• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 Appendices
 Biographical sketch














Title: Personal control beliefs within correctional settings
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00098655/00001
 Material Information
Title: Personal control beliefs within correctional settings
Physical Description: viii, 132 leaves : ; 28 cm.
Language: English
Creator: Freeman, Douglas Joseph, 1948-
Copyright Date: 1977
 Subjects
Subject: Control (Psychology)   ( lcsh )
Prison psychology   ( lcsh )
Prisoners -- Attitudes   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Statement of Responsibility: by Douglas Joseph Freeman.
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 108-112.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098655
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000063584
oclc - 04213251
notis - AAG8783

Downloads

This item has the following downloads:

PDF ( 6 MBs ) ( PDF )


Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
    Table of Contents
        Page iii
        Page iv
    List of Tables
        Page v
        Page vi
    Abstract
        Page vii
        Page viii
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
    Methods
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
    Results
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
    Discussion
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
    References
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
    Appendices
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
    Biographical sketch
        Page 132
        Page 133
        Page 134
        Page 135
Full Text












PERSONAL CONTROL BELIEFS
WITHIN CORRECTIONAL SETTINGS












By

DOUGLAS JOSEPH FREEMAN


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY





UNIVERSITY OF FLORIDA















ACKNOWLEDGMENTS


The author wished to recognize and thank three members of the

supervisory committee--Drs. Landsman, Davis, and Sieler--for their

assistance and cooperation in completing this study. Special appreci-

ation is extended to Drs. Richard M. Swanson and Ben Barger who served

as cochairmen. Dr. Swanson's encouragement, support, and criticism

throughout the course of this study is gratefully acknowledged.

Appreciation is also extended to Mr. Thomas Van Den Abel for

his assistance in the analysis of the data.















TABLE OF CONTENTS


LIST OF TABLES v

ABSTRACT vii

INTRODUCTION 1

The Construct Personal Control 1
Review of the Literature--The Significance of
Variables Related to Personal Control 5
Alternative Conceptualizations of Personal Control 18
Choice of Research Setting 24
Application to Corrections 25
The Research Plan 26
Hypotheses 31

METHODS 34

Subjects 34
Institutions 36
Instruments 40
Procedure 50

RESULTS 56

Internal Properties of the Scales 56
Validity 57
Relation of Personal Control to Other Personality Variables 64
Distribution of Personal Control Within Social Settings 69
Changes in Personal Control Beliefs 73
Environmental Influences 81
Success in the Environment 92

DISCUSSION 101

REFERENCES 108

APPENDICES

A. Attitude Toward Prison Behaviors 113
B. Definition of Situation 117
C. Attitude Toward Deviance 121
D. Alienation 123









E. Interpersonal Trust and Self-Esteem 124
F. Freedom of Movement 128
G. Personal Control 130
H. Ratings of Inmates 131

BIOGRAPHICAL SKETCH 132















LIST OF TABLES


Table Page

1 Internal Properties of Purified Scales 58

2 Intercorrelations Among Attitude Toward
Prison Behaviors Measure 59

3 Intercorrelations Among Institution Oppor-
tunity and Authority Scales 60

4 Intercorrelations Among Remaining Individual
Difference Variables 61

5 Intercorrelations Among All Scales 65

6 Differences in Personal Control Measure
Within Four Male and Three Female
Correctional Institutions 71

7 Ethnic Differences in Personal Control
Measure Across Seven Correctional Institutions 72

8 Ethnic Differences in Personal Control Measure
Within Seven Correctional Institutions 74

9 Correlations Between Personal Control and
Socioeconomic Status 75

10 Correlations Between Personal Control and
Length of Stay Variables 77

11 Differences in Personal Control Among Inmate
Samples Classed by Phase of Sentence Being
Served Within Seven Correctional Institutions 80

12 Differences in Personal Control Among Male
and Female Samples with Maximized Time-In-
Prison Differences 82

13 Measure of Group Consensus Derived from the
Attitude Toward Prison Behaviors Items for
Seven Institutions 84









14 Measure of Group Consensus Derived from the
Attitude Toward Prison Behaviors Items for
Seventeen Living Areas 89

15 Measure of Group Consensus Derived from the
Attitude Toward Prison Behaviors Items
for Ethnic Groups Within Six Institutions 90

16 Personal Control and Population Size of
Seven Correctional Institutions 93

17 Correlations Between Personal Control and
Success in the Environment Measures 94

18 Success of Dichotomized Personal Control
Groups Within Low and High Consensus
Living Units 97

19 Success of Dichotomized Personal Control
Groups Within Low and High Consensus
Correctional Institutions 98














Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy



PERSONAL CONTROL BELIEFS
WITHIN CORRECTIONAL SETTINGS

By

Douglas Joseph Freeman

June 1977

Chairman: Ben Barger
Cochairman: Richard M. Swanson
Major Department: Psychology

The study was designed to examine the construct personal control

within correctional institutions contrasted by their relative opportunity

structures. Cross-sectional samples of inmates from four male and three

female settings served as subjects. Personal control correlated as

predicted with a large number of other personality variables drawn

largely from social learning theory, including freedom of movement,

alienation, attitude toward deviance, self-esteem, interpersonal trust,

attitude toward prison behaviors, and attitude toward institutional

opportunity and authority structures.

Personal control beliefs were distributed over correctional settings

such that 1) inmates from high opportunity correctional environments held

significantly stronger beliefs in personal control than inmates from low

opportunity environments; 2) majority inmates' beliefs were significantly

stronger than minority inmates' beliefs; and 3) there was a significant

positive correlation between personal control and socio-economic status.










Examination of changes in personal control beliefs as a function

of length of stay in prison failed to significantly support the hypo-

thesis of curvilinearity. Instead, a number of significant linear

relationships between personal control and length of stay variables

obtained with various inmate samples.

Group consensus on appropriate behaviors discriminated among

inmate groups by institution, by living unit, and by ethnic group

within institutions. The power of this discrimination was supported

by the finding of positive relationships between personal control and

group consensus in each level of analysis. A nonsignificant positive

relationship obtained between personal control and institution population

size.

While high personal control beliefs were generally predictive of

inmate success in the environment, further analysis indicated that high

personal control inmates in high consensus environments were more

successful than their counterparts in low consensus environments.

Success differences between personal control groups in low consensus

environments tended to be nonsignificant.















INTRODUCTION


The Construct Personal Control

Joe (1971), LeFcourt (1966a, 1972), Phares (1976), Rotter (1966),

and Swanson (1970a) have reviewed the theory and research related to a

hypothetical dimension of personality referred to variously as personal

control, locus of control, internality-externality, effectance, and

personal efficacy. They suggest it is a variable of widespread pre-

dictive utility, noting that it has received an enormous amount of

attention in a wide variety of research settings with a multitude of

subject populations. Rotter (1975) reports well over 600 published

research studies as well as countless unpublished theses, disserta-

tions, and papers.

In essence, the concept of personal control refers to an indi-

vidual's expectations that the consequences of his actions are contin-

gent upon his own behaviors and attributes. It is held that the

individual who perceives these contingencies accurately may learn to

manipulate his environment in order to achieve his desired goals. The

individual who fails to perceive these contingencies is more likely to

attribute the consequences of his behavior to fate, luck, chance,

systematic forces, or powerful others.

Personal control is seen as a variable of great import to the

understanding of an individual's interactions with his environment

and with others. The individual who has perceived the contingencies









between his behavior and his desired rewards would be expected to have

learned to control and manipulate his environment. Thus, there is an

increased possibility that that individual will adapt to more socially

desirable behavior in order to achieve his goals. Likewise, the indi-

vidual who perceives no contingency between his behavior and his

desired rewards may be unwilling, or even unable, to conform to societal

demands. That individual's behavior may become less goal-oriented and

more random.

The largest portion of the research on the construct personal

control has been in the study of individual differences in personality.

Julian B. Rotter is in the forefront in this research effort. In his

monograph describing his research on the topic, he suggested that

"consistent individual differences exist among individuals in the degree

to which they are likely to attribute personal control to reward in the

same situation" (1966, p.1). The personal control concept emerged from

a comprehensive theoretical framework known as social learning theory

(Rotter, 1954; Rotter, 1966; Rotter, Chance, and Phares, 1972), and

was considered to be of "major significance in understanding the nature

of learning processes in different kinds of learning situations"

(1966, p.1).

Before turning to a description of some of the significant research

on the construct personal control, and in order to fully understand its

significance in explaining human behavior, it is now necessary to delin-

eate the theoretical and conceptual framework from which it is derived.

Theoretical Background. Within social learning theory, any behavior is

the result of a selection, choice, or decision process. The choice of








behavior is based upon which action has the highest expectation of

maximizing valued consequences for the actor in that particular

situation. In order to account for the occurrence of any specific

behavior, there must be knowledge of the following: that the behavior

is in the repertoire of the actor (that it has been learned); the ex-

pectations of the actor that that behavior will lead to specific out-

comes; the value of those outcomes For the actor; and the various

outcomes available in that situation.

Thus, the basic concepts in Rotter's theory are: 1) expectation

(E), which refers to the actor's subjective probability that a specific

behavior will lead to the occurrence of a specific outcome(s) or

reinforcementss; 2) reinforcement value (RV), which refers to the

actor's degree of preference for the specific outcomes(s) or reinforce-

ment(s) which are contingently related to behavior; 3) behavior potential

(BP), which refers to the probability of the occurrence of a behavior,

or the relative strength of the tendency to respond in a particular

manner; and 4) the psychological situation, which refers to the actor's

meaningful environment and represents the actor's immediate context of

action.

These fundamental terms generate Rotter's descriptive formula,

stated in its most simple form:

BP = f(E & RV).

The formula reads: the potentiality of any behavior occurring in a

given situation is some function (probably multiplicative) of the expec-

tancy that that behavior in that situation will lead to a particular out-

come and the value of the outcome for the actor in that situation.









Thus, social learning theory regards expectancies as prime deter-

minants of behavior; reinforcement alone does not explain behavior

adequately. In other words, an individual's behavior is determined

not only by the reinforcement received through goal achievement but

also by whether that individual expects his behavior to lead to specific

goals. Moreover, expectancies that certain behaviors will lead to

reinforcement are increased by successes in one's past experiences and

decreased by past failures. Thus, expectancies for behavioral outcomes

are, in large part, learned through past experiences.

Another factor about expectancies should be considered before

returning to a discussion of the construct personal control. Social

learning theory recognizes that behavior is determined both by situation-

specific factors and by general factors. With regard to expectancies,

when individuals are in novel situations, generalized expectancies are

expected to be more potent determinants of expectancy than will be

specific expectancies which are based purely upon prior experience in

that situation. On the other hand, when an individual is greatly

experienced in a particular situation, generalized expectancies should

have relatively little potency and specific expectancies in that

situation should be dominant.

Generalized expectancies are derived from the assumption in

social learning theory that individuals categorize situations along

various dimensions of similarity. It is assumed that individuals can

categorize situations in a limitless number of ways, thus allowing for

the development of a limitless number of generalized expectancies.

Rotter suggests that these hypothetical dimensions of personality take









on "functional properties and make up one of the important classes of

variables in personality description" (1966, p.2). The end result is

that such generalized expectancies, coupled with specific expectancies

and reinforcement values, act to influence the choice of behavior.

One such relatively consistent generalized expectancy is the degree

to which an individual believes in internal or external control of rein-

forcement--whether the individual believes what happens to him is under

personal control or is attributable to luck, fate, chance, powerful others,

etc. It is held that this belief in personal control, dependent as it is

upon one's reinforcement history, is a generalized personality construct

that can be quantified and used in conjunction with other social learning

theory variables to understand and predict complex human behavior. Let

us now focus upon some of the relevant research on this concept.



Review of the Literature--The Significance of Variables
Related to Personal Control

As noted above, the construct personal control has been a particu-

larly robust one for researchers. Some of the relevant findings will

now be discussed. The hypotheses which they lead to will be specified

in more detail at the conclusion of this chapter.

Relation to Personality

Belief in personal control has been shown to be associated with

a wide range of personality variables. Joe (1971), in his extensive

review of the research on personal control, concludes that "the most

significant evidence for the construct validity of the internal-external

control concept lies in the area of personality functioning" (p.634).

One personality variable derived directly from social learning theory









that is very important for belief in personal control is the individual's

perception of the opportunity structures available to him. Rotter called

this variable freedom of movement. If the individual perceives the

environment to be one characterized by opportunity, he will be more

likely to hold high expectations of goal attainment. Thus, this indi-

vidual should have developed a feeling of mastery over his environment.

Jessor, Young, Young, and Tesi (1970), in a cross-cultural study

of Italian drinking behavior in Rome, Palermo, and Boston, reported upon

the relationship between expectations for goal attainment and personal

control. The predicted relationships obtained in all three cities

(Rome, r = .38; Palermo, r = .28; and Boston, r = .30). Jessor, Graves,

Hanson, and Jessor (1968), in a southern Colorado community, found that

in adults the correlation between expectations and personal control was

.28, and for high school students, the correlation between expectation for

academic recognition and personal control was .14. All of the above

correlations, with the exception of the final one cited, were significant

at the .05 level or better. In two other studies, Nelson and Phares

(1971) and Strassberg (1973), using different measures, found that a

lower expectancy of achievement of valued goals was associated with

externality.

While the correlations here reported are somewhat low, it does

appear that persons who believe they control their own outcomes will

also tend to hold high expectations that they will reach their goals.

A concept with some similarity to the construct personal control

is that of alienation. Seeman (1959) described five aspects of this

concept, the first of which appears to be most similar to personal









control and the second of which may be a direct result of the first. The

five aspects of alienation are: 1) powerlessness, the expectancy that

one's own behavior cannot determine the occurrence of his personally

valued outcomes or reinforcements; 2) meaninglesnenss, a low expectancy

that satisfactory predictions can be made about future outcomes;

3) normlessness, a belief that socially proscribed behaviors are re-

quired to achieve valued goals; 4) isolation, a rejection of socially

valued goals; and 5) self-estrangement, the dependence of behavior

upon expected future goals rather than upon current intrinsic ones.

A person who experiences a belief in personal control in their life

would be unlikely to also experience a feeling of alienation. On the

other hand, a person who felt that the rewards of life were independent

of his own actions or characteristics would be likely to feel a sense

of meaninglessness, isolation, and estrangement from others--in short,

a sense of alienation.

In a study of alienation and social learning, Seeman (1963) pre-

sented materials relating to correctional matters to prison inmates.

Inmates low on the alienation measure retained the information pertinent

to parole significantly better than inmates high on the alienation

measure. In the Jessor et al. (1970) study cited above, alienation

predicted expectations for goal attainment better than personal control

(r's ranging from -.37 to -.51). This was also the case with the

Jessor et al. (1968) study in Colorado where the correlation between

expectations and alienation for adults was -.35. In the same study,

internal control correlated negatively with social isolation (r = -.45).

A person who feels able to deal effectively with his environment

should also have a belief in his own abilities and self-worth. He should









develop a sense of power and self-acceptance. In a study of 891 college

students, Hersch and Scheibe (1967) reported that internals described

themselves as more active, striving, achieving, powerful, independent,

and effective. In addition, they reported the following relationships

with externality, all significant at the .05 level or better: Self-

Acceptance, r = -.17; Self-Confidence, r = -.18; and Adjustment, r = -.10.

The Coleman Report (Coleman, Campbell, Hobson, McPartland, Mood,

Weinfeld, and York, 1966) strikingly underlined the significant positive

relationship between belief in personal control and positive conceptions

of own ability and interest in school work among white and black school

children. Studying 144 members of the Detroit labor force, Douvan and

Walker (1956) reported a positive correlation between a "sense of

effectiveness" and satisfaction over outcome of life plans (r = .28).

Fitch (1970) also reported a low significant positive rank-order

correlation between locus of control and self-esteem, with low self-esteem

subjects tending to score as externals. All these findings point

toward the conclusion that persons who believe in personal control also

tend to think more highly of themselves.

Persons who believe they exercise some control over their environ-

ment tend to experience other people in their environment as facilitative

or helpful to them. People are generally seen as benevolent and trust-

worthy. Their behavior is seen as more lawful and predictable. Rotter

defines interpersonal trust as "an expectance held by an individual or a

group that the word, promise, verbal or written statement of another

individual or group can be relied upon" (1967, p.651).

Hamsher, Geller, and Rotter (1968) found a high correlation

between I-E scores and subject's willingness to accept the findings of









the Warren Commission Report. Miller and Minton (1969) found that

externals were more apt to violate experimental instructions than were

internals. They also reported a low significant correlation between

externality and machiavellianism. Therefore, there is evidence that

suggests a relationship between mistrust or suspicion and personal

control. The findings suggest that persons who believe in personal

control will also tend to trust others.

Certain behavioral dispositions and/or styles of interactions

with the environment should be related to belief in personal control.

If a person believes that he will be able to gratify his goals he

should be less willing to engage in behavior that is socially judged

as being deviant in order to maximize his chances of reaching his goals.

He would be more likely to believe that by exercising the conventional

means he believes are open to him, he can avoid the dangers inherent

in employing the deviant neans. Therefore it is suggested that his

attitude toward deviant behavior would be negative in nature.

Jessor et al. (1968) reported no significant relationship between

personal control and attitude toward deviance in either their high

school or adult age samples. However, Wood, Wilson, Jessor, and Bogan

(1966), with a sample of prisoners, supported the hypothesis. They

suggested that it may be that individuals believing in personal control

simply are able to perceive the possibility of punishment for engaging in

deviant behaviors. Even though these findings are inconclusive, logic

continues to suggest that a person who believes in personal control also

tends to be intolerant of deviant behavior. Further research on this


relationship is needed.









Another group of personality variables that should have some

direct relationship with personal control are concerned with an indi-

vidual's attitude toward his own situation. Two aspects of this

attitudinal set are: 1) prescriptions for behavior of self and impor-

tant others in the environment; and 2) how the individual perceives and

defines the nature of his environment or situation. It is expected

that persons who believe in personal control- prescribe behaviors

which increase the likelihood of reward and decrease the likelihood

of punishment. They should do this for themselves as well as for impor-

tant other people in their environment. This has been supported in

previous research by Swanson (1968).

It is expected that a person who believes in personal control

should perceive and define the nature of his environment as one of

greater opportunity with the significant people therein benevolent

and helpful. It has been suggested that a person who perceives his

environment as hostile and its people as malevolent could still believe

in personal control if he felt capable of overcoming the environmental

obstacles. However, it follows that if a person perceives his environ-

ment as more friendly and richer in opportunity, he should be more likely

to develop a belief in personal control. This has been supported by

Wood et al. (1966) with a prison population and by Swanson (1968), in a

partial replication of the Wood et al. (1966) study.

There, as has been suggested in previous research, it appears

that individuals who believe in personal control tend to have more

positive attitudes toward their situation. More specifically, they

tend to prescribe behaviors for themselves and for others that

maximize reward and minimize punishment. And they tend to perceive their

environment as opportunity-rich and the people in it as friendly.









Distribution of Personal Control Wlithin Social Settings

Just as persons who perceive their environments as rich in oppor-

tunity should believe in personal control, it should also be true that

environments that are, in fact, relatively high in opportunity should

contain individuals who believe more strongly in personal control than

environments that are, in fact, relatively low in opportunity.

However, since Jessor et al. (1968) suggested that the institu-

tionalized means of achieving goals emphasized by our culture are

unevenly distributed in our society, "the lower social strata and

certain subgroups, notably racial and ethnic minorities, represent

disadvantaged locations with respect to opportunity" (1968, p.55).

Therefore, it is expected that persons of ethnic/minority or lower

socioeconomic status should believe in personal control to a lesser

extent than persons of majority or higher socioeconomic status.

Much of the work on social antecedents to personal control

supports a relationship with social class and ethnicity. The impli-

cation is that lower social class or minority individuals have relatively

little access to social mobility, significant power, or material advan-

tages and this lack of opportunities manifests itself in higher exter-

nality. Franklin (1963) studied a nationally stratified sample of

1000 cases and reported a significant correlation between socio-

economic class and personal control. The higher the socioeconomic

class the more the person believed in personal control. Battle and

Rotter (1963) studied sixth and eighth grade children and reported

that lower socioeconomic, as well as black, children were more exter-

nal than higher socioeconomic or white children, respectively.









Using six different measures relating to personal control,

Lefcourt and Ladwig (1965, 1966) found black prison inmates more exter-

nal than comparable groups of white prison inmates. However,

Kiehlbauch (1967) was unable to show significant differences between

black and white reformatory inmates, although there was a trend toward

greater externality among blacks.

Coleman et al. (1966) studied a nationally stratified sample of

school children and, among other measures, employed a measure of personal

control. They reported that black children believed in personal control

to a lesser extent than did white children. Moreover, this variable

was far more predictive of school achievement for black children than

it was for white children.

Scott and Phelan (1969), in a tri-ethnic study of unemployable

persons on the welfare rolls, found whites more internal than either

blacks or Mexican-Americans, and Mexican-Americans slightly more external

than blacks. Other studies that support the greater externality of

blacks relative to whites include Lessing (1969), Shaw and Uhl (1971),

Strickland (1972), and Zytkoskee, Strickland, and Watson (1971).

While these studies have all supported the hypothesized rela-

tionship between personal control and minority and socioeconomic status,

other studies have failed to do so. Gore and Rotter (1963) failed to

find a significant association between personal control and socioeconomic

class. They surmised that the limited range of social classes in their

subject population could have been responsible for this finding.

Studying university psychology students, Rotter (1966) also attributed

the lack of significant relationship between social class and personal

control to sample homogeneity. Swanson (1968) studied the resocialization









process in a correctional setting. His sample included four ethnic popu-

lations (Anglo, Black, Indian, and Spanish-American), and he failed to find

any significant differences in personal control among these groups. Jessor

et al. (1968) also reported a failure to find significant differences among

ethnic groups on their personal control measure. While some findings of no

significant difference are no doubt related to sample homogeneity, Jessor

et al. (1963) and Swanson (1968, 1970a, 1970b) suggest an even greater

possibility is the weakness of the personal control measure utilized.

The published findings, therefore, on the distribution of personal

control within the social settings offer both support and disagreement

with the suggested relationships. Possible reasons for the disagreement

have been proposed. It is felt that the questions relating to environ-

mental opportunity differences and minority and social class differneces

bear further testing.

Changes in personal Control Beliefs

The extensive influence of belief in personal control upon person-

ality and behavior has been suggested. It is important to understand

the conditions which influence the development of personal control beliefs.

The simplest reason for change in belief in personal control may stem from

age change. Penk (1969) reports that among children internality typically

increases with age. Crandall, ir.l :., and Crandall (1965), in a study

on academic achievement, noted a tendency toward externality in the third

grade, and increasing internality to a maximum in the eighth and tenth

grades, with a return to externality in the twelfth grade. They suggest

that perhaps the look toward the uncertainties beyond graduation from high

school tends to externalize students' beliefs about personal control.









Kiehlbauch (1967), in a cross-sectional study of reformatory inmates

reported a similar curvilinear relationship between personal control

and length of stay in the reformatory. Inmates were external at the

beginning of their term, more internal in the middle of their stay, with

a reversion to externality just prior to release. His conclusion that

a curvilinear relationship exists between personal control beliefs and

environmental uncertainties seems clear. To support the interpretation

of this curvilinear function, Kiehlbauch noted that he did not find the

return to a more external belief among inmates who had been assigned to

a work release program and had already had some meaningful contact with

the outside world before their release.

There is evidence suggesting that personal control beliefs may be

somewhat dependent upon the social climate of the times. Changes have

been observed in college students' scores on personal control measures

over the past decade. Schneider (1971) and Phares (1976) report that

several studies across the country have shown students to become signi-

ficantly more external since the publication of Rotter's I-E Scale in

1966. Speculation for this drift towards externality has centered on

the sense of alienation on campuses attributable to national issues such

as the Vietnam conflict or Watergate.

Another research thrust underlining the importance of under-

standing personal control belief changes has a more clinical tone.

MacDonald (1971) has suggested that shifting clients' beliefs in the

internal direction is important in rehabilitation work, and Lefcourt

(1966b) has written of the importance of encouraging an internal locus

of control in psychotherapy. Gillis and Jessor (1970) report that









patients showing improvement in psychotherapy also showed greater ten-

dencies toward internal beliefs than their untreated group of patients.

Therefore, it has been shown that personal control beliefs are

flexible and seem to be influenced by a variety of change-agents.

Some research suggests that environmental factors are related to personal

control beliefs. Another line of research supports the conclusion that

interpersonal factors influence changes in personal control beliefs.

Clinicians have shown a preference for developing stronger beliefs in

personal control among their patients and have developed strategies to

facilitate this process. It can be concluded that researchers and

clinicians alike attach a great deal of significance to those factors

that operate so as to effect changes in personal control beliefs

over time.

Environmental Influences

It has been suggested that one of the central factors involved in

a belief in personal control concerns the individual's belief in the

predictability of his environment and the persons in it, such that

as the predictability of an environment increases, personal control

beliefs of persons within that environment should also increase. This

thesis will be discussed from the vantage point of the normative struc-

ture of an environemnt and its suggested relation to personal control

beliefs. In addition, the potential importance of the variable of

population size to the normative structure will be described.

The normative structure of a sociocultural environment refers to

the influence of the reference group in setting standards, or norms, for

the behavior or beliefs or an individual. "Norms are not simply

average or modal behaviors; instead...norms are to be construed as









socially defined standards or expectations about what are appropriate

modes of action in various social situations" (Jessor et al., 1968, p.60).

In an environment where there is considerable consensus or agreement

among its members about appropriate behaviors or beliefs, the individual

would tend to feel more comfortable and attuned to the predictiveness

of that environment.

On the other hand, Durkheim (1951) defined anomie as a property

of a sociocultural system in which norms are no longer operative or

effective guides to action. The individual's sense of social cohesion and

social orderliness is weakened and the predictiveness of his environment

is lessened. The lack of consensus or agreement among members about

the prescriptions and the proscriptions for behavior lead to the condition

where no one knows how to behave or what to expect from others. Since

the individual has difficulty predicting the outcomes or reinforcements

of his behavior, it is difficult to develop a belief in personal control.

Moreover, as Merton (1957) has suggested, in a state of anomie

individuals perceive that they are unable to attain their valued goals

by legitimate behaviors, thus experiencing an increased pressure to

attain those goals by illegitimate means. This produces further tension

and breakdown in the behavioral prescriptive and proscriptive structures.

The importance of the normative structure to the construct personal

control seems considerable. Swanson (1968) compared the amount of con-

sensus on behavioralprescriptions as measured by a six-scale, sixty

item measure of attitudes toward prison behaviors given to all inmates in

a correctional setting. This normative measure was pertinent to behavior

of both staff and inmates. Swanson was most interested in how group

influence or group consensus on these attitudes might influence an inmate









to change in a rehabilitative direction. While he found that majority

group consensus was greater than minority group consensus, and that

clear differences between prison dormitories could be obtained, he

failed to show the predicted relationships with personal control. It

seems important to restudy this question with another sample population.

Therefore, it will be maintained that to the degree that there is norma-

tive cohesion (or consensus on what are appropriate behaviors) in a

social system, to that degree it becomes easier for an individual to

predict the outcomes and reinforcements of his behavior with consis-

tency. The individual thus finds it easier to develop a belief in

personal control. Conversely, to the degree there is normative incon-

sistency, the more difficult it becomes for a person to make those pre-

dictions and the less likely it is the individual will develop a belief

in personal control.

Finally, it is proposed that as the number of the persons within

an environment (the gross population size) increases, the degree of

personal control within that environment decreases. This is suggested

from Zimbardo's (1969) discussion of the theoretical and empirical

underpinnings of a hypothetical construct he calls deindividuation.

His research findings support the contention that group size is directly

related to the experience of anonymity and the lack of social evaluation

and sanction: "If others can't identify or single you out, they can't

evaluate, criticize, judge, or punish you, thus there need be no concern

for social evaluation" (1969, p.255).

It appears then that the person engulfed in a large anonymous

social system would be more likely to adopt an external frame of refer-

ence for understanding the locus of his reinforcements, whereas the









individual residing in a smaller unit would be more likely to experience

the social intimacy with others in that environment and "learn the ropes"

more quickly. Thus, that individual will be more likely to develop a

belief in the predictability of outcomes of his behavior.



Alternative Conceptualizations of Personal Control

Rotter's original conceptualization of personal control was as a

bipolar unidimensional concept with external control at one end of the

continuum balanced by internal control at the other. The terms were de-

fined in this manner: "Internal control refers to the perception of

positive and/or negative events as being a consequence of one's own

actions and thereby under personal control; external control refers to

the perception of positive and/or negative events as being unrelated to

one's own behaviors in certain situations and therefore beyond personal

control" (Rotter, Seeman, and Liverant, 1962, p.473).

However, Rotter (1966) later clouded the above definitions by

including a world view in his definition of personal control (i.e. the

personal belief about the reinforcement contingencies for the generalized

other). A growing number of researchers and theorists have called into

question the unidimensionality of a concept that includes both the personal

as well as the generalized other (Swanson, 1970a). Dies (1968),

suggesting that major aspects of personal control are not measured by

Rotter's I-E Scale, concludes the scale is weighted toward social and

political factors rather than toward personal fa-tors.

Gurin, Gurin, Lao, and Beattie (1969) and Lao (1970) suggest

that the concept of internal-external control has no simplistic applica-

tion to Black youth. They factor analyzed the I-E Scale responses of









1695 Black students and found four independent factors. The first two

they labelled Control Ideology and Personal Control. Respectively, these

referred to the generalized other and personal dimensions of control

suggested above. The third factor, System Modifiability, referred to

the degree the individual believed that war, racial discrimination, and

other worldwide problems could be attenuated. The fourth factor, Race

Ideology, contained most of the race-related items.

Mirels (1970) performed a varimax rotation on the I-E measure

of 316 college students. He also reports the identification of two fac-

tors, one concerned with personal factors of control, the other concerned

with the more general factors of control over political and world affairs.

Taken together, these findings suggest that Rotter's I-E instrument has

some inherent conceptual weaknesses that bring into question the validity

of a unidimensional personal control construct. Instead, they strongly

suggest the notion that personal control may be a multi-dimensional con-

struct, at least in the sample populations studied thus far.

Swanson (1970a), in a comprehensive literature review, has suggested

that additional conceptual imprecisions in the personal control concept

involve the over-generalized definition of control and the assumption

that internality is the polar opposite of externality. Specifically, he

suggests that a person may be internal in one need area and, at the same

time, external in another. Thus, a definition of personal control which

cuts across need areas is necessarily weakened. For example, it is easy

to imagine a very accomplished student who always receives A's as well

as the plaudits of his teachers. This identical student, at the same

time, remains socially isolated and conspicuously unsuccessful in his

heterosexual endeavors. His perception may well be that his academic









accomplishments (need for achievement) are due to his diligent study

habits and his quick mind (internal locus of control), while his hetero-

sexual failures (need for affection) are due to poor luck in meeting the

right girl, the impersonal quality of the university setting, etc. (exter-

nal locus of control). The logic of the position that an individual is

either internal or external in all situations and/or all need areas is

easily refutable.

The final conceptual weakness to be discussed concerns the hypo-

thesized polar opposition of internal and external loci of control.

Rotter assumed that a person who attributed his rewards and his punish-

ments to his own efforts would also deny any influence of luck and chance.

This view oversimplifies human nature. The ingredients of any goal achieve-

ment can, at once, be seen as a degree of individual efforts, a little luck,

some chance, and perhaps the influence of some systematic outside force.

All of these conceptual difficulties have a serious bearing on two

separate issues which are important to this study. The first of these

concerns Rotter's notions of the relationship between adjustment (a term

that shall here be redefined as "success in the environment") and per-

ceived locus of control. The second concerns the difficulties in making

valid and reliable measurements of the construct personal control. Each

of these concerns will now be discussed.

Success in the Environment

One area of the research literature on personal control that is

replete with confusing and contradictory findings is that area con-

cerned with the relationship between personal control and adjustment.

Rotter (1966) has suggested that hypothesis of a curvilinear relationship

between personal control and adjustment. He theorized that those









individuals who were extremely internal may have such a Napoleonic view-

point as to their own control capacities as to be seriously maladjusted.

Likewise, those who were extremely external may believe that everything

is beyond their control and they become vegetatively helpless and hope-

less. Even though the research findings of the past decade have not

supported his hypothesis (Phares, 1976), he recently maintained this is

due to instrument insensitivity and restricted subject samples (Rotter,

1975), rather than to an inadequate theoretical formulation.

Phares (1976) agrees that research samples have not typically

included the extremes in scores and thus there has not been an adequate

testing of Rotter's hypothesis. Moreover, Phares suggests that the

criteria of adjustment have never been agreed upon in various studies

and have too often reflected the researcher's personal biases. He

suggests that future research operationalize its definitions of adjust-

ment in a more concrete, behaviorally oriented manner.

Instead of supporting the curvilinearity notion, much of the

research has suggested a linear relationship between personal control

and measures of adjustment. It appears that those persons who believe

in personal control tend to exhibit more initiative in looking for those

behavioral alternatives which lead to successful positive outcomes in

controlling their environments (Phares, 1965; Seeman, 1963; Seeman and

Evans, 1962). Hersch and Scheibe (1967) reported that internal college

students were more effective as volunteer workers in a mental hospital

setting than were external students. In a prison population, Hood et al.

(1966) found a significant relationship between illegitimate, deviant,

or troublemaking behavior and belief in personal control. These findings

suggest that individuals believing in personal control tend to get into









less trouble in their environments and have a higher expectation that

they will achieve their goals. Moreover, they are likely to be per-

ceived by others in their environment as leaders, or as successful

models, and as more socially inclined. Finally, as has been suggested

earlier, they should be less alienated. Taken together, this cluster

of variables can be viewed as measuring "success in the environment."

While it is concluded that internals are, in general, more

successful than externals, it is suggested that this difference should

be small and reflective of a certain crudeness in conceptualization.

This follows from the earlier discussion of some of the conceptual

weaknesses inherent in treating personal control as a unidimensional

variable operating consistently across need areas and situations.

Swanson (1970a) has suggested that some environments demand of an indi-

vidual a specific internal or external position. He suggests that the

person who is able to represent his situation most veridically may indeed

be the most successful person. For example, in an environment charac-

terized by its externalizing qualities, an internal individual may not

be as successful as an external who represents the situation most ver-

idically. Conversely, in an environment characterized by its internal-

izing qualities, an external should not be as successful as his internal

counterpart. While these two formulations hold the environment constant,

the related formulations with control beliefs constant can be stated in

the following fashion: The internal individual residing in an external

environment should be less successful than the internal in an internal

environment. And, the external individual residing in an internal environ-

ment should not be as successful as the external in an external environ-

ment.









Some examples may make these formulations more clear. Let us

suppose we have two persons in an externalizing situation, one internal

and the other external. For the purposes of illustration, let us further

suppose our two individuals are black men who seek membership in the

Ku Klux Klan. In this situation, a belief denying personal control

may be a totally rational conclusion based upon the objective reality

of the situation, whereas a belief in personal control may be an extreme

illogicism. Clearly, the individual likely to be most successful in

coping with this hypothetical situation should be the one who is able

to represent the objective reality of the situation most veridically.

Another example should suffice. Let us suppose we have two more

individuals, both of whom possess strong beliefs in personal control.

Let us place one in an environment where the objective reality is that

the individual does control the outcomes of his behavior. The other

we shall place in an environment where the objective reality is that

the individual does not control the outcomes of his behavior. Which of

these two should be most successful in their respective environments?

Clearly, the individual whose phenomenal reality departs from the objec-

tive reality would be predicted as being less successful.

It can be seen that these formulations are somewhat in competition

with Rotter's hypothesized curvilinear relationship between personal

control and adjustment. One means of studying the relationship between

personal control and success in the environment from these two theoretical

viewpoints is to employ a normative structure assessment to predict the

environmental demands for specific control positions. Internal environ-

ments could be defined as ones in which there exists normative consensus,

while external environments could be defined as lacking in consensus.









Internal and external scoring individuals could then be located in their

internal or external environments and the subsequent relationships could

be delineated.

Measurement of Personal Control

Some of the conceptual imprecisions in the personal control construct

have been discussed above. In a comprehensive methodological study designed

to clarify some of these ambiguities, Swanson (1970b) developed alternative

control measures and compared their predictive utility and scale charac-

teristics to that of the widely used Internal-External Control Scale,

developed by Rotter et al. (1962). Swanson varied his measures by five

need domains, by locus of control (self, chance, or systematic powerful

forces), and by self versus other orientation in the wording of the scale

items. Using prison inmates as subjects, only modest improvements in scale

characteristics and in prediction of various personality and behavioral

variables resulted with the specificity of the various subscales. There-

fore, it is suggested that a generalized, non-need area specific measure

of internal-external control may be the most suitable personal control

measurement device, at least for use in correctional settings.



Choice of Research Setting

It is the overall aim of this research project to study the inter-

relationships between the personal control construct and a number of

personality, behavioral, and systems variables. To best understand

these interrelationships as they operate in their real-life manner, a

natural community seemed best for experimental study. However, in

order to be optimal for study, it was felt the community must be suffi-

ciently self-contained so as to provide some control of the number of

agents of change influencing its members.









Correctional settings were selected as ideal for this study.

There is nothing contrived about these natural research settings.

Real-life sanctions and reinforcements are at play in their everyday

functioning. Moreover, correctional institutions offer the degree of

control necessary to carry on meaningful research. They are relatively

self-contained, they contain populations small enough to permit data

collection on a large enough scale, and they may contain various sub-

cultures within the larger organization.

Since the influence of contrasting institutional control and/or

opportunity structures upon the construct personal control seemed so

pertinent, six correctional environments were selected for inclusion

in this study. These settings were selected so as to insure variation

in the personal freedoms allowed the inmates and the number of rules and

regulations imposed upon them. Thus the shades of sociocultural variation

are provided by these contrasting correctional settings.



Application to Corrections

That this study achieves practical relevance to the field of

corrections is readily apparent. Among the most pressing social prob-

lems currently facing our complex society are the related concerns of

crime and corrections. Policy-makers and lay citizens alike decry the

rising crime rate and point with particular frustration to the indi-

viduals who come to the attention of the criminal justice system for

the second, third, fourth, fifth...time. With the recidivism rate

unacceptably high, there is almost universal agreement that the insti-

tutions whose mandate is the management and correction of the convicted

criminal offender are largely ineffective in rehabilitating the offender









(Lipton, Martinson, and Wilks, 1975). In recent times, however, the

problem has crescendoed to new heights as the number of convicted

offenders has increased. Prison facilities are often severely over-

crowded and woefully understaffed--a condition often remedied by early

release programs distasteful to large segments of the general population.

Given that there are sidely held doubts about the effectiveness

of current corrections facilities, programs, and techniques, and given

that society will almost certainly continue to demand punishment/

treatment/correction of the offender, those charged with the responsibil-

ity of resocializing the offender would seem to be especially pressured to

develop and refine truly "corrective" environments. In order to devise

innovative and effective procedures that will encourage adjustment to

our society, these policy-makers and practitioners are in need of accurate

knowledge about the correctional environments which they currently manage.

Research is needed to understand, evaluate, and retool the correctional

environment.



The Research Plan

This study contributes to the understanding of correctional environ-

ments by identifying variables important to changing attitudes and behaviors

of the convicted offender. Previous understanding of the criminal justice

system has been most often limited by the lack of theoretical frameworks

and methodological consistency. This research is grounded in a compre-

hensive theoretical design based upon the formulations of Rotter (1954)

and Jessor et al. (1968), as well as the work of Swanson (1968, 1970a,

1970b, 1973a, 1973b) in this area. This study is a part of a larger

research effort, the purpose of which is to embed the understanding of









various correctional strategies in a comprehensive network of personality,

sociocultural, behavioral, and systems variables. The focus of this re-

search endeavor centers upon the construct personal control as it relates

to this complex of variables.

The overall research plan for this study is to compare and con-

trast selected cross-sectional samples of male and female inmates from

contrasting correctional institutions. One criteria for distinguishing

correctional settings is that of opportunity. High opportunity institu-

tions having high staff/inmate ratios, more diverse program alternatives,

and positive incentives in controlling behavior, are contrasted with low

opportunity institutions having low staff/inmate ratios, less diverse

program alternatives, and negative incentives in controlling behavior.

In order to provide a comprehensive background in which the research

hypotheses of this study can be interrelated and understood more clearly,

it is now necessary to describe the theoretical and conceptual underpinnings

from which this study is derived.

Social Learning Theory in the Field

Jessor et al. (1968) have developed a field theory elaboration of

basic social learning theory and have coordinated it with Merton's (1957)

sociological opportunity theory, suggesting that high objective opportunity

for reward in a social system will enhance individual feelings of high

expectation for reward. Including the conceptualizations of Cloward and

Ohlin (1960), Cohen (1955), and Sutherland (1955), Jessor et al. developed

three predictor systems of variables in an attempt to understand and

explain deviant behavior. The three predictor systems were: 1) the

personality system, which included the perceived opportunity structure,

the personal belief structure, and the personal control structure;








2) the sociocultural system, which included the opportunity structure,

the normative structure, and the social control structure; and 3) the

socialization system, which included the parental reward structure, the

parental belief structure, and the parental control structure. As can be

quickly noticed by examining the specific structures contained within

the three systems, these predictor systems were conceptualized in parallel

fashion so that relevant aspects of the personality could be related to

relevant aspects of the environment. The socialization system was con-

strued as including those processes that lie at the interface, and mediate

between, the personality and the sociocultural systems.

This study contains variables quantified from each of these

predictor systems. Let us now turn to a description of the predictor

systems contained within this research.

Following the conceptualization of Jessor et al. (1968), the

personality system consists of three structures: the structure of

perceived opportunity, the belief structure, and the personal control

structure. Each structure is related to the development and occurrence

of anti-social behavior (Jessor et al., 1968; Mobley and Swanson, 1972;

and Swanson, 1973b).

The structure of perceived opportunity includes the needs of the

individual and the expectation he has for fulfilling those needs. Low

expectations for attaining valued goals have been shown to be an impor-

tant determinant of crime, delinquency, and troublemaking in prison

(Mobley and Swanson, 1972; Mobley and Swanson, 1974; Swanson, 1973a;

Swanson, 1973b; Wood et al., 1966). Two perspectives are utilized in this

structure: 1) perceived opportunity for the attainment of long-future

goals such as achievement, love and affection, independence, recognition,








and family life; and 2) perceived opportunity within the prison context,

for s.; things as vocational training, interpersonal development,

understanding of self and relationships with authority.

The belief structure includes three cognitive orientations which

have been conceptualized as important to the understanding of criminality

and conformity. The belief in personal control refers to the person's

attribution of causality to himself and his behavior or to capricious

factors such as luck, fate, or chance. It is held that to the degree

that the individual believes that the consequences of his actions are

dependent upon his own behavior, to that degree he should engage in

planned, directed goal-oriented work (Mobley and Swanson, 1972; Rotter,

1966; Swanson, 1968, 1970b; Swanson, 1973a, 1973b; Wood et al., 1966).

The second belief alienation refers to social estrangement, the feelings

of isolation from others, and lack of meaning felt for traditional life

roles such as marriage, work, and friendship. An individual with such

a cognitive orientation should demonstrate a lessening of productive

activity and a withdrawal from active, goal-seeking behavior (Jessor et al.,

1968; Jessor et al., 1970; Swanson, 1973a, 1973b). The third belief

interpersonal trust refers to the belief that one can rely and depend

upon others to facilitate the attainment of life goals rather than act

as a frustration or a block to the attainment of such goals (Hamsher et

al., 1968; Hochreich and Rotter, 1970).

The final structure contained within the personality system is

the personal control sturcture. Included here are three variables that

directly regulate criminality and conformity. Attitude toward deviance

has been shown to be the most consistent predictor of delinquence among

youth (Mobley and Swanson, 1972; Swanson, 1973a). Future-time perspective









is predictive of goal-oriented contributive behavior. The third variable,

attitude toward earning a living, discriminates between work-release

programs as well as state prison systems that differ in the utilization

of prison industries and vocational training (Swanson, 1973a).

The sociocultural system includes three aspects of the prison

environment, each providing a source of influence for personality and

behavioral change. The formal structure, conceptualized as the oppor-

tunity structure, includes formal program policies, staff attitudes

and behavior, and the objective opportunity provided the inmates for

personal development. The informal structure, termed the normative

structure, includes inmate prescriptions for staff treatment, staff

custody, and staff punishment, as well as for inmate self-improvement,

inmate avoiding trouble, and inmate opposition. Contact with the out-

side community, conceptualized as the social control structure, is

similar to the differential association approach of Sutherland (1955),

in which delinquent or socially acceptable behavior is learned, depending

upon the behavior of one's role models.

The socialization system is that set of variables which stand at

the interface of the personality and the environment. Personaltiy thus

depends upon the process of socialization or enculturation and this

approach is exemplified by such terms as "patterned exposure," "training,"

"modeling," and "learning." The informal socializations which take

place on an incidental basis in everyday living, the consequences of

exposure to formal or informal role models, etc.--these are examples of

the socialization system processes in need of further understanding.









Hypotheses

The literature on the construct personal control has been re-

viewed. A number of related variables have been discussed and their

expected relationships with personal control have been delineated.

Correctional institutions have been chosen as research settings with

inmates as subjects. Hypothesized relationships involving personal

control are ordered into five categories: 1) Relation to Personality;

2) Distribution of Personal Control Within Social Settings; 3) Changes

in Personal Control Beliefs; 4) Environmental Influences; and 5) Success

in the Environment.

The hypotheses for this research are the following:

1. Relation to Personality

A. Belief in personal control should correlate positively

with expectations for goal attainment.

B. Belief in personal control should correlate negatively


with alienation.

C. Belief in personal

with self-esteem.

D. Belief in personal

with interpersonal

E. Belief in personal

with an intolerant

F. Belief in personal


control should correlate positively



control should correlate positively

trust.

control should correlate positively

attitude toward deviance.

control should correlate positively


with prescriptions for behaviors which maximize staff

rewards and negatively with prescriptions for behaviors

which maximize staff punishments.









G. Belief in personal control should correlate positively

with favorable attitudes toward environmental

opportunity and authority structures.

2. Distribution of Personal Control Within Social Settings

A. Persons living in environments defined as relatively

high in opportunity should believe in personal control

to a greater extent than persons living in environments

defined as relatively low in opportunity.

B. Persons of ethnic or minority status should believe in

personal control to a lesser extent than persons of

Anglo or majority status.

C. Persons of lower socioeconomic status should believe in

personal control to a lesser extent than persons of

higher socioeconomic status.

3. Changes in Personal Control Beliefs

A. Belief in personal control bears a curvilinear relation-

ship with a person's length of stay in a novel and con-

fining environment such that belief in personal

control is minimized upon entry and again upon exit

from that environment.

4. Environmental Influences

A. Persons residing in an environment where there is low

consensus on what are appropriate behaviors should believe

in personal control to a lesser extent than persons re-

siding in an environment where there is high consensus.

B. Persons residing in environments which contain small

numbers of people should believe in personal control









to a greater extent than persons residing in environ-

ments which contain large numbers of people.

5. Success in the Environment

A. Belief in personal control should be positively corre-

lated with measures related to success in the environment.

B. Within environments characterized by low consensus on

appropriate behavior, persons who do not believe in

personal control should be more successful than persons

who believe in personal control. Conversely, within

environments characterized by high consensus on appropri-

ate behavior, persons who believe in personal control

should be more successful than persons who do not believe

in personal control.

C. Persons who believe in personal control and reside in

environments with high consensus on appropriate behavior

should be more successful than their counterparts in

environments with low consensus. Conversely, persons who

do not believe in personal control and reside in

environments with low consensus on appropriate behavior

should be more successful than their counterparts in

environments with high consensus.















METHODS


Subjects

Subjects were sampled from six correctional institutions located

in Florida, Colorado, and Kentucky. Five institutions were state facil-

ities, three of which were for male offenders and two for female offenders.

The sixth institution, in Kentucky, was a co-correctional federal

facility.

Data for this study were gathered in conjunction with a larger

research effort entitled "Socialization after Childhood: A Longitudinal

Examination of Contrasting Correctional Institutions for Male and Female

Offenders." This study was based upon data gathered in the second year

of the larger project. During the second year of data collection, a

random sample of inmates, stratified by race, was drawn from population

rosters provided by the institutions. All subjects who had participated

in the first year of the project were included and all new subjects had

been incarcerated for one year or less. The subject sample for this

study included 538 inmates. Of these, 381 were males and 156 were

females. Of the male sample, 56% were White, 37% were Black, and 7% were

non-Black minorities (Mexican American, Cuban American, American Indian).

Of the female sample, 42% were White, 53% were Black, and 5% were

non-Black minorities.

Correctional officers were selected to rate certain aspects of

inmate success in the environment. While the procedures employed to









select correctional officers did not insure a random sample, the sample

was considered to be representative of the overall grouo of officers

for the following reasons. Rating forms were given to supervisory

personnel with instructions that they be completed by the officers who

were most experienced on the various housing units and who knew the

inmates best. Further instructions were given to the effect that offi-

cers participating as raters should represent both day and evening

shifts as well as first line supervisory personnel. Sixty-two correc-

tional officers from five institutions were included in the sample.

Ratings by correctional officers from the sixth institution, the Men's

Correctional Institution of Colorado, were lost during the data

collection process.

Inmates rated by correctional officers constituted a random

sample of those selected for the larger study, with the constraint that

ten or fewer inmates from each housing unit be drawn into the sample.

This was done to keep the number of inmates to be rated by each correc-

tional officer at a manageable level. This resulted in an original

N = 250. Of these, 196 actually participated in the larger study and

were therefore eligible for analysis. The remaining 54 inmates had

declined to participate in the larger study when informed that the study

was of a voluntary nature. Of the 196 inmate subjects who actually did

participate in the larger study, ratings were made on 131 males and

63 females, and as well on 89 Whites, 85 Blacks, and 7 other minority

subjects.









Institutions

Institutions were selected for their contrasting opportunity

structures. An attempt was made in the selection process to maximize

environmental contrasts, so that the research results would be general-

izeable to the largest possible correctional population.

Institution 1

Florida Correctional Institution for Men (FCIM) at Lowell, Florida,

is a medium/minimum security institution. Security measures are relatively

unobtrusive. The facility is surrounded by one fence and a manned control

booth where a single officer is posted to monitor and control individuals

entering or leaving the compound. The average sentence for an inmate is

approximately one and one-half years, predominately for victimless crimes

and property crimes. The institution is geared toward vocational and

educational programs and the stated program goals are toward the reinte-

gration of the inmate into community life. The staff/inmate ratio was

approximately 1/5 at the time of data collection.

The institution has a designed housing capacity of 250. At the time

of data collection, the inmate population was about 300, as a double-wide

mobile home had been set up as an honor unit for those inmates that were

in work-release programs. Other inmates were all housed in the same

building consisting of four dormitories with bunk beds. Inmates were ran-

domly assigned to each living area. The sample included 69 inmates drawn

from all of the four dormitories.

Institution 2

Union Correctional Institution (UCI) at Raiford, Florida, is a

medium/maximum security facility for male offenders. This institution is

surrounded by three fences, one of which is electrified, and manned gun









towers. The average sentence is approximately 10 years and the most pre-

valent types of offense are crimes associated with violence, such as mur-

der, rape, or armed robbery. Most of the programs offered are custodial

in nature with less than 50% of the population engaged in educational or

vocational programs. The orientation of the institution is toward con-

formity and obedience. At the time of data collection, the staff/inmate

ratio was approximately 1/9.

The inmates are housed in two main buildings. One building con-

tains dormitories with single and bunk beds, and individual rooms. Assign-

ment to living areas represents a reward-type structure, based upon staff

recommendations. The institution's designed housing capacity is 1400

inmates. At the time of the data collection, the population was just

under 2000 inmates. Consequently, nearly every housing unit was seriously

overcrowded. The sample at this institution included 196 inmates drawn

from 17 living areas.

Institution 3

Men's Correctional Institution of Colorado (ColoM) at Canon City,

Colorado, is a medium security institution for male offenders. The facility

is surrounded by fence with control towers at the entrance. The institution

was beginning to receive more violent offenders at the time of data collec-

tion and security measures were being increased. Thus, the institution's

previous efforts toward re-education and development were being de-emphasized.

The staff/inmate ratio was approximately 1/10 at the time of data collection.

Inmates are housed in dormitories with single beds or blocks of

single cells. Assignment to living area is by choice. This institution

had a housing capacity of 454 and at the time of data collection, the

population was crowded. The sample from this institution included 61

inmates drawn from two living areas.









Institution 4

Florida Correctional Institution for Women (FCIF) at Lowell,

Florida, is the only correctional institution for female offenders in

Florida. It is a minimum/maximum security institution housing all

female offenders. The facility is surrounded by a double fence. The

average sentence is approximately two years. The orientation is toward

re-education and development and the programs are geared to assist rein-

tegration into community life. There is a diversity of vocational and

educational programs offered. The staff/inmate ratio was approximately

1/4 at the time of data collection.

Inmates are housed in three types of structures: cells, dormi-

tories, and trailers. Housing assignment is based on security risk.

The designed capacity for FCIF is 400 inmates but at the time of data

collection, there were more than 600 inmates. The sample from this

institution included 77 inmates drawn from seven living areas.

Institution 5

Women's Correctional Institution of Colorado (ColoF) at Canon

City, Colorado, is the only correctional institution for female offenders

in Colorado. It contains all security levels and all offender types.

The facility is surrounded by fence. Halls and entrances are monitored

by television cameras. The average sentence is approximately two and

one-half years. The orientation of this institution is toward re-

education and development. However, the programs offered are less varied

than the other female institution sampled in Florida. The staff/inmate

ratio was 1/3 at the time of data collection.

Inmates at ColoF are housed in identical single rooms, with

housing assignments made on a random basis. The institution is designed









to house 90 inmates and at the time of data collection, the population

was 77. The sample from this institution included 32 inmates drawn from

one living area.

Institution 6

Federal Correctional Institution (LexM and LexF) at Lexington,

Kentucky, contrasts markedly with each of the other institutions included

in this study. The population of this federal institution is approximately

600, with 400 males and 200 females (hence, the designations LexM and LexF

for the male and female populations respectively). Security is down-

played in this setting but a taboo against sexual behavior is strictly

enforced. Considerable emphasis is placed upon giving the inmates rich

educational and vocational opportunities. Each inmate is employed or is

in school each day. A large professional social service staff supplements

a well-trained correctional staff. The staff/inmate ratio was approxi-

mately 1/2 at the time of data collection.

Housing in this institution is not crowded and ranges from private

and semi-private rooms to four-person rooms. In addition, therapeutic

communities are offered to those who desire or need this experience.

The sample from this institution included 100 inmates, of which 53 were

males and 47 were females. Inmates were drawn from five living areas,

one of which houses both male and female offenders.

As stated above, six correctional institutions, three male, two

female, and one co-correctional, were selected so as to maximize the

contrasts in their respective opportunity structures. Three members

of the research staff who were involved in the data collection process

in each of these institutions independently ranked the male and the

female institutions "as to the opportunities they provide in such areas









as interpersonal development, vocational development, and academic

development." The following is the average ranking of the environmental

differences among the six institutions. (Note: LexM and LexF were

treated as separate institutions.) The average Spearman rank-order

correlation for ranking male institutions was rs = .984, and for female

institutions, rs = .945 (N's too small to achieve significance).

Male Institutions

Avg. Rank

LexM 1 Defined as the highest opportunity
structure among the male institutions
in the study.

FCIM 2.5 Defined as a medium opportunity
structure.

ColoM 2.5 Defined as a medium opportunity
structure.

UCI 4 Defined as the lowest opportunity
structure among the male institutions
in the study.

Female Institutions

LexF 1 Defined as the highest opportunity
structure among the female institu-
tions in the study.

FCIF 2.33 Defined as a low opportunity
structure.

ColoF 2.67 Defined as a low opportunity
structure.



Instruments

The data collection procedures involved in this study were group

administered questionnaires requiring the subject to self-report about

behavior, personality attributes, formal and informal interactions, as

well as demographic variables. The questionnaire included a large









number of measures previously developed or modified for use with prison

populations. Many of these measures have been derived directly or in-

directly from social learning theory, as developed by Rotter (1954).

The scales to be used include the following:

Attitude Toward Prison Behaviors

This scale was developed by Swanson (1968, 1970b, 1973a) to mea-

sure the shared attitudes of inmate groups about the appropriateness

of different behaviors in a correctional setting. In addition, the

scale may be utilized as a measure of individual orientation toward

particular prison behaviors. Six subscales represented inmate-

recommended, or inmate-prescribed, behaviors for staff and inmates in

each of three contrasting types of correctional institutions: those

oriented toward punishment of the inmate, those oriented toward custody

of the inmate, and those oriented toward treatment of the inmate (Street,

1965). Twenty-nine items described staff and inmate behavior appropriate

for each of the three situations. Subjects were asked to respond to these

29 behavioral descriptions on a five-point continuum: absolutely should,

should, may or may not, should not, and absolutely should not. Items

were scored so that a high score meant a prescription (recommendation)

of its orientation and a low score, a proscription. The six subscales

arranged in pairs appropriate to each of the three situations were as

follows: Staff Punishment-Inmate Opposition, Staff Custody-Inmate

Avoiding Trouble, and Staff Treatment-Inmate Self-Improvement. The items

composing these paired scales represented behaviors ranging from less to

more cooperative. The six scales are listed below and are followed by

an example of an item from that scale. Scott's (1960) Homogeneity

Ratio (H.R.) and Cronbach's (1954) Alpha, as reported by Swanson









(1968, 1970b) in previous research with these scales, are included as

an indicator of the internal consistency of these measures. (A copy of

the complete instrument can be found in Appendix A.)

Staff Orientation

1. Staff Punishment (SPU)

Staff members see to it that inmates have a hard time
here to make up for what they did on the outside.
(H.R.'s were .28 and .30 in previous research and Alpha's
were .79 and .75.)

2. Staff Custody (SCU)

Staff members see an inmate as someone to be controlled.
(H.R.'s were .20 and .16 in previous research and Alpha's
were .68 and .53.)

3. Staff Treatment (STR)

Staff members try to teach an inmate skills that will
help on the streets.
(H.R.'s were .32 and .32 in previous research and Alpha's
were .82 and .78.)

Inmate Orientation

1. Inmate Opposition (IOP)

Inmates cause as much trouble as they can.
(H.R.'s were .38 and .38 in previous research and Alpha's
were .86 and .81.)

2. Inmate Avoiding Trouble (IAT)

Inmates try to get along by keeping their mouth shut
around the staff.
(H.R.'s were .16 and .17 in previous research and Alpha's
were .58 and .58.)

3. Inmate Self-Improvement (ISI)

An inmate really tries to learn something in work release
that will be of use later.
(H.R.'s were .31 and .29 in previous research and Alpha's
were .82 and .74.)









Definition of Situation

This measure was developed by Wood et al. (1966) and modified by

Swanson (1968, 1970b, 1973a). Containing 36 items, this measure reflected

six attitudinal orientations toward the institutional environment. Sub-

jects were asked to respond to each item on a five-point continuum:

strongly agree, agree, neither agree nor disagree, disagree, and

strongly disagree. Four scales with 22 total items represented the

individual's definition of the institutional opportunity structure.

These included: Opportunity to Develop a Personal Identity, Opportunity

for Social/Interpersonal Development, Definition of Personal Commitment,

and Value of Commitment. The sum of these four measures yielded an over-

all definition scale. The two remaining scales, with a total of 12 items,

reflected the orientation toward the institution as an authority structure.

These were the Attitude Toward Rules and Regulations and the Attitude

Toward Authority Figures scales. The sum of these two scales yielded an

Attitude Toward Authority in General measure. The 36 items were presented

in scrambled order with five buffer items interspersed among them. Half

of the items were reverse-worded so as to balance the effect of acqui-

escence set. The scales were scored so that a high score represented a

favorable attitude and a low score represented an unfavorable attitude.

The four opportunity and two authority subscales are listed below

followed by an item from that scale and by H.R.'s and Alpha's obtained

in previous research (Swanson, 1968, 1970b). (See Appendix B for a copy

of this instrument.)

The opportunity items were divided among four subscales:









1. Opportunity to Develop a Personal Identity (PID)

By the time most people leave this place they have a
better idea of what they want out of life than they
had before.
(H.R.'s were .23 and .36 in previous research and Alpha's
were .64 and .74.)

2. Opportunity for Social/Interpersonal Development (SID)

If you don't learn anything else here, at least you learn
how to get along with others.
(H.R.'s were .21 and .20 in previous research and Alpha's
were .61 and .60.)

3. Definition of Personal Commitment (DPC)

An institution like this is a place where a person can
get his or her feet on the ground and begin to make a
fresh start.
(H.R.'s were .16 and .41 in previous research and Alpha's
were .53 and .74.)

4. Value of Commitment (VOC)

Whether you like it or not, at least you have a chance to
learn something useful while you're here.
(H.R.'s were .27 and .26 in previous research and Alpha's
were .68 and .68.)

The overall definition scale (DEF) was the sum of the above four

subscales. H.R.'s were .21 and .26 in previous research and Alpha's

were .86 and .68.

The authority items were divided between two subscales:

1. Attitude Toward Rules and Regulations (ATR)

Like most other things, there are probably some good
reasons for most of the regulations on the book.
(H.R.'s were .25 and .33 in previous research and Alpha's
were .66 and .75.)

2. Attitude Toward Authority Figures (ATA)

Officers are trying to help inmates even though they're
getting after them all the time.
(H.R.'s were .27 and .31 in previous research and Alpha's
were .69 and .72.)










These two subscales were summed to form an Attitude Toward

Authority in General (AGA) scale. H.R.'s for this scale were .22 and

.28 in previous research and Alpha's were .77 and .82.

Attitude Toward Deviance (ATD)

This scale measured the degree to which the respondent tolerated

or proscribed deviant behavior (Mobley and Swanson, 1972). It consisted

of 36 behavioral descriptions of socially defined deviant acts of varying

severity. Respondents read each statement and checked one of eight

possible responses indicating their degree of tolerance for the behavior

described. Responses ranged from "serious enough for the death penalty"

to "not serious at all, nothing wrong with it." Responses were scored

so that the higher the score, the more intolerant of socially defined

deviance.

An example of an item from the Attitude Toward Deviance scale is

listed below and is followed by the internal consistency data obtained

in previous research. (See Appendix C for a copy of this instrument.)

Kidnapping someone and holding them for ransom. (check one)
1. Serious enough for the death penalty.
2. Serious enough to put a person in prison for life.
3. __ Serious enough to put a person in prison for several
years.
4. _Serious enough to put a person in prison for about
a year.
5. Serious enough to put a person on probation for
about a year.
6. Serious enough to give a person a good warning.
7. Not serious enough to punish a person in any way.
8. Not serious at all, nothing wrong with it.
(H.R. obtained in previous research was .33 and Alpha
was .94.)

Alienation (AL)

This scale measured social estrangement, the feeling of isolation

from others,and lack of meaning felt for traditional life roles









(Jessor et al., 1968). It included 8 items which elicited the respondent's

level of agreement with a statement indicative of alienation. Responses

were indicated on a four-point continuum: strongly agree, agree, disagree,

and strongly disagree. The higher the score, the more alienated the

subject.

An example of an item from this scale is listed below and is

followed by the internal consistency data obtained in previous research

with this scale. (See Appendix D for a copy of this instrument.)

I often find it difficult to feel involved in the things
I'm doing.
(H.R. in previous research was .25 and Alpha was .80.)

Interpersonal Trust (IPT) and Self-Esteem (EST)

Items from these two scales were combined into one 30 item ques-

tionnaire. Fourteen items composed the Interpersonal Trust measure and

8 items composed the Self-Esteem measure. Half of the items for each

scale were reverse-worded to balance for effects of acquiescence set.

Eight additional items were included as buffer items designed to obscure

the purpose of the test.

The Interpersonal Trust scale was an additive measure in which

a high score indicated the individual believes he can rely and depend

upon others to facilitate, rather than frustrate, the attainment of his

goals. The scale was modified from Rotter's 1967 scale in order that

it would be appropriate for use with prison inmates (Hamsher et al., 1968;

Nelson, 1975; Rotter, 1967).

The Self-Esteem instrument, originated by Rosenberg (1962), was

a unidimensional scale measuring self-acceptance. This scale was also

scored so that the higher the score, the greater the subject's self-

esteem.










Both scales employed Likert categories of strongly agree, mildly

agree, agree and disagree equally, mildly disagree, and strongly dis-

agree. Example items from both scales are included below along with

internal consistency data obtained in previous related research with

these measures. (See Appendix E for a copy of this instrument.)

Interpersonal Trust (IPT)

In dealing with strangers a person is better off being
very careful until you are sure you can trust them.
(H.R. obtained in related research was .17 and Alpha
was .82.)

Self-Esteem (EST)

I feel that I have a number of good qualities.
(Rosenberg, 1962, reports a scale reproducability coef-
ficient of .92 and a test-retest reliability over two
weeks of .85.)

Freedom of.Movement (FM)

This scale measured a generalized expectancy of goal attainment.

It was developed by Jessor et al. (1968) and modified and expanded by

Swanson (1970b, 1973a). This instrument was an 11 item scale, with each

item asking the subject how sure he was about something in the future.

The subject responded on a four-point continuum: very sure, pretty

sure, not too sure, and not sure at all. The higher the score, the more

certain was the subject of attaining his goals.

An example item from the FM scale is listed below and is followed

by internal consistency data obtained in previous related research.

(See Appendix F for a copy of this instrument.)

When you think about what you really expect to happen in
the future, how sure are you that your life will work out
the way you want it to?
(H.R.'s in previous research were .26 and .36 and Alpha's
were .85 and .90.)

Personal Control (PC)

This scale measured the degree to which the individual saw his








rewards as contingent upon his behavior and/or attributes. It was a

modified version of a scale developed by Swanson (1970b) as a "general-

ized non-need area specific measure of internal-external control,

including two types of externality: chance and systematic powerful

others." Following each of eight statements, the subject responded on

a five-point continuum: strongly agree, agree, neither agree nor disagree,

disagree, or strongly disagree. The scale was scored so that the higher

the score, the greater the perceived internality of the subject.

An example item for the PC scale is listed below and is followed

by internal consistency data obtained in previous research.

Most of the unhappy things in my life have happened because
I was unlucky enough to be in the wrong place at the wrong
time.
(H.R.'s in previous research were .20 and .22 and Alpha's
were .93 and .73. This compares favorably with data
reported by Swanson, 1970b, using Rotter's I-E Scale:
H.R. = .11 and Alpha = .73.)

(See Appendix G for a copy of this instrument.)

In addition to these instruments, inmates completed forms request-

ing extensive demographic data. This information included age, sex, race,

family background, educational background, and prior criminal record, etc.

The internal consistency of these instruments has been demonstrated

in previous research. The validity of these instruments has been examined

in previous research as well. Rosenberg (1962) reported convergent

validity on the Self-Esteem scale as it correlated from .65 and .83

with several similar measures and with clinical assessment. Swanson

(1968), in studying the relationships obtained among the Attitude Toward

Prison Behaviors measures, the Definition of Situation measures, and the

Personal Control measure, established a strong sense of construct validity









of these measures. He reported that 78% of the interrelationships among

the scales followed the predicted result, while all of the nonpredicted

relationships obtained with the Staff Punishment and the Staff Custody

measures.

Swanson (1970b) established the construct validity-of a number of

control measures, including a form of the present Personal Control mea-

sure. He reported the measures were predictive of the following person-

ality measures included in this research: Attitude Toward Prison Behav-

iors, Definition of Situation, Freedom of Movement, and Attitude Toward

Deviance. Finally, McCown (1975) examined the construct validity of most

of the present instruments. She studied the obtained relationships be-

tween Personal Control and the following measures: Attitude Toward

Prison Behaviors, Definition of Situation, Freedom of Movement, Attitude

Toward Deviance, Alienation, and Interpersonal Trust. All correlations

obtained in the predicted direction. Of 18 correlation, 12 were signifi-

cant. All nonsignificant relationships obtained with the Definition of

Situation measures. Thus, a strong sense of construct validity has

typically been the case in previous work with these scales.

Correctional officer data, here construed as pertinent to inmate

success in the environment, was collected by means of ratings booklets,

each containing no more than ten inmate names to be rated. Four dimen-

sions were rated: Maladjustment, Association, Leadership, and Anonymity.

The last dimension was used to eliminate those ratings of inmates unknown

to the correctional officer. Six items provided information on the four

dimensions: three items made up the Maladjustment scale and one item

for each of the other three scales. The three items of the Maladjustment

scale asked for ratings of inmates on three five-point scales, ranging









from Model Inmate to Troublemaker, Very Well Adjusted to the Institution

to Very Poorly Adjusted, and Would Adjust Very Well to Society if Released

to Would Adjust Very Poorly. The Association item asked how the inmate

spends his time, Always Alone to Always with Others. The Leadership

item asked whether the inmate Is Never a Leader to Always Leads, and

the Anonymity item asked how well known the inmate was to the rater,

Very Well to Hardly At All.

Each inmate was rated by three officers on the six item scale.

Ratings for each dimension were then collapsed across raters yielding a

Maladjustment scale of nine items (three items rated by each of three

officers), and Association, Leadership, and Anonymity scales of three

items each (one item rated by each of three officers). Of the 196

inmates rated by the staff and also participating in the larger study,

those who obtained an average score of "Hardly at All" or higher were

dropped from the analysis, resulting in a final N = 178.

These scales were developed and utilized in previous research

utilizing the population used in this study (Morton, 1976), where the

internal consistency of the four collapsed scales was reported as

follows: Maladjustment (H.R. was .509 and Alpha was .902); Association

(H.R. was .282 and Alpha was .541); Leadership (H.R. was .398 and Alpha

was .664); and Anonymity (H.R. was .262 and Alpha was .513). (See

Appendix H for a complete copy of the rating scales.)



Procedure

The data were collected by a research team who administered an

elaborate self-report questionnaire to subjects in each of the six

correctional institutions. Testing sessions were approximately one and









one-half to two hours in length and included approximately 20-30 inamtes.

Sessions were held in large multi-purpose rooms within the institutional

confines. Each section of the questionnaire was carefully explained to

the inmate groups and individuals who required it received individual

assistance in understanding the instructions. Inmates were told that

their participation was voluntary and that the confidentiality of their

responses was assured.

Staff ratings were acquired by giving the correctional staff

ratings booklets accompanied by a cover letter explaining the purpose

of the research and asking their assistance. The names of ten selected

inmates were contained in each booklet, two names to a page, with the

pages randomly sorted in the various booklets and the names randomly

sorted on the pages. All booklets were returned by correctional staff

within the two to three days allotted for the completion of the ratings.

Before studying the hypothesized relationships among the scales,

the internal consistency of all measures was examined in terms of scale

homogeneity (Scott's Homogeneity Ratio, 1960) and scale reliability

(Cronbach's Alpha, 1954). Where necessary, scales were purified by

removing those items that correlated with the total score on the re-

maining items at a magnitude of less than .15.

The discriminant validity of the purified scales was examined by

comparing the purity of each measure with other related measures. The

rationale of the method employed follows. If two measures were related

while, at the same time, somehow different from each other, then it was

expected that their respective reliability coefficients would be greater

than their correlation coefficient. In other words, to the degree that

a measure had a reliability coefficient (Alpha) greater than its average









correlation (Pearson product-moment correlation coefficients) with related

constructs, to that degree was it measuring an independent construct.

The construct validity of the purified scales was examined by

studying the obtained Pearson product-moment correlation coefficients

among the scales. For example, the construct validity of the personal

control measure was examined by studying its relationships with the

other variables. (Hypotheses 1A through 1G dealt with the predicted

relationships between personal control and these variables.)

The effects of the various institutional environments upon the

personal control measure (hypothesis 2A) was determined by comparing

inmates from the four male settings and the three female settings.

Differences in mean scores on the personal control measure were assessed

by means of t-tests.

The effect of ethnic or minority status upon belief in personal

control (hypothesis 2B) was determined by comparing inmates of the various

ethnic groups. Differences in mean scores on the personal control measure

for White, Black, and non-Black minority group members (Mexican American,

Cuban American, and American Indian) were assessed by means of t-tests.

These differences were examined among inmate populations within and

across correctional institutions.

The analysis by socioeconomic status (SES) was conducted by means

of Pearson product-moment correlation coefficients (hypothesis 2C). A

SES Index was computed for each subject using a two-factor formula

(Hollingshead, 1957). The SES Index resulted from assigning factor

scores to subject questionnaire responses about educational and occupa-

tional history, multiplying these factor scores by assigned factor

weights, and summing the results. The SES Index was then correlated









with Personal Control scores. Analyses were conducted by sex, by race,

and by institution.

The relationship between changes in personal control beliefs and

length of stay within the institutional confines (hypothesis 3A) was

examined in three separate analyses. In the first analysis, the length

of time at present prison was correlated with Personal Control scores.

This analysis was done by sex, by race, and by institution.

The remaining two analyses were conducted by comparing cross-

sectional samples of inmates. In the first of these, the ratio of length

of time at present prison/length of present sentence was computed for

each subject. These ratios were then trichotomized yielding three

cross-sectional samples: inmates in the beginning phase of their sentence,

inmates in the middle phase of their sentence, and inmates in the final

phase of their sentence. Trichotomization was conducted by institution.

Differences in mean Personal Control scores for these groups were

examined by t-tests.

In the final analysis of the relationship between personal control

beliefs and time within the institutional setting, subject samples were

composed of inmates whose length of stay in the institution had been

six months or less, whose length of stay was within six months of the

expected time remaining in prison, and whose expected time remaining

in prison was six months or less. Group differences in mean Personal

Control scores for these subjects were examined by means of t-tests.

Analyses were conducted by sex.

Measuring group consensus on appropriate behaviors was important

for hypotheses 4A, 53, and 5C. In analyzing consensus, the major concern

was the strength of the group consensus regardless of its directionality.









This was measured by percent of respondents in the modal category for

each item contained in the Attitude Toward Prison Behaviors measure and

by the amount of variance around the group's item mode, or subcultural

norm. (This measure w'.as Jessor et al.'s (1968) operational definition

of anomie.) Summing the variance for all items yielded the group consen-

sus measure--the measure of anomie based on the subcultural norm.

The measure of group consensus was computed by institution, by

living unit, and by ethnic group within institutions. Within each anal-

ysis, groups were ranked ordinally by degree of consensus. The relation-

ship of these consensus rankings with mean Personal Control scores for

the respective groups (hypothesis 4A) was assessed by means of Spearman

rank-order correlation coefficients.

Analysis of the relationship between population size and personal

control beliefs (hypothesis 4B) was examined by ranking institutions by

population size. The relationship of these rankings with mean Personal

Control scores for each institution was assessed by means of Spearman

rank-order correlation coefficients.

Success in the environment was conceptualized to include the

following variables: freedom of movement, alienation, number of times

rewarded by staff in the past year, number of disciplinary reports

received in the past year, and staff ratings of maladjustment, association,

and leadership. The relationship of each of these seven measures with

the Personal Control measure was analyzed by means of Pearson product-

moment correlation coefficients (hypothesis 5A).

Subjects, within institutional or dormitory environments ranked

ordinally by group consensus on appropriate behaviors, were classed as

high personal control subjects or low personal control subjects by






55


whether their Personal Control score fell above or below the median for

subjects within that environment. Environments were also dichotomized

by median split on the degree of consensus. Differences in mean success

scores for the two personal control samples in both high and low consen-

sus environments were analyzed by means of t-tests (hypothesis 5B).

Additionally, mean success scores for high personal control

subjects from high consensus environments were compared with mean

success scores for high personal control subjects from low consensus

environments. In the same manner, low personal control subjects from

low consensus environments were compared with low personal control

subjects from hiqh consensus environments. These analyses of differences

were conducted by t-test (hypothesis 5C).















RESULTS


Internal Properties of the Scales

The internal consistency of the scales utilized in the present

research was examined with regard to scale homogeneity and reliability.

Scott's (1960) Homogeneity Ratio (H.R.) was used to obtain a measure

of scale homogeneity. This statistic, which is not influenced by scale

length, indicates the degree to which each item is measuring the same

attribute and is computed as the weighted average interitem correlation.

An H.R. of .33 provides for the optimal discrimination among subjects

if the items are split with a probability of .5 of passing or failing.

As the H.R. becomes greater than .33, this indicates redundance of

scale items. On the other hand, as the H.R. becomes smaller than .33,

this indicates the scale items may be measuring different attributes

altogether.

Cronbach's (1954) Alpha was used to obtain a measure of scale

reliability. This statistic, which is sensitive to both scale length

and homogeneity, provides an indication of the precision with which the

scale measures the intended attribute. Since Alpha treats each item

as a separate test of the attribute being measured and computes the

degree of correlation between them it can be viewed as an estimate of

the scale's correlation with an equivalent test.

Finally, the item score to scale score correlation for each

item was computed to determine the contribution of each item to the









particular scale. Items with correlations to scale scores of a magni-

tude less than .15 were eliminated in order to strengthen and purify

the scales. One item was removed from each of the following scales:

Interpersonal Trust, Self-Esteem, and Personal Control. The internal

properties of the purified scales are reported in Table 1. It can be

seen that all scales have sufficient internal consistency and reliabil-

ity, with the possible exception of Interpersonal Trust, which had a

somewhat weaker H.R. of .173, but with thirteen items an adequate

Alpha of .73.



Validity

The discriminant validity of the scales was established by exam-

ining the relationships anong the scales. The purity of each measure was

compared with all other related measures. If two measures were related

while, at the same time, somehow different from each other, then it was

expected that their respective reliabilities would be higher than their

correlation. To the degree that a measure had a reliability coefficient

greater than its average correlation with related constructs, to that

degree was it considered to be measuring an independent construct.

Tables 2, 3, and 4 show the comparison of the estimate of reliability

(Alpha) with the average interscale correlation for related measures for

the Attitude Toward Prison Behaviors measures, the institutional oppor-

tunity and authority measures, and the remaining individual difference

measures, respectively. (It should be noted that in Table 3, the Overall

Definition of Situation scale as well as the Attitude Toward Authority

in General scale have been omitted since these measures are composed of

items pooled from the other Table 3 scales.)












Table 1. Internal Properties

No. Items
N Included

533 5


Note: N's vary due to missing data.


of Purified

Cronbach's
Alpha

.744

.727

.587

.616

.728

.674

.712

.669

.750

.729

.904

.756

.734

.819

.933

.817

.730

.776

.901

.723


Scales

Scott's
H.R.

.370

.350

.263

.245

.353

.294

.383

.340

.429

.354

.358

.442

.409

.363

.294

.360

.173

.318

.452

.275












Table 2. Intercorrelations Among
Attitude Toward Prison Behaviors Measures


A (.744)


B .311 (.727)


C .497


D .252


E -.388


.109 (.587)


.429


-.305


.310 (.616)


-.121


-.211


(.728)


F -.271 -.386 -.039 -.173 .454 (.674)


Average Interscale Correlation:

.344 .308 .215 .275 .296 .265


where A = Staff Punishment

B = Inmate Opposition

C = Staff Custody

D = Inmate Avoiding Trouble

E = Staff Treatment

F = Inmate Self-Improvement


Note: Estimate of reliability appears in parentheses.











Table 3. Intercorrelations Among
Institutional Opportunity and Authority Scales


A B C D E F


A (.712)


B .637 (.669)


C .706 .563 (.750)


D .751 .525 .522 (.729)


E .403 .303 .484 .369 (.756)


F .552 .378 .626 .467 .554 (.734)


Average Interscale Correlation:

.610 .481 .600 .547 .423 .515


where A = Opportunity to Develop a Personal Identity

3 = Opportunity for Social/Interpersonal Development

C = Definition of Personal Commitment

D = Value of Commitment

E = Attitude Toward Rules and Regulations

F = Attitude Toward Authority Figures


Note: Estimate of reliability appears in parentheses.












Table 4. Intercorrelations Among
Remaining Individual Difference Variables


A (.933)


B .030 (.817)


C .115


D -.066


E .079


-.197 (.730)


-.443


-.329


(.776)


(.901)


F .045 .408 .161


Average Interscale Correlation:


.389 .273 (.723)


where A = Attitude Toward Deviance

B = Alienation

C = Interpersonal Trust


D = Self-Esteem

E = Freedom of Movement

F = Personal Control


Note: Estimate of reliability appears in parentheses.









The intercorrelations of the Attitude Toward Prison Behaviors

scales are reported in Table 2. In all cases the estimate of reliability

for each scale is greater than the average correlation with the remaining

scales. Moreover, in all cases the estimate of reliability for each

scale was greater than each correlation with remaining scales. The

average estimate of reliability for the six measures was .679 and the

average mean scale intercorrelation was .284. All of these factors

point toward substantial discriminant validity for each of these scales.

The intercorrelations of the institutional opportunity and author-

ity measures are reported in Table 3. As was true with the scales in

Table 2, the estimate of reliability for each scale in this grouping

was greater than the average correlation with the remaining scales.

However, closer examination of Table 3 reveals some overlap between the

scales. Opportunity to Develop a Personal Identity, column A in Table 3,

correlated with Definition of Personal Commitment, the Opportunity

for Social/Interpersonal Development, and the Value of Commitment scales

at approximately the same magnitude as its estimate of reliability and

thus failed to achieve discriminant validity when compared with these

other scales. By the same criteria, the Opportunity for Social/

Interpersonal Development scale was not distinguishable from Opportunity

to Develop a Personal Identity, nor was Definition of Personal Commitment

distinguishable from Opportunity to Develop a Personal Identity, nor

was Value of Commitment distinguishable from Opportunity to Develop a

Personal Identity.

Among these four scales, which represent the institutional

opportunity structures, the average estimate of reliability and the

average mean scale intercorrelations was .634. The weak discriminant









validity among these measures suggests that they could successfully

be treated as one scale. On the other hand, examination of the remaining

two scales shows that they possessed stronger discriminant validity.

Both the Attitude Toward Rules and Regulations scale and the Attitude

Toward Authority Figures scale were distinguishable from each of the

other scales.

The intercorrelations among the remaining scales dealing with

personality attributes utilized in this study are reported in Table 4.

In all cases, the estimate of reliability for each scale was greater than

each correlation with the remaining scales. The average estimate of

reliability for these six individual difference measures was .813 and

the average mean scale intercorrelation was .214. Thus, Table 4

indicates these measures possessed substantial discriminant validity.

The construct validity of these scales was established by exam-

ining the intercorrelations among the scales. For purposes of this exam-

ination, the six subscales contained within the Attitude Toward Prison

Behaviors measure were treated as indicators of individual orientation.

The institutional opportunity and authority measures were all scored in

the direction of positive adaptation, i.e., a high score reflected a

definition of opportunity or benign authority, as were the Personal

Control measure, the Attitude Toward Deviance measure, the Interpersonal

Trust measure, the Self-Esteem measure, the Freedom of Movement measure,

and the positive attitude scales, Staff Treatment and Inmate Self-

Improvement. The only exceptions were the Alienation measure and the

following four negative attitude scales: Staff Punishment, Inmate

Opposition, Staff Custody, and Inmate Avoiding Trouble. Therefore, it

was expected that all those scales scored in a direction of positive









adaptation would correlate directly with each other and inversely with

the four negative Attitude Toward Prison Behaviors subscales and the

Alienation measure.

Table 5 reports the intercorrelations among all the scales utilized

in this research. Of these 190 correlations, the hypothesized pattern

obtained in 87% of the cases. Of the remainder, the majority (20 of 25,

or 80%) obtained in the relationships within the Staff Punishment and

Staff Custody scales. Since the predicted relationships obtained in

the majority of cases, strong support was found for the construct

validity of these measures.



Relation of Personal Control to Other Personality Variables

Research hypotheses are restated here and the results are

reported.

Hypothesis 1A. Belief in personal control should correlate
positively with expectations for goal attainment.

Table 5 shows the obtained correlation coefficient for the rela-

tionship between personal control belief and the freedom of movement,

or expectation of goal attainment, variable was .273 (492), p<.01.

Those inmates who believe they control their own outcomes and reinforce-

ments also tend to expect they will succeed in reaching their goals.

Hypothesis IB. Belief in personal control should correlate
negatively with alienation.

As predicted, an inverse correlation obtained for the relationship

between belief in personal control and alienation, with r = -.408 (497),

p<.01. This was the coefficient of largest magnitude of any of the

relationships between Personal Control and the other personality vari-

ables. The inmate who believes that the rewards of life are independent







Table 5. Intercorrelations Among All Scales


SPU 10P SCU IAT STR ISI PID SID DPC VOC DEF AIR ATA AGA ATD AL

SPU .744


.311 .727
(533)
.497 .109 .587
(533) (536)
. 52 ..'29 .310 .616

- 38 -.305 -.121 -.211 .728
(533) (536) (536) ( 35)
-.271 -.-3 6 -.039 -.173 .45 .674
(533) (534) ( 34) 3) (534)
.101 -.202 .?30 .213 .112 .237 .712
(527) (530) (530) ('29) (530) (528)
.013 -.164 .145 -. 50 .116 .2 4 .637 .69
(527) (530) 533) (52?) (530) (5?, ) (S31)
.170 -.292 .211 -.269 .125 .19 .7C5 .5- 3
(526) (E 9) (529) (528) (52 ) (527) (530) (530)
100 -.iS8 .200 -.197 .055 .1 3 .751 .525
(5:9) (532) (53?) (E31) (5,2) (533) (530) (530)
.078 -.248 .2 6 -.47 .123 .251 .903 786
(530) (533) (533) (532) (5 3) 531) (531) (531)
-.007 -.3 7 .073 -.376 .1'5 1, 03 .303
('26) (529) (529) (' 8) (529) (527) (530) (530)
-.767 .179 -.250 C 3 ..1.. .552 .37
( ;6) ;523) (52a| ,;525) (5 s9) (527) (530) ( 3")
.C17 -.273 .142 -.358 .1.15 .110 .539 .3 1
(526) (529) (; 529) (5 3; (5 9) (527) (i30) (53.)
.120 -.311 .1 7 -. 56 .207 242 .150 .17
(510) (5 0 ) (5 () (519) (520) (519) (516) (516)
.145 .104 .138 .281 -.111 -.0?0 -.191 -.174
(522) (523) (523) (522) (523) (522) (519) (519)
.187 -.124 .133 -.204 -.067 -.047 .286 .177
(499) (501) (501) (500) (501) (500) (499) (499)
-.156 -.059 -.161 -.217 .094 .066 .118 .114
(498) (500) (500) (499) (500) (499) (498) (498)
.024 -.143 .064 -.147 .091 .148 .268 .211
(493) (495) (495) (494) (495) (493) (494) (494)
-.168 -.213 -.289 -.336 .124 .094 .107 .097
(497) (499) (499) (498) (499) (497) (497) (497)


.622 729
(530)
.851 r5
(53C) (533





(5.1 ) (51 )
.214 .11 2
(516) (519)
.124 -.004
(519) (522)
.347 .271
(499) (501)
.060 .049
(498) (500)
.204 .216
(494) (496)
.096 .073
(497) (499)


IPT EST FM PC


.457 .7E6

.5 4 701

S59 )1 C 6C .0 19
(5~0) (5<7) (530)
103 1 .157 .: 6 .93
(519) (516) ( 16) (516)
.167 -.'42 -.165 -.234 .030 .817
(522) (519) (519) (519) (517)
.109 .330 .405 .416 .115 -.197 .730
(501) (499) (499) (499) (494) (499)
.102 .090 .051 .O 2 .066 -.443 -.031 .776
(500) (498) (498) (498) (493) (498) (501)
.252 .142 .164 .173 .079 -.329 .129 .519 .901
(496) (494) (494) (494) (488) (493) (485) (484)
.116 .314 .119 .250 .045 -.08B .161 .389 .273 .723
(499) (497) (497) (497) (492) (497) (488) (488) (492)


Note: Estirate of reliability on diagonal, N In parentheses.


-









of his own actions or characteristics tends also to experience a feeling

of social isolation and estrangement from others.

Hypothesis IC. Belief in personal control should correlate
positively with self-esteem.

The correlation between the Personal Control measure-and the

Self-Esteem measure obtained as predicted, with r = .389 (488), p<.0l.

This correlation coefficient ranks second only to that between Personal

Control and Alienation in terms of its magnitude. Inmates who exper-

ience a belief in their own capabilities and who thus feel able to deal

effectively with their environment also tend to hold self-accepting

postures. The tend to value themselves more highly and feel more

positively about their own abilities.

Hypothesis 1D. Belief in personal control should correlate
positively with interpersonal trust.

The correlation for the relationship between Personal Control

and Interpersonal Trust obtained as predicted, with r = .161 (488),

p<.01. This finding suggests that inmates who believe in personal con-

trol tend also to see others as trustworthy and reliable. While this

relationship obtained for all inmates, it was of greater magnitude among

female inmates than among male inmates, with r = .309 (138), p<.01 for

females and r = .102 (349), p<.05 for males.

Hypothesis 1E. Belief in personal control should correlate
positively with an intolerant attitude toward
deviance.

The predicted relationship between belief in personal control

and attitude toward deviance failed to occur. Instead, the correlation

coefficients were near zero, indicating a random nonsignificant rela-

tionship. For all inmates, r = .045 (492); for males r = .074 (351);

and for females, r = -.026 (139). These findings failed to support










the prediction that inmates holding a belief in personal control would

proscribe involvement in criminal behavior.

Hypothesis 1F. Belief in personal control should correlate
positively with prescriptions for behaviors
which maximize staff rewards and negatively
with prescriptions for behaviors which maximize
staff punishments.

Treating the six subscales contained in the Attitude Toward

Prison Behaviors measure as indications of individual orientation, we

will now examine their relationships with the Personal Control measure.

The Staff Treatment and the Inmate Self-Improvement scales represent

behavior of staff and inmates, respectively, appropriate for a correc-

tional institution oriented toward treatment of the inmate. An inmate

who prescribes these behaviors would be oriented towards maximizing

staff rewards for himself. On the other hand, the Staff Custody and

Inmate Avoiding Trouble subscales represent staff and inmate behaviors

oriented toward confinement of the inmate, while the Staff Punishment

and the Inmate Opposition subscales represent behaviors of an institu-

tional environment organized toward punishment of the inmate. An inmate

who prescribes the orientations of these four scales would be prescribing

behaviors which maximize punishments.

All correlations with belief in personal control obtained in the

predicted fashion. Personal Control correlated positively with Staff

Treatment, with r = .124 (499), p<.01, and also correlated positively

with Inmate Self-Improvement, with r = .094 (497), p<.05. Personal

Control correlated negatively with Staff Custody, with r = -.289 (499),

p<.01, and correlated negatively with Inmate Avoiding Trouble, with

r = -.336 (498), p<.Ol. Personal Control also achieved a significant

inverse relationship with Staff Punishment, with r = -.168 (497), p<.01,








and with Inmate Opposition, with r = -.213 (499), p<.01. These findings

strongly support the prediction that individual inmates who maintain a

belief in personal control will prescribe, or recommend, behaviors for

themselves and for the correctional staff that will tend to increase

the likelihood of cooperation and reward and decrease the likelihood

of opposition and punishment.

Hypothesis 1G. Belief in personal control should correlate
positively with favorable attitudes toward
environmental opportunity and authority structures.

Attitudes toward the opportunity structures of the correctional

environments studied were measured by five different scales. The fifth

scale was composed of all items from the other four. Opportunity to

Develop a Personal Identity correlated with Personal Control at the

level of r = .107 (497), p<.05. Opportunity for Social/Interpersonal

Development correlated at the level of r = .097 (497), p<.05. Defini-

tion of Personal Commitment correlated at the level of r = .096 (497),

p<.05, while the correlation between Personal Control and Value of

Commitment failed to achieve significance, with r = .073 (499). The

correlation between Personal Control and the Value of Commitment scale

did achieve significance for female subjects, with r = .189 (139), p<.05.

The overall Definition scale correlated with Personal Control at the level

of r = .116 (499), p<.01. These results indicate that inmates who hold

beliefs in personal control also tend to see their environments as

richer in opportunity.

Inmate attitudes toward the authority structures of their correc-

tional environments were measured by three scales. These were the

Attitude Toward Rules and Regulation, the Attitude Toward Authority

Figures, and the Attitude Toward Authority in General scales. The final









scale was composed of the items of the previous two. Attitude Toward

Rules and Regulations correlated with Personal Control at the level

or r = .314 (497), p<.Ol. Attitude Toward Authority Figures correlated

with Personal Control at the level of r = .119 (497), p<.01. The

overall authority measure also correlated positively with belief in

personal control at the level of r = .250 (497), p<.01. The results

suggest that, among inmate populations, belief in personal control is

associated with positive attitudes toward environmental authority

structures and the authority figures therein.

Seen in their total, the results of these correlational analyses

suggested a strong sense of construct validity for the Personal Control

measure. Belief in personal control of reinforcements was shown to he

associated with a wide range of other personality variables, including

expectations for goal attainment, alienation, self-esteem, and inter-

personal trust. It was also associated with prescriptions for behaviors

which maximize rewards and minimize punishments. Finally, it was associ-

ated with perceptions of enriched environmental opportunities as well as

benign environmental authority.



Distribution of Personal Control Within Social Settings

Hypothesis 2A. Persons living in environments defined as rela-
tively high in opportunity should believe in
personal control to a greater extent than persons
living in environments defined as relatively low
in opportunity.

Based upon the rankings of the environmental opportunities reported

earlier, the four male and the three female correctional institutions

were defined as to the level of opportunities available to their residents.

It will be remembered that among male institutions, LexM was defined as









the highest in opportunity, UCI was the lowest, with FCIM and ColoM

falling in the intermediate range. Among the female institutions,

LexF was defined as the highest in opportunity while FCIF and ColoF

were both defined as low opportunity environments.

It was expected that mean personal control beliefs within the

respective institutions would correspond with these various defined

levels of opportunity, such that those institutions defined as high

in opportunity would differ most markedly with those institutions

defined as low in opportunity. The results are contained in Table 6.

Among male institutions weak significant differences in mean

Personal Control scores obtained between the LexM sample and the FCIM

and UCI samples, respectively. Other differences between male samples

were all nonsignificant. Among female institutions, the pattern of

differences was somewhat more clear-cut. The LexF sample differed with

the FCIF sample at the .005 level and with the ColoF sample at the .10

level. These findings tended to support the hypothesis, especially

when the differences in environmental opportunity were in the sharpest

relief.

Hypothesis 2B. Persons of ethnic or minority status should
believe in personal control to a lesser extent
than persons of Anglo or majority status.

The distribution of belief in personal control among ethnic groups

was examined among inmate populations within and across the seven correc-

tional institutions. Table 7 contains the results of this examination

when all subjects were grouped together. White inmates exhibited the

strongest belief in personal control, with the non-Black minority groups

and the Black sample following in that order. As predicted, significant

differences obtained between the White and Black groups (p<.001) and








Table 6. Differences in Personal Control Measure Within
Four Male and Three Female Correctional Institutions


Means and Standard Deviations


Male Institutions


Significance


LexM FCIM ColoM


1 1
vs vs
2 3V
2 3


24.2364
4.6227
(N=b5)


24.0030
4.8975
(N=191)


t=1.55 t=1.04 t=1.59
p<.10 N.S. p<.10


t=-0.48 t=-0.26 t=0.33
N.S. N.S. N.S.


Female Institutions


LexF FCIF ColoF


24.5357
4.8951
(N=30)


1 1
vs vs
2 3


t=2.99 t=1.46 t=-0.78
D<.005 p<.10 N.S.


Note: All t-tests are two tailed.
N.S.--difference not significant at the .10 level or better.


X 25.1667
o 4.3380
(N=46)


23.8257
4.6432
(N=65)


X 26.1112
S 3.5634
(N=44)


23.6712
5.1187
(N=70)









Table 7. Ethnic Differences in Personal Control Measure
Across Seven Correctional Institutions


Means and Standard Deviations


White

X 25.5157
F 4.4922
(N=254)


Black Non-Black Minority


23.0511
4.5875
(N=184)


23.3889
4.9470
(N=36)


vs
B
t=5.60
p<.01


Significance
W
vs
NB
t=2.44
p<.05


Note: All t-tests are two tailed.
N.S.--difference not significant at the .05 level or better.


B
vs
NB
t=-0.38
N.S.










between the White and non-Black minority group sample (p<.05). The

difference between the Black inmates and members of the other minorities

sample failed to achieve significance.

The results of the analysis within correctional environments are

contained within Table 8. Within all institutions except LexF, the

differences between majority and minority ethnic samples on the Personal

Control measure are in the predicted direction. However, in only the

three Florida institutions did these differences achieve significance

(all at the .01 level). It should be noted that only in the Colorado

institutions did the sample size of the non-Black minorities group allow

inclusion in the analysis.

Hypothesis 2C. Persons of lower socioeconomic status should
believe in personal control to a lesser extent
than persons of higher socioeconomic status.

The relationship between Personal Control and socioeconomic status

was examined by means of Pearson product-moment correlation coefficients.

The results are contained in Table 9. A significant coefficient obtained

when all inmates were included in the correlational analysis, with

r = .112 (437), p<.05. Further analysis showed that the power of this

correlation was largely contributed by the female sample, with r = .183

(122), p<.05, whereas the relationship for males was not significant.

Analyses by race and by institution also found nonsignificant coefficients

of correlation.



Changes in Personal Control Beliefs

Hypothesis 3A. Belief in personal control bears a curvilinear
relationship with a person's length of stay in
a novel and confining environment such that
belief in personal control is minimized upon
entry and again upon exit from that environment.











Table 8. Ethnic Differences in Personal Control Measure
Within Seven Correctional Institutions


Means and Standard Deviations


Significance


Non-Black
White Black, Minority


FCIM X 25.5428
o 4.2032
(N=35)

UCI X 25.2766
o 4.3906
(N=94)

ColoM X 24.5555
o 5.0622
(N=36)

LexM X 25.9655
o 4.1703
(N=29)

FCIF X 26.5000
o 4.8812
(N=24)

ColoF X 25.5714
o 5.3308
(N=14)

LexF X 25.8333
o 4.2183
(N=18)


22.2900
4.3408
(N=23)

22.8562
5.0667
(N=76)

23.2000
4.0249
(N=5)

25.2823
3.4525
(N=13)

21.8861
4.5218
(N=41)

25.1667
2.7869
(N=6)

25.8750
3.4429
(N=24)


22.8000
3.5839
(N=10)


t=2.83
p<.Ol


t=3.29
p<.01


t=0.68 t=1.24 t=0.19
N.S. N.S. N.S.


t=0.55
N.S.


t=3.78
p<.01


24.0000
4.1952
(N=6)


t=0.22 t=0.71 t=0.57
N.S. N.S. N.S.


t=-0.03
N.S.


Note: All t-tests are two tailed.
N.S.--difference not significant at the .05 level or better.











Table 9. Correlations Between Personal Control
and Socioeconomic Status


Values of r

All Inmates (N=437) .112 (p<.05)

By Sex
Males (N=314) .085 (N.S.)
Females (N=122) .183 (P<.05)

By Race
Whites (N=235) .008 (N.S.)
Blacks (N=158) .073 (N.S.)
Non-Black Minorities (N=33) .138 (N.S.)

By Institution
FCI, (N=59- .217 (N.S.
UCI (N=172) .070 (N.S.)
ColoM (N=48) .237 (N.S.)
LexM (N=35) -.269 (N.S.)
FCIF (N=64) .113 (N.S.)
ColoF (N= 23) .299 (N.S.)
LexF (N= 35) .109 (N.S.)

Note: N.S.--coefficient not significant at the
or better.


.05 level









Before examining the above hypothesis in light of the results

of this study, it seemed instructive to first examine the relationships

between personal control beliefs and a number of variables concerned

with an inmate's length of stay within correctional environments.

Table 10 contains the obtained correlations between Personal Control

scores and four such variables. Analyses were done by sex, by race,

and by institution.

The four length of stay variables included in this analysis

were: 1) length of time at present prison (LT); 2) expected time

remaining in prison (ET); 3) length of present sentence (LS); and

4) percent of sentence served (PERSEN). Examination of Table 10

reveals that when all subjects were grouped for analysis, only the

correlation between Personal Control and ET achieved significance,

with r = -.121 (436), p<.05. Among all inmates, as the expected time

remaining in prison decreases, personal control beliefs tend to in-

crease. This relationship held for male subjects with r = -.136

(306), p<.05, and as well as for Blacks, with r = -.222 (160), p<.01,

but failed to hold for females, Whites, and the non-Black minority

sample. Among female inmates the only relationship achieving signifi-

cance was between Personal Control and LT, with r = -.178 (138), p<.05.

For females, it appears that as length of time in prison increases,

personal control beliefs decrease. Two significant correlations obtained

with White subjects. Personal Control correlated with PERSEN at the level

of r = -.152 (247), p<.05, and with LS at the level of r = .133 (248),

p<.05. These results suggest that Caucasian inmates with longer sen-

tences tend to hold stronger beliefs in personal control and that as

the percent of their sentence served increases, their beliefs in personal

control become weaker.












Table 10. Correlations Between Personal Control
and Length of Stay Variables


Values of r


All Inmates



By Sex
Males



Females



By Race
Whites



Blacks



Non-Black
Minorities

By Institution
FCIMS



UCI



ColoM



Lex l


LT

-.013
(N=478)
N.S.


.055
(N=339)
N.S.

-.178
(N=138)
p<.05


.060
(N=252)
N.S.

-.066
(N=181)
N.S.

.152
(N=34)
N.S.

.054
(N=62)
N.S.

.047
(N=182)
N.S.

.228
(N=53)
N.S.

.187
(N=42)
N.S.


ET

-.121
(N=436)
p<.05


-.136
(N=306)
p<.05

-.086
(N=129)
N.S.


-.059
(N=237)
N.S.

-.222
(N=160)
p<.01

-.077
(N=30)
N.S.

-.272
(N =58)
p<.05

-.156
(N=159)
p<.05

-.007
(N=51)
N.S.

-.351
(N=38)
p<.05


LS

.022
(N=470)
N.S.


.009
(N=333)
N.S.

.054
(N=136)
N.S.


.133
(N=248)
p<.05

-.072
(N=176)
N.S.

-.117
(N=35)
N.S.

-.045
(N=59)
N.S.

.029
(N=181)
N.S.

.201
(N=51)
N.S.

-.426
(N=42)
p<.01


PERSON

-.059
(N=467)
N.S.


-.046
(N=330)
N.S.

-.090
(N=136)
N.S.


-.152
(N=247)
p<.05

.029
(N=176)
N S.

.030
(N=34)
N.S.

.076
(N=59)
N.S.

-.054
(N=-179)
N.S.

-.270
(N=51)
N.S.

.229
(N=41)
N.S.












Table 10 continued


Value,

LT ET

FCIF -.159 -.093
(N=69) (N=65)
N.S. N.S.

ColoF -.103 .232
(N=27) (N=26)
N.S. N.S.

LexF -.214 .026
(N=42) (N=38)
N.S. N. S.


Note: N.S.--ccefficient not significant
or better.


s of r

LS

.134
(N=69)
N.S.

.192
(N=26)
N.S.

.001
(N=41)
N.S.


PERSON

-.103
(N=69)
N.S.

-.256
(N=26)
N.S.

-.237
(N=41)
N.S.


at the .05 level









The analysis by institution reveals that for three of the four

male institutions, correlations between ET and Personal Control achieved

significance at the .05 level or better, failing to achieve significance

only in the ColoM sample. The only other significant correlation was

within the LexM subject sample where a relatively strong inverse rela-

tionship existed between Personal Control and LS, with r = -.426 (42),

p<.01.

We will now turn to an examination of the research hypothesis

stated above. Table 11 contains results of an analysis of differences

in personal control beliefs among cross-sectional samples of inmates

within the seven institutional environments. Inmates within each

institution were classified in one of three groups according to the

percentage of their sentence served so that there were approximately

equal numbers of inmates in each group. Differences in personal control

beliefs among subjects in the beginning phase, in the middle phase, and

in the final phase of their sentences were assessed by means of t-tests.

Only two of the seven institutions demonstrated a significant

difference in Personal Control between inmate samples in different

phases of their sentences. Within FCIM, inmates in the beginning phase

of their sentence held significantly weaker beliefs in personal control

than did inmates in the middle phase of their sentence (t = -2.2485,

p<.05). This was the only significant difference supportive of the hypo-

thesis. Within the LexM institution, inmates in the final phase of their

sentence held significantly stronger beliefs in personal control than

did inmates in the middle phase of their sentence (t = -2.4391, p<.05).

One additional analysis was accomplished in order to shed further

light on the research hypothesis. In this analysis, an attempt was made












Table 11. Differences in Personal Control
Among Inmate Samples Classed by Phase of Sentence
Being Served Within Seven Correctional Institutions


Means and Standard Deviations

1 2 3

1st Phase 2nd Phase 3rd Phase


Significance


FCIM X 22.1176
g 4.1515
(N=17)

UCI X 24.7667
o 5.3563
(N=60)

ColoM X 25.0000
c 4.3925
(N=18)

LexM X 25.5113
o 4.6456
(N=15)

FCIF X 24.9091
c 4.6282
(N=22)

ColoF X 26.0000
o 4.0927
(N=9)

LexF X 26.8750
o 4.1292
(N=16)


25.5714
5.3157
(N=21)

23.7667
4.7439
(N=60)

22.6250
2.9637
(N=16)

23.6923
3.4733
(N=13)

23.3187
6.3152
(N=23)

25.6667
4.2426
(N=9)

25.8791
4.1103
(N=l )


23.8571
4.0036
(N=21)

23.7724
4.5749
(N-59)

24.1176
6.0092
(N=17)

26.9231
3.2777
(N=13)

22.7500
4.5325
(N=24)

21.7500
5.8979
(N=8)

25.2857
2.3674
(N-14)


Note: All t-tests are two tailed.
N.S.--difference not significant at the .05


t=-2.25 t=-1.30 t=1.18
p<.05 N.S. N.S.


t=1.08 t=1.09 t=-0.01
N.S. N.S. N.S.


t=1.87 t=0.49 t=-0.91
N.S. N.S. N.S.


t=1.18 t=-0.94 t=-2.44
N.S. N.S. p<.05


t=0.97 t=1.60 t=0.35
N.S. N.S. N.S.


t=0.17 t=1.71 t=1.55
N.S. N.S. N.S.


t=0.62 t=1.31 t=0.43
N.S. N.S. N.S.


level or better.









to maximize the differences among three cross-sectional samples in terms

of their time in prison. The first sample contained those inmates whose

time in prison had been six months or less. The second sample included

inmates whose time in prison was within six months of their expected

time remaining. The third sample was composed of those inmates whose

expectation was that they would be released within six months. In order

to be classified within any of the three groups, the additional require-

ment was made that an inmate had to have a sentence of 13 months or more.

This prevented any one inmate from being classified in more than one

group. Table 12 contains the results of the analysis of differences in

Personal Control among male and female inmates within these three

samples.

The predicted pattern of change in personal control beliefs was

that they would be strongest in the middle phase sample and more exter-

nalized in the entry and exit phase samples. Looking at Table 12, we see

that this pattern held for both male and female samples. However, only

one significant difference occurred between the respective samples.

Among male inmates, Personal Control differed significantly between the

entry sample and the middle phase sample with the middle phase sample

holding stronger beliefs in personal control (t = -1.9832, p<.05). All

other difference tests failed to achieve significance.



Environmental Influences

Consensus on appropriate behaviors was measured by the percent

of respondents in the modal category and by summing the variance around

each item mode for all items in the Attitude Toward Prison Behaviors

measure. The measurement of consensus was accomplished by institution,












Table 12. Differences in Personal Control
Among Male and Female Samples
with Maximized Time-In-Prison Differences


Means and Standard Deviations


Significance


Males X 24.0156
o 4.7123
(5=64)

Females X 24.2963
c 5.1506
(N=27)


2

25.5625
3.5425
(N=48)

25.0476
5.0147
(N=21)


3

25.2335
3.5345
(N=20)

23.8000
5.9789
(N=20)


1 1
vs vs
2 3


t=-1.98 t=-1.24 t=0.35
p<.05 N.S. N.S.


t=-0.51 t=0.30 t=0.72
N.S. N.S. N.S.


Note: All t-tests are two tailed.
N.S.--difference not significant


at the .05 level or better.


1 = Inmates whose time at present prison was 6 months or less.
2 = Inmates whose time at present prison was within six months
of the expected time remaining in prison.
3 = inmates whose expected time remaining in prison was 6 months
or less.










by living unit, and by ethnic group within institutions. We will now

look at these measures of consensus within the three analyses.

Institutions

Looking at Table 13, we see the measures of consensus for the

seven institutions. The scale item is first presented and is followed

by the item mode, the percent of respondents in the modal category,

and the variance around the group's own mode. For example, for item

one "Inmates cause as much trouble as they can," all seven institutions

were highly proscriptive of this behavior. FCIF had the highest percen-

tage of inmates in the modal category while LexF had the lowest amount

of variance around its own mode. UCI showed the greatest amount of

variance and thus demonstrated the least amount of consensus among the

seven institutions on this item. Looking across the items, note that

the institutions remained fairly consistent in their choice of group

modes, but varied in their degree of consensus. Looking now at the

total variance on the Inmate Opposition items, we see that ColoF in-

mates exhibited the least amount of variance around their own modal

category, while FCIM inmates exhibited the greatest amount of variance.

Thus, on the behaviors described within the Inmate Opposition scale,

ColoF residents had the most consensus and the FCIM sample had the

least consensus. Looking at the overall measure of group consensus on

the last page of Table 13, we see that ColoF inmates had the lowest

total amount of variance around all item modes while FCIF inmates had

the highest amount. Thus, within this analysis, the ColoF institution

was characterized as the environment with the greatest consensus on

appropriate behaviors while the FCIF institution was characterized as

the environment with the least consensus on appropriate behaviors.











Table 13. Measure of Group Consensus Derived from the
Attitude Toward Prison Behaviors Items
for Seven Institutions


Inmate Opposition


FCIM


1. Inmates cause as
much trouble as they
can.

8. Inmates lie to an
officer if they can get
away with it.

14. Inmates goof off
while they are in
school.

28. Inmates work it
out so they can con
the staff.

34. An inmate tries
to get around as many
of the rules as pos-
sible.


4
47
1.132

3
36
1.246

4
51
1.087

3
40
1.132

4
39
2.179


UCI ColoM L

5 4,5
44 38
.198 .968 1

4 3
39 43
.258 1.286

4 4
46 50
.716 .677 1


4
49
1 .031

4
47
1.272


Total Variance on Inmate Opposition:
6.776 5.475


3
41
1.051

4
39
1.443


exM FCIF

5 5
49 57
.226 .908

3 5
41 34
.941 3.026

5 4
40 46
.327 .671

3 4
46 47
.923 .797

3 4
45 41
.961 .878


ColoF

4
45
.548

3
57
.500

4
55
.824

4
48
.533

4
53
.567


5.425 5.378 6.280 2.972


Staff Treatment


3. Staff members try
to help inmates take a
new look at their life.

9. Staff members try
to understand an in-
mate's problems.

15. Staff members
take a personal inter-
est in inmates here.

22. Staff members
help inmates to plan
their future on the
outside.


1 1
7 45
.985 1.197


2
41
1.324

2
43
1.493

2
37
1.471


2 1 2
44 54 51
.825 .904 .711


1 1,2 1
5 42 56
1.207 .984 .962

2 2 1
35 37 42
1.052 1.517 1.250

2 2 2
45 42 45
.979 .729 .667


LexF

4
46
.533

3,4
40
.851

4
63
.717

4
53
.533

3
37
1.217



3.851


2 2
43 39
.733 .674


1
42
1.742

2
48
1.387

2
35
2.065


1
55
1.117

2
38
1.636

2
37
1.440


1



1












Table 13 continued


Item FCIM UCI ColoMl LexM FCIF ColoF LexF

29. Staff members try M 2 1 2 1 1,2 2 1
to teach an inmate % 45 53 45 53 42 42 53
skills that will help V .836 1.125 .783 .980 .763 .903 1.128
on the outside.

Total Variance on Staff Treatment:
6.109 5.560 4.838 4.763 5.667 6.830 4.053

Inmate Avoiding Trouble

4. Inmates try to get M 1,2 3 2 3 2 2 4
along by keeping their % 27 31 33 31 32 33 27
mouth shut around the V 1.636 1.492 1.825 1.615 1.667 1.800 2.356
staff.

16. An innate tries M 2 3 3 4 4 3 4
to steer clear of the % 31 35 34 40 35 58 41
staff. V 1.691 1.128 1.242 1.038 2.230 .889 1.478

23. Inmates try to M 3 3 3 3 4 3 3
find the easiest job % 39 45 51 55 37 55 51
they can. V 1.043 .984 1.467 .804 1.547 1.000 .778

30. An inmate keeps M 3 3 3 3 3 3 3
to oneself as much as % 39 38 39 41 38 53 40
possible. V 1.015 1.157 1.017 .824 1.173 .667 1.222

35. Inmates act like M 3 4 4 4 4 3,4 4
their stay here is just % 40 44 47 39 42 32 41
a matter of waiting out V 1.106 1.108 1.133 .902 1.278 1.161 1.130
time.

Total Variance on Inmate Avoiding Trouble:
6.491 5.769 6.684 5.183 7.895 5.517 6.964

Staff Custody

2. Staff members only M 4 4 5 5 4 4 4
concern themselves with % 40 29 33 42 32 33 56
keeping the inmate from V 1.788 1.897 21635 2.170 2.342 1.500 1.422
causing them trouble.

5. Staff members act M 2 2 2 2 2 2 2
like their main job is % 42 38 32 34 34 37 33
to keep things running V 1.776 1.862 2.667 2.509 1.895 2.833 1.848
smoothly.












Table 13 continued


Item

10. Staff members act
as if their main job
is preventing escape.

17. Staff members see
an inmate as someone
to be controlled.

Total Variance on Staff


FCIM UCI ColoM LexM FCIF


4
32
1.809

4
42
1.167


4
30
1.859

4
36
1.391


4
42
1.306

4
47
1.186


Custody:
6.540 7.009 7.794


4
35
1.462

4
37
1.385


4
41
1.539

4
41
1.413


ColoF LexF

4 5
38 31
1.226 4.178

4 4
55 42
.806 1.000


7.526 7.189 6.365


Staff Punishment


6. Staff members push M
inmates till they %
break. V

11. Staff members are M
rough with inmates to %
show them who's boss. V

18. Staff members see M
to it that inmates have %
a hard time here to V
make up for what they
did on the outside.

24. Staff members re- M
mind inmates that they %
are in here to pay for V
their crimes.

31. Staff members send M
an inmate to segregation%
even for little things. V


5
48
1.652

4
40
1.412

4
39
1.130



4
38
1.191


4
49
.791


4 4 4 5
50 52 48 48
.742 .661 .519 1.792


4 5
41 44
.808 1.607


5
52
1.585


5
43
2.344


4 4
45 44
.937 1.148


4,5 5
40 49
.759 2.493


5
42
2.077



4
31
1.255


5
57
1.195



5
43
2.622


4 4 4 5
42 42 43 39
.691 .814 .569 2.373


Total Variance on Staff Punishment:
6.176 4.763


6.574 5.189 10.475 4.516


Inmate Self-Improvement


7. An inmate really M
tries to learn something%
in work release that V
will be of use later.


1 1 1
57 53 63
.884 .705 1.032


1 1 1 1
55 63 58 74
.902 .592 .516 .261


8.448


4 5
58 46
.419 1.457


5
45
1.613


5
54
1.043


4 5
52 57
.742 1.227


4
55
1.161


4,5 4,5
42 42
.581 .837


5.541












Table 13 continued


I tem


19. Inmates do the
best work they can when
they are on a work
detail.

25. Inmates try to
figure out how to get
along with other in-
mates while they are
in here.

32. Inmates do their
best to cooperate when
they are assigned to
work with another in-
mate.

36. Inmates spend a
lot of time thinking
while they are in here
about how to get along
on the outside.


FCIM UCI ColoM LexM


FCIF ColoF LexF


2 2 2 2 2 2 2
43 54 47 44 46 42 42
.812 .802 .633 .923 .645 .677 .756


1,2 2 2 1
41 48 58 46
.721 .725 .475 1.200


1 2 2
49 58 56
.747 .419 .511


2 2 2 2 1 2 2
46 53 54 51 47 57 62
.746 .637 .508 .549 .773 .433 .378


1
46
1 ,000


2 2 1
43 44 53
.824 .729 1.078


Total Variance on Inmate Self-Improvement:
4.163 3.693 3.377

Measure of Group Consensus (Overall Variance):
36.255 32.369 34.692


1
52
1.027


2 1
65 60
.452 .956


4.652 3.784 2.497 2.862


32.691 41.290 28.697 31.719


Note: M = Group Mode, where 1
2
3
4


= absolutely should
= should
= may or may not
= should not


5 = absolutely should not.
% = Percentage of group in modal category.
V = Variance around the group's item mode.









Living Units

Seventeen living units, each with 15 or more respondents, were

included in this analysis. Table 14 presents the total amounts of vari-

ance around the mode of all items in each of the six subscales for each

living unit. Inspection of Table 14 reveals considerable variation in

the amount of variance around the item modes within various institu-

tional living areas. FCIF Stowe had the least amount of total variance

while FCIF Tubman had the greatest amount of total variance. Thus,

these two living units represented the extreme conditions of group

consensus on appropriate behaviors within this sample. Therefore, among

these living units, FCIF Stowe was characterized as an environment

where there was the greatest consensus on appropriate behaviors (or

was the least anomic unit) and FCIF Tubman was characterized as an en-

vironment where there was the least consensus on appropriate behaviors

(or scored highest on this index of anomie).

Ethnic Groups Within Institutions

The group consensus measure was computed for ethnic groups within

six institutions. Groups from the ColoF institution were excluded due

to insufficient sample sizes. Each of the twelve ethnic samples included

in this analysis had 10 or more respondents. A Black sample and a White

sample were included from five of the institutions. From ColoM, a Uhite

sample and a sample composed of non-Black minority subjects were included.

Table 15 contains the variance calculations for each ethnic sample.

As in Table 14, Table 15 shows only the total amounts of the variance

around the modes of all items in each of the six subscales. It can be

seen that in all institutions, with the exception of LexM, the variance

total was greater for the minority sample than for the majority sample.












Table 14. measure of Group Consensus Derived from the
Attitude Toward Prison Behaviors Items
for Seventeen Living Areas

Overall
IOP STR IAT SCU SPU ISI Variance


5.457
6.473
4.509
7.610
5.873
5.900
8.028
3.797
4.924
,.936
6.216
8.732
2.867
5.709
2.716
3.630
3.691


6.497
7.696
4.518
8.330
4.347
5.750
11.516
3.372
3.454
6.207
5.864
8.957
3.308
9.533
6.893
7.433
4.103


5.469
8.660
5.476
7.269
7.656
6.845
6.094
6.370
5.865
6.148
8.886
10.747
5.000
6.681
4.692
6.222
4.351


6.763
9.771
12.591
7.007
10.311
9.984
7.586
6.666
5.910
11 .742
6.640
11.200
7.067
15.734
6.365
10.730
6.222


5.885
8.620
6.336
3.564
5.143
6.400
12.664
3.075
4.408
6.342
7.333
14.513
3.362
12.200
4.516
5.631
5.533


3.178
4.927
3.345
4.329
5.373
4.000
3.371
2.443
3.121
5.298
3.286
2.002
3.001
4.067
2.497
3.983
3.593


33.249
46.147
36.775
38.109
38.703
38.879
49.259
25.723
27.682
40.673
38.225
56.151
24.605
53.924
27.679
37.629
27.493


Note: Dorm 1
2
3
4
5
6
7
8
9


= FCIM Dorm C
= FCIM Dorm D
= UCI East 2T
= UCI East A Floor
= UCI East D Floor
= UCI East I Floor
= ColoM Dorm
= ColoM Cell House
= LexM Antaeus


= LexM Younity
= FCIF Bethune
= FCIF Tubman
= FCIF Stowe
= FCIF Keller
= ColoF Single Rooms
= LexF Honor Alley
= LexM-F Renaissance


Dorm
















Group


Table 15. Measure of Group Consensus Derived from the
Attitude Toward Prison Behaviors Items
for Ethnic Groups Within Six Institutions

Overall
IOP STR IAT SCU SPU ISI Variance


FCIM-White 4.973 4.229 5.618 4.334 5 657 2.943 27.754
FCIM-Black 8.559 8.920 7.898 8.645 9.512 4.713 48.253


4.054 4.789 5.022 7.487
7.139 7.353 8.097 13.300


4.215 3.366 28.933
6.873 5.267 48.029


ColoM-White 3.527
ColoM-NonBlack 10.134
Minority


4.612 4.909 7.500
6.266 9.067 14.738


5.972 2.684
10.800 3.267


LexM-White 4.903 5.019 5.148 7.775 5.644 4.613 33.102
LexM-Black 3.677 4.258 5.472 7.765 5.526 4.779 31.477


FCIF-White 6.000 2.208 6.083 4.500 4.000
FCIF-Black 6.668 10.606 9.606 11.624 12.145


LexF-White 2.416 3.787 6.700 4.932
LexF-Black 5.060 4.691 8.176 10.894


3.459 26.250
3.712 54.361


4.465 3.145 25.446
5.814 3.154 37.789


UCI-White
UCI-Black


29.204
54.272









Thus, minority groups tended to demonstrate less consensus on appropriate

behaviors than did White samples. Among White samples, LexF had the

highest consensus ranking while Lexll had the lowest. Among minority

samples, LexM-Black had the highest consensus ranking while FCIF-Black

and ColoM-NonBlack minorities had the lowest.

We will now return to an examination of the hypothesized rela-

tionship between personal control and consensus.

Hypothesis 4A. Persons residing in an environment where there
is low consensus on what are appropriate behaviors
should believe in personal control to a lesser
extent than persons residing in an environment
where there is high consensus.

In order to examine this hypothesis, the measures of group consen-

sus within each analysis were ranked ordinally by degree of consensus.

The relationship of these consensus rankings with mean Personal Control

scores for the respective groups was assessed by means of Spearman

rank-order correlation coefficients.

In the analysis by institution, degree of consensus and Personal

Control correlated at the level of rs = .750, p<.05. In the analysis

by living unit, degree of consensus correlated with Personal Control

at the level of rs = .630, p<.01. In the analysis by ethnic group

within institutions, degree of consensus correlated with Personal Con-

trol at the level of rs = .740, p<.0l. Thus, there is strong, consistent

support for the hypothesis.

Hypothesis 4B. Persons residing in environments which contain
small numbers of people should believe in person-
al control to a greater extent than persons resid-
ing in environments which contain large numbers
of people.

Institutions were ranked by population size. The relationship of

these rankings with mean Personal Control scores for the institutions









was examined by Spearman rank-order correlation coefficients. Table 16

contains the results of this analysis. The coefficient of correlation,

rs = .572, failed to achieve significance at the .05 level.



Success in the Environment

Hypothesis 5A. Belief in personal control should be positively
correlated with measures related to success in
the environment.

Success in the environment was defined to include the following

variables: freedom of movement (FM), alienation (AL), number of times

rewarded by staff in the past year (REWARD), number of disciplinary

reports received in the past year (DR), and staff ratings of maladjust-

ment (MALAD), association (ASSOC), and leadership (LEADER). Each of

the seven measures of success were correlated with the Personal Control

measure. Table 17 contains the Pearson product-moment correlations

which obtained.

Personal Control correlated most highly with FM and with AL, with

r = .273 and r = -.408, respectively (both significant at the .01 level

or better). Significant relationships in the predicted direction (at the

.05 level or better) also obtained between Personal Control and DR

(r = -.088) as well as between Personal Control and MALAD (r = -.156).

Relationships with REWARD, ASSOC, and LEADER failed to achieve signifi-

cance. The results demonstrated that inmates who believe in personal

control tended to hold high expectations of goal attainment, feel less

alienated, receive fewer disciplinary reports, and were more likely to

behave in ways to be seen as model inmates by the correctional staff.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs