DOCUMENT RESUME

ED U78 166 VT 020 330

TITLE Experimental Manpower Laboratory for CorrectionsPhase III. Final Report for the Period September,1971 to February, 1973.

INSTITUTION Rehabilitation Research Foundation, ,Elmore, Ala.SPONS AGENCY Manpower Administration (cm), Washington, E.C.

Office of Research and Development.REPORT NO DLMA-82-01-69-06-8PUB DATE Feb 73NOTE 46p.AVAILABLE FROM National Technical Information Service, Springfield,

VA 22151 (DLMA-82-01-69-06-8, MF $0.95, BC $3.00)

EDRS PRICE MF-$0.65 HC-$3.29DESCRIPTORS *Employment; Evaluatibr; Followup Studies; Job

Placement; Males; Manpower Utilization; *Prisoners;Programed Instruction; *Rehabilitation; *ResearchProjects; Statistical Analysis; *VocationalDevelopment

ABSTRACTThis report covers the objectives, procedures, and

findings of four major projects directed by the Experimental ManpowerLaboratory for Correction (EMLC): (1) Token Economy (Ecology)Project, (2) Correctional Officer Training Project, (3) LongitudinalFollow-Up Studies, and (4) Information Dissemination and UtilizationProject. EMLC grew out of an earlier project conducted to test thefeasibility of using programed instruction in basic education classesfor male felons. Included are an overview of the project, adiscussion of the problems encountered in its operation anddevelopment, and announcements of proposed directions for phase IVstudies. (Author/SN)

k'V.off

2

.1'48W4w z?

14'

wx0

t3ti"

eVei

14,

4415'

(

,` 4

5!

-113361 .f.A141:z1

ti"V4rS

h

kOb

,

t.tct,g

t

0;44-

1.4541r4.11:7"t -7"" je."7446,7,54::

van=

MIN

Mor

tto

..1.3

.4*1

ST

AN

DA

RD

TIT

LE P

AG

EF

OR

TE

CH

NIC

AL

RE

PO

RT

S1.

Rep

ort N

o.=

A82

-01-

69-0

6-8

,i0,,.

akee

tt440

,,,,..

.4

,;!,'!

",`R

..il4

r,3.

Rec

ipie

nt's

Cat

alog

No.

I. T

itle

and

Sub

title

Exp

erim

enta

l Man

pow

er L

abor

ator

y fo

r C

orre

ctio

nsPh

ase

111

Fina

l Rep

ort

S. R

epor

t bat

eFe

brua

ry 1

973

6. P

erfo

rmin

g O

rgan

izat

ion

Cod

e

,i. A

utho

r(s)

LR

ehab

ilita

tion

Res

earc

h Fo

unda

tion

S. P

erfo

rmin

g O

rgan

izat

ion

Rep

t.N

o.

9. P

erfo

rmin

g O

rgan

izat

ion

Nam

e an

d A

ddre

ss

Reh

abili

tatio

n R

esea

rch

Foun

datio

nP.

O. B

ox 1

107

Elm

ore,

Ala

bam

a 36

025

10. P

roje

ct/T

ask/

Wor

k U

nit N

o.

11.

Con

trac

t/Gra

nt N

o.

DL

82-

01-6

9-06

12. S

pons

orin

g A

genc

y N

ome

and

Add

ress

U.S. Department of Labor

Manpower Administration

Office orRetearch and Developaent

1111 20th St., R.W.

Washington, D.C.

20210

13. T

ype

of R

epor

t & P

erio

dC

over

ed F

inal

rep

ort

Sept

. 197

1 -

Feb.

197

314

.S

pons

orin

g A

genc

y C

ode

15. S

uppl

emen

tary

Not

esi

Non

e16

. Abs

trac

ts

Thi

s re

port

cov

ers

the

four

maj

or p

roje

cts

of th

e E

xper

imen

tal M

anpo

wer

Lab

orat

ory

for

Cor

rect

ions

(E

ML

C):

(1)

Tok

en E

cono

my

(Eco

logy

) Pr

ojec

t, (2

) C

orre

ctio

nal O

ffic

erT

rain

ing

(CO

T),

(3)

Lon

gitu

dina

l Fol

low

-Up

Stud

ies,

and

(4)

Util

ilatio

n. P

roje

ct o

bjec

tives

t

and

a su

mm

ary

of p

roce

dure

are

incl

uded

, as

are

find

ings

and

dat

a fo

r th

ose

proj

ects

in w

hich

ana

lysi

s is

com

plet

e. A

bri

ef h

isto

ry o

f th

e E

ML

C is

pre

sent

ed in

the

intr

oduc

tion

to th

is r

epor

t; th

e co

nclu

ding

sec

tion

caps

uliz

es p

roje

ct f

indi

ngs,

pro

blem

s en

coun

tere

d,an

d pr

opos

ed d

irec

tions

for

Pha

se I

V s

tudi

es. T

he v

ario

us p

rodu

cts

com

plet

ed in

Pha

se I

IIar

e lis

ted

and

anno

tate

d.

17.

Key

Wor

ds a

nd D

ocum

ent A

naly

sis.

17o

. Des

crip

tors

Adj

ustm

ent

Man

pow

er U

tiliz

atio

nE

duca

tion

Mot

ivat

ion

Em

ploy

men

tPr

ogra

mm

ed I

nstr

uctio

nE

valu

atio

nR

ehab

ilita

tion

.

Ince

ntiv

esR

einf

orce

men

tL

earn

ing

Stat

istic

al A

naly

sis

Lite

racy

Man

pow

erM

ales

17b.

loen

tifie

rs/O

pen-

End

id T

erm

s

Con

tinge

ncy

Man

agem

ent

Tok

en E

cono

my

(Eco

logy

) Pr

ojec

tC

orre

ctio

nal O

ffic

er T

rain

ing

(CO

T)

Util

izat

ion

Indi

vidu

ally

Pre

scri

bed

Inst

ruct

iona

l (IP

I) S

yste

m

. ..

Elm

ore,

Ala

bam

a36025

DL 82-01-69-06

12.

Spo

nsor

ing

Age

ncy

Nam

e an

d A

ddre

ssU.S. Department of Labor

Manpower Administration

Office orRetearch and Development

1111 20th St.,

R.W

.Washington, D.C.

20210

13. T

ype

of R

epor

t & P

rrio

dC

over

edFi

nal r

epor

tSe

pt. 1

971

- Fe

b. 1

973

14. S

pons

orin

g A

genc

y C

ode

Is.

Sup

plem

enta

ry N

otes

N o

ne16

. Abs

trac

ts

Thi

s re

port

cov

ers

the

four

maj

or p

roje

cts

of th

e E

xper

imen

tal M

anpo

wer

Lab

orat

ory

for

Cor

rect

ions

(E

ML

C):

(1)

Tok

en E

cono

my

(Eco

logy

) Pr

ojec

t, (2

) C

orre

ctio

nal O

ffic

erT

rain

ing

(CO

T),

(3)

Lon

gitu

dina

l Fol

low

-Up

Stud

ies,

and

(4)

Util

izat

ion.

Pro

ject

obj

ectiv

esan

d a

sum

mar

y of

pro

cedu

re a

re in

clud

ed, a

s ar

e fi

ndin

gs a

nd d

ata

for

thos

e pr

ojec

tsin

whi

ch a

naly

sis

is c

ompl

ete.

A b

rief

his

tory

of

the

EM

LC

is p

rese

nted

in th

e in

trod

uctio

nto

this

rep

ort;

the

conc

ludi

ng s

ectio

n ca

psul

izes

pro

ject

fin

ding

s, p

robl

ems

enco

unte

red,

and

prop

osed

dir

ectio

ns f

or P

hase

IV

stu

dies

. The

var

ious

pro

duct

s co

mpl

eted

in P

hase

III

are

liste

d an

d an

nota

ted.

T. K

ey W

ords

and

Doc

umen

t Ana

lysi

s. 1

7e. D

escr

ipto

rs

Adj

ustm

ent

Man

pow

er U

tiliz

atio

nE

duca

tion

Mot

ivat

ion

Em

ploy

men

tPr

ogra

mm

ed I

nstr

uctio

nE

valu

atio

nR

ehab

ilita

tion

Ince

ntiv

esR

einf

orce

men

tL

earn

ing

Stat

istic

al A

naly

sis

Lite

racy

Man

pow

erM

ales

17b.

Ioen

tifie

cs/O

pen-

End

ed T

eem

s

Con

tinge

ncy

Man

agem

ent

Tok

en E

cono

my

(Eco

logy

) Pr

ojec

tC

orre

ctio

nal O

ffic

er T

rain

ing

(CO

T)

Util

izat

ion

Indi

vidu

ally

Pre

scri

bed

Inst

ruct

iona

l (IP

I) S

yste

m.

; Ind

ivid

ualiz

ed R

eadi

ng S

yste

m f

or A

dults

(IR

SA)

' Lon

gitu

dina

l Fol

low

-Up

Stud

ies

174.

CO

SA

TI F

ield

/Gco

up5J

& S

IC11

1. D

istr

ibut

ion

Sta

tem

ent

Distribution is unlimited.

Available from

National Technical Information Service, Springfield

Va., 22151.

....

-.

19. S

ecur

ity C

lass

(T

his

Rep

ort)

21. N

o. o

f Pos

es

20. s

ecur

ityC

sifie

errir

i541

----

--'

121 clailf101............a92._____.

IJSC

OM

IA. D

C 1

1100

2 P7

0

TABLE OF CONTENTS

Page

Introductionx

3

Token Economy (Ecology) Project 9

Correctional Officer Training (COT) 21

Longitudinal Follow-Up Studies 27

Utilizaticn 37

In Conclusion 47

-.0

Rehabilitation begins in i'Irtitlitions like this . .

4

introduction

The EMLC

The present Experimental Manpower Labora-tory for Corrections (EMLC) grew out of an earlyprogrammed learning project conducted by the Re-habilitation Research Foundation (RRF) at DraperCorrectional Center in Elmore, Alabama. Thisproject, which operated between 1961 and 1964,tested the feasibility of using programmed instructionin basic education classes for adult male felons. Thisproject was successful; several inmates even beganwriting their own programs for areas in which no suchinstruction existed.

Once working in the institution, however, theRRF found that basic education was only one areain which the inmates were deficient. Vocationaltraining was also a critical need. But because prisoninmates had not been considered a part of theunemployed labor force at which MDTA was aimed,no such training programs existed in the institution.

Between 1964 and 1968 the RRF contractedwith the Department of Labor to conduct a seriesof projects at Draper which were later known as theDraper Project. These were similar in nature andpurpose to those conducted at the federalreformatory in Lorton, Virginia, and at the city jailat Rikers Island, New York. These projects tested thefeasibility of operating MDTA training programs inthe respective types of correctional institutions.

The projects at the three institutions accom-plished to some degree what they had proposedto do. MDTA training was found to be operable incorrectional institutions, with some modificationnecessary to accommodate the nature and specialneeds of the inmate population., The modificationthat evolved helped to structure amendments(Section 251) to MDTA in 1966 which authorizedmanpower training for a broad number ofinstitutionalized offenders.

r3

It had become apparent, however, thatlong-range assessment of various program componentswas necessary, an assessment which would require avehicle of a slightly different nature than that of theE&D approach. The RRF contracted with theManpower Administration to conduct a number ofstudies with the prison population, assessing theneeds of this population, evaluating programeffectiveness, and exploring additional possibilitiesfor rehabilitation programs. This was the beginningof the EMLC.

Many of the EMLC projects in Phase I (1968-70)examined the relationship of the released offenderto the community, particularly in terms ofadjustment problems. One project centered onemployment barriers the offender encountered,determining which of these could be overcome in theinstitutional training programs. Another projectprovided small grants to ex-offenders to assist themin establishing themselves in the community, butfound that the availability of these funds made littledifference in training-related placement. Intensivefollow-up services for a small group of trainees wereprovided in another project, studying theenvironmental support available to the ex-offender inthe community.

Two other Phase I studies laid the foundationfor projects which have continued through Phase III.One of these investigated the effect of the prisonenvironment on the training program. A correlationwas found between the amount of accurateirformation institution personnel had about theprogram and their attitude toward it, a finding whichled to the correctional officer training program begunin Phase II.

The other study was the '69 Follow-Up, whichevolved from the project dealing with communityenvironmental support. The Follow-Up study soughtto develop a methodology for systematic longitudinal

follow-up which would establish a behavioraldemogiaphy of the released or paroled offender,evaluate MDT training, and develop instrumentspredictive of law violation, criminal behavior, andrecidivism. Data analysis for this study continued intoPhase III of th e Lab.

The Phase II (1970-71) projects were muchmore institutionally oriented than those of Phase I,in an attempt to forestall postrelease adjustmentproblems through institutional treatment programs.These projects also sought to incorporate many ofthe behavior modification concepts which wereproving successful in other settings, e.g., mentalinstitutions. The MDT training project wasindividualized to better meet the needs of thetrainees, resulting in greater training efficiency. Theproject had originally proposed to completely train95 men and partially train 20 others; instead, theproject was able to completely train a total of 123men.

Part of this efficiency was attributed to theconcurrent contingency management study whichexplored w,ys of motivating our experimental groupsof prison inmates, comparing time-contingent payoffto performance-contingent payoff. The study datasuggested that the more payoff was contingent onperformance, the higher the level of that performancewould be.

Recognizing the important influence ofinstitutional experiences on the offender, the EMLCsought to control these experiences and provide atotal environment for rehabilitation by establishinga token economy in the institution. This EcologicalUnit occupied one cellblock in the institution anddemonstrated that the token economy was a

successful alternative to aversive control as a meansof managing and motivating prison inmates.

To capitalize on the influence of theinstitutional staff upon the inmates, the EMLC

4

conducted two cycles of correctional officer training(COT) in Phase 11. This training introduced theofficers to the use of behavior modificationtechniques as an alternative to punishment andaversive control. Ultimately, these trained officerswould be able to administer, with the propersupervision, a program of institutional treatmentbased on behavior modification principles.

The '69 Follow-Up from Phase I was extendedinto Phase II, then followed by a more refined study,the '71 Follow-Up. This study was designed toprovide replication for the earlier study and evaluatethe effect of various EMLC projects components onrecidivism and employment.

Phase III of "t'e EMLC, the period covered inthis report, expan al refined the studies begunin the previous p. emphasizing even more thebehavioral orientation of the Lab. The tokeneconomy study in Phase II hau been limited tobehaviors on the Ecological Unit, but in Phase IIIwas expanded to include behaviors occurring in otherareas of the institution. The contingency managementstudy became a sub-study, providing the basiceducation component of the Ecological Unit.Additionally, an institutional check system was begunby the Draper staff to motivate farm performance.The token economy section of this report begins onpage 12.

Correctional officer training continued, with thethird cycle providing the tryout for a package ofself-instructional booklets designed to teach theofficers the use of behavior modification techniques.Evaluation of this method of instruction wascomplicated when difficulties arose concerning thepracticum phase of the training. This was the phasein which the officers were to have had an opportunityto use their new skills under Lab supervision. COTis described in the section beginning on page 12.

4

A key activity throughout all the phases of theEMLC has been utilization. While the Lab's chiefcommitment is to the design, implementation, andevaluation of experimental programs, an increasinglyimportant effort is that of dissemination forutilization. Throughout the existence of the Lab,products slich as reports, "how-to" manuals,professional papers, and programmed lessons havebeen prepared and distributed to influence positivechange within the context of correctional manpowerproblems. Two educational systems, the IndividuallyPrescribed Instructional System for basic educationand the Individualized Reading System for Adults,have been developed and are being revised. Moreregular and continuing means of disseminationinclude orienting visitors on site, deliveringpresentations in the community and at professionalmeetings, and university teaching. Two newslettershave also aided in dissemination. These activities aredescribed in more detail in the section starting onpage 12,

A general discussion of some of the findings andproblems encountered in Phase III, plus a briefdescription of the proposed studies for Phase IV, arefound in the section beginning on page 12.

fl

Punch cards typify the token economy techniques used for behavior change and institution management.

OA economy(ccollogg)

Amoject

s'e

The Token Economy in Corrections..

The EMLC's decision to explore the feasibilityof applying the token economy model in correctionswas made in response to the success of this modelin other settings, e.g., the military and mentalinstitutions. The studies in these settings haddemonstrated the effectiveness and efficiency of thetoken economy in changing behavior: once the targetbehavior was selected and procedures institutedwhich followed the principlds of behavior, behavioralchange occurred. In addition to being efficient andeffective, the tot:en economy was economical, takingadvantage of existent reinforcers.

For the correctional institution, with its meagerbudget, these advantages are particularly attractive.De efficiency of the token economy model is ineffect doubled in the correctional setting, since theinstitution's goals of treatment and custody can bemet-simultaneously: behavior can be,,both motivatedand managd by token economy procedures.Additionally, this model can use reinforcers presentlyavailable in the institution, many of which arecost-free (e.g., admittance to a work-release program).And, as a spinoff, the token economy is the kindof systcm that, once dcmonstrrted and establishedin the institution, can be administered by the presentline staff, with appropriate supervision. Thus theapplication of the token economy in corrections canbe seen to have many advantages, but prior to theLNILC's study it was not known if this application'was feasible.

First Cycle Accomplishments

During Phase II the EMLC developed andimplemented the first token economy in an adult

9

correctional institution, housed in a separatecellblocic of the prison. The first cycle of this sti .ywas designed as a test of the feasibility of establishinga token economy within the institution and focusedon three general types of behaviors: vocationaltraining, basic education, and self-management skillsdevelopment. These behaviors v.:re identified,measured, and then treated by manipulating theirconsequences. The behavior mod:fieation proceduresapplied were found to be effective in inducing thedesired behaviorsbehavicts which are traditionallyinsured by threat of punishment and aversive control.Thus the token economy was demonstrated to be aviable alternative method of managing inmatebehavior. The first cycle also provided Ce foundatio,for the development of a model which could beexported for use in other correctional settings.

Objectives of the Second Cycle

Having demonstrated the feasibility ofestablishing a token economy in this setting, the Labproposed to expand the scope of the token economyin Phase III to include more behaviors which appearcritical to postrelease success. This expansion wouldprovide a basis for determining the practicality ofadopting the model on an institution-wide basis,employing the diverse resources of the institution nsback-up reinforcers. Three preliminary objectives forthe second cycle of the token economy werepostulated:

I. Work k,signments. The use of aversirecontrol procedures to stimulate inmates to completework assignments and the use of work assignmentsas punishment would be replaced by motivationaltechniques more consistent with those used in thefree world. It was proposed that inmates

participating in the token economy hold institutionaljobs and earn an !lowly "wage" (via the medium ofthe token economy). Data Mould be collected on anumber of work criteria (punctuality, productivity,etc.), and promotions and demotions (in terms ofboth responsibility and "pay") would be based uponwork performa-Ace. This procedure would provide acontinuity of oehavior from the institution to thecommunity and allow the inmate to experiencerecognition and encouragement for initiative anddiligence in work performance.

2. Basic Education. The structure of the tokeneconomy would be applied on a wide scale to themanagement and motivation of basic education.Inmates residing in the experimental cellblock andenrolled in the basic education training programsconducted by the EMLC would earnperformance-contingent "wages" (via the medium oftoken economy). The educational program would beself-paced, using the Individually Prescribed Instruc-tional (IPI) System developed by the Lab.

3. Expanding the Reinforcement Menu. Toincrease the exportability of the token economy, therange of back-up reinforcers would be expanded totake advantage of ',lore of the resources available inthe institution. It was proposed that all aspects ofinstitutional life which could serve as cost-freeconsequences of behavior be incorporated into thestructure of the token economy. Such privileges asimproved job assignments, reduced custody status,and admission to work-release programs would bepotentially powerful additions to the reinforcementmenu of the token economy. If direct control ofthese %ariables was possible, it would be exercised.Otherwise, letters of recommendation to thoseagencies controlling such reinforcers would be madeperformance- contingent.

10

4. Development of Self-Maintenan«- SkillsEssentially in teplication of an objective of the earliercycle of the study, behavior modification procedureswould be directed at certain skills deemed importantfor successful postrelease adjustment. Variousreinforcers would be made contingent upon thesebehaviors to determine the most effective means ofdeveloping such skills. The expansion of thereinforcement menu, mentioned above, would bedirectly linked to this objective.,

As the second cycle of the token economy studyprogressed, the degree of emphasis among the originalobjectives was adjusted, and certain entirely newobjectives were added. Among these was a brief studyin the effects of punitive control, an evaluation ofan alternative token economy attempted by theinstitution itself, and an extended study inengineering the token economy for success.

Operating the Token Economy

The second cycle of the token economy studycovers a period of nearly 400 days. More than 125inmates participated in the study for varying lengthsof time. The participants were randomly selectedfrom a pool of eligible inmatesthe criteria foreligibility being that they be first-time adult felons,under 30 years of age, physically able to work onthe prison farm, and eligible for release or admissionto vocational training or work release within threemonths of the termination of the cycle. The racialdistribution approximated that of the general prisonpopulationabout half black and half white.Additional subjects were assigned to the Unit by theprison administration. The experimental cellblock,called the Ecological Unit, housed a maximum of 40men, with an average population of about 30. Themen earned tokens (points) by engaging in

preselected activities, and exchanged these points forsuch reinforcers as items from the Unit store or timein recreational areas.

Behavior occurring on the Unit was a target forintervention in both fycles. The activittc., weredivided into two types considered to be necessarycomponents of successful functioning in the "freeworld's and/or the correctional institution: (1)self-maintenance skills, which include such things asmaintaining a neat personal appearance, bed making,keeping the immediate living area neat and clean, andcontributing to the maintenance of the commonliving area; and (2) educational behaviors, which wereactivities specifically related to learning, whetherreading a book or participating in a basic educationprogram.

In the first cycle intervention was limited toUnit activities, but the project design of the secondcycle was expanded to include activities off the Unit.The Unit residents divided their day between theUnit, the project's basic education component, andthe institution farm; their behavior was observed inall three places. The performance of self-maintenancetasks was reinforced on the Unit, educationalactivities were reinforced in the basic educationcomponent, and performance on institutional jobassignments was reinforced on the farm. Thus theinfluence of the token economy was expanded tomanage behavior occurring in the institution as wellas on the Unit.

In the first cycle a checkbook system wasusedthe points the men earned for socially adaptivpbehaviors were in a checking account, and the menwrote the -ks for the reinforcers they selected. Thesecond cycle-replaced the checkbook system with apunch card system. Each day a new card was issuedto t.--ch man. His card was punched as he earned

II

points and marked as he exchanged these points forvarious reinforcers. Any points remaining at the endof the day were transferred as "savings" to the nextday's card. This system provided immediate, tangiblereinforcement, simplified record keeping, and enabledthe men to determine their "balance" at any timeby looking at their cards.

Data collection for the second cycle was begunon January 22, 1972, on performance in theEcological Unit and the institution farm, and onMarch 21, 1972, in the basic education program.This was a baseline (pre-experimental) period priorto the experimental/treatment phase, whichcommenced July 10, 1972.

The Institution's Farm Credit System

As the projected starting date for theintervention phase of the token economy studyapproached, it was learned that the institution wasplanning a token-economy-like procedure to motivatefarm work performance. On eight occasions duringthe day, correctional officers supervising the farmsquads were to rate each inmate as to whether ornot he was working at that particular time; ifworking, he earned one institutional credit.Additionally, each inmate received one credit eachtime he arrived promptly at the back gate of theinstitution for the twice-daily checking out to thefarm assignments. Inmates working on the farm couldthus earn a maximum of 10 institutional credits eachday.

These institutional credits were to beexchangeable for certain reinforcers which areavailable in the institution and which had beendetermined to be meaningful to inmates. Forinstance, a total of 100 credits could be exchangedfor: (1) a day off from work, (2) a conference with

ra

the warden or classification officer, or (3) permissionto make a telephone call. These opportunities wouldagain be presented to the inmate following eachaccumulation of 100 credits. The third time heearned 100 credits, he might choose even moredesirable reinforcers: (1) a scheduled job change or(2) a meeting with the Custody Board if he wereeligible for a custody change. If he met with theCustody Board and had his security classificationreduced to "minimum," he would then be eligiblefor participation in the institution's work-release andhome furlough programs. Thus work performance onthe farm had the potential of earning the inmatesome very desirable rewards.

in the light of this unanticipated farm creditplan of the institution, it was decided to modify the

plans laid for the research strategy of this cycle ofthe 'project, postpone the introduction of theintervention phase, and first evaluate the effectivenessof the institutional farm credit system. The revisedexperimental design, which employed a multiplebaseline technique to make this evaluation, is shownin Chart 1.

As can be seen from this chart, inmates in theEcology Project had the potential of earning 12institutional credits per day: 4 on the farm (no creditswere awarded to Unit residents for promptness at theback gate), 4 on the Unit, and 4 in the basiceducation program. Unit residents' institutional creditaccounts were kept separately from their EMLCtoken economy accounts.

CHART 1Experimental Design for Evaluation of Effectiveness

of Institutional Farm Credit System

Phases

Areas of InmatePerformance

Being Observed

Baseline

Phase (beganMarch 22,

1972)

Phase [ (began May 22, 1972)Start institutional point system

and continue for durationof project. Phase II (began June 5, 1972) Phase III (began June 19, 1972)

Farm No creditsawarded

4 credits contingent uponon-task behavior*

4 credits contingent uponon-task behavior (nochange in contingencies)

4 credits contingent uponon-task behavior (nochange in contingencies)

Ecological Unit No creditsawarded

4 credits awarded non-contingently

3 credits contingent uponcompletion of conveniencebehaviors and I credit con-tingent upon Unit mainte-nance behaviors (change incontingencies)

3 credits contingent uponcompletion of conveniencebehaviors and I credit con-tingent upon Unit mainte-nance behaviors (no changein continzencies)

Basic EducationProgram

No creditsawarded

4 credits awarded non-contingently

4 credits awarded non-contingently (no change incontingencies)

4 credits contingent uponon-task behavior (changein contingencies)

The credits arc awarded on a per man. per day basis.

12

1

In order to make an evaluative comparison ofthe token economy as administered by the institutionwith the one operated by the Lab, the same multiplebaseline technique was used to implement the EMLCtoken- economy in each of the three observationareas. The FMLC points were exchangeable for avariety of back-up reinforcers which were availableon the Unit. For example, inmates were charged onepoint for every 100 minutes spent in the variousreinforcing event rooms, and one point for every 50minutes spent off the Unit. In the token economy"store," each point was worth about 5 cents, witha cup of coffee (the least expensive item) selling for

one point and a pack of cigarettes (the mostexpensive item) selling for nine points. A maximumof 8 points per day could be earned on the farm,8 on the Unit and 8 in the basic education program.

The institution credits were applied sequentiallyto the farm work, performance on the Unit, andperformance in the basic education area. Only on thefarm did a clear effect appearthe proportion of timethe men spent working at the assigned task declinedsteadily throughout the time the credits were awardedcontingent upon the men remaining at their task.

In contrast, when the EMLC points wereintroduced, again sequentially (see Chart 2), to the

CHART 2Research Design of the EMLC Token Economy Project

Phases

Areas of InmatePerformance

Being Observed

Phase I (began July 10,1972) EMLC TokenEconomy began and

institutional pointsystem continued.

Phase II (began

August 7, 1972)Phase III (began Sep-tember 5, 1972)

Phase IV (began Oc-tober 2, 1972)

Farm EMLC tokens con-tingent upon on-taskbelidvior

EMLC tokens con-tingent upon on-taskbehavior (no changein contingencies)

EMLC tokens con-tingent upon on-taskbehavior (no changein contingencies)

LMLC Token Econ-omy and institu-tional point systemcontinue; projects de-signed to deal withspecial problems be-gin

FLulogcal Unit EMLC tokens award-ed non-contingently

EMLC tokens con-tingent upon com=pletion of conve-

nience behaviors and

Unit maintenance be-haviors (change in

contingencies)

EMI C tokens con-.

tingent upon coin-pletion of conve-

nience behaviors andUnit maintenance be-haviors (no changein contingencies)

13.isic Fducation

Program

EMLC tokens award=

ed non-contingentlyEMLC tokens award-ed non-contingently(no change in con-tingencies)

EMLC tokens con-

tingent upon on-taskbehavior (change incontingencies)

13

farm work, performance on the Unit, andperformance in the basic education area, an entirelydifferent trend developed. Performance on the farmreturned to its original, relatively high level;performance of the Unit self-maintenance tasksexhibited a striking increase; and performance in thebasic education area, already at a fairly high level,also increased somewhat.

Perhaps the marked difference in the effect ofthe two systems results from the choice of potentialreinforcers, as there are others available within theinstitution, e.g., opportunity to participate in prisoneducation and training programs. Individualizing thepotential reinforcers by providing a reinforcementII menu if is also a possibility.

More likely, though, the less-than-systematicapproach of the farm credit system is largelyresponsible for its ineffectiveness. Project staffobserved many instances of the failure of thecorrectional officers to follow the procedures whichhad been established for awarding points, indicatingan unfamiliarity with the procedures or anunawareness of the importance of following theseprocedures. Peihaps the planning for the farm creditsystem did not include enough provPsion for stafftraining and supervision. The apparent success of theEcology Project may be due, in, large part, to itssystematic approach and a high level of staffcommitment.

Contingency Management in Basic Education

The contingency management sub-studyprovided the basic education component of the tokeneconomy, seeking to determine the most efficientmeans of motivating study behavior. The principalmeasures of this behavior were: (1) on-taskbehavior--the percentage of times the student was

14

observed in study position, (2) efficiencyquotient- -the ratio of tests passed divided by teststaken, and (3) learning ratethe actual time spentstudying a module of work divided by the timeestimated for that module by the IPI System. Thefirst two of these are investigated as dependentvariables; the learning rate serves as an indicator ofa continued stable state.

When the points available in the token economywere made contingent upon the student remainingin study position, the on-task measure (thoughalready at a relatively high rate) increased. When thecontingency was removed, the rate returned toapproximately its original level. When the points weremade contingent upon passing the module tests thefirst time they were taken, the number of tests passedon first attempt increased somewhat, and the increasewas reversed when the contingency was removed.

The average number of points earned each dayby the students in the basic education programdropped severely during the phase of the studyduring which earnings were contingent upon passingposttests the first time taken., None of the learningmeasures suffered. In terms of efficiency (response/cost), therefore, the more desirable of the twocontingencies was payoff on the basis of tests passedrather than remaining on-task in the study area. Thestudents learned just as well at a decreased cost to thetoken economy.

Self-Management Skills

As in the first cycle of the token economystudy, certain self-management skills were designatedas target behaviors in the motivation and managementsystem, The skills included making the bed, keepingthe immediate living area neat and clean, keepingone's personal appearance presentable, andvolunteering for and completing a Unit clean-up job.

The procedure for awarding points for completionof these tasks was designed to provide immediatefeedback to the resident. Each morning, beforeleaving the Unit for his farm assignment or for thebasic education area, the resident would complete thespecified tasks and report their completion to thestaff member on duty. The staff member wouldcheck tie job, praise the resident for the tasks whichwere completed, and punch the appropriate numberof points onto the resident's card (two points foreach of the four self-management skills areas). TheEMLC point system clearly demonstrated its abilityto motivate behavior change, It also served themanagement function of a Olken economy--theEcological Unit was considered by the institutionaladministration to be of exemplary cleanliness.

Expanding the Reinforcement Menu

During this cycle of the token economy study,an in-depth analysis of the available back-uprcinforcers within this particular setting wasconducted. The reinforcers provided by the Unitstore had been shown to be effective, but theexportability of the token economy model would beincreased if the resources of the institution wereavailable for use as back-up reinforcers. Since theenvironmental contingencies which appeared mostdesired by the inmates were controlled by therelatively autonomous correctional system, the mostinfluential route available to the token economy staffwas that of providing performance-contingent lettersof recommendation. Whenever a participant in thetoken economy would request a letter ofrecommendation from the staff, a statement aboutthe quality of his performance in each of the threeobseryatioriareason the farm, on the Unit, and inthe education areawould be prepared. This conceptis in a preliminary developmental stage, and an

15

evaluation of its effectiveness is not now available.However, it is clear that a great number of potentialback-up reinforcers are available in the institution,either unused or applied in an indiscriminate manner,with little or no relation to behavior change.

Evaluating Punitive Control

Early in the baseline period of the tokeneconomy study, project staff had the opportunity toanalyze the effectiveness of the institution'straditional approach to problem behavior, that is,punitive control. A lassaiz-faire approach to theself-maintenance behaviors was taken by the tokeneconomy staff during the baseline period, and it wassoon clear that the base rate for making , beds,cleaning the living areas, etc. was quite low with thispopulation. The prison administration wasparticularly concerned about the unmade beds anddecided that the residents would be forced to beginmaking their beds. A correctional officer assigned tothe Unit for security purposes was instructed to"write up" a disciplinary report on anyone who failedto make his bed. If the inmate received a seconddisciplinary report, he would be transferred toanother institution to be placed in solitaryconfinement on a restricted diet foi a maximum of21 days.

The results were immediatepractically all bedswere made on the days this particular correctionalofficer was assigned to work on the Unit. However,a look at other measures of the inmates' behaviorreveals what appear to be possible "side effects" ofthis control procedure. During the period of punitivecontrol, the percentage of men volunteering for andcompleting Unit maintenance tasks decreased. Also,the number of disruptive incidents (fights, verbalabuse, etc.) increased. After termination of thethreatening situation (the officer formerly assigned

to the Unit was reassigned), these measures returnedto their baseline rates.

This is not, of course, a definitive study of theeffects of punitive control, but it is an example.Voluminous reports of the inadequacies and sideeffects of aversive control in a wide range of settingsare available. However, in this situation, thetechnique was seized upon as the immediate solutionto the problem. Indeed, the correctional officer inquestion doubtlessly considered it to be successful,since he had made no provision to observe its possibleside effects, and was not present to observe thedeterioration in the bed-making behavior immediatelyupon the removal of the threat. In a sense, thecorrectional officer was reinforced for his use ofaversive control: both by positive reinforcement (theimmediate increase in the number of beds made) andnegative reinforcement (termination of the wrath ofthe prison administration).

Engineering the Economy for Success

A key to the successful implementation of abehavior treatment regimen like the token economyis commihilent to maintaining the power andprecision of the system, even though this mayincrease the administrative work involved. A situationmight arise in which a decision about procedurescould lii.ten the administrative burden but wouldalso reduce the effectiveness of the treatment.Experience and experimental evidence dictate thatadherence to the principles of behavior must be thefirst concern if the mechanics of the interventionprocedure are to achieve the objectives of theprogram.

A ease in point would be the solution to a designweakness encountered in the second cycle of thetoken economy, The original procedure for paying

16

for the use of the reinforcing event rooms (television,Ping-Pong, and pool) and spending time off the Unitproved to be inadequate. Many residents were findingthat when the points used in the reinforcing eventareas and off the Unit were totaled, they had usedmore points than they had earned. This was the resultof a record-keeping procedure in which the men,turned in time cards showing their use of thereinforcing event areas only when the cards were full,rather than each day: Men who were "overdrawn"were prohibited from using the token store until their"debts" were paid.

A change in the procedure then required theresidents to turn in the time cards each day beforethe store opened, whether these cards were filled ornot. This resulted in considerably more work for theproject staff, but the number of men prohibited fromusing the token store diminished to near zero.

Other administrative problems are alsosusceptible to this sort of a functional analysis, andfrom these controlled manipulations, valuableinformation concerning the establishment of a tokeneconomy in other diverse settings may be generated,

Summary of Progress

The token economy study should be evaluatedalong four criteria: effectiveness, efficiency,generality and utilization, and stimulation ofcontinued research. The following summarizesprogress along these lines:

(1) Both the immediate and the long-termdimensions were considered in evaluating this study.The immediate effect is clear. Working with behaviorswhich are generally accepted as critical to postreleasesuccesswork performance, educational advancement,and self-management skills developmentnotableimprovements in behavior within the institution were

generated. The long-term effectiveness will bedetermined by the Lab's follow-up studies, which willshow whether or not these behavioral changesgeneralize to the "free" environment and assist theparticipants in obtaining and holding jobs, taking partin productive social relations, and refraining from theactivities which precipitated their incarceration.

(2) A major emphasis of the token economystudy was to seek out and utilize those cost-freereinforcers which were already available in theinstitution, These are quite often the most powerfulconsequences of behavior, and quite often the leastused. A preliminary analysis of such reinforcers wasmade which should aid greatly in increasing theefficiency of the token economy.. In addition, it wasdemonstrated that when such back-up reinforcers asa token store are used, the economy of operationis often determined, in large part, by the precisionof the contingency management techniquesemployed.

(3) The management/motivational systemdeveloped in this study was designed in such a waythat it might be utilized in a wide range of situations.The final technical report will provide more detailsof the organization of the token economy, along witha practical list of the "do's and don'ts" of operatinga token economy in a correctional institution. Theattempt of the Draper administration to implementthe system was faulty, but encouraging. It was clearthat the prison administration recognized the powerof the system; unfortunately, they did not have aguide to the development of a token economy, onlyrespect for accomplishments they had observed inanother token economy.

(4) Perhaps the most impressive outcome of thepresent study is its indication of positive researchavenues. Principal among these are the development

17

of an institution-wide token economy; the harnessingOf powerful no-cost reinforcers such as parole, workrelease, and custody reductions; the application ofthese techniques to juvenile offenders before theybecome embittered and stigmatized by imprisonment;

and community-based intervention programs whichcould use the token economy model to parallel andcorrelate with institutional programs.

In summary, the present study has shown effectsnow and has pointed the way toward future increasedeffectiveness; it has begun analyzing the reinforcerswhich are available in the correctional institutions;it has provided the basis for the development of anexportable "how-to" model; and it has suggestedpromising directions for further research.

Products

Products generated by the token economy andits basic education component, the contingencymanagement study, include the following:

Token economy technical report. Gives detailsof the study along with a list of practicalconsiderations in operating a token economy ina corrrectional institution.An ecological experiment in corrections: Aprogrammed environment for behaviormodification. Describes the design of the tokeneconomy operated by the EMLC as a steptoward merging the objectives of both custodyand treatment under one set of behavioralcontrol techniques.Contingency management in a correctionalinstitution. Reports on the EMLC's use ofcontingency management primarily in basiceducation and the token economy study, butalso as part of the content of the instruction

provided in the correctional officer trainingproject.

The use of contingency management to affectlearning performance in adult institutionalizedoffenders, Discusses the use of contingencymanagement techniques in basic education andvocational training classes at Draper. One of thesix studies reported compares results obtainedby contingency contracting with inmates andwith academically deficient freshman nursingstudents.Imprisoned resources-- Innovative techniques ineducating prison inmates. Describes the designof a study which used educationally advancedinmates as one-to-one basic education tutors forinmate students, presenting implications forimitate educational programs.

18

Correctional officers can change inmate behavior; the first step is training.

coneetionalofficertwining(0 T)

Previous Training Cycles

The influence of the correctional officer oninmates has generally been overlooked in treatmentplans for the institution. Often he lacks the skillsnecessary to capitalize upon his potential role as aneffective change agent working with the treatmentteam. Given the appropriate training and supervision,however, correctional officers should be able toeffectively operate rehabilitation programs, forexample, an institutional token economy, Todetermine the best way to aid the officers inacquiring the necessary skills, the EMLC hasconducted three cycles of Correctional OfficerTraining (COT), teaching selected officers theprinciples of behavior modification.

Two cycles of training were completed in PhaseII of the Lab. All the officers in the institution hadindicated an interest in the training and were thusconsidered volunteers. From these men, a controlgroup of 15 men and two 15-man training groupswere then selected at random. Additionally, twoofficers from other Alabama correctional institutionsjoined the second cycle training group. The officertrainees spent several weeks in the classroom portionof the training before beginning the practicumexercises, which were conducted by the trainees inon-the-job situations arising from their institutionalassignments. Considering the time spent informallyin the practicum assignments, each man spentopproximately 90 hours in the training program.These first two cycles of training are described inan interim report which was prepared in May, 1971.

Third Cycle Preparations

Preparation for the third cycle of COT begannear the end of Phase II, when a series of bookletswas planned which would present the classroomcontent of the training in a self-instructional,

21

illustrated format. Considerable effort was devotedto developing an appropriate curriculum which wouldincorporate both textbook materials and knowledgegained from experience in the previous cycles oftraining. Work on the booklets carried over intoPhase III; ten booklets dealing with the history ofcorrections and the principles and techniques ofbehavior modification were completed for tryout inthe third cycle training.

Subject selection was also completed early inPhase III by a committee composed of the warden,classification officer, prison psychologist, COTproject director, and two shift commanders (fellowofficers) who had previously completed training. Thetrained officers were able to use their trainingexperience in selecting the officers most likely tobenefit from the training. The random selection ofthe earlier trainees had failed to take into accountthe amount of inmate contact each man had, animportant consideration in making the trainingmaximally effective. For the third cycle training, menwhJ had such institutional assignments as tower dutyand truck driving were eliminated in order to includeother officers whose assignments allowed them morecontact with the inmates. Twelve experimental andten control subjects were chosen.

Pretraining assessment measures were taken todetermine the characteristics of the traineepopulation. Measures which were to be administeredonly once, prior to training, included an IQ test, twoachievement tests, and a demographic data gatheringinstrument. Some of the officers appeared asschOuled and took the IQ and achievement tests,but several officers were so apprehensive about thetests that they refused to take them. Since the EMLCstaff believed that enough educational data had beenobtained from the other cycles of training to establisha suitable level of approach for the training, the plansfor IQ and achievement testing were discarded rather

than risk jeopai dizing the relationship between theofficers and the project staff by forcing the matterof testing.

Information concerning' the officers'relationships with the inmates was also collected priorto training to provide a basis for posttrainingassessement. The officers' interactions with inmateswere recorded prior to training by two observersusing the Behavioral Observation Index (BOI), achecklist instrument used to collect data on thefrequency and kind of officer-inmate interactions.Pretraining assessment also included the use of the"M" technique in which a randomly selected groupof 30 inmates ranked the correctional officer traineesalong the dimensions of general caliber, fairness,concern about inmates' welfare, and punitiveness.

Characteristics of the Third CycleOfficer Trainees

Demographic characteristics for the officersparticipating in the third cycle training were foundto be as follows: trainees and controls ranged, inrank, from cntry-level officer to shift lieutenant (aposition next in rank to the captain of the guards).Ninety-five percent of the officers had lived thegreater portion of their lives in Alabama; 53% inElmore County (the location of Draper) and 47% inbordering counties. most of which, like ElmoreCounty, are agrarian oriented.

Officers ranged in age from 23 to 67 years, witha median age of 50. Ninety-five percent of theofficers were married.. and one was single. Theiryearly incomes ranged from about $4,500 to $6,000.The reported educational levels of the correctionalofficers ranged from completion of the seventh gradeto college graduate level. The median grade level was11.

Training Progress

The actual classroom training of correctionalofficers began on May 31, 1972, after an addresspresented earlier by the Commissioner of theAlabama Board of Corrections. The classroomtraining lasted approximately five weeks, ending onJuly.; 5. Twelve men wcrc to have been trained, butattrition for several reasons reduced this number to10. Nine officers completed all ten of theinstructional booklets; the remaining officercompleted six.

The schedule of classroom instruction requiredeach officer to complete two instructional bookletsper week (one hour allowed for each booklet), todemonstrate that he had gone through each bookletsystematically rather than randomly, and to attenda two-hour discussion session each week to commenton the materials being presented and any difficultieswhich might have arisen. Most trainees needed onlyan average of 30 minutes to complete a bookletrather than the full hour allotted, but the discussionsessions generally lasted the entire two hours. Eachofficer, as part of an incentive program unique tothis cycle of training, could earn a maximum of $5per week if he complied with the training schedule.

Each officer trainee took a pretest (called abaseline check) prior to beginning each booklet. Aftercompleting the booklet, he took the same test as aposttest (called a progress check). The word "test"was purposely avoided because of the anxietyexhibited earlier by the officers in regard to the IQand achievement tests. Score comparison on thesetests provided part of the training assessment; theoverall gain from pre- to posttest on all ten bookletswas 30.2%.

The classroom training was to have beenfollowed by a four-month period during which eachofficer trainee would conduct a practicum exercise

in the institution or on the farsn. These exercises wereplanned as a means of maximizing the effects oftraining in an on-the-job situation while aiding in theevaluation of the officers' ability to apply,. behaviormodification techniques in the correctional setting.The practicums would also demonstrate that what theofficers learned could indeed have a positive effectupon inmate behavior.

However, unforeseen changes in the institutionadministration and a shift in institutional concernsoccurred near the end of the classroom portion ofthe training. Manpower in the Alabama correctionssystem, and especially at Draper, dropped to a criticallow, and the officer trainees were needed to mancritical security posts or to augment the farm crews.The final stages of COT were halted, and the officerswere not allowed to begin their practicum exercisesor the additional booklets which had been prepared.The additional booklets dealt with graphing andcontracting for behavior, information which wasnecessary for the practicum exercises.

In late November, 1972, the Alabama Board ofCorrections requested that the Lab evaluate the stepswhich would be required to complete the training.A plan was submitted to the Board in earlyDecember, 1972, which would involve the selectionof 5 of the original 10 trainees to receive furtherclassroom training and conduct practicum exercises.At this time no specific arrangements for trainingcompletion have been made, pending the approval ofthe Draper administration.

Final Report Prepared

Results of the third cycle of COT are containedin a final report which describes the experiences andevaluation outcomes for all three cycles of training.The first two cycles were evaluated on the basis ofthe practicum exercises and demonstrated that the

23

officers were capable of applying behaviormodification techniques. Ad -ally, severalofficers from the first two cycles fare seen by staffobservers and inmates as having changed after trainingin the direction of more frequent and more positiveinteractions with inmates. For example, when theinmate evaluation group ranked officers acrossselected measures of effectiveness, 16 of the 28trained officers (57%) were mentioned"spontaneously" as contrasted with 18 of the 101officers who had not received training (18%). Ofthese 16 trained officers, 10 (63%) were seen by theinmate group as increasing in their overalleffectiveness in dealing with the inmate population.The available pre- and post-training data for the thirdcycle were analyzed, but this cycle of training wasof necessity evaluated differently, since there wereno practicums.

In addition _to reporting the evaluationoutcomes, the COT final report pinpoints specificproblem areas in implementing a behaviorallyoriented training program. Several solutions to theseproblems are offered as suggestions for institutionsand agencies interested in conducting such training.

The Training Package

The number of booklets in the training packagehad originally been set at 10, and thew booklets wereprepared for the classroom portion of the training.However, as the training progressed, the plans wererevised to include more of the training content inthe booklets. This would be a decided advantage inimplementing such training in other correctionalinstitutions. The total number of booklets is now setat 20, plus an instructor's guide which will give adetailed account of how to implement the COTprogram. Each of the booklets is also illustrated forgreater learning involvement.

MI 20 of the booklets have been written. 15have been printed, and the instructor's guide is beingdrafted. Revisions have begun on the first 10

booklets, based on questions and comments notedin the tryout.

Products

Several products have been prepared inconnection with the Lab's experiences with COT.These include the following:

Correctional officer training in behaviormodification: An interim report. Describes the firsttwo cycles of COT; includes examples and results ofspecific projects dealing with inmate behavior.

A behavioral observation index (801) designedto evaluate training of correctional officers in a prisonsetting. Describes an instrument designed toobjectively evaluate the interaction with inmates ofcorrectional officers who have participated in COT.Includes instructions for its use and an explanationof behavioral principles.

Training correctional officers in behaviormanagement techniques: Methods, application, andeffect. Reports on the fist two cycles of COT andthe then-present third cycle, emphasizing the specialfeatures of the third cycle training.

Correctional officer training package. Designedto provide correctional officers with a backgroun 1in the history of corrections and a workingknowledge of behavior modification principles andtechniques, including graphing and contracting forbehavior. Twenty self-instructional, illustratedbooklets with an instructor's guide.

COT final report. Reports on the three cyclesof correctional officer training conducted by theEMLC at Draper, using two methods of instruction.Evaluates the training by using several measures,including practicum exercises for the first two cycles.

24

/ /.

/

^

The Need for Follow-Up

The ultimate measure of the effectiveness of aninstitutional treatment program is the behavior of thereleased offender--do,:s his postrelease behaviorindicate that he is indeed rehabilitated? Long-rangefollow-up of the released offender is necessary toanswer this and other questions. What combinationof events and behaviors, both in the institution andthe community, differentiates the ex-offender whomakes a successful societal adjustment from the onewho returns to prison? Which institutionalintervention and treatment programs are the mosteffective? What approach and what instruments canbe used to best collect the data to answer thesequestions?

Despite the obvious need for information on theex-offender's activities in the community, littlesystematic research has been conducted with regardto postrelease behavior. One reason for this is aconfusion about what to measure and how. TheEMLC uses a behavioral approach in its follow-upstudies, seeking to de-ielop and apply a methodologyfor measuring and assessing the behavior patterns ofreleased offenders, as well as evaluating and validatinginstitutional intervention programs. -To measurebehavior, instruments are being developed and refinedwhich are predictive of the presence or absence oflaw violation, criminal behavior, and recidivism., Inthe process of developing these instruments, effortsare made to identify postrelcase behaviors leading torecidivism and those conducive to successful societaladjustment. Additionally, it has been necessary toanalyze the criterion of law violation and criminalbehavior into its functional components. At the sametime, identifying the factors involved in recidivismand criminal behavior provides specific bases forfuture intervention, treatment, and retrainingprograms, both in the institution and thz community,

27

by indicating the behavioral areas in which these areneeded.

'69 Follow-Up

The first follow-up study conducted by the Lab,the '69 Follow-Up, was the initial step in thedevelopment of a systematic follow-up procedure.This study had three main objectives: (1) to establisha behavioral demography of the released offender; (2)to evaluate institutional MDTA training; and (3) todevelop and refine instruments which would predictlaw violation, criminal behavior, and recidivism. Datawere collected through face-to-face interviews withthe study subjects, using an interview guide and otherfollow-up instruments.

The total number of releasees in the study was173: 106 MDT trainees and 67 non-trainee controlswho had applied for training. These men, locatedwithin a 200-mile radius of Draper, were interviewedprior to release and at 3-, 6-, 12-, and 18-monthintervals after release. A small group of the studysubjects, both MDT trainees and non-trainee controls,had settled within a 50-mile radius of Draper andparticipated in a more intensive part of the study;these men were interviewed once weekly over aperiod of 18 months.

Data analysis and interpretation were completedlate in 1972, and a draft of the technical report wasissued in January, 1973. The study found thatemployment was one of the most significant factorsin postrelease adjustment. A released offender ismuch less likely to commit a law violation if he isemployed and somewhat more likely to be employedif he has had institutional MDT training. While theeffects of such training are small, there is a tendencyfor MDT trainees to work more and earn more moneythan men who did not receive training, at least inthe first six months after release. The study alsodeveloped a systematic method of data collection and

record keeping which included the use of instrumentspredictive of law violation and recidivism.

Design Changes in the '71 Follow-Up

As a result of experience gained in developingthe methodology of the '69 study, the '71 Follow-Upincorporated several changes in the study design toallow for more efficient use of staff time. Forexample, the geographic area covered by the '71Follow-Up was limited to a 50-mile radius ofMontgomery and Birmingham, two areas whereDraper's releasees usually locate. This eliminatedmuch of the time spent by the follow-up team indriving through rural areas 4.1 locate a few subjects.The subjects who settled in the study area wereadministered three interviews (prerelease, 3-6 month,and 12-15 month) rather than the five interviewspreviously used. Additionally, the postreleaseinterview guide was condensed in December, 1971,to structure a shorter and more systematic interview.

In the more intensive level of the study, somesubjects were contacted and interviewed monthly inaddition to the three standard interviews. The areafor this portion of the study was reduced to a 25-mileradius of Montgomery., A procedure was institutedin January, 1972, in which these subjects were paid52 for reporting to the downtown office for eachof their monthly interviews. This change resulted infurther reduction of travel expense as well as savingstaff time, allowing the interviewers to concentrateon the more reluctant subjects.

To compare the effects of the various treatmentprograms, the '71 Follow-Up expanded thegroups of the '69 study (MDT and non-MDT) to fivegroups. Three of these were experimentally treated:MDT, ecological, and combination MDT-ecological.These were compared to a group of subjects whoreceived training at the J. F. Ingram Trade School,located at nearby Frank Lee Youth Center, and to

28

another group which received no training orexperimental treatment.

Subjects

All of the 142 subjects in the '71 Follow-Upwere Draper inmates who were released or paroledbetween October, 1970, and January, 1972. The menwere quite typical of the Draper and Alabama prisonpopulation in terms of demographic characteristics.The overall mean age of the study subjects was 25,58% were black, and 18% were married. They re-ported educational levels which averaged 9.4 years.Fifty-six percent of the men were serving time forcrime against property; 44% were recidivists. Specificdemographic data for each group are shown inTable 1.

TABLE 1

Comparison of Basic Demographic Datafor the Five Study Groups

DemographicData

Groups

MDTN = 58

EcologyN = 13

MDT &EcologyN = 16

StateTradeN = 20

ControlN = 35

Percent black 55 31 63 50 74

Percent married 17 31 19 20 14

Percent priorfelonies 50 31 56 40 34

Percent paroled 69 54 88 90 63

Percent whosecrimes wereagainst property 66 62 38 50 49

AgeMean 25 27 26 21 25

Median 23 23 25 23 23

Range 17-46 20.54 20-34 1942 1947Educational level(reported bysubject)Mean 9.3 7.1 9.8 9.6 10.3Median 9.0 6.0 9.5 10.0 10.0Range 5-12 1-12 6.12 2-14 6-13

Follow-Up Instruments and Measures

Experience gained in the '69 Follow-Up studyindicated a need for refinement of existing behavioralassessment instruments and development of others tocover additional aspects of behavior. The two primaryinstruments used in both follow-up studies were theEnvironmental Deprivation Scale (EDS) and theMaladaptive Behavior Record (MBR), The EDS is a16-item scale which focuses on environmental inputfrom a variety of areas such as occupation,organizations, and interpersonal relationships; theMBR, which also contains 16 items, deals withbehavioral output in terms of maladaptive responses.Both have proven to be highly predictive of lawviolation and recidivism. The revised postreleaseinterview guide (IG) was also use& It contains over100 items dealing with work experience, societaladjustment, criminal behavior, finances, familyaffairs, housing, and public acceptance.

Other behavioral assessment instruments weredeveloped and refined during EMLC Phase III. TheWeekly Activity Record (WAR) measures"free-world" behavior in terms of the wad theex-offender allocates his time on a weekly basis. Thisinstrument was introduced late in the '69 study andindicated some differences in the way recidivists andnon-recidivists spend their time. Another instrument,the Behavioral Incident Inventory (BII), focused onsignificant persons in the early life and developmentalbehavioral sequence of the individual, The need wasalso recognized for an overall diagnostic instrumentwhich would delineate areas of potential strengthwhile pinpointing areas of behavioral deficit andexcess in which intervention and retraining wererequired. Some initial planning was made for thedevelopment of such an instrument.

In response to the many inquiries about theEMLC's follow-up instruments, a series of manuals

29

for their use was begun. The first of these, a manualfor the use of the EDS, was completed in October,1972. The MBR manual and a behavioral interviewguide were drafted and will be completed during thenext phase of the Lab. Technical reports on thevarious instruments are also planned.

In analyzing the study data, the need becameapparent for some means of categorizing lawviolation. The method adopted was the use of acontinuum of degrees of law violation, with severitybeing reflected by the length of the sentenceimposed. The continuum is divided into three maincategories: no law violations, minor law violations(misdemeanors), and major law violations (felonies,i.e., recidivism). These categories have proven usefulin the Lab's follow-up studies, but the continuumis being examined for possible revision andrefinement.

Data Collection

All data were collected in behavioral interviewsand were voluntarily contributed by the subjects.Interviews were conducted at the subjects' homes,places of employment, or other convenient locations,e.g., restaurants, and lasted 60-90 minutes. Subjectsliving in the Montgomery area were often interviewedat the Lab's downtown office there. Payment for the3-6 and 12-15 month interviews was $5; for thosesubjects in the intensive study area, payment was $2for each additional monthly interview conducted atthe downtown office.

The intensive interviewing was terminatedAugust 15, 1972, and the downtown office wasclosed. The cut-off date for all other interviewing wasJanuary 1, 1973. As of January 1, 1973, theinterviews given totaled: 142 prerelease interviews,123 interviews at the 3-6 month interval, 89interviews at 12-15 months, and 280 monthlyinterviews.

Preliminary Findings and Early Trends of the control group are similar to those of thetreated groups.The data from this study have been keypunched

for computer analysis and some initial analysis hasbeen done. Computer analysis will be completed inPhase IV and a final report prepared at that time.The present findings should be consideredpreliminary, for these are subject to change a-analysis continues.

All of the 142 subjects eligible are included inthis data discussion. Subjects are considered ineligiblewhen they have absconded, recidivated, died, or leftthe study area. Those who return to the area are thenconsidered eligible again. Another limitation oneligibility is that some subjects may not have beenreleased long enough to be interviewed at thespecified time intervals. Thus the Ns in the followingdiscussion may vary somewhat as a result of subjecteligibility.

One of the main concerns of the 1971Follow-Up was the effect of treatment on postreleaselaw-violating behavior. A cumulative record of lawenforcement encounters was kept, and allinformation was verified by checking with theAlabama State Board of Corrections and local city,county, state, and federal law enforcement agencies.Disposition of cases was verified through therespective courts and records. Table 2 shows the lawviolation status of the study groups as of January 1,1973.

The data in Table 2 indicate that the ecologicalgroup tends to be the lowest in major law violationsalthough high in minor violations. The N for thisgroup is small, however, and many of these men havenot been released as long as the other subjects. MDTtraining and the combined MDT-ecological treatmentappear to be least effective in terms of all lawenforcement encounters, but none of the trends arestatistically significant. The law violation percentages

30

TABLE 2

Cumulative Percentage of Law Violationand Encounters as of January I, 1973*

Groups

Law EnforcementEncounter

Major law violators(recidivists) (in-cludes absconding)

Major and minorlaw violatorscombined

All law enforcement encounters(includes aboveplus pick-up onsuspicion, chargesdropped, etc.)

MDTN = 54

EcologyN = 11

MDT &EcologyN = 13

StateTradeN = 19

ControlN = 33

50% 18% 46% 32% 36%

65% 64% 54% 53% 52%

81% 73% 77% 63% 64%

*Subjects had been released for 11 to 26 months.

Employment was another concern in this study,and all of the follow-up instruments contained itemsdirectly reflecting this concern. Employment statusat 12 months for all eligible subjects is shown inTable 3. Men attending school more than half-timewere not included in the percentage data. Thecombined MDT-ecological group has the highestpercentage employed (100%); 33% of these wereworking on training-related jobs. The state tradeschool group followed with 82% employed, of whom33% were working on training-related jobs. Only 17%of those employed in the MDT group had jobs relatedto their training. Because the Ns in this table aresmall, the data should be interpreted with caution.

FABLE 3

Percent of &Nei% Employed at the Endof their First Year in the Free World

Groups Percent Employed

MDT (N = 39) 62

Ecology (N = 6) 67

MDT & 1.colovy(N = 7) 100

State Trade (N 12) 82

Control (N = 25) 67

TABLE 4

Employment Data Collected on Subjectsat their 12.15 Month Interview

Items

Groups

MDTN=39

EcologyN=6

MDT &Ecology

N = 7

StateTradeN = 12

ControlN = 25

Total incomeMean $3,255 $3,923 S4,097 55,039 53,825Median $3,800 53,943 54,610 $5,776 $3,500Range SO- 51,390- $2,076- 5654- $300-

10,000 6,532 5,976 8,300 11,000

Amount savedMean S78 $207 5240 $145 $312Median SO 550 $30 SO $0Range $0-700 $0-1,000 S0-750 $0.750 $O -5,000

Amount of debtMean $529 5769 $1,164 $930 $1,019Median $250 S35 5120 $400 S310Range S0-3,000 $04,000 S0-5,525 $0-3,100 $0-9,475

Percent of timeemployed fulltime

Mean 65 78 80 77 76Median 75 93 98 88 86Range 0 100 0.100 25.100 20-100 0.100

Number of full-time jobs held

Mean 2.5 2.8 2.7 2.6 2.4Median 20 2.5 2.0 2.0 2.0Ranee 0-6 0 0-7,0 1-6.0 1-5.0 1-6.0

31

The particular details of employment werecollected with the postrelease interview guide andincluded related aspects, such as money management.Table 4 presents data collected at the 12-15 monthinterview for the five study groups. Again, the Nsfor some groups are relatively small. It should benoted that the first item represents total income forthe period rather than salary.

TABLE S

Employment Data Collected on MDT Traineesalter Release 12 -IS Months Based

on Vocational Trade Area

Items

Vocational Area

ButchersN= 14

BarbersN = 6

Welders

N = 18

RefrigerationRepairN = 8

Total incomeMeanMedianRange

$2,714$2,838

SO-5,440

$3,749$3,286

$O -7,000

$3,642$4,076$276-

$3,700$3,470$2,076-

10,000 5,976

Amount savedMean $10 $213 $75 $245Median SO $30 $0 $6Range 50-90 S0 -700 SO-500 50.750

Amount of debtMean 5332 $846 $494 $1,271Median 5212 $120 5100 $600Range $0-1,745 SO 2,200 SO-3,000 50.5,525

Percent of timeemployed fulltime

Mean 59 76 70 69Median 52 87 75 64Range 0-100 0-100 6.100 25.100

Total number offull-time jobsheld

Mean 2.2 2.8 2.5 29Median 2.0 25 2.0 2.0Range 0-5 0-6 1-6 2.5

The MDT group was also examined in terms ofthe specific vocational training areas, as shown inTable 5. The data indicate that butchers wereemployed less than any other group, with barbersbeing employed the largest percentage of time. Thegroups who saved the most (barbers and refrigerationrepair) also had the largest debts. None of the trendsin Tables 4 and 5 are statistically significant,however.

Other data and trends may be summarized asfollows:

1. Employment continues to be one of themost significant factors in postreleaseadjustment.

2. Institutional programs (MDT training andthe token economy study) thus far haveshown only small effects on measuredpostrelease behavior.

3. A clear-cut relationship betweenlaw-violating behavior and scores on theEDS and MBR is shown. Major lawviolators have much higher scores thannon-law violators, with the minor lawviolators falling in the middle.

Preliminary Study of the 'Convict Culture'

As a spin-off from the '71 follow-up study, theLab began an initial analysis of the "convict culture"in terms of the influence of institutional experiencesupon postrelease behavior. By measuring the degreeof an inmate's adherence to the value system of theinmate population, it may he possible to predict thesuccess of his postrelease adjustment and, ultimately,recidivism.

On the basii of interviews with experiencedprison staff, a preliminary checklist was set up and

32

executed by interview with inmates. It dealt withbehavioral adherence to the so-called "convictcontra-culture" which is assumed to operate at crosspurposes to rehabilitation efforts and the rules of thefree world. This checklist measured in large part theinmate's interactions with people, other inmates,prison staff and administrative personnel, andfree-world people.

The checklist items were in the form ofstatements, e.g., "I am not treated fairly by theprison staff." The respondent indicated on a scaleof 1 to 5 the degree of his agreement with thestatement. The maximum score was 185, with a highscore indicating greater adherence to the convictculture.

The checklist was administered initially to 15inmates and later to 80 others. These men had beenselected by the prison staff as being "solid" (adheringto the inmate value system) and "non-solid." Thescores of the "solids" compared to those of the"non-solids" were not significantly different at areasonably low level of probability (.05 to .10),raising some question as to the validity of the useof institutional staff ratings of "solidness."

The group of 80 men who had been interviewedwith the checklist were followed up after release. Theinitial data on their postrelease law violationscompared to their checklist scores were as follows:

Non-Law Law

Scores (Maximum = 185) Violators Violators

lii (98 and above)Lo (97 and below)Total

2327 10

50 30

Thus the scores on the checklist moderately butsignificantly predicted postrelease law-violatingbehavior.

The next step in the study was the developmentof an instrument based on behavioral data whichcould be gathered in an intensive interview or bydirect observation. An interview guide, the Recordof Institutional Behavior (RIB), was designed torecord and evaluate in-prison behavior and the effectsof this behavior on postrelease success. Thepreliminary form of this guide covered four majorareas of in-prison behavior: (1) "getting-along"behaviors, (2) friends, (3) drug use, and (4) sexualinteractions.

The interview guide has been administered tosubjects in the '71 Follow-Up, and, while data areinsufficient to support any major conclusion, earlyresults are encouraging. There seems to be a wide,if not mutually exclusive, disparity in the in-prisonbehaviors of recidivists and non-recidivists. Continuedrefinement of the interview guide is planned beforeit will be tested as a potential predictor of postreleasesuccess.

A short technical report on the preliminaryfindings of the convict culture study has beendrafted.

Products

Many papers and technical reports have beenprepared describing the methodology and design ofthe follow-up studies and the instruments developedand used in these studies. The following arerepresentative:

Prolegomena to the measurement anda.)ses,znent of human behavior. Focuses on theevaluation of gross human behavior, using a four-stageapproach to behavioral analysis. Discusses theapplication of the EDS to the prediction ofrecidivism.

The measurement and prediction of criminalMitzi for and recidivism: The Environmental

Deprivation Scale (EDS) and the Ma ladaptiveBehavior Record (MBR). Describes the use of theEDS and MBR in the EMLC's follow-up studies.Includes data which demonstrate the capacity ofthese instruments to measure and predict criminalbehavior and recidivism.

Evaluation of criminal behavior: Theory andpractice.. Preliminary report on a longitudinalfollow-up study conducted by the EMLC (the finalreport on this study is the following entry). Presentedin two parts: "The Functional Analysis of CriminalBehavior" and "Long-Range Follow-Up Validation inthe Measurement of Criminal Behavior.,"

A longitudinal follow-up investigation of thepostrelease behavior of paroled or released offenders.Reports the results of a three-year postreleasefollow-up study conducted by the EMLC. Describesthe study design, methodology, and the instrumentsused. A technical appendix is being prepared for thisreport.

The Maladaptive Behavior Record (MBR): Ascale for the analysis and prediction of communityadjustment and recidivism of offenders. Describes theuse of the MBR in the EMLC's follow-up studies toassess behavior and predict recidivism. Includes acopy of the instrument.

A manual for the use of the EnvironmentalDeprivation Scale (EDS) in corrections: Theprediction of criminal behavior. Explains the use ofthe EDS to predict law violation and resultantrecidivism. Includes information on behavioralinterview techniques, scoring the instrument,interpretation of score, and validity and reliabilitydata. A copy of the instrument and interviewexcerpts are included.

Other anticipated products include a guide forbehavioral interviewing and a "how-to" manual forthe use of the MBR.

_ t

33

-Itsmet06.

4,

"4" . at aw...au"

4

.]

Dissemination for utilizationa first step in the process of making project findings available to others.

utilization

Communication of Findings

The objective of the EMLC's Utilization Divisionis to communicate the Lab's findings, often itsmethodology, to a variety of target groups. Theseprimarily include policy makers, administrators, andpractitioners in the fields of corrections, education,psychology, and sociology, but also include theprisoners themselves and college and universitystudents. The Lab makes every effort to ensure theuse of its findings and procedures by using a broadrange of media for dissemination. The various formswhich this utilization effort takes are described inthe following sections.

Products Completed

In an -effort to close the time gap between aproject's completion and the report on its findings,the Lab in Phase III worked to provide progressrepot is on all projects, including preliminary findingsand analysis of data trends whenever possible. Threesuch progress reports were completed, as well as thefinal report describing Phase II. The Lab's bimonthlynewsletter, Pacesetter, reported current progressby the EMLC and in the field of corrections ingeneral. Additionally, technical reports on individualprojeAs and instruments were prepared, some ofwhich were printed during Phase III. The presentreport is also a part of this effort to provideup-to-the-minute reporting on Lab projects andfindings. Several of the products completed inPhase III were listed for each project in the earliersections of this report. Additional products are asfollows:

- RRF '72. An annotated publications listdesigned to facilitate selection of pertinent

37

information by those requesting thisinformation., It has been widely disseminated,both by mail and as handouts.Employment Service (ES) Guide. Givesguidelines for ES counselors in developing jobsfor ex-offenders, developing and placing traineesfor those jobs, and providing follow-up in theform of supportive services and record keeping.EMLC Findings Sheets and Briefs. One-pagereports designed to give an overview of someof the Lab's projects and their findings. Twofindings sheets summarized the Lab's bondingproject and the results of a nationwide hoi!:efurlough study. An EMLC brief on theprogrammed instructional materials developedby the RRF was also printed.Bonding Project Final Report. Describes theLab's participation in a national bonding projectwhich led to legislation providing for fidelitybonding for ex-offenders.Individually Prescribed Instructional (IPI)System Revisions. Changes in the system r..ndmaterials available required the revision of theEstablishing Guide and preparation of the IPIPrescribing Catalog supplement. Notices of thechanges in the IPI System and the availabilityof the supplement were mailed to all knownholders of the system.Due to an increased demand that staff time be

used in other kinds of dissemination, the newsletterIntervene was temporarily discontinued after twoissues were prepared in Phase III. The audienceserved by this newsletter, i.e,, Draper staff andinmates, was judged too small to warrant the timenecessary to prepare Intervene on a bimonthly

schedule. Instead, staff energies have been more fullydevoted to Pacesetter, which has a much broaderaudience.

Products in Progress

The nature of the Lab's projects often preventstheir completion coincidental with the end of aparticular phase of the Lab's funding. Severalproducts which are currently in progress have beendescribed in earlier sections of this report, e.g., thetechnical report on the token economy study andcorrectional officer training. Descriptions of othermajor products which are in progress follow:

'71 Follow-Up Study Report. Describes thesecond of the Lab's longitudinal follow-upstudies, using three experimentally treated andtwo control groups.Correctional Surreys. Reports to be prepared onthe results of four correctional surveys whichsought information concerning: (1) Use ofbehavior modification in adult corrections, (2)study release programs for adult felons, (3)employment of ex-offenders in corrections, and(4) "good time policies in adult correctionalinstitutions. The survey questionnaires were sentto 50 state correctional systems, to Guam,Puerto Rico, the Virgin Islands, the Canal Zone,and to the District of Columbia and New YorkCity Departments of Corrections. The results arepresently being tabulated.lndiriduali:ed Reading System for Adults( 1RSA ). A graduated system of six self-pacedtracks designed to cover a broad range of readingneeds, with initial track placement determinedby the Tests of Adult Basic Education (TABE)

38

locator test. On-site tryout of the system hasshown dramatic reading gains with mostsubjects, but revisions are needed before thesystem is ready for release.

Publication Distribution

The Lab disseminates its publications by mailand as handouts on site and at professional meetings.In Phase III the number distributed was 6,896, ofwhich 4,900 were mailed and 1,906 given out ashandouts. These figures do not include the Pacesetterwhich was mailed as part of the regular newslettermailing, IPI and related materials, or the 8,500 copiesof the RRF '72 publications list which were sent asa special mailing.

Those who received publications were largelyrepresented by the following areas: research(educational and psychological); corrections (adult),criminal justice; educators (including vocational andMDT) and students; community lrvices, mentalhealth; and juvenile corrections and services.

Visitors

A total of 826 people visited the EMLC duringPhase III, receiving an orientation to previous RRFwork and current Lab projects through fihRtqlidepresentations, and a tour of the facilities. Visitorsoften requested orientation to specific projects,notably the token economy, correctional officertraining, the IPI System used in the contingencymanagement study, and IRSA, the reading program.

Many of these visitors were educators and theircollege classes, representing the fields of lawenforcement, psychology, and education.Correctional planners and administrators, many from

.

a

other states, also toured the Lab and talked with thestaff. These visits provide an opportunity forinterested groups to see the prison setting and to thusgain an nppreciation and understanding of thecomplications involved which no prowess report canfully convey.

Technical Consultation

The Lab's many publications usually providesufficiently detailed information to satisfy the needsof most programs and agencies which ask for thisinformation. Some products are specifically designedto give "how to" instructions, e.g., the guide whichwas developed for Employment Service counselors.Some groups, however, need special assistance inapplying Lab experiences and findings in theirparticular problem areas and request technicalconsultation., During Phase III, the Lab has providedtechnical consultation in regard to the followingareas:

. . . work-release program planning and legislation. . longitudinal follow-up methodology and

instruments. behavior modification program planning and

implementation. . use of the IPI System to deliver adult basic

education

Some specific examples of the Lab's technicalconsultation activities involved the following groups:

. Draper Correctional Center staff with regard toguidelines for implementing a farm credit systemto promote greater production on the institutionfarm.

. University of Alabama, concerning a survey ofmental health needs in the Alabama prisonsystem.

. , Experimental Manpower Laboratory operatedby Mobilization for Youth, Inc. in Ncw YorkCity to discuss possible future directions fortheir behavior modification programs.

. . Dallas County Juvenile Project, Selma, Alabama,regarding the use of the 1P1 System as part oftheir dropout prevention work.

. . . Montgomery's WSFA television viewers,concerning work-release and its benefits to thepublic.

. North Carolina correctional system on setting upa token economy.

. . Alto Correctional Facility in Athens, Georgia,as part of a team reviewing the academic andvocational programs and activities at the ficilit:..

. . Georgia prison system with regard to thepossible implementation of EMLC evaluativetechniques and follow-up instruments to designtraining programs for the entire prison system.These examples illustrate the diversity of the

Lab staff's technical consultation, which providesinformation and assistance to a broad range ofagencies and programs. Such consultation is animportant part of the utilization effort.

Presentations to Professional Meetings

Many of the papers in the EMLC's publicationslist, RRI: '72, were first prepared for presentationto various professional meetings. Such presentationsare an important part of the Lab's dissemination

effort and are followed by numerous request, for thepapers presented. Some of the major professionalmeetings at which Lab staff made presentationsduring Phase III are listed below.

American Psychological Association (APA),September, 1971, in Washington, D. C, andSeptember, 1972, in Honolulu, Hawaii. Lab staffspoke at these meetings on the EMLC'scorrrectional officer training project, use ofcontingency management in institutionaleducation programs, and the measurement andassessment of human behavior. Two otherpapers presented dealt with i comparison offormats for the presentation of programmedinstructional material and the role of thepsychologist consultant in educational design.Annual Meeting of the Hawaiian CorrectionsAssociation, October, 1971, in Honolulu. TheLab director made a keynote address on futuredirections for corrections and presented atechnical paper describing the application ofbehavior theory to correctional practice,drawing on the Lab's projects and experiences.Alabama Council on Crime and Delinquency(ACCD), September. 1971, in Birmingham. TheEMLC's training coordinator presentedinformation about the correctional officertraining project and the self-instructionaltraining package being prepared.National Society Jri Programmed structi,n(NSP1)11arch, 1972. in New Orleans. Membersof the Lab staff presented two papers describingthe IPI System and its use in Lab studies.American Personnel and Guidance Association(.4 PGA). March, 1972. in Chicago, The EM LC

40

director and the research psychologist presentedpapers dealing with the token economy as abasis for correctional management and the useof contingency management in an individuallyptescribed instructional system.Southeastern Psychological Association (SEPA),April, 1972, in Atlanta. Staff memberspresented four papers at this meeting, two ofwhich described the theory and procedures usedin the longitudinal follow-up studies conductedby the EMLC. The other papers discussed thetoken economy as an alternative to aversivecontrol and the use of individually prescribedinstruction for academically deficient prisoners,relating these topics to Lab projects.The 102nd Congress of Correction of theAmerican Correctional Association, August,1972, in Pittsburgh, Pennsylvania. The EMLC'straining coordinator discussed the three cyclesof correctional officer training, accompanyinghis paper with a slide presentation,

Conferences, Seminars, and Workshops

Another important utilization activity ispresentation of and participation in conferences,seminars, and workshops, particularly those in thefields of corrections, psychology, and education.These meetings provide an opportunity to informlarge groups of people about the Lab's work and inthe case of workshops, to instruct them in the useof specific procedures and instruments developed bythe Lab. The following are some examples of theLab's participation in this form of dissemination:

- WI Workshops. A total of nine workshops wereheld, both in Alabama and out of state, to

demonstrate the use of the IPI System.Workshop participants, mostly representing thefields of education and corrections, came fromtwelve states.

Behavior Modification Institute. Several bf-theworkshops for the 5th Annual BehaviorModification Institute held in Tuscaloosa,Alabama, in March, 1972, were conducted byLab staff. These workshops dealt with the useof contingency management in basic educationand vocational training, behavior modification incorrections,. the evaluation of behaviormodification programs, and teaching behaviormodification techniques to paraprofessionalpersonnel. EMLC staff members also attendedother workshops.

Ore America Workshop. The EMLC had thedistinction of being chosen to host a workshopin August, 1972, for One America, a projectwhich provides aid for female offenders recentlyreleased from prison. The Lab was chosenbecause of its extensive uork in adultcorrections, much of which has been directedtoward assisting the offender once he is releasedto the community.Seventeenth Southern Assembly. The Lab'sdirector was a participant in the conferencepresented by the American Assembly, ColumbiaUniversity and Tulane University at Biloxi,Mississippi, in January, 1973. The conferencetopic was "The American Correctional System:its Problems."Gulf Coast Invitational Conference onMeasurement in Education. The Lab's directorwas one of the keynote speakers at this

41

conference, presented in October, 1972, at theUniversity of South Alabama in Mobile. !Hsaddress dealt with the IPI System and its usein the EMLC's programs.Regional Utilization Network Conferences. Latstaff have participated in three conferences, thefirst of which was held at Oak Ridge, Tennessee,by the Oak Ridge Associated Universities. Thesecond and third were held at the regionaloffices of the Department of Labor in Atlanta.Participants represented the Research andDevelopment (R&D) projects in the southeast.The conferences were held to examine thefeasibility of, and :hen to establish, a utilizationnetwork to promote the utilization of projectfindings by existing agencies and institutions.This network would also help to identifyregional resources and facilitate interchangebetween manpower R&D groups, practitioners,and utilization planners.

Regional Seminar on Adult Basic Education.The Lab director presented a paper describingthe use of hardware and softwa-e in basiceducation programs in correcticns at the seminarheld in Pomona, California, in May, 1972.Seminar on Corrections. The EMLC's trainingcoordinator presented a three-hour semir-r forthe Department of Psychology, University ofAlabama at Tuscaloosa in September, 1971. 't hisseminar dealt with the history of corrections andthe problems of contemporary corei!tions,thawing heavily upon the work done by theEMLC at Draper.

Books and Book Chapters

EMLC staff members are sometimes invited tocontribute chapers for books of readings beingcompiled dealing with corrections and appliedpsychology. In this phase work was begun on achapter entitled "Principles of ilehavior ModificationApplied to Corrections," one of approximately 30chapters in the Handbook of Criminology edited byDr. Daniel Glaser. The book will be published byRand McNally & Company.

The Lab's director has been asked to preparea chapter for Training in Behavior Modification,edited by Dr. Martha E. Bernal. The book will bepublished in 1973 by Brooks/Cole and will deal withthe kinds of training programs that are available foruse with various kinds of problems which occur ina range of settings. The EMLC's contribution willdescribe the correctional officer train.ng projectconducted at Draper.

The EMLC was contacted by Scott, Foresmanand Company for permission to reprint severalexcerpts from Lab publications dealing with thecontingency management studies, including charts,figures, and tables. These will be part of a collegetext entitled Control of Human Behavior: InEducation, Volume III authored by Roger Ulrich,Thomas Stachnik, and John Mabry. The text will bea collection of readings on techniques and newprograms of behavioral education.

Notice has also been received that McGraw-Hillwill publish a book entitled The Design of BehavioralExperiments written by EMLC research coordinator,Dr. W. 0. Jenkins. and Dr. Nolan Hatcher. Many ofthe statistical methods used in the analysis of theLab's follow-up data arc described in this book.

42

Articles in Newsletters and Magazines

The Lab's publications and the papers presentedby staff members at professional meetings oftenprovide information for articles in newsletters andpopular magazines. For example, the December 1972issue of the newsletter Offender Employment Reviewcarried a feature article describing the EMLC and itsprojects as bridging the gap between theory andpractice in offender rehabilitation. The same issuealso announced the availability of the EMLC's Guidefor F=-Iployment Service Counselors in CorrectionalMDT Programs.

The Lab was also included in an article dealingwith behavior modification which appeared in theMarch 1973 issue of Playboy. The article, entitled"Zap! You're Normal," describes the Lab's tokeneconomy study and briefly mentions the correctionalofficer training project and the effect the EMLCprograms have had upon the institution.

Additionally, the editors of Psychologl Todayare planning an article based on the paper "NewDirections in Corrections," the Lab director'skeynote address to the Hawaiian CorrectionsAssociation. The dissemination of Lab findings inpopular magazines such as these reaches a broadaudience of readers who would not otherwise beaware of the Lab's correctional research.

Other Utilization Efforts and Directions

Audioridual Presentations

Films and slide presentations dealing with thework of the EMLC were used in on-site orientationof visitors, as are other films related to correctionswhich are owned by the Lab. In addition, these

audiovisual materials were shown to civic groups,college and university classes, workshop participants,Land correctional officer trainees. Cooperating agencypersonnel also borrowed these materials for use invarious presentations.

University Teaching

As in earlier phases of the Lab, several membersof the professional staff taught on a part-time basisat the Montgomery campus of Auburn University.The courses taught included psychology, sociology,penology, and criminology, Dissemination of theEMLC's findings through the channel of theuniversity classroom has generated considerableinterest in the Lab's work, involving a number ofstudents in this work.

Several other staff members also taught on apat-time basis in the Alexander City State JuniorCollege program for inmates which began at Draperlate in Phase III of the EMLC. Two years of creditcould be earned toward a full four-year college degreeuy participation in the program. The classes weretaught in the evenings and used some of the Lab'sphysical facilities..

College Corps Program

To give college and university students anopportunity for practical experience in correctionalresearch, the EMLC has continued its College CorpsProgram. College Corpsmen work at Draper for aquarter or semester, by arrangement with theirrespective schools, and receive credit for their work.Six Corpsmen participated in the program during thisphase.

Graduate Thesis and Disso tation

EMLC follow-up data provided the basis for a-master's thesis completed at Auburn University which

43

evaluated the effects of MDT training by focusingon an intensive examination of the communityperformance of a small _ample of released offenders.Son, of the results of this study were included inthe t hnical report on the '69 Follow-Up.

The EMLC also provided the data and staffassistance for a Ph.D. candidate from Florida StateUniversity for his dissertation dealing with thepostrelease effects of MDT training, using computerand multivariate discriminant analysis. The computeroutcomes from this dissertation were included in partin the technical report on the '69 Follow-Up andwere found to agree with those resuiting from theLab's data analysis.

Course Development

The EMLC's training coordinator was given aspecial two-week assignment to organize and writethe Military Occupational Specialty (MOS) Course inCorrections for the United States Army Reserve. Thecourse is designed as a Program of Instruction (P01)and will be utilized within the Third Army area inthis coming year with option for inter-Army areausage.

Professional Papers Submitted for PublicationA paper prepared by the Lab's training

coordinator and research psychologist, entitled"Survey of Home Furlough Practices in AmericanCorrectional Agencies," has been accepted byCriminology for publication in the May 1973 issue.

Additionally, the paper which the Lab directorcoauthored with Dr. Jeanne Moon and presented tothe 1971 meeting of the American PsychologicalAssociation has been submitted for publication. Thepaper is entitled "The Psychologist Consultant inEducational Design."

s

Some ( Inments on Phase III studies and

. ,

,

a preview of Phase IV.

meow in°mu concluoion

J

Findings Within Study Limits

The Lab's projects have not been designed asthe ultimate answers to the problems of corrections.Rather, they have been purposely limited to deal withspecific problems within the restrictions imposed bythe institution. This approach has seemed realistic inview of the staggering range and depth of the needsof the prison inmateneeds for vocational skills,personal-social skills, academic instruction, counsel-ing, and job development, for example.

The projects have been further limited by beinginstitutionally based, resulting in such unexpectedevents as unscheduled delays due to shakedowns (i.e.,searches) for weapons and drugs. Often institutionalconcerns for security and harvesting of farm cropsoverrode project concerns for testing and datacollection, Again, these are realistic limitations whichare largely typical of many institutions in which arehabilitation program might be based.

Thus the projects of the Lab, while experimentalin design, have also had this practical aspect, withutilization of the findings always being an ultimategoal. In this context, the individual projects shouldbe considered both in terms of what wasaccomplished and what was learned.

The Token EconomyOne of the major and most ambitious projects

of the Lab was the ecological studythe first tokeneconomy implemented in an adult penal institution.The study demonstrated that the token economymodel is a Pasible and effective alternative to aversivecontrol procedures to manage inmate behavior.

The token economy encountered certaindifficulties in its implementation, however. Like the

47

special education program in an elementary school,the Unit found itself viewed by the prisonadministration as a handy place to assign thediscipline problems, and project staff protests aboutsuch assignments were sometimes ignored. Thus theUnit population varied somewhat from that specifiedby the study design. The effect of this upon the datacollected is difficult to determine.

Likewise, the institution's farm credit systemmay have affected the study data, particularly sinceLab staff members were responsible for reporting thecredits earned on the Unit and in basic education.The loss of credibility when the institution failed todeliver its reinforcers and to systematically followprocedures may have generalized to the Lab's study.Such effects, however, are difficult to assess.

The unsuccessful farm credit system did pointout one thinga certain amount of professionalexpertise and a great deal of staff commitment arevital to an effective token economy procedure. Ifthese are missing, the results most likely will bediscouraging; indeed, the target behavior may becomeeven more of a problem if the treatment is

misapplied.

The attempt of the token economy to increaseits effectiveness by expanding the reinforcementmenu was met with opposition by the institution,despite the fact that many potential reinforcers arereadily available within the correctional setting. Thusthe effect of such reinforcers as performance-contin-gent parole remain unexplored at this time. The tokeneconomy has proven itself to be both feasible andeffective, but further study is needed to increase itseffectiveness.

Correctional Officer Training

All three cycles of correctional officer trainingconducted by the Lab have shown that correctionalofficers can, indeed, learn the basic principles ofbehavior modification as these were presented in thetraining. Moreover, this learning can take place in arelatively short time. If the average time needed perbooklet remains at 30 minutes after the completeseries has been tried out, then each officer could beexpected to spend 10 hours in the classroom tocomplete the booklets and approximately 20 hoursin discussion sessions. The total 30 hours, spread overa ten-week period, should allow the correctionalofficers of most institutions sufficient time in whichto arrange a comfortable schedule for work and

Scheduling is perhaps a lesser problem, however.The major difficulty the Lab encountered was thelack of any control of strong positive reinforcerswithin the institutional system which would fullyinvolve the officers in their training. Encouragementand positive reinforcement from the institution'sadministration did not occur on a regular basis.Instead, officers were simply "told" to attend thesessions, received little recognition for having done so,and were expected to give their custody and securityduties priority over training. At times, the officerswere even verbally discouraged from participating inany way more active than mere attendance. Theeffects are predictable.

Ideally, correctional officers would not only beencouraged by their superiors to attend training andbe recognized for having done so, but they shouldreceive promotions in rank and salary increases forsuccessful completion of training--which would

48

include completion of practicum exercises. Totalcommitment by the institution is necessary forimplementation of a successful training program.

Longitudinal Follow-Up StudiesSystematic follow-up studies of the released

offender in the community had not been conductedprior to the Lab's studies. The EMLC found thatlongitudinal follow-up can be conducted using abehavioral model to evaluate the effectiveness ofinstitutional intervention programs. Additionally,,instruments have been developed which can measureand predict criminal behavior, The follow-up studiesconducted by the Lab can be considered to providea foundation for program assessment and planning.

One of the major difficulties encountered in thefollow-up studies was providing the training necessaryto use the behavioral interview techniques andinstruments, since these are relatively new and littlewritten information is available, One-to-one trainingof staff was used, a method which certainly cannotclaim to be highly efficient in terms of t timenecessary. "How-to" manuals are being developed toaid in training interviewers, and the possibility ofconducting training workshops is being considered.

Analysis of the data collected also presentedsomewhat of a problem, since the data-gatheringinstruments were new and just being tried out andvalidated. Refinement of the instruments has beencontinuous as analysis proceeds, increasing thepredictive capacity of several instruments. The inter-view guide, which gathers primarily demographicdata, has also been revised to deal with questionsmost pertinent to the assessment of program effec-tive less,

Utilization

The dissemination for utilization effort of theEMLC has been quite extensive, in spite of thelimited time and funds available. Correspondence,newsletters, personal appearances by staff at meetingsand speaking engagements, as well as the preparationof numerous papers, have created a great interchangeof information from the Lab to inquiring people incorrections and related fields of study.

Specific Lab products, e.g., the IPI System, theMaladaptive Behavior Record, the EnvironmentalDeprivation Scale, and the Employment ServiceGuide, to name a few, have been accepted andimplemented by people in many states throughout

.... the United States. In short, the utilization effort hasbeen quite successful, including changes which haveaccrued within the Alabama correctional systemrelated to input by EMLC staff. However, severalthings have emerged as keys to improved utilization,

Determining the specific audience to whom thepublications speak has been a limitation to moreeffective implementation of Lab findings throughdissemination. Related to the matter of audiences isthe difficulty which has been encountered in attemptsto generate effective dissemination plans for newpublications.

One of the most significant facts which hasevolved is the need for intensive work withadministrative and operational staff at the beginningof a project if the findings of the research studiesare to be implemented on a regular basis. This refersnot only to the immediate host agency, but also tothe related agencies which are involved in workingwith offenders, e.g., Probations and Parole,Employment Service, Vocational Rehabilitation, etc.

449

The Lab has done a great deal of this kind of workto involve the significant groups and agencies, butthere is an increasing need to explore feasiblestrategies to assure the effective implementation ofresearch findings.

Proposed Directions for the EMLC

Based on what has been learned in previousEMLC projects, certain general areas for furtherdevelopment and study have emerged. The continuedrefinement of the behavior modification model rankshigh amongst projects which warrant further explora-tion. The behavior modification model at the EMI,C'spresent site has more than adequately demonstratedthat institutional inmate behavior can be managedthrough these means, while simultaneously servingthe needs of institutional control and custody. Thenext step is to provide the conditions for broaderexpansion of the techniques throughout the institu-tion and also to correlate the target behaviors forinstitutional treatment more closely to those free-world behaviors which can keep the released inmateemployed and productive. To do this, it is highlydesirable, if not absolutely mandatory, that theenvironment for carrying out such a correlation allowthe most detailed analysis of behavior possible.

A proposed site for carrying out the next phasein the development of the behavior modificationmodel is an institution for youthful offenders locatednear the present site of the EMLC. The site will notonly offer the controlled conditions necessary foroptimal study findings, but the technical supportstaff of the institution will also augment the effectiveimplementation of the next, more sophisticated stageof the program.

Another area for innovative work, based onmany of the findings of the EMLC follow-up studies,is the development of an effective manpower modelfor community treatment of released offenders. Whathas become increasingly clear over the years is thatprograms which attempt to effectively treat thereleased offender in the free-world communityenvironment must take a look at the individual andhis particular behavioral strengths and weaknesses.Historically, programs have dealt with offenders interms of gross behaviors and gross behavioral deficits.Vocational programs, counseling programs, andothers have all too often attempted to fit allex-offenders into-a mold of pioblems and treatmentsolutions, regardless of individual variations ofbehavior. Yet, of course, what causes one man tocommit a crime is not necessarily what causes anotherman to commit a similar crime, One man steals formoney, another steals for prestige, another steals forpeer acceptance, another may steal because he hasbeen trained to do so, and still others may stealbecause it is the only way they know of getting whatthey want. For some of these cases, gaining a voca-tional skill and earning a living may be a solution, ora partial solution. For others, a vocational skill andwhatever monies it can offer may miss the markentirely.

A proposed affiliation with the psychologydepartment of University of Alabama in the nearfuture will allow the implementation of such anindividualized community program. Selected EMLCstaff are currently planning to maintain an office onthe university campus and to begin developing thecommunity manpower model for released offenders.The university site seems extremely advantageous

50

because of the resources which it can offer, e.g., thecomputer analysis of data, technical support staff,and concurrently developing criminal programs. Thepotential of better access to national funding sourcesand increased utilization of Lab findings also makethe university affiliation desirable.

The overall goal of this project is to developan effective community model for the employmentof youthful ex-offenders in the community workforce. Specifically, the objective is to develop amanpower model for receiving, motivating, training,placing, and following up youthful ex-offenders forjobs in the community work force. Sub-objectivesare: (1) to develop a model for the cooperativeparticipation of the University of Alabama in thismanpower crime prevention and control program and(2) to develop procedures for the effectiveparticipation of employers in the project for the jobtraining and placement of ex-offenders.

Plans for the next phase of the EMLC alsoinclude continuing work in the adult correctionalinstitutionat Draper Correctional Center. Twoprojects are planned; one is a continuation fromPhase III and the other is a new project. Thecontinued project is the completion of thecorrectional officer training package. Approximatelyfive of the original trainees from the just completedcycle of training will review the materials which theyhave already been through, then complete theremaining booklets (10), and finish up their trainingby conducting practicum exercises.

The new project which will be carried out atDraper will be the assessment of the work-releaseprogram of the Alabama Board of Corrections. Theassessment will focus on the ex-offender's specific

postrelease behavioral adjustment. The instrumentsfor doing such an assessment have been developedby the EMLC in its own follow-up work. Theyinclude the Environmental Deprivation Scale (EDS)and the Maladaptive Behavior Record (MBR) andothers.

The proposed work-release project is expectedto yield a definitive, objective, and systematicmethod of collecting and comparing data. Theanalysis of these data will supplement thework-release staff's knowledge so that moremeaningful decisions can be made regarding theprogram's operation. Intervention, e.g., counseling,pievocational training, and vocational training, canthen be made more systematic. Based on these data,more appropriate changes in participant selectioncriteria will also be possible. Information from thepostrelease follow-up will be made available to anumber of related state and local agencies and to thework-release staff so that ally postrelease interventioncan take place with specific, behavioral referencepoints. The project is intended to establish anindefinite ongoing program assessment for theAlabama Board of Corrections.

Concurrent with the major research efforts ofthe Lab, two specific learning systems meritcontinued work. The Individually PrescribedInstructional (IPI) System, a basic education deliverysystem developed by the Lab and already widelyutiliied, will continue to be refith.d. The system willhe used in the basic education component of thebehavior modification program described earlier,

A more recent development of the Lab, theIndividualized Reading System for Adults (IRSA),will also be further developed. IRSA hasdemonstrated its effectiveness in creating remarkablechanges in reading level for Draper trainees, and it

51

can be adapted very well to the similar populationat the Alabama Industrial School (AIS), theinstitution for youthful offenders which is beingconsidered as a site for some of the projects in thenext phase of the EMLC.

In conjunction with the development of thelearning systems, a great deal of work has also beendevoted to generating an effective motivationalsystem to complement IPI and IRSA. The systemwhich has been explored, the contingencymanagement system, will be implemented in futureuses of the IPI System and IRSA., As is the case withoffender trainees at Draper, the population at theAlabama Industrial School will also be able to benefitfrom individualized motivation procedures which canproduce learning achievement in a minimum of time.

The completed systems, both IPI and IRSA, aswell as the methodology for contingencymanagement, are highly exportable products from theLab. The WI System has already been exported toa considerable extent, and IRSA and the contingencymanagement procedures show the same promise.

In summary, the EMLC in the next phase willbe continuing to explore the research areas whichhave shown promise in preceding work. The researchduring this proposed phase will be carried out at threedifferent sites: at Draper, where the EMLC has beenfor the last three phases of operation; at Mt. Meigs,Alabama, the site of the Alabama Industrial School;and at the University of Alabama campus, in

conjunction with the Department of Psychology. Thebehavior modification program, probably in the formof a token economy, will be at AIS, the communitytreatment project will be operated with theuniversity, and the correctional officer trainingproject and the work-release evaluation will be carriedout at Draper.