGI
I
CI O I [
XI IH 1I
~§ : „*vhll-1)-I>·ll•1P·4|¤(h|b4!4I-II¤Ih|)•I!-|!·|I-II-|b(b|!-1l-I>·(h(I-Ild>-4IUIHKHLJ>-4>-<>-(><|-<!·4>·4!·4>1>(>4l-|I-1>-4I-I>IP|)·4>-l!-1I~II-•gg
II Q
gt I u. gl
2* ·-hl I
.
II * IN
IQ :
Ih
gI tcI·
« N T•
gl UI I
•-et I tc
b' I In E-S‘§ ‘” I'II
:; I‘ ‘
sz I *ut
I III
gtI IA
[,-4
II II
-4II I
Ei I *.
>-[I II/Ix I t
II IG
MalI IN
>• I II•-I I I
GII I
GII I
II I
II I
'. Z*,“‘?
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIN
I I I I I I I I I I
G III Q Ie'1 Q Ih Q Ih Q
“‘" " °
‘? ‘?T T
'I‘
I-II •$-4 U GO x E
Q 4-4 $-4 $-4Q :;.;.14-I .
2 U U aa cu ug@0 LI cz. cpu cc
EI ¤I·I-I oppowv ~I—¤.I3- SU--4 ..:4 QJQJ--4·I-4¤.u.;u¢/> <¤-ru p 1I>+.I4.Ip3: I.I5 E-3 1I ·.:--4.::up r¤>„<cI-I I-4 _U\ ·-‘
I··4OI-IT-EIJ¢¤C1)•-II-I $-4U} 3)-*;-*%
$-4·I-•~-4GJ O$« 1).:1 *4-44-I ·•-•:E,:a-I-7 .-’Z‘¢¤.:d&-I.Zr—IOU>»UC‘J-—41);/;-I,:;'}:},)':,.(}·-4<1 0:4-Ic·¤0~-44.II-43c¤-><pc·-I<J;;vI—-IcQJ _¢'J U CT! G3 ,Q _T'J U :-4 CU _CU _’„U _€¤ --I O TU 413 U 11 -4 zu
2 .Jx:¤3Q:·•<.‘=-I-Ir•rII—-.;ILJEZ<<:::.*wZ
• 4 •
-
• • • I • • • • • • • • • I • •
-4 —4:'\1¢^\7'II^~CY‘®3*¢¤LJC.I;1£n;DI-1:A
{¤xx=:tttx=:::::=•I:¤::=:::::g:¤:¤:}¤¤xtzzsxxxzgtxnxxätzxxnn{ I 2**
-2 I {-J' I I
dl I 2S} I {*-—! I ID
1· I I.
IA! I [N>• I :'{ I I- 2an I :
{ I {I :Inä? I ..
F{ I‘ {”‘
Z: ¤ I gI
g'
Q: I 4 :1 I 2
„I'. I'I.?
>2 I {"-2 I 2-l· I 'S ·4: I c {'§· · Ix 4 IIIN-
I I.
-4: In I tcIx: I :0: oI ::::_ ; Ih {
{ I *~z I II,S: II- : Igz I I
-
gpq
•
II = I°. I .*9
I I II. I ,4, .ä‘
E I:2
QI I?
E; 4¤"° I g?I
8 I
gi E’ I
„: I II‘ I III"‘ 2‘.;
g: :{I
ä? : 2C2 I 2ZI E I2‘$‘{ I {TI-4: :
=EHI I .“{ I {{ I {IIIJlllillflllIIXIIIIll!l'lIllll;i¥lll(IIII¥lll¥$l$IIllIlI!Il-H;
: { : : : • : : : I
Q In Q In Q In Q III QI4] IIS A Q Q Q Ä -4
~
183
L.•
In C) Cc 2 EI-I II-I In I-I..I
nn’
4-J CQC LS
:1 ;I aa 1; rc ’.IJ"3*.2-C II-I L :I-Is-I :I:
EI <'JH·-I CI-II-IOQJZU --*22L! :11-- At —1I:J·-•-1::..;.IZuZÜ €7}·—ICU L4 §)>$-II•-IL.3: IJ=¤ 2-3 1I :~I-I.:;Is-I :¤>«<:3 H U\
·—‘ I-IOI-INSGQJI-I«—· InC/3 U-IE-I}-I Z-I C.JI¤·1J.L."2·-L-ILJS'„1.:'>I·-IO·—<·-I 1) O$-· 1).-SC- ‘4—«é·J ··-I3E.C4-II-7 .:5¤.¥I-I.:£·—IG1)>¤<J2)·-I1JI—4:.OQJ¤·C··-I< :!J•-I:·--•O···~I-I·—·3¤·—·><&-•:·-•£;;Ju1·—IELJ =I:;I:¤<¤Q_¢1I:~—¤¤_¤.I_:I_r¤;-·::Ir:;;v1-I=¤
E ....................
I-4liIIglltllzlllllzlllllztltlI?IIIIIilllItglltitgltttlgllltüiII I I
.I· I ~dl I Iut I Int I IE'. I TC?ux I IN>I In I I.„I I I
'. I‘.
2I I I{ I {
El 0 #1 n I·InExIE:
I Iu.I I I
Ei I II I II I IQ
>I I I-I I II I I
QT I T<I I I*{ I: I " Qäs I .wnt I IQ
ää I N 'ZI I{
M; I éß·H•-••·4¤•-IHHI-••·<I-••-••-¢»-•=>«»·|•·4r·d•·-4»<>•><I·<><»1><I-<I><»·<>-INI-Ib-•I-er-er-In<>•>•>-<>··1»•»-••-Iv··••-•t·<•·•»•><I·4•-<>• •éI I
HiI2%I .
su Q IEf II .I ."!
u. I Q
I I I'. 9* I Z •
g= I In: : rn IQIIsI ~ II
I,... I-“= I e }S; I nI-Ii I N %"!
Si I {T.4I I I
Ef I TnI I {Gi ' !•
I *.9E;.I IN
>; I II,..3 I I
E'. I TDI I I: I I•
I PI?IIIIIIIIIIIIIIII-IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIN
3 I I I I I I I I I
¤ In Q In Q In Q In Q·¢ -*
-‘ II I? II I II I·
I-I •
E C1) EQ ,& E
T tf 55 I-IZ!)F
5:1 I: 2. :2-E gg
—· ‘" "·—I·I—•..1-J
‘
QI-I—«¢.—•3 11 C~-•.:.'u2—• m>„<:;V1 1- _11\ —-I =—I0I-I:c:¤¤a.II.I—I 2.2
Q1I—~r$-I 5-I JI-I'1J.T.2-—II-I;J:::;*,_;:I..Ig;J ‘•:'*I 11 CP 1).:.:1.
-:-¤——‘<*-*—%—•Ol1>~<1:1·-I:J:n—I.:o4I:.o~-•Q J21-1IZ··-I:>--II.I-I3:q,-I>;;;_„;;.„,;)£J0-}_,_•;:·_J1 _I§11I'JI5_O_€‘JU;—•<'J_Q_:U:\;-—Q5~gr\;;;y;„.Igg;
E QZ23Jr·<‘.-=Jrr·•5-C=I-1.JZZ<<.‘<2‘1Z=*:
IIII:lIIII=IIlIl=IIIII=IIIII;IIIII=IIIIl=IlIII;lIIII=IIIIII'1
: : 2*I
Z.? 2 Q.Ix: I
•uu I
•-II I IE‘2 I *.9III: I nwZ} I {
N: E
·A
• z {I I I
Ex, :nm
EE 2 E ° ·~ VIII: I ;E; " I
‘° ‘an I {
{_ {6gi Q -I :,-Ä
n I {Z'. I 2
=E E E§.2 : · 22vet I iczu I x2{ I {..- < { I
..Q : Q·I«•«¤>-an¤•-••-••-«¤I-«•-«•-••-«—aI-••-•¤•-••-•»«»«•-«»•I-«I-«I-•E»-II-II-•>-•¤I-••«¤>-II-«I-•¤•-«I-«»·4•-«•-••-•»-I»«•-••-«I-•I-•»«»••« -9
Q. I .EQ 2 „ Q9. I '
."!
ä“ I EI
; — I » ° 'Is · 2 I E.2; Q '° {.2äi I I
I iE{ ° I {9*. I
‘E“!
Sg“‘ g„ Q2
§·. : 2I-In I x
252 I {6,* I *.9Int I IN>x I :I614 I {zu I xSi I i• I s
x I•I
'E I **9In x • • x x x t n I
o UI 6 Ih c Ih o Ih 622 .2 .2 6 6 6 .; .2 6
I I I I I
185
$-4 •
$-4 GJ CI0 ac E
_ I-} N N $-44-JJ2 Q CJ O L:-C
L4 U} L-4 3 4.4 :13CJ GJ Q) Q 1J"U
Q :0 L-4 ':.. ::4-44-I :::_ Q4—4#-4 OHHCUQ) -I-ICDr· C‘.1J··-• ..>A :4.4c;4··-•~-4;:.4.4,44.4441 :$—I¤J 4-4 :J>4-44-44-Q: 4.4V3 Z-I3 GJ I:-·-·:;;4-4 «¤>„4Q L-4 _1J\ I-4 L-4434-4¤¤:¤:¤Q;4.4I—I 4.4:4 14-4:-·L-4 L- UL-4J.4’:¤-·-4:-ICJC-1:.::4-40U1
~-•·I·— 1; QL- :;,::4... 44-444 ·-4:$E;4J_ .:Q.;c·-4;¢··-31J>¤LJ¢.)~—1lJUJ·-I,:.'OQ.)3D•—4-4 W•-4C·-·3·¤-4-J4—43QI-<><L-C·-·•CJQJ¤fJ·I-•'§«‘
_; • • • • • • « • • • • • • • • • • • • •
4 „-•:\Ir*'v<I•·m„¤I\coc~<::¤lJ:‘.::J::.;::..—;;;
I I I I I I : I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIA1
=
QI~
-: : Q(I 4 IUI 4 I·-—I 4 I·-I 4 IO
444I 4 IN>I I I—-I I I
*. I '.NI : I
{ I {Ef z I *."I•-I ¢· 4 Id$‘!I
'I
ZI ~4 I
sul I-5* : „ :Q: 4 0 I
I 4 IGgg. I ..>g I -4 la
{ I {‘I 4 I
(I 4 I•—IN :
I
§* ‘ : *-1-«{ 4 IGII 4 IGI 4 0 III Q I,
4 'I I”'{ I {¤;•·1¤•-4¤¤•-•>-4•-4•-•r<•·I•-••-•»••-4•-4•«•-•n·«¤·•••>-4>·1•-••-1•1><:an•·1•·••-••-er-er-er-41-1•-•¤¤>·1¤•-4•·«•«r·••1•-•>-1•··I•-••-«•-4>•ä:
= · §·Q. ” Q .pg 4*§ ‘ I°. I .9
4 u. G4
•
E · I. ° I .
..* I I
6: : Q»«I I U IO;.. I --;I Q 4A IT
6* ·‘· Q
I ¤p«· g 4 4III
•
6: : Q°{ „ ,L ."f
S: : :1*"{ I {EI I
I.•.I I I"’{
I {¤
EI. I I6;>I I I4L-I I IEI I Iu4· I ~QI 4 I
E 4 Q„„IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIGI
I I I I I I I I I I
G O nn G en Q G
ev -• ·¥ 6 ci 6 Ä ··· ev
186
H$1-e •U C
0 g E*" $-* $-4 $-11.1
V7 m L1.:Q
3) N wgI-1
;~___M
.¤*-·•-^:>1-wäasäZ':¢$"" J DSJ--•·1-1 ::1.1.1 *1.1
1-11)—· -·C-·:0:<:;1L1.1<c1_,
-1 i'J·-•r·L« L-.CI:11:.;—
Jl•-•’.Z····3·-•s.11-3:',-;4j;,.·_;·.„.1L);;y·l,,_„-
g §ää,;!_'-E;5‘ä;j¢g_11_c.1_:·:;-1¤§<¤uv1—1EIn..-4.J
• • • • • • • • • • • • , ,_
Q Q I I=
I I II·IIl IIIII IIIIIIIIIIIIIIIII·IIIIIIIIIIIIIIIIIIIIIIIIIIIIHI
E’ ‘“1
- ·
{
ue1
g
Z:I
*o
$6:2:1‘“
:1
{
"{ ° I {
gg. E *.111:
eu 1 2::-•
E:1 =
E- I '.-1:
1g
Q: znn
=
Q'-I *.9
_E E ·=· E"
::· „. ‘ E” °·
:1
ä: 1 Ä.N-
1..
-4:1 Q IQ
ää I {z: 1
n
rl•-• 1E
’ ä·•-•r·••·e•«•-1-1-·••-•>•>-er-nv-•>+•-«•·<•-1•·«•-••·4»1»•1-1•-•»•»••-••-1E»••-•-••-1-eu-••-4>-«>-••-••-••-•>-••-1•-••«•-•»-••-••-••-•»·<•-•re•·n•-•|
7gi =§‘
• 1.
1 e ;-4
19, I
."!
§ I '?1 „ ¥
gi · 1 ¢-•=
A O 0 : uio
:' 1 Ih {-Ä
si‘ 1·
:1 E ·: Q 1
7
E: IM | I
°-I
*.“E
äi E —= E7ET I
{
:{ I {1/1:
1 :
Q:. :IQ
1.11:1
zu
Z.{ I {'sx
1 :
ät. 1:~
{ I {:
1nm
:::::1:::t:•:1¤:::::::::::::::;:x::::::::1•::::::•t:::::::::'¤1:
• t• I : : : s 1
Q un Q 1h Q uh Q 1n Q
6 .-2 -2 6 6 6 -1 . 111
187
Lg •
I- 1) I':C .;c E4.: I- I- I-LI
m<‘¤ O C Cr-II
I., gr; I-I 3 M TUQ; Q, C.: CU 1:1
3 :5 L. L ¤I-LI ::1:2..... CI-II-IGCJZJ --:1:
L- X;—;....;·; L. ;,I>..:4.::-G: 4.:
y} Z.-3 Q C-:-.’:U$-I YU>~€
Q;_, ;,\ -—- L. O I- C-D3 PI'! 3)MI—I I-I
L. L. 1,; L- 1J.I2·-I-ILQI ILM 0;;; ........ ; QL. ;,.:L #-4-: ·—<Z‘IEL·-I
;::.;-·;:-';b>I:41;·~UUI-I.:Q31:$¤·_;<_; I_,:.... :....,_I-3:'_;—•><jII-£.'··I;J.JCf}·-·—( T; DISC ?'Z1.·-<IJ l)$··•O¤€UUUi·:-T'};; :.:;:.~:>3;+<:=.-&«I—r-¤-:.>.i‘Z<:‘.<.*1-=-<W_; • • • • • • • • • • • • • • • • • • • •
4I
I { I I I I IIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIUIIIIIIIIIIIIIIIIIIIII INS I :*8ät. I !
ul I
=Z;I§•I {N
*: :·N; 0 I ‘.
I I :·-{ I z ‘IIIE; I '..Q: IN =—
I I‘;5'*.I *.I-AI IQ: I
=gl I :QIU I o Ph
{AI ( I [‘I
I [äf I °b~I I {S: r~ I im2*.; I "' "SI I Q =¤QU we I 3:U
I 3‘x I
‘Q: I S~S I i'Q I
~ I IE! ' '•-IIIGII In
ä I In g?· I •
6. · I S•-If· I Ia 0 to
<t: ITI Ihäi · Spg. I Iu.I Q I
[ I6· · SUI Z : Ill!gl I Il-!I IH IIAI I IEf I
‘.D-nl I [>—I I II/It I IU I UQ[MI I IN>I I II¤<I I IZI I IGI I II I lI I II
I IlhIIIIIIIIIIIIIlIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIN
I I I I I I I I I IQ Ih Q Ih Q Ih Q Ih QN rd •·•
Q O O rl rn N
APPENDIX I
SUBJECT WEIGHTS MEASURING THE IMPORTANCE OF EACH DIMENSION
189
SUBJECT NEIGHTS
DIMENSIONSUBJECT PLOT HEIRO— 1 2 3 4
NUPBER SYFBOL NESS
1 1 0.1877 0.3716 0.2503 0.4464 0.32822 2 0.1767 0.4793 0.3403 0.5799 0.40913 3 0.2084 0.5110 0.2687 0.4130 0.28144 4 0.1988 0.4216 0.3176 0.5415 0.47325 5 0.0650 0.3590 0.2926 0.3526 0.30146 6 0.2590 0.3282 0.6232 0.3559 0.44447 7 0.1156 0.4973 0.4876 0.4996 0.32298 8 0.0632 0.4052 0.3391 0.3866 0.36569 9 0.0665 0.4579 0.5095 0.4763 0.4559
10 A 0.1127 0.3443 0.4140 0.4482 0.367511 8 0.0891 0.5161 0.3954 0.4328 0.339412 C 0.0684 0.4040 0.3602 0.4138 0.386013 0 0.1459 0.3981 0.5373 0.3883 0.362014 E 0.0911 0.3813 0.3442 0.4297 0.370915 F 0.1014 0.3868 0.3792 0.4717 0.351916 G 0.1770 0.4817 0.3720 0.5530 0.311017 H 0.1963 0.5765 0.2922 0.4108 0.385118 I 0.2055 0.5252 0.3363 0.3041 0.266519 J 0.0706 0.5075
A0.5353 0.4345 0.4282
20 K 0.1228 0.3102 0.3693 0.3409 0.372521 L 0.1476 0.3666 0.4057 0.4518 0.471322 M 0.1321 0.3653 0.4232 0.3624 0.434523 N 0.1662 0.5047 0.3095 0.3410 0.292924 0 0.0913 0.4850 0.4662 0.3884 0.458125 P 0.3411 0.6803 0.4484 0.2086 0.321226 G 0.2145 0.3622 0.5614 0.3255 0.411727 R 0.0993 0.4240 0.3477 0.4616 0.385328 S 0.0415 0.4560 0.4028 0.3859 0.386629 T 0.1764 0.4733 0.3667 0.5372 0.299630 U 0.3011 0.6885 0.2493 0.3911 0.397131 V 0.1761 0.4577 0.5165 0.5772 0.312632 N 0.2338 0.4153 0.5937 0.3045 0.471933 X 0.0495 0.3516 0.2853 0.2977 0.279034 Y 0.0736 0.4036 0.4459 0.4110 0.329535 Z 0.3258 0.5980 0.2989 0.5917 0.226736 1 0.1555 0.4197 0.5226 0.3548 0.459437 2 0.3107 0.6402 0.2707 0.3869 0.257238 3 0.1628 0.5400 0.5247 0.4082 0.299739 4 0.1299 0.4600 0.4352 0.5916 0.422840 5 0.1838 0.2539 0.4160 0.3183 0.296241 6 0.1551 0.4747 0.2897 0.4534 0.374242 7 0.3024 0.2757 0.4437 0.4521 0.610443 8 0.0401 0.4248 0.3873 0.4219 0.383144 9 0.0747 0.4312 0.4022 0.4753 0.4150
19 O
45 A 0.1782 0.5352 0.5772 0.4462 0.300646 8 0.1602 0.4102 0.5874 0.4174 0.398947 C 0.0554 0.4109 0.4046 0.4315 0.399148 0 0.1958 0.3821 0.5872 0.3810 0.468449 E 0.1477 0.4894 0.3184 0.3152 0.336850 F 0.1401 0.2927 0.3600 0.4093 0.353351 G 0.0771 0.4600 0.4883 0.4449 0.348852 H 0.1709 0.2972 0.4013 0.4707 0.365253 •L 0.2180 0.2888 0.4214 0.5397 0.399954 J 0.1429 0.5488 0.3921 0.3458 0.347255 K 0.1286 0.3283 0.2467 0.3548 0.248356 L 0.0477 0.3732 0.3967 0.3908 0.343657 M 0.2458 0.3585 0.6090 0.3297 0.401258 N 0.0568 0.3753 0.2984 0.3115 0.283459 0 0.2372 0.6464 0.3264 0.3497 0.384560 P 0.1741 0.4547 0.3610 0.3920 0.512061 O 0.1320 0.3544 0.3958 0.4744 0.322262 R 0.1376 0.5456 0.3360 0.4224 0.403663 S 0.1363 0.4267 0.5433 0.3785 0.398064 T 0.2200 0.3447 0.5278 0.3917 0.532665 U 0.3848 0.3123 0.6753 0.3703 0.216466 V 0.1362 0.6053 0.5053 0.3759 0.413167 H 0.1234 0.4015 0.3146 0.4495 0.371268 X 0.0835 0.4047 0.3487 0.4109 0.394369 Y 0.1412 0.4224 0.4791 0.3949 0.280070 Z 0.2301 0.3742 0.5006 0.3934 0.220071 1 0.1769 0.5584 0.3769 0.5852 0.351472 2 0.2601 0.5396 0.2482 0.4987 0.294673 3 0.2185 0.4838 0.2596 0.4391 0.469974 4 0.2610 0.6445 0.4116 0.3239 0.279575 5 0.1907 0.3656 0.5001 0.5222 0.303976 6 0.2161 0.5706 0.3570 0.5147 0.273377 7 0.1007 0.4381 0,4982 0.3744 0.401078 8 0.2347 0.5762 0.5555 0.2990 0.324979 9 0.1368 0.3263 0.3593 0.3972 0.406780 A 0.1203 0.3529 0.3113 0.4206 0.290381 8 0.1570 0.3999 0.4953 0.3376 0.304782 C 0.1755 0.3233 0.4089 0.5061 0.425383 0 0.2143 0.4764 0.4907 0.4694 0.226884 E 0.0793 0.3986 0.4036 0.3612 0.285585 F 0.0815 0.4393 0.3881 0.4683 0.339786 G 0.2275 0.3375 0.3997 0.6040 0.471487 H 0.1438 0.5641 0.3568 0.3831 0.390688 I 0.1077 0.4912 0.5724 0.4671 0.382889 J 0.0919 0.4846 0.4256 0.3690 0.320990 K 0.1834 0.4693 0.5790 0.4722 0.296591 L 0.0443 0.4252 0.4192 0.4147 0.404192 H 0.1108 0.2822 0.2479 0.3338 0.249593 N 0.1957 0.5362 0.6149 0.3551 0.368794 O 0.1838 0.3556 0.2277 0.4017 0.321095 P 0.2651 0.5085 0.3602 0.3348 0.597596 0 0.1249 0.4486 0.5536 0.4182 0.369597 R 0.1506 0.4212 0.3164 0.4056 0.445198 S 0.0561 0.3602 0.3180 0.2846 0.288099 T 0.1028 0.2579 0.2907 0.3085 0.2901
100 U 0.1686 0.5113 0.4438 0.5328 0.2851101 V 0.2529 0.6176 0.4376 0.3505 0.2478102 N 0.2161 0.3484 0.6057 0.4131 0.3807103 X 0.1977 0.3758 0.2730 0.4898 0.3351104 Y 0.3390 0.3729 0.5947 0.2513 0.6050
191
105 Z 0.2039 0.5890 0.2993 0.4487 0.3460
106 1 0.2073 0.5818 0.3080 0.4768 0.3218
107 2 0.0807 0.4050 0.4656 0.4481 0.4222
108 3 0.1605 0.4391 0.2964 0.4214 0.4358
109 4 0.0604 0.4335 0.4025 0.3486 0.3688
110 5 0.1951 0.5125 0.2747 0.3451 0.3001
111 6 0.2411 0.5718 0.3645 0.5343 0.2503
112’7°
0.1339 0.5787 0.4204 0.3681 0.3875
113 8 0.3165 0.3665 0.2782 0.3769 0.5817
114 9 0.0958 0.5382 0.4335 0.3994 0.3506
115 A 0.0808 0.3089 0.3327 0.3614 0.2780
116 8 0.0562 0.4701 0.3824 0.4262 0.3483
117 C 0.3582 0.5287 0.6520 0.1752 0.3788
118 D 0.2185 0.4663 0.5594 0.2830 0.4564
119 E 0.1593 0.5862 0.3428 0.4151 0.3868
120 F 0.1195 0.3496 0.4428 0.3372 0.3587
121 G 0.1547 0.6023 0.3570 0.4261 0.4156
122 H 0.1405 0.3608 0.2292 0.3196 0.3056
123 I 0.1425 0.5652 0.3426 0.4566 0.3872
124 J 0.1356 0.3992 0.2774 0.3680 0.2495
125 K 0.2010 0.3523 0.5290 0.6186 0.4441
126 L 0.1213 0.4789 0.3617 0.4400 0.4670
127 M 0.0831 0.2960 0.3097 0.3518 0.2766
128 N 0.0870 0.4610 0.3760 0.3650 0.2998
129 0 0.0255 0.3989 0.3799 0.3657 0.3162
130 P 0.1373 0.3541 0.4075 0.4574 0.2932
131 Q 0.1547 0.3186 0.4571 0.3868 0.2984
132 R 0.2276 0.5291 0.2990 0.5036 0.2773
133 S 0.1917 0.3465 0.3544 0.3822 0.4788
134 T 0.1200 0.4960 0.3520 0.3425 0.3257
135 U 0.0510 0.4845 0.4152 0.4355 0.4305
136 V 0.2446 0.3334 0.5552 0.4098 0.2663
137 H 0.0966 0.5174 0.3965 0.3884 0.3377
138 X 0.0863 0.3880 0.4155 0.4643 0.3573
139 Y 0.1075 0.4323 0.4108 0.3866 0.4574
140 Z 0.2610 0.3875 0.5695 0.2932 0.2876
141 1 0.1110 0.3828 0.4950 0.4375 0.3684
142 2 0.1135 0.4015 0.3824 0.4870 0.3348
143 3 0.2186 0.3757 0.3724 0.4863 0.5536
144 4 0.0404 0.3834 0.3896 0.3571 0.3488
145 5 0.2484 0.3779 0.3848 0.5611 0.2477
146 6 0.0318 0.5046 0.4971 0.4647 0.4508
147 7 0.1422 0.4758 0.3207 0.4885 0.3614
148 8 0.1812 0.5162 0.5894 0.3677 0.3470
149 9 0.1996 0.5228 0.5661 0.3238 0.3420
150 A 0.0577 0.4518 0.3902 0.4014 0.3191
151 8 0.0765 0.4376 0.4713 0.3571 0.3996
152 C 0.2754 0.2588 0.6000 0.4904 0.5128
153 D 0.1800 0.5712 0.4107 0.4094 0.2769
154 E 0.3631 0.6762 0.2393 0.3584 0.2570
155 F 0.1291 0.2992 0.3059 0.3232 0.3573
156 G 0.1780 0.4150 0.5448 0.3840 0.3032
157 H 0.1437 0.4561 0.3961 0.4056 0.5055
158 I 0.1367 0.4816 0.5139 0.4974 0.3110
159 J 0.2918 0.3299 0.4534 0.4285 0.6457
160 K 0.2590 0.5080 0.4211 0.2960 0.5854
161 L 0.1396 0.4949 0.5997 0.4149 0.3981
162 H 0.0529 0.4468 0.4671 0.4020 0.3842
163 N 0.2357 0.5953 0.3034 0.3828 0.2915
164 0 0.0529 0.5143 0.4123 0.4510 0.4163
192
165 P 0.1144 0.4298 0.5505 0.4324 0.4280166 Q 0.2459 0.6507 0.4423 0.3294 0.2984167 R 0.1341 0.5775 0.4012 0.3877 0.3621168 S 0.2454 0.5374 0.3215 0.3046 0.5022169 T 0.2875 0.4582 0.6791 0.3496 0.2962170 U 0.0880 0.3812 0.3698 0.3931 0.4037171 V 0.1004 0.3530 0.3728 0.4406 0.3378172 N 0.0903 0.5112 0.4085 0.3688 0.4003173 .X 0.1944 0.5746 0.3033 0.4423 0.3381174 Y 0.1522 0.3955 0.5840 0.4623 0.4411175 Z 0.1702 0.6052 0.3656 0.3896 0.3661176 1 0.0766 0.4432 0.5118 0.4353 0.4025177 2 0.1019 0.4742 0.3354 0.4374 0.3832178 3 0.2082 0.3014 0.5008 0.4514 0.4814179 4 0.0619 0.4507 0.3833 0.4405 0.3387180 5 0.2993 0.3492 0.4169 0.3862 0.6386181 6 0.1168 0.5406 0.3622 0.4225 0.4267182 7 0.1205 0.5695 0.3728 0.4439 0.3884183 8 0.1950 0.6392 0.3886 0.4265 0.3221184 9 0.2486 0.4525 0.2164 0.3216 0.2174185 A 0.0429 0.4421 0.3826 0.3908 0.3261186 8 0.1276 0.5111 0.5031 0.3806 0.5019187 C 0.1135 0.3180 0.3004 0.3943 0.2980188 D 0.0352 0.4798 0.4065 0.4312 0.3728189 E 0.2006 0.4111 0.6391 0.4341 0.3744190 F 0.1219 0.4489 0.3378 0.4874 0.3774191 G 0.0876 0.3520 0.3764 0.4024 0.2914192 H 0.1255 0.3660 0.2625 0.3713 0.2651193 I 0.0762 0.3607 0.3070 0.3761 0.3301194 J 0.1334 0.5184 0.3270 0.4356 0.3525195 K 0.2682 0.3358 0.4364 0.3108 0.5424196 L 0.3327 0.6474 0.3015 0.4742 0.2066197 M 0.3530 0.4135 0.6431 0.1977 0.5257198 N 0.1529 0.4547 0.5353 0.3404 0.3722199 O 0.0314 0.4928 0.4408 0.4221 0.3822200 P 0.0864 0.4397 0.3521 0.3573 0.3907201 G 0.1343 0.4498 0.5753 0.4091 0.4547202 R 0.1117 0.4535 0.3307 0.4590 0.3546203 S 0.0907 0.4357 0.3622 0.4307 0.4226204 T 0.1852 0.4471 0.5814 0.3404 0.3868205 U 0.2108 0.3979 0.5904 0.3364 0.3874206 V 0.1339 0.3900 0.4814 0.4761 0.4903207 H 0.0565 0.5092 0.4100 0.4716 0.4199208 X 0.2995 0.3187 0.6860 0.4499 0.3306209 Y 0.1348 0.3822 0.3060 0.3948 0.4053210 Z 0.2814 0.4269 0.4508 0.2897 0.5973211 1 0.0390 0.4472 0.4385 0.4658 0.3886212 2 0.1456 0.4934 0.4021 0.4923 0.2898213 3 0.0468 0.3843 0.3842 0.3345 0.3152214 4 0.2767 0.2937 0.5805 0.3645 0.2982215 5 0.1348 0.4957 0.3278 0.4603 0.4320216 6 0.0963 0.3716 0.3827 0.4543 0.3754217 7 0.1569 0.3779 0.4266 0.5321 0.3351218 8 0.1096 0.4281 0.3065 0.4023 0.3743219 9 0.0412 0.4511 0.4517 0.4427 0.4254220 A 0.0660 0.5168 0.4015 0.4189 0.3992221 B 0.1875 0.3865 0.4935 0.3403 0.4886222 C 0.2888 0.2965 0.5602 0.3494 0.5497223 0 0.2552 0.2644 0.3052 0.4906 0.4132224 E 0.1001 0.4143 0.4919 0.4283 0.4514
193
225 F 0.0740 0.4214 0.3339 0.3922 0.5040
226 G 0.1092 0.5869 0.4996 0.4525 0.4097
227 H 0.1471 0.5775 0.3405 0.4962 0.5661
228 I 0.1777 0.5459 0.5231 0.4287 0.2802
229 J 0.2484 0.6116 0.5059 0.3547 0.2549
250 K 0.5085 0.2806 0.5489 0.5941 0.6098
251 L 0.1960 0.3589 0.2843 0.4926 0.5542
252 M- 0.5562 0.6510 0.2871 0.5758 0.2105
255 N 0.1079 0.4084 0.3256 0.4457 0.5514
234 0 0.2098 0.6505 0.3230 0.4143 0.5615
255 P 0.2052 0.6288 0.3999 0.4200 0.2965
256 Q 0.0707 0.4186 0.4175 0.5598 0.5605
257 R 0.2542 0.4402 0.2516 0.5085 0.2676
258 S 0.0568 0.4511 0.4485 0.4947 0.4171
259 T 0.4575 0.6939 0.2568 0.2571 0.2291
240 U 0.0911 0.4066 0.5715 0.4599 0.4001
241 V 0.0668 0.4845 0.4507 0.4745 0.5514
242 N 0.2199 0.4059 0.5022 0.5715 0.4991
245 X 0.1595 0.5551 0.5271 0.4168 0.5684
244 Y 0.0955 0.4554 0.5214 0.4327 0.5777
245 Z 0.5624 0.8398 0.1949 0.2616 0.1956
246 1 0.0982 0.4294 0.5421 0.4581 0.5656
247 2 0.1947 0.2959 0.5006 0.4648 0.4168
248 5 0.1251 0.2844 0.2796 0.5620 0.3054
249 4 0.1021 0.5042 0.5606 0.4654 0.4240
250 5 0.2597 0.5537 0.5995 0.4105 0.5519
251 6 0.1055 0.4868 0.5409 0.3753 0.5765
252 7 0.0975 0.5505 0.2627 0.2815 0.5024
255 8 0.1017 0.5797 0.5621 0.4109 0.2725
254 9 0.5152 0.5015 0.6251 0.5878 0.2694
255 A 0.0409 0.4517 0.3658 0.4089 0.3505
256 B 0.2541 0.4059 0.4972 0.5954 0.6128
257 C 0.2596 0.6815 0.4050 0.5590 0.5299
258 D 0.1247 0.4056 0.5857 0.5005 0.4591
259 E 0.1565 0.4125 0.5589 0.4712 0.2887
260 F 0.5645 0.7046 0.2678 0.5209 0.2776
261 G 0.0761 0.5094 0.4134 0.5908 0.5545
262 H 0.2046 0.3075 0.4894 0.5459 0.4017
265 I 0.1165 0.2697 0.2145 0.2685 0.2711
264 J 0.0981 0.4950 0.5476 0.4404 0.5788
265 K 0.1065 0.2978 0.3407 0.5822 0.5052
266 L 0.1968 0.4044 0.5725 0.4096 0.5056
267 M 0.1452 0.5055 0.4700 0.5471 0.4854
268 N 0.1550 0.5862 0.5876 0.4098 0.5529
269 0 0.1868 0.4282 0.6525 0.4971 0.5574
270 P 0.1255 0.3526 0.2559 0.2599 0.5074
271 Q 0.2442 0.6291 0.5551 0.5061 0.2859
272 R 0.0725 0.4612 0.4807 0.5145 0.5886
275 S 0.1107 0.5854 0.2976 0.4055 0.5559
274 T 0.2407 0.2800 0.5617 0.4223 0.4475
275 U 0.1103 0.4255 0.2849 0.5589 0.2954
276 V 0.5556 0.5505 0.6859 0.5072 0.5467
277 N 0.2448 0.5566 0.2464 0.5500 0.5507
278 X 0.0885 0.5166 0.5506 0.5795 0.2871
279 Y 0.1565 0.3542 0.5315 0.4759 0.5945
280 Z 0.1598 0.5771 0.4420 0.5416 0.5441
281 1 0.1717 0.4739 0.5462 0.5202 0.3996
282 2 0.1475 0.5470 0.5264 0.4407 0.5718
285 5 0.0652 0.3478 0.5048 0.5519 0.2650
284 4 0.1559 0.5546 0.4002 0.4159 0.5151
191+
285 5 0.0939 0.4632 0.4339 0.4330 0.3045286 6 0.0865 0.3652 0.3183 0.4036 0.3335287 7 0.2090 0.3674 0.5429 0.3542 0.4955288 8 0.1070 0.3670 0.4121 0.4395 0.4193289 9 0.2270 0.2386 0.3214 0.4293 0.3773290 A 0.3094 0.2594 0.6574 0.4242 0.3844291 B 0.0434 0.3713 0.3253 0.3587 0.3288292 C 0.1428 0.3370 0.3498 0.3898 0.4197293 .0 0.0828 0.4738 0.4621 0.5481 0.4134294 E 0.1337 0.4504 0.2953 0.4211 0.3170295 F 0.1187 0.4876 0.3834 0.4466 0.2967296 G 0.1263 0.3786 0.3859 0.4536 0.2852297 H 0.1596 0.4137 0.4397 0.5752 0.5137298 I 0.1113 0.3746 0.3354 0.4494 0.3457299 J 0.1114 0.4138 0.5262 0.4632 0.3751300 K 0.1402 0.5374 0.4055 0.3341 0.3466301 L 0.1049 0.4626 0.4871 0.3908 0.4690302 H 0.2009 0.4514 0.6283 0.4736 0.3193303 N 0.1605 0.3241 0.3133 0.4547 0.3455304 O 0.1018 0.4812 0.3688 0.3849 0.4290305 P 0.1118 0.5787 0.3980 0.4298 0.4401
OVERALL IMPORTANCE OF EACH DIMENSION 0.2051 0.1812 0.1742 0.1431
195
APPENDIX J
PLOTS OF INDIVIDUAL SUBJECT WEIGHTS
FOUR-DIMENSIONAL INDSCAL SOLUTION
196
I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIQ
E §"{ {-1 at:14 IU
2: :=·»-III gw· .>* N 1'{ {NI- IO
I: ä°EI I
9{ {wat gEI INOI O
=A U D:
=@· „ O {‘
{ 0 D N I__> A GO
i'g{ ¤ MF ‘“ 2 *°I Q N><X22{„ „"‘ ‘„ ‘
*SI U Z 1 M U x 0 äma N Igi 09* 1I3 0 A \I\ III. N Oälhz „.4 •·• g Nu. 0uJ QA •-• Q2 D¤ Q NN M
*-gßrt
> J zäh- N 0E- I! I- GZ SWE: A R nu U ·>-N 14/1A nn ¢•-« Q A N >0 <-< 0Z ••
>· G 0 xu. Dtw BO') nu I ucsu Q 0
Y- N1 Q 8>N> XMJEUX I# -6se ¤ -• NN unna: vl §Q > N uh 040 ¤0 0 pnI G mv-•4 2· I(MXYNAam
> '> U g{
·· Za U
zäzu
IMV)- 9* U N Z ztn. •-•- 0 >-1 A Qä
•-x N rn 2 N Q•-t 1 U ln Qt-gg er :EI IN*5: IQ
Iä: .GI IWI I>I I;•E1wat IQQ:=I I
g I{ {1 IQIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIOI I I I I I I I I I
U 0 N 0 In 0 M N •·• QO O O O O O O O O
I I I I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IQ
E VI II I
"I I;JI IV
UI IOAI I>-I IRI I
>I N IUI I
I IMI IQ
I IOI I
EI II
I
QI II I
sul III IN
QI D I- IQI U L I
„{ {>- I .4 QG ·
I 0 N Q I_: Q A §JI ä ä 0(· N 0 ··
I N Z IQI -« EI :r~ 0 {I 00 N Q
NI•-• x I I
¤· NQ 2 ·RI N ( XO x VIQ R N U2 "I( QQ D K Il
• N '3 ~~N uu. Q 0 ua Q
A N Q 01I- XON -4 QI R er gun :.4zI-ä• VIE! CIM! !—0· -u. >M (NA•-• o~ AA ¤>· Q NQ R n ·-ID K2 Q äzw 2 I O
• Ngu .IQ 1.•
Y > u u o•-• III N IM 0-40VI¤ C Z¤ In NU 00QN >-.4 SA HII2 KI
• I >~G¤>I IH Mu. X GL•
L9 ( u ac*7me x >
u Z uxÖ; zNMpg•-• u. Q Q
ä;N I-x M I ¤
A Q ( +-. III -¢
I- N3* ..WI *9
I I§‘ 'GI I‘$‘{ {EI
IA
Wg IQQI
I‘ .I I{ {3
IQIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIO
I I I I I I I I I I
0 3 r~ a In Q M N A Q
Q Q Q Q Q Q Q Q Q Q
I I I I I I I I I IIIIIIIIIIIIII IIIIIIIIIII IIIII IIIII IIIIIIIIIIIIIIIIIIIQ
ä i"In xlI I
·I Iästl tvUI IGMx IEl {*2 ~ i”x •'I IQ
I IQI I
ZI IQt I
9{ {wt IZI u~
INQI D b- IO
I UA M1 I{ {
2- Q G t.; ·I Q N I_I O LEN R 0
1· Z N ·•EI L11 NZ IQI B. N! -4 II A I
¤··QIN:
{MglIh ~0 UX! N¤Hzi
B! Q! { 3:D Ih· :^?„ ‘6
rl .1 ON ¤> M ¤E „ .§ =~:·sx5•A GO MZIQE 0 ·M -4> 1 IM N >4 .1 •-eq
9 gz "'25¤ aääz 8: «·g- Q In >g Z NC.JIu *3 ~öMI M Q > ZÜYAM äßlhQ: >·Q A .1 täte M3 12 3: N
· Ih Ih ! >·IL >8¤¤8M ·E -2 ": "„ °«
‘;>. {
X··I U 2: 1 U; 3 Intn- U M N IL O
--•—
M L Q
5; 1 n- nn Nt §M U ¤ < mr-
*2 ‘ 2P-Q N
7: =Ggl IVII I,QI IMAI I>t I
;* *.*3wI :O°E 2
I II II IQ
IVIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIOI I I I I I I I I I
° *2 'T °“! 3T "!
"‘ "‘°Q Q Q Q Q Q Q Q Q Q
1 1 1 1 1 1 1 1 1rl
.4: ¤¢U, 15:: äcz:
*_
g: {, IO"‘,{ IGg IS: =$1 =‘£§ ¤~.4— ··no°{ X :»— > uIg‘. 1zuE<= =
“* . :zg Id
U *6
*" >« 8 Q? E‘ ’“
§S: ¤ .4 0~NN; n- UG wc UU- ac.: ·-« ·g; 5 ua >~Z I•-
M> :.7 W'. un ~~•<.a ¤¤v~ ··z r~ cow .4 S ¤"‘; °‘ > N E8 1:. ¤ w ¢—s>· N L ~5 :: N ••·~ o >¤z
" : §.4°us;° cm:Q 8 o ¤ oz ac.: ¢. 0
•~ cxizxgag ··E uu? < o,., am <O M
° "G., "‘B'“ä§ 2 E. az wx U ¤. ·N ILwwwzzu. IM
V,. N ¤ .4 n4 z:- z ·~,., 5: J 4: z ¤m¤n:>-x SG
ääX mn uzm u. N;N. z •-1 D I2 ·W I-II
N {QS: 5mx .2: ‘ EE! ·.·t..4: {Qcg ,
I‘~¤ {=
:5
I I I I I I I I I
° °'“
° "7 Q *7 °€ "T Q5 5 5 o o c 5 ¤ ¤ Q
I I I I I I I I IIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIID
E Fi £I I
Z} ä,d·
..
UIIQ
-•I g
n-·IgZI g
•u· .>I (—-I I
I IQ
I IOI I
SI II I
9*‘
I IWI I
EI IN
¤ X
=>
•—D
Io
I IgI 4 I
I I nu I__; wa Q I:¤~4t!Z o<· mo 4
-•
t >- Q Q3 Q IoäI G> •- ¢ IN; u
•-U I
¤· X L ¤< g•
an r~ 1 en Q0 P- I >Wz rl >·
>•n•-1
-0
>¤• •·11 an
. Q QM Q NN N N.•
Ju; Z NU h Q 0zu E >·0 NE U
• 0•
'I >-L 3 >~ ä~=§gg lll I 0 JNGQ! LQ QI Q 0~g¤14ä$¤.äJ .6><
ML eq
ci LM
L4§„1·2z \s>U‘-•uQ~§
§
I-m gäz-••-eh
g•·• ·
1; un QNZQ ux II sn fl 8 0 I
”* z m¤> 2;:2 m N 4‘M
Ü'°
za J Ä _, 1: oäi „ « 2—,~ EEi ¤ x
Dxp- 0UI l-1 N
aa]: N IQ
§{‘
WI ;}I
äi :;* V'?wg IQÖ: {
• f{ {g Ig IQI-IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIO
I I I I I I I I I I
0 Q •~ 0 nh rnNÖ
Ö Ö Ö o o Ö o c Ö
I I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIQ
i i"¤' JI I‘I
I=I IFUI IQ»-«I I>-I Ian I>I {_:
I'I IQI :6{ I
Sl IHI. I.
gl Iu4I IEI IIITQ} IG
{ {gl l'
II
J ax Io~7 N IOgg z N M > I I
N: In In« Q (J ·äg Q Z NN >-D ° EInS gS Ih. EQ •-•N
··M U > 'A 2,.Is .4,4 N QQ .I
~a¤n~Dän• v~ nn.4 oä X n„<.I ·E .4 0 näÄ"‘ ”
E22°$‘é
82*2II\ I GIhn I-¤„ .I>- g>- N8< u.;0N N Q
Q ° !'^° ° gf ¤"'¤! =¤ ¤Q Q .°‘ °zäQ "‘§„‘%{ E S; ä I
A GI x rn Zi n um In. 0.. x > · rnI/I· N! GII Z 0 ·öE = ^ — iQ,. A. I- N¥. > g{Sg ¤ ~W.IäII U I
Iä. ;.
IS{ gZ{ ¤_·1Ei •¤ce: Q
I *_
5 {
·
I¤ Q9I.=IIIII:IIIII=IIIII:IIIII:IIIII:IIIII:IIIII:IIIII;IIIII:I¢
' ¤ " ° "! Y "E 'Y 'T 9ci ö o c c Q o c o Q
APPENDIX K
CORRELATIONS FOR DEMOGRAPHIC VARIABLES
203
QQQr-I
Ä
·k·!<0
}
E
Q
3Q QQ Ü
¤\
<¤ v
Q Ih
N·*5i
0 N
¤¤
I-• xa
IN4
aa
3 N··I
:
Q C?c
·«-I
Q
Q"‘ I
Q rs
U\
Io
¤¤ •"
ä V
S $ g 9
„
.Q
B4
O
M
•
Q
N
‘ ‘^
4 cn
O>¢W N!
Q ,.4ln Ih
·•-•0
Q ,.,O
O°U"‘
O
· . '
.
O
ä ä
•—•?
?'
2:Ä:
»·s
Ih
eq <¤ 4
- ‘cn IN
> V
Ä? Ä;2 ==“
9O Ih
Q cO
U
O~0 •-•
„'
•"*
O r-•.
Q O
‘ä.9 6 9
tu^
•-•ÄB Ü
~¤
O4; M 6* g
E‘*=
·¤ Q o °‘.
0
4; ..•U4 O O
O
¤
o :1 Ü° ' 5 I
I6-I
O cn .? I
O
¤ ‘O
I
/'\• QM
|-Ü
: xa
NN'?
O
Mrd""
fiO IN
Q O
IJ
¤— ·—c Q .
·?
N
WO F-•
.O O
•-•
QN
Q.
OI I
9O N
· O I
I-•^ Q
· o
>I
In •-1
- Ou
>sL}
m•-•
Q xa •—•
U¤
QC
oI.,
E ‘; ·•0’U
VVU °
Cu
ß ä6 ·=“:€ Jil'? “,,
E ZQ
LT 3 3;·„£" §*;‘,E‘.:"·»=
Q 3,, °~‘Q
"" V:1.. ·-ev
Q
BN·¤V wm 3 Ä WE ii “:
20l+
APPENDIX L
COMPUTER PROGRAM FOR DETERMINING ACCURACY SCORES
205
DATA ONE;INPUT IDA 1-5 ID 6-9 TAPE1 10 RATEE1 11F1 12 FZ 13 F3 14 F4 15 F5 16 F6 17F7 18 F8 19 F9 20 F10 21 F11 22 F12 23 P1 24 P2 25 P3 26 P4 27 P5 28P6 29 P7 30 P8 31 GLOBAL1 32 IDZ 33-41 TAPEZ 42 RATEE2 43 F13 44 F14 45F15 46 F16 47 F17 48 F18 49 F19 50 F20 51 F21 52 F22 53 F23 54 FZ4 55P9 56 P10 57 P11 58 P12 59 P13 60 P14 61 P15 62 P16 63 GLOBALZ 64 Q2ID3 1-9 TAPE3 10 RATEE3 11 FZ5 12 FZ6 13 FZ7 14 F28 15 F29 16 F30 17F31 18 F32 19 F33 Z0 F34 21 F35 Z2 F36 23 P17 24 P18 25 P19 26 P20 27PZ1 28 PZZ 29 PZ3 30 P24 31 GLOBAL3 32 ID4 33-41 TAPE4 42 RATEE4 43F37 44 F38 45 F39 46 F40 47 F41 48 F42 49 F43 50 F44 51 F45 52 F46 53F47 54 F48 55 P25 56 PZ6 S7 PZ7 58 P28 59 P29 60 P30 61 P31 62 P32 63GLOBAL4 64;t1=5.4; t2=3.4; t3=5.3; t4=2.9; t5=3.S; t6=3.3: t7=3.2; t8=3.2;t9=4.6; t10=4.4; t11=4.3; t1Z=4.7; t13=4.0s t14=4.0; t15=4.2;t16=4.5; t17=4.1; t18=4.0;t19=4.1; tZ0=3.9; t21=3.3; t2Z=3.4;t23=3.3; t24=3.6; t25=Z.4; t26=Z.5; t27=2.0; t28=1.5; tZ9=2.2;t30=2.3; t31=1.6; t32=1.7sTF1=5.1; TF2=4.3; TF3=4.4; TF4=4.2; TF5=4.1; TF6=5.3; TF7=6.2;TF8=4.4; TF9=5.8; TF10=4.4s TF11=3.9; TF12=4.4;TF13=5.7; TF14=5.9;TF15=6.1; TF16=6.7;TF17=6.1; TF18=5.8;TF19=3.9; TF20=6.4; TF21=6.7sTFZZ=3.1;TF23=5.4; TFZ4=6.4;TF2S=5.6;TF26=4.0; TF27=6.1; TFZ8=6.1;TF29=5.0; TF30=5.3s TF31=6.5; TF32=6.0; TF33=5.2; TF34=4.6; TF35=3.3;TF36=5.6; TF37=4.3; TF38=2.0; TF39=3.4;TF40=4.1; TF41=3.3; TF4Z=4.1;TF43=6.1; TF44=2.3; TF45=3.6; TF46=2.2; TF47=2.6; TF48=3.0;If tapel= 0 then tape=1;If ta¤e1=1 then tape=2;If ratee1=0 then r1=1;If ratee1=1 then r1=2;If ratee1=2 then r1=3;If ratee1=3 then r1=4;If ratee2=0 then r2=1sIf ratee2=1 then r2=2;if ratee2=2 then r2=3;If ratee2=3 then r2=4;If ratee3=0 then r3=1;if r¤tee3=1 then r3=2;if ratee3=2 then r3=3;if ratee3=3 then r3=4;If ratee4=0 then r4=1;If ratee4=1 then r4=2;If ratee4=2 then r4=3;If ratee4=3 then r4=4;ARRAY A P1-P32 F1—F48;
do over A;A = A + 1;end;
206
ARRAY REVERSEF5 F4 F5 F7 F8 F9 F12 F15 F16 F17 F19 FZ9 F21 FZ4FZ7 FZ8 F29 F31 FSZ F35 F56 F39 F40 F61 F45 F44 F45 F48TF3 TF4 TF5 TF7 TF8 TF9 TF12 TF15 TF16 TF17 TF19 TFZ0 TFZ1 TFZ4TF27 TFZ8 TFZ9 TF31 TF32 TF33 TF36 TF39 TF40 TF41 TF43 TF44 TF45 TF48;DO OVER REVERSE;REVERSE=8—REVERSE; END;
XPERFORMANCE RATING MODIFICATIONS HERE;XFOLLONING ARE MEANS (PERF RATINGS) FOR EACH RATEE;PMRATEE1=MEAN(P1„P2,P5•P4;P5„P6„P7;P8);PMRATEEZ=MEAN(P9„P10;P11„P1Z,P13,P14„P15,P16);PMRATEE3=MEAN(P17,P18;P19;PZO;PZ1„PZZ„PZ5;PZ4);PMRATEE4=MEAN(PZ5;PZ6;PZ7,P28„PZ9;P30,P31;P32);¥FOLLONING ARE MEANS (PERF RATINGS) FOR EACH DIMENSION;PMDIM1=MEAN(P1,P9;P17,P25);PMDIMZ=MEAN(PZ;P10;P18;P26);PMDIM3=MEAN(P3,P11;P19,P27);PMDIM4=MEAN(PQ„P1Z;P20„PZ8);PMDIM5=MEAN(P5,P15,PZ1,P29);PMDIM6=MEAN(P6;P14;P22„P30);PMDIM7=MEAN(P7„P15;P23;P31);PMDIM8=MEAN(P8;P16;PZ4;PSZ);¤TRUE SCORE (PERF) MODIFICATIONS HERE;TMRATEE1=MEAN(T1;TZ;T3;T4;T5;T6„T7;T8);TMRATEEZ=MEAN(T9;T10;T11;T1Z;T15;T14;T15;T16);TMRATEE3=MEAN(T17,T18„T19;T20;TZ1„TZZ,TZ3;TZ4);TMRATEE4=MEAN(TZ5„TZ6;TZ7,TZ8;TZ9„T30;T31;T3Z);TMDIM1=MEAN(T1,T9;T17;TZ5);TMDIM2=MEAN(T2,T10„T18;TZ6);TMDIM3=MEAN(T3,T11;T19;TZ7);TMDIM4=MEAN(T4;T12,TZO;T28);TMDIM5=MEAN(T5;T13;T21,TZ9);TMDIM6=MEAN(T6;T14„TZ2„T50);TMDIM7=MEAN(T7,T15;TZ5;T31);TMDIM8=MEAN(T8;T16„TZ4;T3Z);PMOVERAL=MEAN(P1;PZ;P3;P4;P5;P6„P7;P8;P9;P10;P11;P12,P13,P14,P15;P16;P17;P18,P19;P20„PZ1;PZZ;PZ$;PZ4„P25„PZ6;PZ7;PZ8;PZ9„P50,P31,P3Z);TMOVERAL=MEAN(T1;TZ;T3;T4,T5;T6;T7;T8„T9„T10;T11;T12„T13;T14,T15;T16;T17„T18•T19„TZO;TZ1;TZZ;TZ5;TZ4,TZ5;TZ6;TZ7„T28,T29„T30„T31,T3Z);XNEXT COMPUTE P' AND T' FOR DIFFERENTIAL ACCURACY;PPRIME1=(P1—PMRATEE1—PMDIM1+PMOVERAL);TPRIME1=(T1—TMRATEE1-TMDIM1+TMOVERAL);PPRIME2=(PZ—PMRATEE1-PMDIMZ+PMOVERAL);TPRIMEZ=(TZ—TMRATEE1—TMDIMZ+TMOVERAL);PPRIME3=(P3—PMRATEE1-PMD1M5+PMOVERAL);TPRIME3=(T3-TMRATEE1—TMDIM3+TMOVERAL);PPRIME4=(P4—PMRATEE1-PMDIM6+PMOVERAL);TPRIME4=(T4—TMRATEE1—TMDIM6+TMOVERAL);PPRIME5=(P5-PMRATEE1—PMDIM5+PMOVERAL);TPRIME5=(T5—TMRATEE1—TMDIM5+TMOVERALI;PPRIME6=(P6—PMRATEE1—PMDIM6+PMOVERAL);TPRIME6=(T6-TMRATEE1-TMDIM6+TMOVERAL);PPRIME7=(P7-PMRATEE1-PMDIM7+PMOVERAL);TPRIME7=(T7—TMRATEE1—TMDIM7+TMOVERAL);PPRIME8=(P8—PMRATEE1·PMDIM8+PMOVERAL);
207
TPRIME8=(T8·TMRATEE1—TMDIM8+TMOVERAL);PPRIME9=(P9-PMRATEE2—PMDIM1+PMOVERAL);TPRIME9=(T9—TMRATEEZ—TMDIM1+TMOVERALl;PPRIME10=(P10—PMRATEE2—PMDIM2+PMOVERAL);TPRIME10=(T10·TMRATEE2·TMDIM2+TMOVERAL):PPRIME11=(P11—PMRATEE2·PMDIM3+PMOVERAL);TPRIME1l=(T11—TMRATEE2—TMDIM3+TMOVERAL);PPRIME12=(P12-PMRATEE2—PMDIM4+PMOVERAL);TPRIME12=(T12—TMRATEE2—TMDIM4+TMOVERAL);PPRIME13=(P15-PMRATEE2—PMDIM5+PMOVERAL);TPRIME13=(T13·TMRATEE2-TMDIM5+TMOVERAL)sPPRIME14=(P14-PMRATEE2—PMD!M6+PMOVERAL);TPRIME14=(T14-TMRATEE2—TMDIM6+TMOvERAL);PPRIME15=(P15—PMRATEE2—PMDIM7+PMOVERAL);TPRIME15=(T15-TMRATEE2-TMDIM7+TMOVERAL);PPRIME16=(P16—PMRATEE2—PMDIM8+PMOVERAL);TPRIME16=(T16·TMRATEE2—TMDlM8+TMOVERAL);PPRIME17=(P17—PMRATEE5—PMDIM1+PMOVERAL);TPRIME17=(T17—TMRATEE5—TMDIM1+TMOVERAL)sPPRIME18=(P18—PMRATEE3—PMDIM2+PMOVERAL);TPRIME18=(T18-TMRATEE3-TMDIM2+TMOVERAL);PPRIME19=(P19—PMRATEE5·PMDIM5+PMOVERAL);TPRIME19=(T19—TMRATEE3—TMDIM3+TMOVERAL);PPRIME20=(P20—PMRATEE}-PMDIM4+PMOVERAL);TPRIME20=(T20—TMRATEE3-TMDIM4+TMOVERAL)sPPRIME21=(P21—PMRATEE5-PMDIM5+PMOVERAL);TPRIME21=(T21—TMRATEE3-TMDIM5+TMOVERAL);PPRIME22=(P22—PMRATEE5-PMDIM6+PMOVERAL);TPRIME22=(T22—TMRATEE3—TMDIM6+TMOVERAL);PPRIME23=(P23—PMRATEES—PMDIM7+PMOVERAL)STPRIME23=(T25—TMRATEE3-TMDIM7+TMOVERAL);PPRIME24=(P24—PMRATEE3-PMDIM8+PMOVERAL);TPRIME24=(T24-TMRATEE3—TMDIM8+TMOVERAL);PPRIME25=(P25-PMRATEE4—PMDIM1+PMOVERAL);TPRIME25=(T25-TMRATEE4—TMDIM1+TMOVERAL);PPRIME26=(P26—PMRATEE4—PMDIM2+PMOVERAL);TPRIME26=(T26-TMRATEEG-TMDIM2+TMOVERAL);PPRIME27=(P27-PMRATEE4—PMDIM3+PMOVERAL):TPRIME27=(T27-TMRATEE4-TMDIM5+TMOVERALl;PPRIME28=(P28—PMRATEE4—PMDIM4+PMOVERAL);TPRIME28=(T28—TMRATEEQ—TMDIM4+TMOVERAL):PPRIME29=(P29—PMRATEE4-PMDIM5+PMOVERAL); ,TPRIME29=(T29—TMRATEE4—TMDIM5+TMOVERAL)sPPRIME30=(P50—PMRATEE4—PMDIM6+PMOVERAL);TPRIME50=(T30—TMRATEE4—TMDIM6+TMOVERALlsPPRIME31=(PS1—PMRATEE4—PMDIM7+PMOVERAL);TPRIME31=(T31-TMRATEE4-TMDIM7+TMOVERAL);PPRIME32=(P32—PMRATEE4-PMDIM8+PMOVERAL);TPRIME32=(T32—TMRATEE6—TMDIM8+TMOVERAL);DIFF1=(PPRIMEl—TPRIME1)¤¤2:DIFF2=(PPRIME2·TPRIME2)X¥2;DIFF3=(PPRIME3-TPRIME3)X¥2;DIFF4=(PPRIMEQ—TPRIME4)¤¤2;
{ DIFF5=(PPRIME5—TPRIME5)¤¥2:DIFF6=(PPRIME6—TPRIME6)¤¥2;
208
DIFF7=(PPRIME7—TPRIME7)XX2;DIFF8=(PPRIME8-TPRIME8)xx2;DIFF9=(PPRIME9-TPRIME9)xx2;DIFF10=(PPRIME10-TPRIME10)¤¥2;DIFF11=(PPRIME11-TPRIME11)X¤2;DIFF12=(PPRIME12-TPRIME12)¤¥2;DIFF13=(PPRIME13—TPRIME13)xx2;DIFF14=(PPRIME14-TPRIME14)**2;DIFF15=(PPRIME15—TPRIME15)¥¤2;DIFF16=(PPRIME16—TPRIME16)**2;DIFF17=(PPRIME17—TPRIME17)¥¤2;DIFF18=(PPRIME18—TPRIME18)¤!2;DIFF19=(PPRIME19—TPRIME19)¤X2;DIFF2U=(PPRIMEZÜ•TPRIME20)*!2iD1FF21=(PPRIME21—TPRIME21)xx2;DIFF22=(PPRIME22—TPRIME22)¤¥2:DIFF23=(PPRIME23-TPRIME25)X¤2;DIFF24=(PPRIME24-TPRIME24)xx2;DIFF25=(PPRIME25—TPRIME25)¥¤2;DIFF26=(PPRIME26—TPRIME26)X¤2;DIFF27=(PPRIME27—TPRIME27)¥¤2;DIFFZ8=(PPRIME28*TPRIME28)X¥Z;DIFFZ9=(PPRIMEZ9—TPRIME29)xx2;DIFF30=(PPRIME30-TPRIME30)¥¥2;DIFF51=(PPRIME31·TPRIME31)§¤2;DIFF32=(PPRIME52—TPRIME32)**2;DAPERF=(DIFF1+DIFF2+DIFF5+DIFF4+DIFF5+DIFF6+DIFF7+DIFF8+DIFF9+DIFF10+DIFF11+DIFF12+DIFF15+DIFF14+DIFF15+DIFFl6+DIFF17+DIFF18+DIFF19+DIFF20+DIFF2l+DIFF22+DIFF23+DIFF24+DIFF25+DIFF26+DIFF27+DIFF28+DIFF29+DIFF30+DIFF31+DIFF52)/52;DAP=SQRT(DAPERF);*COMPUTE DACOR FOR PERF RATINGS AS FOLLONS;¤P=PPRIME1;¤T=TPRIME1;*0UTPUT;¤P=PPRIME2;¤T=TPRIME2;¥0UTPUT;¤P=PPRIME3;¤T=TPRIME3;¤OUTPUT;¥P=PPRIME4;¥T=TPRIME4;¥0UTPUT;¤P=PPRIME5;*T=TPRIME5;*0UTPUT;¥P=PPRIME6;¤T=TPRIME6;¥0UTPUT;¥P=PPRIME7;¥T=TPRIME7;¥0UTPUT;¤P=PPRIME8;¤T=TPRIME8;¥0UTPUT;¤P=PPRIME9;*T=TPRIME9;¥0UTPUT;*P=PPRIME10;*T=TPRIME10;¤0UTPUT;XP=PPRIME11;¥T=TPRIME11;¥OUTPUT;¥P=PPRIME12;¤T=TPRIME12;¤OUTPUT;¥P=PPRIME15;¤T=TPRIME15;¥OUTPUT:¥P=PPRIME14;XT=TPRIME14;XOUTPUT;¤P=PPRIME15;¤T=TPRIME15;¤0UTPUT;¥P=PPRIME16s¤T=TPRIMEl6s*0UTPUT;¤P=PPRIME17;¤T=TPRIME17;*OUTPUT;¤P=PPRIME18;¤T=TPRIME18;XOUTPUT;¤P=PPRIME19;¥T=TPRIME19;¥OUTPUT;¤P=PPRIME20;¤T=TPRIME20;x0UTPUT;xP=PPRIME21;¤T=TPRIME21;¥OUTPUT;*P=PPRIME22;¤T=TPRIME22;¥0UTPUT;¤P=PPRIME23;¤T=TPRIME23;¥0UTPUT;
209
¤P=PPRIME24;XT=TPRIME24;¤0UTPUT;¤P=PPRIME25;¥T=TPRIME25;*0UTPUT;xP=PPRIME26;XT=TPRIME26;*0UTPUT;¤P=PPRIME27;¥T=TPRIME27;*0UTPUT;¤P=PPRIME28;¤T=TPRIME28;¤0UTPUT;¤P=PPRIME29;¤T=TPRIME29;¥0UTPUT;¥P=PPRIME50;¤T=TPRIME30;XOUTPUT;¥P=PPRIME31;*T=TPRIME51;*0UTPUT;¤P=PPRIME32;!T=TPRIME32;*0UTPUT;¥CORRELATE P AND T LATER;*MODIFICATIONS FOR FREQUENCY RATINGS HERE;FMRAT1=MEAN(F1;F2„F3;F4;F5;F6;F7;F8;F9,F10;F11;F12);FMRAT2=MEAN(F13;F14;F15;F16„F17,F18;F19;F20„F21;F22;F23;F26);FMRAT5=MEAN(F25„F26;F27;F28„F29„F30„F51„F32;F33;F36„F35,F36);FMRAT4=MEAN(F57;F38;F39;F40;F41;F42;F43„F44;FQ5„F46;F47,F48);FMDIM1=MEAN(F1;F13;F2S;F37);FMDIM2=MEAN(F2;F14„F26,F38);FMDIM5=MEAN(F3;Fl5;F27,F39);FMDIM4=MEAN(F4„F16;F28;F40);FMDIM5=MEAN(F5;F17;F29;F41);FMDIM6=MEAN(F6„F18;F30,F42);FMDIM7=MEAN(F7„F19„F51;F43);FMDIM8=MEAN(F8,F20„F32;F44);FMDIM9=MEAN(F9;F21,F55,F45);FMDIM10=MEAN(Fl0,F22;FS4;F46);FMDIM11=MEAN(F11,F23,F35„F47);FMDIM12=MEAN(F12,F24;F36„F48);xTRUES FOR FREQUENCIES GO HERE;TFMRAT1=MEAN(TF1,TF2;TF3„TFO,TF5;TF6;TF7;TF8;TF9;TF10;TF11„TF12);¥;g§AT2=MEAN(TF13,TF14„TF15,TF16,TF17,TFl8„TF19,TF20,TF21,TF22,TF25,TE§§ATé;2EAN(TF25,TF26,TF27,TF28,TF29;TF50;TF31„TF32;TF35;TF34;TFMRAT4=MEAN(TF37,TF38„TF39;TF40;TF41;TF42;TF4$„TF44;TF45;TF46;TF47„TF48);TFMDIM1=MEAN(TF1,TF13,TF25;TF37);TFMDIM2=MEAN(TF2;TF14,TF26;TFS8);TFMDIM3=MEAN(TF3;TF15;TF27,TF39);TFMDIM4=MEAN(TF4,TF16„TF28;TF40l;TFMDIM5=MEAN(TF5;TF17;TF29;TF41);TFMDIM6=MEAN(TF6;TF18„TF30,TF42);TFMDIM7=MEAN(TF7,TF19,TF31;TF43);TFMDIM8=MEAN(TF8,TF20;TF52;TF44);TFMDIM9=MEAN(TF9,TF21;TF33,TF45):TFMDIM10=MEAN(TF10;TF22;TF34;TF46);‘TFMDIMl1=MEAN(TF11,TF25;TF35;TF47);TFMDIM12=MEAN(TF12;TF24;TF36;TFQ8);FMOVER=MEAN(F1„F2;F3;F4;F5;F6;F7;F8;F9,F10;F11„F12,F13,F14,F15,F16,F17;F18;F19;F20,F2l;F22,F25,F24„F25;F26;F27;F28,F29,F50,F31,F32,F33,
TFMOVER=MEAN(TF1;TF2;TF5„TF4„TF5„TF6,TF7,TF8„TF9,TF10„TF11,TF12;TF15;TFl4;TF15;TF16;TFl7„TF18;TF19;TF20;TF21,TF22;TF23;TF26;TF25,TF26;TF27„TF28„TF29„TF50,TF31„TF32„TF33„TF34,TF35„TF36,TF37;TF38,TF39,TF40;
*COMPUTE P' AND T' (FREQUENCY RATINGS);
2lO
FPR1=(F1—FMRAT1-FMDIM1+FMOVER);FPR2=(F2—FMRAT1—FMDIM2+FMOVER);FPR3=(F3—FMRAT1—FMDIM5+FMOVER);FPR4=(F4-FMRAT1-FMDIM4+FMOVER)SFPR5=(F5—FMRAT1—FMDIM5+FMOVER);FPR6=(F6·FMRAT1—FMDIM6+FMOVER);FPR7=(F7—FMRAT1—FMDIM7+FMOVER);FPR8=(F8—FMRAT1—FMDIM8+FMOVER);FPR9=(F9—FMRAT1—FMDIM9+FMOVER);FPR10=(F10—FMRAT1—FMDIM10+FMOVER);FPR11=(F11-FMRAT1—FMDIM1l+FMOVER);FPR12=(F12—FMRAT1—FMDIM12+FMOVER);FPR13=(F15'FMRAT2'FMÜIM1+FMÜVER)BFPR14=(F14—FMRAT2·FMDIM2+FMOVER);FPR15=(F15—FMRAT2—FMDIM3+FMOVER);FPR16=(F16-FMRAT2~FMDIMQ+FMOVER)sFPR17=(F17—FMRAT2—FMDIM5+FMOVER)sFPR18=(F18—FMRAT2—FMDIM6+FMOVER);FPR19=(F19'FMRÄT2'FMDIM7+FMÜVER)BFPR20=(F20—FMRAT2-FMDIM8+FMOVER)2FPR21=(F21—FMRAT2—FMDIM9+FMOVER);FPR22=(F22-FMRAT2—FMDIM10+FMOVER);FPR23=(F25—FMRAT2—FMDIM11+FMOVER);FPR24=(F24-FMRAT2—FMD1Ml2+FMOVER);FPRZ5=(FZ5-FMRAT5—FMDIM1+FMOVER);FPR26=(F26—FMRAT5—FMDIM2+FMOVER);FPR27=(F27-FMRAT3-FMDIM3+FMOVER)sFPR28=(F28-FMRAT5—FMDIM4+FMOVER);FPR29=(F29—FMRAT3-FMDIM5+FMOVER)sFPR30=(F30-FMRAT3·FMDIM6+FMOVER);FPR31=(F51-FMRAT5—FMDIM7+FMOVER);FPR52=(F52-FMRAT3-FMDIM8+FMOVER);FPR35=(F33—FMRAT3—FMDIM9+FMOVER);FPR34=(F34—FMRAT5·FMDIM10+FMOVER);FPR35=(F55·FMRAT5—FMDIM11+FMOVER);FPR56=(F36—FMRAT3—FMDIMl2+FMOVER);FPR57=(F57-FMRAT4-FMDIM1+FMOVER);FPR58=(F38-FMRAT4-FMDIM2+FMOVER);FPR59=(F39—FMRAT4—FMDIM3+FMOVER);FPR40=(F40-FMRAT4—FMDIM4+FMOVER);FPR41=(F41-FMRAT4—FMDIM5+FMOVER)sFPR42=(F42-FMRAT4—FMDIM6+FMOVER);FPR63=(F43-FMRAT4-FMDIM7+FMOVER);FPR44=(F44-FMRAT4—FMDIM8+FMOVER);FPR45=(F45-FMRAT4—FMDIM9+FMOVER);_FPR46=(F46·FMRAT4—FMDIMl0+FMOVER);FPR47=(F47-FMRAT4—FMDIM11+FMOVER);FPR48=(F48—FMRAT4—FMDIM12+FMOVER)ETFPR1=(TF1—TFMRAT1—TFMDIMl+TFMOVER);TFPRZ=(TF2—TFMRAT1-TFMDIM2+TFMOVER);TFPR5=(TF3-TFMRAT1-TFMDIM3+TFMOVER);TFPR4=(TF4-TFMRAT1—TFMDIM4+TFMOVER);TFPR5=(TF5—TFMRAT1—TFMDIM5+TFMOVER)E_TFPR6=(TF6—TFMRAT1·TFMDIM6+TFMOVER);.TFPR7=(TF7—TFMRAT1—TFMDIM7+TFMOVER);
2ll
TFPR8=(TF8—TFMRAT1—TFMDIM8+TFMOVER);TFPR9=(TF9—TFMRAT1-TFMDIM9+TFMOVER);TFPR10=(TF10-TFMRAT1—TFMDIM10+TFMOVER);TFPR11=(TF11-TFMRAT1-TFMDIM11+TFMOVER)zTFPR12=(TF12-TFMRAT1—TFMDIM12+TFMOVER);TFPR13=(TF15·TFMRAT2-TFMDIM1+TFMOVER);TFPR14=(TF14-TFMRAT2—TFMDIM2+TFMOVER);TFPR15=(TF15-TFMRAT2-TFMDIM5+TFMOVER):TFPR16=(TF16—TFMRAT2-TFMDIM4+TFMOVER);TFPR17=(TF17-TFMRAT2—TFMDIM5+TFMOVER)sTFPR18=(TF18—TFMRAT2—TFMDIM6+TFMOVER);TFPR19=(TF19-TFMRAT2—TFMDIM7+TFMOVER);TFPR20=(TF20—TFMRAT2—TFMDIM8+TFMOVER);TFPR21=(TF21—TFMRAT2—TFMDIM9+TFMOVER);TFPR22=(TF22—TFMRAT2—TFMDIM10+TFMOVER);TFPR25=(TF23-TFMRAT2—TFMDIM11+TFMOVER);TFPR24=(TF24—TFMRAT2-TFMDIM12+TFMOVER);TFPR25=(TF25-TFMRAT3-TFMDIM1+TFMOVER);TFPR26=(TF26—TFMRAT3—TFMDIM2+TFMOVER);TFPR27=(TF27-TFMRAT5-TFMDIM3+TFMOVER);TFPR28=(TF28—TFMRAT3—TFMDIM4+TFMOVER);TFPR29=(TF29—TFMRAT3—TFMDIM5+TFMOVER);TFPR50=(TF30-TFMRAT3—TFMDIM6+TFMOVER);TFPR31=(TF31—TFMRAT5-TFMDIM7+TFMOVER);TFPR52=(TF32—TFMRAT5-TFMDIM8+TFMOVER1;TFPR53=(TF33·TFMRATS—TFMDIM9+TFMOVER);TFPR54=(TF34—TFMRAT3—TFMDIMl0+TFMOVER);TFPR35=(TF35—TFMRAT3—TFMDIM11+TFMOVER);TFPR36=(TF56—TFMRAT3—TFMDIM12+TFMOVER);TFPR37=(TF57—TFMRAT4-TFMDIM1+TFMOVER);TFPR58=(TF38-TFMRAT4-TFMDIM2+TFMOVER);TFPR39=(TF39—TFMRAT4—TFMDIM5+TFMOVER);TFPR60=(TF40-TFMRAT4—TFMDIM4+TFMOVER);TFPRQI=(TF61—TFMRAT4-TFMDIM5+TFMOVER):TFPR42=(TF42—TFMRAT4-TFMDIM6+TFMOVER);TFPR43=(TF43—TFMRAT4—TFMDIM7+TFMOVER);TFPR44=(TF44—TFMRAT4-TFMDIM8+TFMOVER);TFPR45=(TF45—TFMRAT4—TFMDIM9+TFMOVER);TFPR46=(TF46—TFMRAT4—TFMDIM10+TFMOVER);TFPR47=(TF47-TFMRAT4·TFMDIM11+TFMOVER);TFPR48=(TF68-TFMRAT4-TFMDIMl2+TFMOVER);FDIFF1=(FPR1—TFPR1)xx2;FDIFF2=(FPR2-TFPR2)xx2:FDIFF3=(FPR3—TFPR3)xx2;FDIFF4=(FPR4-TFPR4)xx2s
_FDIFF5=(FPR5-TFPR5)¥¤2;FDIFF6=(FPR6-TFPR6)xx2:FDIFF7=(FPR7—TFPR7)xx2:FDIFF8=(FPR8-TFPR8)x¤Z;FDIFF9=(FPR9—TFPR9)x¤2;FDIFF1U=(FPR1U'TFPR1U)¥¥2;FDIFF11=(FPR11-TFPR11)¤*2;FDIFF12=(FPR12—TFPR12)**2:FDIFF15=(FPR13—TFPR15)Xx2;FDIFF14=(FPR14-TFPR14)xx2;
212
FDIFF15=(FPR15—TFPR15)¤¥2§FDIFF16=(FPR16—TFPR16)XX2:FDIFF17=(FPR17-TFPR17)xx2sFDIFF18=(FPR18—TFPR18)x¥2;FDIFF19=(FPR19—TFPR19)xx2;FDIFF20=(FPR20—TFPR20)xx2;FDIFF21=(FPR21-TFPR21)xx2;FDIFF22=(FPR22-TFPR22)xx2;FDIFF25=(FPR25—TFPR23)¤x2;FDIFF24=(FPR24-TFPR24)xx2;FDIFF25=(FPR25-TFPR25)xx2;FDIFF26=(FPR26—TFPR26)xx2;FDIFF27=(FPR27—TFPR27)xx2;FDIFF28=(FPR28—TFPR28>xx2;FDIFF29=(FPR29-TFPR29)**Z;FDIFF50=(FPR30-TFPR30)¥§2;FDIFF31=(FPR$1—TFPR31)¥x2;FDIFF32=(FPR32-TFPR32)**2;FDIFF53=(FPR53-TFPR55)¤¤2:FDIFF34=(FPR54—TFPR34)xx2;FDIFF$5=(FPR55'TFPR55)¥¥Z§FDIFF36=(FPR56-TFPR56)¥*2;FDIFF37=(FPR37-TFPR37)xx2;FDIFF38=(FPR38-TFPR58)xx2iFDIFFS9=(FPR39·TFPR59)¤¤2;FDIFF40=(FPRQO-TFPR40)¥x2;FDIFF41=(FPR41-TFPR41uxxz;FDIFF42=(FPR42—TFPR42)xx2;FDIFFQ3=(FPR43—TFPR63)xx2;FDIFF44=(FPR44—TFPR44)xx2;FDIFF45=(FPR45-TFPR45)X¥2§FDIFF46=(FPRQ6—TFPR46)x¥2sFDIFFQ7=(FPR47-TFPR67)¤*2;FDIFF48=(FPR48-TFPR48Jxx2:DAFREQ=(FDIFF1+FDIFF2+FDIFF3+FDIFF4+FDIFF5+FDIFF6+FDIFF7+FDIFF8+FDIFF9+FDIFF10+FDIFF11+FDIFF12+FDIFF13+FDIFF14+FDIFF15+FDIFF16+FDIFF17+FDIFFl8+FDIFF19+FDIFF20+FDIFF21+FDIFF22+FDIFF25+FDIFF2Q+FDIFF25+FDIFFZ6+FDIFF27+FDIFF28+FDIFF29+FDIFF30+FDIFF31+FDIFF32+FDIFF33+FDIFF34+FDIFF55+FDIFF36+FOIFF37+FOIFF58+FDIFF39+FDIFF40+FDIFF41+FDIFF42+FDIFF43+FDIFF44+FDIFF45+FDIFF46+FDIFF47+FDIFF48)/48;DAF=SQRT(DAFREQ);XCOMPUTE DACOR FOR FREQ RATINGS AS FOLLONS;¥F=FPR1;¤FT=TFPR1;¤OUTPUT;¥F=FPR2;*FT=TFPR2;¥OUTPUT;¤F=FPR3;¤FT=TFPR5;¤OUTPUT;¤F=FPR4;¤FT=TFPR4;¥OUTPUT;¤F=FPR5;¤FT=TFPR5;¤0UTPUT;¤F=FPR6;¤FT=TFPR6;¤0UTPUT;¤F=FPR7;¥FT=TFPR7;XOUTPUT:¤F=FPR8;¥FT=TFPR8;*OUTPUT;¤F=FPR9;*FT=TFPR9;*OUTPUT;*F=FPR10;*FT=TFPRlU;*0UTPUT;¤F=FPR11;¤FT=TFPR11;*0UTPUT;¤F=FPR12;¥FT=TFPR12;*0UTPUT;¤F=FPR15;*FT=TFPR13;*OUTPUT;
213
¤F=FPR14;¥FT=TFPR14;¥0UTPUT;¥F=FPR15;*FT=TFPR15;¤0UTPUT;¤F=FPR16;¥FT=TFPR16;¥0UTPUT;¤F=FPR17;¤FT=TFPR17;¤OUTPUT;¤F=FPR18;¥FT=TFPR18;¤0UTPUT;¤F=FPR19;¤FT=TFPRl9;*0UTPUT;¤F=FPR20;*FT=TFPR20;¥0UTPUT:¥F=FPR21;*FT=TFPR2l;XOUTPUT;¤F=FPR22;¤FT=TFPR22;¤0UTPUT;¤F=FPR25;XFT=TFPR25;¤0UTPUT;¤F=FPR24;*FT=TFPR26;XOUTPUT;¤F=FPR25;¤FT=TFPR25;*OUTPUT;¤F=FPR26;*FT=TFPR26;*0UTPUT;¥F=FPR27;¥FT=TFPR27;XOUTPUT;¤F=FPR28;¤FT=TFPR28;¤OUTPUT;¤F=FPR29;¤FT=TFPR29;¤0UTPUT;¥F=FPR30;¥FT=TFPR50;x¤UTPUT;¥F=FPR31;*FT=TFPR31;¥0UTPUT;¤F=FPR52;*FT=TFPR32;¤0UTPUT;¤F=FPR53;¥FT=TFPR35;*0UTPUT;¤F=FPR34;¥FT=TFPR54;¥OUTPUTs¤F=FPR55;¥FT=TFPR35;¤0UTPUT;¤F=FPR56;¤FT=TFPR36;*0UTPUT;xF=FPR57;XFT=TFPR37;¤OUTPUT;*F=FPR38;¤FT=TFPR38;¥0UTPUT;¤F=FPR59;¤FT=TFPR39;¥0UTPUT;¤F=FPR40;¤FT=TFPR40;¤0UTPUT;¤F=FPR41;*FT=TFPR41;*0UTPUT:!F=FPR42;¤FT=TFPR42;*0UTPUT;¥F=FPR45;¤FT=TFPR45;¤0UTPUT:¤F=FPR44;¥FT=TFPRü4;¤0UTPUT;¤F=FPR45;¤FT=TFPR45;¤0UTPUT;XF=FPRG6;¥FT=TFPR46;XOUTPUT;*F=FPR47;¥FT=TFPR47;¤0UTPUT;¥F=FPR48;¤FT=TFPR48;*0UTPUT;XCORRELATE F AND FT LATER;äelevation is calculated as follows;EL=(PMOVERAL-TMDVERAL)¤¤2;EP=sqrt(e1);ädifferential slevation is calculatad as followssDELEV1=((PMRATEE1-PMOVERAL)-(TMRATEEl—TMOVERAL))x¥2:DELEV2=((PMRATEE2—PMOVERAL)—(TMRATEE2—TMOVERAL))¤¤2;DELEV3=((PMRATEE3—PMOVERAL)—(TMRATEE5—TMOVERAL))x¥2;DELEV4=((PMRATEE4—PMOVERAL)—(TMRATEE4—TMOVERAL))xx2sDIFELEV=MEAN(DELEVl,DELEV2,DELEV3,DELEV4);‘DEP=SQRT(DIFELEV);*stereotypic accuracy is calculated as follows;STER1=((PMDIM1—PMDVERAL)-(TMDIM1—TMDVERAL))¤x2;STER2=((PMDIM2—PMOVERAL)-(TMDIM2—TMOVERAL))¥x2;STER3=((PMDIM5-PMOVERAL)—(TMDIM3-TMOVERAL))!¥2;STER6=((PMDIM4—PMOVERAL)—(TMDIM4-TMOVERAL))¤¤2;STER5=((PMDIM5-PMOVERAL)-(TMDIM5—TMOVERAL))¥¤2:STER6=((PMDIM6—PMOVERAL)—(TMDIM6-TMOVERAL))**2;STER7=((PMDIM7*PNOVERAL)—(TMDIM7—TMOVERAL))XX2§STER8=((PMDIM8—PMOVERAL)—(TMDIM8—TMOVERAL))**2;
214
STEREOT=MEAN(STER1„STER2„STER3,STER4„STER5„STER6,STER7,STER8)ESAP=SORT(STEREOT);XFOLLONING ARE THE CALCULATIONS FOR FREQUENCY RATINGS;äelevation is calculated as follows;ELEVA=(FMOVER-TFMOVER)X*2;EF=sqrt(e1eva);¥differentia1 elevation is calculated as follows;DEL1=((FMRAT1—FMOVER)—(TFMRAT1-TFMOVER))xx2;DEL2=((FMRAT2-FMOVER)—(TFMRAT2—TFMOVER))¤x2;DEL3=((FMRAT3-FMOVER)—(TFMRAT5—TFMOVER))xx2;DEL4=((FMRAT4—FMOVER)—(TFMRAT4—TFMOVER))xx2;DIFEL=MEAN(DEL1,DEL2,DEL5,DEL4);DEF=SORT(DIFEL);lstareotypic accuracy is calculated as follows;STE1=((FMDIMI-FMOVER)—(TFMDIM1-TFMOVER))¤x2;STE2=((FMDIM2—FMOVER)-(TFMDIM2-TFMOVER))¤¤2;STE3=((FMDIM3-FMOVER)-(TFMDIM5—TFMOVERl)x¤2;STE4=((FMDIM4—FMOVER)—(TFMDIM4-TFMOVER))¤X2;STE5=((FMDIM5—FMOVER)—(TFMDIM5—TFMOVER))**2;STE6=((FMDIM6-FMOVER)—(TFMDIM6-TFMOVER))xx2;STE7=((FMDIM7—FMOVER)—(TFMDIM7—TFMOVER))x¤2;STE8=((FMDIM8—FMOVER)—(TFMDIM8—TFMOVER))**2;STE9=((FMDIM9—FMOVER)-(TFMDIM9—TFMOVER))xx2;STE10=((FMDIM10-FMOVER)—(TFMDIM10—TFMOVER))**2;STE11=((FMDIM11-FMOVER)-(TFMDIM11—TFMOVER))xx2:STE12=((FMDIMl2—FMOVER)—(TFMDIM12—TFMOVER))**2;5TEREO=MEAN(STE1,STE2„STE3,STE6„STE5„STE6„$TE7,STE8,STE9,STE10„STE11„STE12);SAF=SQRT(STEREO);XCOMPUTE SACOR FOR BOTH PERF AND FREQ RATINGS AS FOLLONS;¤SP=PMDIM1;XST=TMDIMl;¥OUTPUT;¥SP=PMDIM2;¥ST=TMDIM2;¥0UTPUT;*SP=PMDIM5;¥ST=TMDIM3;*OUTPUT;¤SP=PMDIM4;¤ST=TMDIM4:XOUTPUT;¥SP=PMDIM5;¤ST=TMDIM5;¥OUTPUT;¤SP=PMDIM6;*ST=TMDIM5;¥0UTPUT;¤SP=PMDIM7;XST=TMDIM7;¤OUTPUT;¤SP=PMDIM8;¤ST=TMDIM8;*OUTPUT;*SF=FMDIM1;¥STF=TFMDIM1;¤OUTPUT;¥SF=FMDIM2;¤$TF=TFMDIM2;¥0UTPUT;¥SF=FMDIM3;¥STF=TFMDIM3;*0UTPUT:¤SF=FMDIM4;¥STF=TFMDIM4;¤OUTPUT;¤SF=FMDIM5;XSTF=TFMD!M5;*OUTPUT;¤SF=FMDIM6;XSTF=TFMDIM5;*OUTPUT;¥SF=FMDIM7;¤STF=TFMDIM7;*OUTPUT;
_*SF=FMDIM8;*STF=TFMDIM8;¤0UTPUT;¥SF=FMDIM9;*STF=TFMDIM9;XOUTPUT;XSF=FMDIM10;*STF=TFMDIM10;*OUTPUT;XSF=FMDIM11;¤5TF=TFMDIM11;¤OUTPUT;¤SF=FMDIMl2;*STF=TFMDIMl2;*0UTPUT;¤COMPUTE DECOR FOR BOTH PERF AND FRE0 RATINGS AS FOLLOHS;XDECP=PMRATEE1;¥DECT=TMRATEE1;¤0UTPUT;*DECP=PMRATEE2;*DECT=TMRATEE2;¤OUTPUT;¤DECP=PMRATEE3;%DECT=TMRATEE5;¤0UTPUT;XDECP=PMRATEE4;¤DECT=TMRATEE4;¤OUTPUT;
215
¤DECF=FMRAT1;¤DECFT=TFMRAT1;¤0UTPUT;¤DECF=FMRAT2;¥DECFT=TFMRAT2;¤0UTPUT;xDECF=FMRAT3;¤DECFT=TFMRAT5;¤OUTPUT;¤DECF=FMRAT4;¤DECFT=TFMRAT4;¤0UTPUT;¤COMPUTE OVERAL ACCURACY AS FOLLOHS;GLOB1=(P1—T1)¤¤2;GLOB2=(P2—T2)¤¤2;GLOB3=(P5—T3)¤¥2;GLOB4=(P4-T4)¤¥2;GLOB5=(P5—T5)¤¥2;GLOB6=(P6—T6)¤x2:GLOB7=(P7—T7)¤¥2;GLOB8=(P8-T8)¤¥2sGLOB9=(P9-T9)x¥2sGLOB10=(P10—T10)¤¤2:GLOB11=<P11-T11)xx2;GLOB12=(P12—T12)x¤2;GLOB13=(P13—T15)xx2;GLOB14=(P14-T14)¤x2;GLOB15=(P15—T15)¥¤2;GLOB16=(P16-T16)¥¥2;GLOB17=(P17—T17)x¤2;GLOB18=(P18—T18)x¤2;
UGLOB19=(P19—T19)¤x2;GLOB20=(P20-T20>xx2;GLOB21=(P21—T21)xx2;GLOB22=(P22—T22)¥¤2;GLOB23=(P23—T25)xx2;GLOB24=(P24—T24)¤¤2;GLOB25=(P25—T25)¤¤2;GLOB26=(P26-T26)*¤2;GLOB27=(P27·T27)xx2;GLOB28=(P28—T28)xx2;GLOB29=(P29—T29)¤x2;GLOB50=(P50—T30)xx2;GLOB51=(P51—T51)¤*2;GLÜB52=(P52'T$2)§XZ}TOTACCP=MEAN(GLOB1,GLOB2,GLOB3,GLOB4,GLOB5,GLOB6,GLOB7,GLOB8,
GLOB9,GLOB10,GLOB11.GLOB12.GLOB15,GLOB14,GLOB15,GLOB16,GLOB17,GLOB18„GLOB19,GLOB20,GLOBZI.GLOB22,GLOB25,GLOB24,GLOB25„GLOB26,GLOBZ7„GLOB28,GLOB29.GLOB30,GLOBSI,GLOB52l;
CKPERF=TOTACCP—(STEREOT+DIFELEV+EL+DAPERF);GLOBAL1=(F1-TF1)xx2;GLOBAL2=(F2—TF2)¤x2;GLOBAL5=(F3—TF3)xx2;GLOBAL4=(F4—TF4)xx2;GLOBAL5=(F5-TF5)xx2;GLOBAL6=(F6-TF6)xx2;GLOBAL7=(F7—TF7)xx2;GLOBAL8=(F8—TF8)x¤2sGLOBAL9=(F9—TF9)xx2;GLOBAL10=(F10—TF10)xx2;GLOBAL11=(F11—TF11)xx2;UGL0BAL12=<Fl2—TF12)xx2;.GLOBAL13=(F13—TF15)xx2;
216
GLOBAL14=(F14-TF14)xx2;GLOBAL15=(F15-TF15)xx2;GLOBAL16=(F16—TF16)xx2;GLOBAL17=(F17—TF17)¤x2;GLOBAL18=(F18-TF18)x¤2;GLOBAL19=(F19·TF19)¥¥2;GLOBAL20=(F20-TF20)xx2;GLOBAL21=(F21—TF21)¤x2;GLOBAL2Z=(F22-TF22)¤x2:GLOBAL23=(F23-TF23)**2;GLOBAL24=(F24—TF24)x¤2;GLOBAL25=(FZS—TF25)¤¥2;GLOBAL26=(F26-TF26)xx2sGLOBAL27=<F27-TF27)xx2sGLOBAL28=(F28—TF28)x¤2;GLOBAL29=(F29—TF29)¤¤2;GLOBAL30=(F30-TF30)xx2;GLOBAL31=(F31—TF51)xx2;GLOBAL32=(F32—TF32)xx2;GLOBAL35=(F53—TF33)¤§2;GLOBAL34=(F34—TF34)x¤2;GLOBAL55=(F35—TF35)xx2;GLOBAL56=(F36-TF36)x¤2;GLOBAL37=(F37·TF37)xx2;GLOBALS8=(F38-TF38)xx2;GLOBAL39=(F39-TF59)¤¤2;GLOBAL40=(F40—TF40)xx2;GLOBAL41=(F41—TF41)xx2;GLOBAL42=(F42-TF42)¤¤2;GLOBAL43=(F45-TF45)x¤2;GLOBAL44=(F44-TF46)¤x2;GLOBAL45=(F45—TF4S)x¤2;GLOBAL46=(F46-TF46)xx2;GLOBAL47=(F47-TF47)x¤2:GLOBAL48=(F48-TF48)¤¤2;TOTACCF=MEAN(GLOBAL1„GLOBAL2,GLOBAL5„GLOBAL4,GLOBAL5,GLOBAL6;GLOBAL7,GLOBAL8;GLOBAL9,GLOBAL10,GLOBAL11,GLOBAL12,GLOBAL13;GLOBAL14,GLOBAL15,GLOBAL16„GLOBAL17,GLOBAL18„GLOBAL19,GLOBAL20;GLOBAL21;GLOBAL22;GLOBAL23„GLOBAL24,GLOBAL25,GLOBAL26,GLOBAL27;GLOBAL28,GLOBAL29„GLOBAL30,GLOBAL31,GLOBAL32,GLOBAL33,GLOBALSG;GLOBAL55;GLOBAL36,GLOBAL57;GLOBAL38,GLOBAL39„GLOBAL40,GLOBAL41,GLOBAL62;GLOBAL43„GLOBAL44,GLOBAL45,GLOBAL46,GLOBAL47;GLOBAL48)SCKFREO=TOTACCF—(STEREO+DIFEL+ELEVA+DAFREQ);KEEP ID DAPERF DAFRE0 ELEVA EL DIFELEV DIFEL STEREOT STEREODAP DAF SAP SAF EP EF DEP DEF TOTACCP TOTACCFCKPERF CKFREO;cards;
2l7
PROC SORT DATA=THREE; BY ID;DATA BOTH;MERGE ONE TNO THREE; BY ID;PROC PRINT; VAR REP DIMS DAP DAF SAP SAF EP EF DEP DEFDARP DARF SARP SARF DERP DERF TOTACCP TOTACCF CKPERF CKFREO;BY D;PROC FREQ;TABLES PAYJOB TITLE MGNOMG ORGFUNC SEX DIMS;PROC CORR DATA=BOTH;VAR PAYJOB YRSNORK MGNOMG AGE YRCOLL SEXREP DIMS DAP DAF SAP SAF EP EF DEP DEF DARP DARF SARP SARFDERP DERF TOTACCP TOTACCF;PROC RANK DATA=BOTH OUT=LAST GROUPS=5;VAR YRSNORK;RANKS EXPER;PROC ANOVA DATA=LAST;CLASS EXPER;MODEL DEP DEF=EXPER;¥MODEL DAP DAF SAP SAF EP EF DEP DEF=EXPER;MEANS EXPER/TUKEY;PROC SORT;BY EXPER;PROC MEANS;VAR DAP DAF SAP SAF EP EF DEP DEF;BY EXPER;XPROC CORR DATA=ONE;*VAR P T;*BY ID;XPROC CORR DATA=ONE;*VAR F FT;XBY ID;¤PROC CORR DATA=ONE;*VAR SP ST;*BY ID;¤PROC CORR DATA=ONE;*VAR SF STF;*BY ID;*PROC CORR DATA=ONE;*VAR DECP DECT;*BY ID;¥PROC CORR DATA=ONE;¥VAR DECF DECFT;*BY ID;x///
218
top related