184
37? /Vg/c/ THE VALUE SYSTEMS OF INCARCERATED EMBEZZLERS: THE IMPLICATIONS FOR SOCIOLOGICAL PRACTICE AND VALUE CLARIFICATION PROGRAMS FOR CORRECTIONAL INSTITUTIONS DISSERTATION Presented to the Graduate Council of the North Texas State University in Partial Fulfillment of the Requirements For the Degree DOCTOR OF PHILOSOPHY By Michael Welch, M.A., M.S. Denton, Texas August, 1987

37? THE VALUE SYSTEMS OF INCARCERATED EMBEZZLERS: …/67531/metadc330949/... · criminological theories fail to provide a cogent understanding of embezzlement. In light of this, there

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

  • 3 7 ?

    /Vg/c/

    THE VALUE SYSTEMS OF INCARCERATED EMBEZZLERS: THE

    IMPLICATIONS FOR SOCIOLOGICAL PRACTICE AND VALUE

    CLARIFICATION PROGRAMS FOR CORRECTIONAL INSTITUTIONS

    DISSERTATION

    Presented to the Graduate Council of the

    North Texas State University in Partial

    Fulfillment of the Requirements

    For the Degree

    DOCTOR OF PHILOSOPHY

    By

    Michael Welch, M.A., M.S.

    Denton, Texas

    August, 1987

  • Welch, Michael, The Value Rvstams of Incarcerated

    Embezzlers: The Implications for Sociological Practice and

    Value Clarification Pm^rami for Correctional Institutions.

    Doctor of Philosophy (Sociology), August, 1987, 178 pp., 12

    tables, references, 62 titles.

    An empirical investigation at a southwestern minimum

    security federal correctional institution was designed to

    assess the value systems of incarcerated embezzlers (N = 31)

    as they compared to a matched offender control group (N =

    31). Based on their responses on the Rokeach Value Survey

    (RVS). no statistically significant differences between

    these groups were found. Therefore, this finding suggested

    that these embezzlers possessed similar value systems held

    by those inmates convicted of other crimes.

    When the responses of the embezzler sample were

    combined with their matched offender control group, a few

    differences were revealed between the combined inmate group

    and the general population norms (National Opinion Research

    Center [N0RC3» Rokeach, 1968, 1973). Simple comparisons of

    the composite medians of the male inmate group and the NORC

    showed differences on the following survey items: "a world

    at peace," "equality," and "national security." Among the

    items which demonstrated differences between the female

    prison group and the NORC were "an exciting life," "wisdom,"

    "independent," "intellectual," "logical," "a world at

  • peace," and "national security." The findings provided

    partial support for Cochrane's (1971) conclusions that

    prisoners are self-centered, and place low importance on

    those values which do not have immediate or personal

    relevance. However, because many of the inmates' responses

    emulated the NORC data, it was concluded that their value

    systems resembled the general population more than other

    prison populations.

    In addition to the empirical analyses, this project

    addressed the practical implications of value systems

    research by proposing value clarification programs for

    correctional institutions. The selection of value

    clarification programs was inspired by the implications of

    the emerging perspective of sociological practice.

    Sociological practice was described as it relates to these

    programs as well as to sociology in the larger context.

  • TABLE OF CONTENTS Page

    LIST OF TABLES . . . .

    Chapter

    I. INTRODUCTION

    Statement of the Problem Purpose of the Study Definitions of Terms Basic Assumptions Review of Related Literature Values The Rokeach Value Survey Value Clarification The Embezzler Significance of the Study

    II. METHOD 18

    Hypotheses Male Embezzlers Versus Their Matched Control Group

    Female Embezzlers Versus Their Matched Control Group

    Fort Worth Male Inmates Versus the NORC Sample

    Fort Worth Female Inmates Versus the NORC Sample

    Scope Instrument The Sample Research Design Procedures for Analysis of Data and Testing of Hypotheses

    Summary

    III. RESULTS 30

    Subjects Procedure Data Analysis

    iii

  • TABLE OF CONTENTS—CONTINUED

    Page

    The Testing of the Hypotheses Interpretation Further Analyses Additional Analyses Summary

    IV. DISCUSSION 63

    Summary of the Findings Methodological Considerations Implications Sociological Practice Value Clarification Conclusions

    APPENDICES 83

    REFERENCES 171

    xv

  • LIST OF TABLES

    Table Page

    1. Composition of Embezzlers at Forth Worth . . . 31

    2. Composition of Non-Embezzler Matched Control Group (Fort Worth) 32

    3. Composition of Samples 33

    4. Value Rankings of Male Prisoners and Matched Control Groups (Terminal) 44

    5. Value Rankings of Male Prisoners and Matched Control Groups (Instrumental) 45

    6. Value Rankings of Female Prisoners and Matched Control Groups (Terminal) 46

    7. Value Rankings of Female Prisoners and Matched Control Groups (Instrumental) 47

    8. Fort Worth and NORC Medians for Terminal Value . 50

    9. Fort Worth and NORC Medians for Instrumental Value 51

    10. Frequency Distributions of Education 56

    11. Frequency Distributions of Religion 58

    12. Frequency Distributions for Income 60

    v

  • CHAPTER I

    INTRODUCTION

    This chapter is presented to introduce the study of

    value systems and embezzlers by stating the research

    problem, the purpose of the study, the definition of

    research terms, and the basic assumptions. Furthermore, a

    review of the literature relating to value systems and

    embezzlers is included.

    This author examined the social psychology of

    embezzlers insofar as it related to their value systems, and

    this study was designed to address some basic research

    issues. First, it was recognized that traditional

    criminological theories fail to provide a cogent

    understanding of embezzlement. In light of this, there is a

    dearth of recent studies on the embezzler despite a recent

    trend to incarcerate white-collar criminals. Second, the

    Rokeach Value Survey (RVS) (Rokeach, 1968), an instrument

    which measures value systems, has been applied to numerous

    populations. However, few studies have focused on the

    criminal population, and this author has not found an

    application of the RVS to white-collar offenders. Third,

    assuming that criminality was related to the inmates' value

    systems, the task of this research was to extend value

  • systems research to the development of value clarification

    programs for correctional institutions. Therefore, the

    examination of the inmates' value systems has direct

    theoretical and practical implications.

    Statement of the Problem

    The study addressed the following research questions:

    (1) Are the value systems of incarcerated offenders

    different from those of the general population, and if so,

    how? (2) Are the value systems of the incarcerated

    embezzlers different from those of other offenders who are

    incarcerated for crimes other than embezzlement, and if so,

    how? Considering these potential differences, can a related

    intervention strategy be introduced in correctional

    institutions to remedy antisocial behavior?

    Purpose of the Study

    The ultimate purpose of this investigation was to

    replicate the value system research of Cochrane (1971) who

    employed the Rokeach Value Survey (RVS) and found

    significant differences between inmates and the general

    population. This author hypothesized similar findings and

    should these results be replicated, they should provide

    adequate reason to propose the development of value

    clarification programs for inmates. The development of such

    interventions was recommended by Cochrane, who maintained

    that a value-oriented strategy may reduce anti—social

    behavior. In essence, this study was designed to continue

  • the value systems research of Cochrane and to explore its

    practical implications.

    Definitions of Terms

    The following terms have restricted meaning and thus

    were in this investigation as follows:

    1. Value: "an enduring belief that a specific mode of

    conduct or end-state of existence is personally or socially

    preferable to an opposite or converse mode of conduct or

    end-state of existence" (Rokeach, 1973, p. 5).

    2. Value System: "an enduring organization of beliefs

    concerning preferable modes of conduct or end-states of

    existence along a continuum of relative importance"

    (Rokeach, 1973, p. 5).

    3. Rokeach Value Survey (RVS): an "instrument [which]

    assesses a respondent's hierarchical arrangement of two

    kinds of values: instrumental and terminal. Instrumental

    values refer to preferable modes of conduct; terminal values

    refer to preferable end-states of existence" (Robinson &

    Shaver, 1973, p. 547).

    4. Value Clarification: the organization of various

    exercises designed to assist individuals in re-examining

    their value systems. It rests on the assumption that value

    systems are dynamic and capable of being restructured

    (Kirschenbaum, 1977).

  • Embezzlers: those offenders "who use a position of

    trust to appropriate company assets for their own personal

    use" (Coleman, 1985, p. 81). In this study, the embezzlers

    represented four categories of embezzlement: "Bank,"

    Postal, Benefit Plan," and "Other" (including Savings and

    Loan Associations) violations.

    Basis Assumptions

    The basic assumption of this research was that a source

    of motivation (criminal intent) is the value system.

    According to Rokeach (1973), the value concept should be

    clearly distinguished from other concepts such as attitude,

    social norm, and need. However, all these other concepts

    are related to values. Rokeach identified five major

    assumptions regarding the nature of human values.

    (1) the total number of values that a person possesses

    is relatively small; (2) all men everywhere possess the

    same values to different degrees; (3) values are

    organized into value systems; (4) the antecedents of

    human values can be traced to culture, society and its

    institutions, and personality; and (5) the consequences

    of human values will be manifested in all phenomena

    that social scientists consider worth investigating and

    understanding (1973, p. 3).

    This study was based on Rokeach's general assumption that

    the value concept should occupy a central position across

    all social sciences.

  • It was assumed that, while value systems at the

    societal level remain somewhat stable, the value systems at

    the individual level are more easily changed, particularly

    when they contrast with the prevailing values. Therefore,

    individual value systems should be more subject to

    clarification and alteration through an intervention. This

    assumption was based on the empirical and theoretical

    findings of Rokeach (1968, 1973) and Grube (1982). This

    author assumed that the restructuring of value systems can

    have a positive social impact on inmates, particularly

    embezzlers. Sutherland and Cressey (1978) noted that the

    embezzler typically has a low rate of recidivism because it

    is unlikely that he or she will again be placed in a

    position of trust. However, it was assumed that if the

    embezzler does not confront these value systems while in

    prison, he or she may turn to other anti-social behaviors

    because the values system that may have influenced his or

    her criminality was not adjusted. Although the embezzlers

    were the target group for this study, it was assumed that

    all inmates may benefit from value clarification programs.

    Finally, this project was based on the notion that value

    system research can provide an empirical base for the

    development of value clarification programs.

  • Review of Related Literature

    The literature review for this research focused on four

    general areas: (1) those works concerned with the

    conceptual and theoretical development of values; (2) the

    research employing the Rokeach Value Survey: (3) those

    projects endorsing the use of value clarification programs;

    and (4) traditional and recent studies of the embezzler.

    Values

    Values have been the topic of numerous social sciences

    analyses. One of the most frequently cited researchers in

    value studies is Milton Rokeach (1960, 1968, 1973, 1979).

    Rokeach's work provided the theoretical and conceptual

    framework for this study. This section includes a review of

    the literature as it relates to Rokeach's emphasis on

    values.

    Rokeach, a social psychologist, was influenced by

    various sociologists, psychologists, philosophers, and

    anthropologists. One of his primary influences was Robin

    Williams (1968) who emphasized the importance of research on

    values. Williams noted that a person's values serve as "the

    criteria, or standards in terms of which evaluations are

    made. . . . Value-as criterion is usually the more important

    usage for purposes of social scientific analysis" (1968, p.

    283). Williams proposed that individuals have a relatively

    small number of values. This idea has theoretical and

    methodological implications in that the "tasks of

  • identifying them [i.e., values] one-by-one and measuring

    them become considerably easier, and it also becomes easier

    to grapple with theoretical problems. . (Rokeach, 1973,

    p. 4). In sum, it was through the theoretical foundations

    of Williams that Rokeach was able to articulate the concept

    of values as a central and dynamic feature of human

    functioning and sociological research.

    Rokeach (1973) discussed the major developments in his

    analysis of values. First, he suggested that a value is

    enduring, and provides the standards and criteria for

    behavior and personal evaluation. Rokeach claimed that

    without the stability of values, the continuity of

    personality and society would be impossible. On the other

    hand, he demonstrated that values are not completely stable,

    and they are apt to change. Without the flexibility of

    values, individual and social change would also be

    impossible.

    Rokeach conceptualized values as being either

    instrumental or terminal. The instrumental values are those

    which are concerned with desirable modes of conduct. These

    instrumental values are classified as moral values or

    competence values. Moral values tend to be interpersonal

    while competence values encompass a personal-oriented system

    which is related to self-actualization. For example,

    honesty and love are typical moral values while imagination

  • 8

    and logic are viewed as competence values. The terminal

    values are those which refer to end-states of existence and

    are divided into personal and social values. Personal

    values are self-centered and intrapersonal. Salvation and

    peace of mind, for example, are personal values, while world

    peace and brotherhood are social values.

    For this study, the most relevant areas of research

    were those inquiries integrating values and delinquency.

    Several studies suggested an important link between the two.

    For example, Hudak, Andre and Allen (1980) claimed that

    social adjustments and deviance were related to social

    values. They reported that developmental failures can be

    identified in the adoption of problematic value systems.

    Saari (1983) discussed the value systems of juvenile

    delinquents and argued that values influence behavior

    insofar as delinquents tend to expect immediate returns on

    their actions. These value systems led to poorly planned

    action which resembled criminal patterns of social behavior.

    Saari also mentioned the utility of interventions that focus

    on preparing adolescents to understand the consequences of

    their action.

    Heather (1979) and Braithwaite and Braithwaite (1981)

    attacked some of the conclusions from this research on value

    systems and delinquency because some of these studies failed

    to account for the influence of social class. Braithwaite

    and Braithwaite stated that many theories about values and

  • 9

    delinquency are contradictory, inconsistent with empirical

    evidence, and detrimental to the development of social

    policy. To be effective, they believed that intervention

    programs must be based on sound value systems research.

    The Rokeach Value Survey

    The measurement of values is a popular area of social

    psychological research. In fact, according to Subonen

    (1985), empirical value research is enjoying a renaissance,

    and the concept of values has acquired new prominence in the

    social sciences. Because recent studies were influenced by

    the theoretical work of Rokeach, many social scientists used

    the Rokeach Value Survey (RVS) in their analyses. For

    example, McKernan (1982) and McKernan and Russell (1980)

    applied the gVS to an adolescent population in Northern

    Ireland to appraise differences in religion and race as

    compared to Americans. McKernan (1982) found that

    minorities in Northern Ireland compared to American

    minorities because they endorsed the value of equality and

    did not place a high priority on hedonism. Similarly,

    McKernan and Russell concluded that political climate and

    environmental pressures affect value systems. For example,

    in Northern Ireland, Catholics rated equality and freedom

    much higher than their Protestant counterparts.

    Teahan, Adams, and Podany (1980) also reported the

    importance of environmental conditions and values. Teahan

  • 10

    et al. investigated the values of police officers and

    demonstrated that as crime rates increased, the officers

    became more emotionally detached, less forgiving of others,

    and more concerned with achieving a sense of inner harmony.

    Other researchers examined the impact of values and

    political behavior. For example, Thomas (1981) used the

    structure of value systems in his prediction of political

    activism among Protestants and Catholics during the Boston

    school desegregation controversy. Many of the politically-

    oriented studies were influenced by Rokeach (1973), who

    identified four basic political orientations: conservative,

    socialist, fascist, and communist.

    The area of cross-cultural research further

    demonstrated the utility of the RVS. Reynolds (1984) found

    that Germans were competence-oriented in their modes of

    conduct, and Americans were generally morality-oriented.

    Moreover, Germans were society-oriented and interpersonal in

    their preferred end states, and Americans were more self-

    centered and intrapersonal. Among the numerous other

    applications of the RVS were studies involving adolescent

    socialization (Skeel, 1976), medical students (Feather,

    1982), corporate managers (Clare & Sanford, 1979), and

    differences among spouses (Medling & McCarrey, 1981).

    In a project especially relevant to this research,

    Cochrane (1971) examined a Michigan prison population using

  • 11

    the RVS to measure the value systems of male and female

    inmates. Cochrane found significant differences between the

    value systems of these inmates and the general population.

    Cochrane concluded that "Prisoners appear to have a shorter

    time perspective and value those things which have immediate

    and personal relevance . . . .Prisoners value the

    characteristics of 'wisdom' and 'self-controlled' relatively

    high[ly] possibly because they see these lacking in their

    lives (1971, p. 79). Cochrane interpreted the value

    systems of these prisoners as hedonistic, self-centered, and

    unconcerned for the plight of others. This Michigan study

    served as the central base for this research insofar as a

    replication was designed.

    Value Clarification

    Because one of the major features of this research was

    the extension from an empirical study on value systems to

    the development of an intervention program for inmates, it

    was appropriate to describe value clarification. The value

    clarification movement was initiated by the work of Raths

    (1968), which paralleled the subsequent conclusions of

    Rokeach (1973) and Grube (1982) that values are subject to

    re-evaluation and change. Value clarification is the

    process by which individuals are allowed the opportunity to

    re-examine their value system and, through various

    exercises, are able to change their current set of values.

  • 12

    Value clarification is a broad term which refers to

    numerous strategies designed to help individuals re-examine

    their value systems. In fact, Simon, Howe, and Kirschenbaum

    (1978) introduced 79 different strategies for value

    clarification sessions. These strategies ranged from group

    exercises to individual tasks. Although these strategies

    implemented different techniques, they all employed the

    seven processes developed by Raths, Harmin, and Simon (1966)

    to assist participants in their search for value clarity.

    They are to (1) choose from alternatives; (2) thoughtfully

    consider the consequences of alternatives; (3) choose

    freely; (4) prize and cherish; (5) publicly affirm; (6) act

    repeatedly; and (7) act with a pattern or consistency.

    Kirschenbaum (1977) emphasized that these seven processes

    could be interpreted as the development of the "criteria"

    for a "value." He reiterated that the task is not only to

    clarify one's values but to become fully aware of the

    consequences (both personally and socially) of one's

    adoption of a given value.

    The primary contributors to value clarification are

    Kirschenbaum (1977), Simon, Kirschenbaum, and Howe (1978),

    and Harmin, Kirschenbaum and Simon (1973). The value

    clarification movement was supported by Rokeach, who

    strongly endorsed this approach to intervention

    (Kirschenbaum, 1977). In one value-oriented program,

    Thoresen (1984) integrated value clarification with

  • 13

    sociological practice and demonstrated its utility as a

    teaching technique. Thoresen noted the importance of

    identifying values which directly influence social

    interaction. She stated that the task of implementing value

    ication with sociological practice and demonstrated

    its utility as a teaching technique. Thoresen noted the

    importance of identifying values which directly influence

    social interaction. She stated that the task of

    implementing value clarification techniques was "to

    sensitize students to the effect of values on their own

    choices, and to identify the social categories around which

    individuals develop values" (1984, p. 137).

    The Embezzler

    Donald R. Cressey is one of the leading contributors to

    research on embezzlement. In Other People's Money: £ Study

    in t]ie Social Psychology of Embezzlement. (1953), Cressey

    described the embezzler as an employee who was granted a

    position of trust in a bank or company. After a

    considerable period of time, this trusted employee began to

    borrow" money from the bank or company to reduce a personal

    debt. This situation of "borrowing" was identified by

    Cressey as a solution to an "unshareable problem." The

    employee s unshareable problem" was one that he or she

    believed could not be resolved by seeking the help of other

    people. Therefore, it was theorized that the employee kept

  • 14

    his or her problem a secret, but embezzled money to reduce

    the debt. Cressey linked the debt to an extravagant

    lifestyle.

    Embezzlement is a unique topic in the criminological

    literature because traditional theories fail to adequately

    explain violations of trust. Sutherland and Cressey (1978)

    noted that embezzlers are rarely "psychopathic, feeble-

    minded, residents of deteriorated slum areas, or in other

    ways personally or situationally pathological" (pp. 266-

    267). On the contrary, embezzlers are usually educated,

    employable, and most violations of trust are committed by

    employees who have held positions of financial

    responsibility for several years. In fact, Sutherland and

    Cressey mentioned that few embezzlements are committed by

    new employees. Furthermore, most embezzlers do not have a

    criminal history, have lived a respectable life, and are

    middle-class and middle-aged males (Hagan, 1986).

    Cressey's (1953)) work was criticized from different

    angles. Schuessler (1954) pointed out that Cressey's

    findings, based on 133 interviews, were difficult to

    generalize because they were ex-post facto and only those

    incarcerated were represented in the study. Nettler (1974)

    criticized Cressey for the introduction of the "unshareable

    problem." He found no support for this "unshareable

    problem," and viewed the embezzler as motivated by greed,

  • 15

    temptation, as well as having the opportunity to violate the

    company's trust.

    These criminological theories were consistent with

    Akerstrom (1985) who stated that criminals are condescending

    toward rule-bound norms and highly value individualism and

    excitement. With the exception of Akerstrom, recent studies

    of the embezzler are rare. Most current research has

    categorized all white-collar offenders together. This is a

    fundamental mistake because there are distinct differences

    among several types of white-collar criminals, such as the

    corporate criminal whose victim is the consumer and the

    intra-organizational offender who victimizes a company

    (Coleman, 1985; Welch, 1987).

    The current literature on embezzlement contributed

    little to the theoretical base of this study. For the years

    1980 through 1986 the Criminological and Penological

    Abstracts listed only a few studies under the heading

    Embezzlement, and no such citations were listed in

    Sociological Abstracts for the same period. The few studies

    included under this heading offered sweeping generalizations

    for the white collar offender with limited links to

    criminological theory. Rather, they attempted to market an

    approach to crime that would provide preventative

    instruments such as personnel assessments. Their attempts

    to reduce crime appeared somewhat theoretically naive and

    their supportive evidence was flawed.

  • 16

    Those studies categorized under the heading

    "Embezzlement" were not relevant to the criminological

    definition of embezzlement, that is, an employee's violation

    of a position of trust. The literature included research

    which is more appropriate for "Employee Theft" studies where

    crimes were committed by employees who were not in a

    position of trust. Consequently, it was difficult to apply

    the generalizations of these investigators to embezzlement.

    Although employee theft research is meaningful, the

    conclusions of these studies were not relevant in the

    attempt to understand the social psychology of the

    embezzler. These studies included employee theft in

    convenience stores (Brown, 1986; Terris & Jones, 1982), drug

    stores (Brown & Pardue, 1985), restaurants (Hawkins, 1984),

    home improvement centers (Jones & Terris, 1983) and

    insurance brokerage firms (Smith, 1981). Therefore, this

    author concluded that these investigations focused on an

    entirely different aspect of white-collar crime.

    Significant of the Study

    It was the hope of this author that by approaching the

    act of embezzlement from a perspective involving value

    systems, a more convincing explanation of this crime could

    be formulated. Moreover, criminological theory of

    embezzlement could be clarified by identifying the

  • 17

    embezzler's possible source of motivation (that is, his or

    her value system).

    The empirical dimensions of this study were included to

    contribute to the practical applications of these research

    findings. In the spirit of sociological practice this

    research was designed to offer direct intervention

    implications by suggesting value clarification programs for

    incarcerated offenders, particularly embezzlers. Cochrane

    underscored the importance of value systems to intervention

    by suggesting that "knowledge of individual prisoners' value

    system could be used for counseling and rehabilitation

    purposes (1971, p. 79). There is a greater need for such

    value oriented programs due to the increase in prison

    sentences for white-collar offenders. This author suggests

    that value clarification programs be developed because there

    is a dearth of intervention programs for those convicted of

    white-collar criminal offenses. Because prisons are assumed

    to be "correctional" it is not difficult to defend the

    development of a value clarification program to adjust

    problematic value systems. By participating in such

    programs it is assumed that the inmate would be less likely

    to be involved in further anti—social behavior.

  • CHAPTER II

    METHOD

    This chapter is presented to address the components of

    the research design by: (1) describing the rationale for

    the hypotheses; (2) listing the hypotheses; (3) discussing

    the scope of the research, the instrument, and the sample;

    and (4) articulating the procedures for data analysis and

    the testing of the hypotheses.

    Hypotheses

    This author set out to replicate Cochrane's (1971)

    research on the value systems of inmates using the Rokeach

    V a l u e S u r v e r (RVS)• It was hypothesized that there would be

    a difference between the value systems of the incarcerated

    inmates at the Federal Correctional Institution, Fort Worth

    and the general population. Cochrane found that inmates

    maintain "a shorter time perspective and value those things

    which have immediate and personal relevance" when contrasted

    with the general population (1971, p. 79). In addition to

    this hypothesis this study assessed the hypothesized

    differences between the incarcerated embezzlers and their

    matched offender control group (those inmates convicted of

    crimes other than embezzlement).

    18

  • 19

    While the embezzlers and their matched control offender

    group were expected to endorse similar value system

    patterns, it was hypothesized that the embezzlers would

    place more importance on some selected items of the RVS.

    This author proposed that the differences in value rankings

    may be related to the fundamental differences between

    embezzlers and those convicted of other crimes. It was

    assumed that those convicted of crimes other than

    embezzlement, such as violent acts or drug-related offenses,

    had more complicated motivations. For example, the criminal

    convicted on drug charges may have been responding to

    various factors such as peer pressure, addictions,

    discrimination, poor education, and few employment

    opportunities.

    These characteristics are in sharp contrast to those of

    the embezzler, who is typically middle class, employable,

    and does not have a criminal history. Hence, it was assumed

    that the act of embezzlement is much more likely to be

    influenced by the value systems of the offenders than by

    other social pressures. Based on this assumption it was

    hypothesized that the responses to certain items on the RVS

    would show that the embezzler was more self-centered than

    his or her matched offender control group.

    The hypotheses were based on the findings of Cochrane

    (1971) who found statistically significant differences

  • 20

    between the responses of a prison population and the general

    population (National Opinion Research Center, NORC, Rokeach,

    1968). The NORC data set was comprised of a national area

    probability sample of adult Americans (N = 1409).

    Consistent with the methodology of Rokeach, Cochrane

    stratified his sample by gender. He found differences

    between the male inmate group as contrasted with the NORC

    data on the following items: "pleasure," "wisdom," "self-

    controlled," "capable," "a world at peace," "equality,"

    "national security," "salvation," and "honest." Cochrane

    also discovered differences between the female inmate group

    as contrasted with the NORC data on the following items:

    'an exciting life," "a sense of accomplishment," "freedom,"

    inner harmony," "wisdom," "broadminded," "capable,"

    "independent," "intellectual," "logical," "self-controlled,"

    a world at peace," "national security," "salvation,"

    helpful," and "honest."

    The hypotheses for this study were designed to

    replicate Cochrane's findings by testing the same survey

    items for which he found statistically significant

    differences. The hypotheses were presented in four

    sections. First, data from the male embezzlers were

    contrasted with data from their matched offender control

    group. Second, data from the female embezzlers were

    contrasted with data from their matched offender control

  • 21

    group. Third, data from the Fort Worth sample of male

    inmates (embezzlers combined with the Fort Worth sample of

    female inmates (embezzlers combined with the matched

    offender control group) were contrasted with the data from

    the NORC.

    Mile Embezzlers Versus Their Matched Control Group

    The embezzler group will place more importance on the

    following items when contrasted with their matched control

    group:

    H 1: Pleasure

    H 2: Wisdom

    H 3: Self-Controlled

    H 4: Capable

    The embezzler group will place less importance on the

    following items when contrasted with their matched control

    group:

    H 5: A world at peace

    H 6: Equality

    H 7: National security

    H 8: Salvation

    H 9: Honest

    F e m a l e Embezzlers Versus Their Matched Control Group

    The embezzler group will place more importance on the

    following items when contrasted with their matched control

    group:

  • 22

    H 10: An exciting life

    H 11: A sense of accomplishment

    H 12: Freedom

    H 13: Inner harmony

    H 14: Wisdom

    H 15: Broadminded

    H 16: Capable

    H 17: Independent

    H 18: Intellectual

    H 19: Logical

    H 20: Self-controlled

    The embezzler group will place less importance on the

    following items when contrasted with their matched control

    group

    H 21: A world at peace

    H 22: National security

    H 23: Salvation

    H 24: Helpful

    H 25: Honest

    Eort Worth Male Inmates Versus the NORC Sample

    The inmate group will place more importance on the

    following items when contrasted with the NORC:

    H 26: Pleasure

    H 27: Wisdom

    H 28: Self-controlled

    H 29: Capable

  • 23

    The inmate group will place less importance on the

    following items when contrasted with the NORC:

    H 30: A world at peace

    H 31: Equality

    H 32: National security

    H 33: Salvation

    H 34: Honest

    Fort Worth Female Inmates Versus the NORC Sample

    The inmate group will place more importance on the

    following items when contrasted with the NORC:

    H 35: An exciting life

    H 36: A sense of accomplishment

    H 37: Freedom

    H 38: Inner harmony

    H 39: Wisdom

    H 40: Broadminded

    H 41: Capable

    H 42: Independent

    H 43: Intellectual

    H 44: Logical

    H 45: Self-controlled

    The inmate group will place less importance on the

    following items when contrasted with the NORC:

    H 46: A world at peace

    H 47: National security

  • 24

    H 48: Salvation

    H 49: Helpful

    H 50: Honest

    Scope

    This author examined the value systems of embezzlers

    and their matched control group at the Federal Correctional

    Institution at Fort Worth who were incarcerated at the time

    of the data collection (April, 1987). The data from the

    general population was limited to the data set of the

    National Opinion Research Center (Rokeach, 1968, 1973).

    The measurement was limited to the inmates' reported

    value systems based on the Rokeach Value Survey and the

    following demographic variables: offense type, gender, age,

    race, education, religion, and income. This author

    appraised the statistically significant differences between

    the embezzlers and their matched offender control group by

    employing the median test of significance. Further analyses

    involved contrasting the composite medians of the responses

    of the inmates at Fort Worth on the RVg to the general

    population (NORC).

    Instrument

    The Rokeach Value Survey (RVS) appraises the

    hierarchical arrangement of instrumental and terminal

    values. Each of these two sets of values was presented in

    an alphabetized list of 18 items which represent the

    instrumental and terminal values. The instrumental values

  • 25

    are those values related to preferable modes of conduct.

    The terminal values represent preferable end-states of

    existence. The participants were instructed to arrange

    these items in order of personal and relative importance.

    The RVS was revised in 1982. Rokeach (1968, 1973) and

    Cochrane (1971) employed the earlier version (Form D)

    whereas this author used the most recent edition of the

    survey (Form G). Form G was selected for this study because

    it was the only form available. The difference between

    these forms is that Form G includes "health" and "loyal,"

    and Form D includes "happiness" and "cheerful." This

    alteration of items led this author to caution against the

    precise comparison of results using these two forms.

    However, it was assumed that this change of items had

    improved the survey. This survey was self-administered in

    groups and required from 10 to 30 minutes to complete. The

    RVS was used by permission of Rokeach and its publisher.

    The Sample

    The participants in this study were those inmates

    incarcerated at the Federal Correctional Institution at Fort

    Worth at the time of the data collection (April, 1987).

    These subjects were selected from this prison for the

    convenience of this study because this author was employed

    at this facility. Two groups were the focus of this

    research. The first group consisted of embezzlers (N = 31).

  • 26

    The second group included a matched control group of those

    convicted of crimes other than embezzlement, who matched the

    embezzlers by gender, age, and race (N = 31). The control

    group was matched according to the precision (pairwise or

    one-to-one) technique. The responses of these inmates were

    contrasted with the responses of a national area probability

    sample of adult Americans obtained by the National Opinion

    Research Center (NORC) (N = 1409) (Rokeach, 1968, 1973).

    Research Design

    Because value systems research is not always amenable

    to experimental design, a quasi-experimental design was

    selected for this study. While experimental designs are

    appropriate for stimulus-response studies this research

    relied on a quasi-experimental design because of the nature

    of values. As they relate to criminality, values constitute

    property-disposition data. A property-disposition

    relationship involves a background characteristic (for

    example, criminality) as it relates to an orientation (such

    as the criminal's value system). Property-disposition

    relationships are different from stimulus-response

    relationships on four basic research issues: time interval,

    degree of specificity nature of comparison groups, and time

    sequence of events (Rosenberg, 1968; Nachmias & Nachmias,

    1981). The quasi-experimental design in this study was the

    contrasted groups design which was used to assess the value

  • 27

    systems of the target groups: the embezzlers, their matched

    control group, and the NORC sample.

    Participation was encouraged but strictly voluntary and

    forms of informed consent were signed by the participants.

    The inmates received one hour of program credit for their

    participation. Program credit allows the inmate to gain a

    favorable impression from the staff who assume that

    participation in studies is indicative of inmate

    cooperation; hence, this served as incentive. Program

    credit is commonly used in prison studies because the

    inmates participation is viewed favorably by parole boards

    and may improve the inmates' chances for parole.

    This researcher guaranteed confidentiality and

    anonymity to the participants. In fact, the responses were

    treated as aggregate data in which composite rank orders and

    median ranks were compiled and contrasted accordingly.

    Therefore, individual responses were not identifiable. This

    information remained in the sole possession of this

    researcher. Upon compiling the composite rank orders and

    median ranks on each of the 36 RVS items (18 items per set),

    each of these items was ranked based on the composite

    median.

    Procedures for Analysis of Data and Testing of Hypotheses

    Once the composite rank order and median ranks were

    tabulated for the responses of embezzlers and their matched

  • 28

    control groups, the median test was employed to determine

    the existence of statistically significant differences

    between these groups.

    The median test is a nonparametric test of significance

    for ordinal data and "a procedure to determine whether two

    independent groups differ in central tendencies" (Siegel,

    1956, p. 111). The median test provides information

    relating to the likelihood that two independent groups have

    been drawn from the same population. In this study the

    median test was applied at the .05 level of significance

    (one-tailed test). The median test was selected as the

    nonparametric test best suited for these data because of the

    small sample size, the ordinal level of measurement (ranking

    of items), and the fact that this researcher could not

    assume that the responses would be normally distributed.

    A different method of analysis was used when the

    medians of the inmate group (embezzlers combined with their

    matched control group) were contrasted with the medians of

    the NORC data set. Simple comparison of the composite

    medians was used to determine the differences between the

    groups. It was decided that a difference of 1.0 between the

    medians would constitute a difference sufficient for

    rejection of the null hypothesis. The median test could not

    be used at this stage of the analysis because nonparametric

  • 29

    tests require the individual identification of the responses

    and the NORC data set did not provide this information.

    Furthermore, the differences in sample sizes impeded the use

    of the median test for comparisons (Fort Worth group, N =

    62; the NORC sample, N = 1409).

    Further analyses involved a presentation of the medians

    of the groups stratified by offense category, gender, age,

    race, education, religion, and income. These analyses were

    used to assess the impact of the demographic variables as

    well as to search for patterns among the value systems.

    Summary

    This chapter presented the research design and

    methodology of the study. In sum, the details of the

    hypotheses and the scope of the study were addressed. In

    addition to a description of the Rokeach Value Survey. the

    procedures for gathering the data from the prison sample

    were explained. The contrasted groups design was presented

    as the research design and the methods of analyzing the data

    were described as they related to the hypotheses.

  • CHAPTER III

    RESULTS

    The findings of this study are revealed in this

    section, as well as is a description of the method,

    procedure, and composition of the samples. The tested

    hypotheses are also presented with their respective results

    and interpretations. Furthermore, the influences of age,

    race, education, religion, and income on the value systems

    of the inmates are discussed.

    Sub.iects

    Two groups of incarcerated offenders were drawn from

    "kh® inmate population at the Federal Correctional

    Institution (F.C.I.) at Fort Worth. One group consisted of

    all the male and female embezzlers (U = 31) who were

    incarcerated at this institution at the time of the data

    collection (April, 1987). The second group, the embezzlers'

    matched control group, included incarcerated offenders

    convicted of crimes other than embezzlement who were also

    incarcerated at Fort Worth at the time of the data

    collection.

    The method of sampling for this study was the precision

    matching technique (also known as pairwise or one-to-one

    30

  • 31

    matching). These embezzlers were matched by gender, race,

    and age. For example, for every embezzler who is female,

    white, and 25 years old, an inmate with the same

    characteristics, who was convicted of a crime other than

    embezzlement, was included in the matched control group. In

    the case where there was more than one inmate who possessed

    these characteristics, a random selection technique was used

    (for example, coin tossing or the table of random digits).

    For this study, inmates selected for the matched control

    group were those whose crimes least resembled white-collar

    offenses. For the distribution of the types of embezzlement

    these inmates represent, refer to Table 1. The offense

    categories for the matched control group are presented in

    Table 2.

    Table 1

    Composition of Embezzlers at Fort Worth

    Embezzlement Category Codes Number

    Postal (100) x

    Bank (101): False Entries, Misappropriation of Bank Funds

    Other (102): Including Savings and Loan Association

    Benefit Plan (103): Employee Benefit Plan 1

    TOTAL 31

  • 32

    Table 2

    Composition of Non-Embezzler Matched Control Group (Ft. Worth)

    Offense Category Number

    Drug Violations 16

    Alien Smuggling 4

    Fire Arms 3

    Bank Robbery 2

    Contempt 1

    Depriving Civil Rights 1

    Escape/Harbor 1

    Fraud 1

    Government Reservation Violation 1

    Kidnaping 1

    TOTAL 31

    The samples were stratified by gender, race, and age in

    order to create a group of inmates that could be compared to

    Cochrane's (1971) sample of Michigan inmates. Cochrane

    compared his Michigan prison sample to a sample assumed to

    be representative of the general population. Cochrane

    matched his prison sample by gender, race, and age with a

    sample drawn from the National Opinion Research Center

    (NORC) (Rokeach, 1968). The design and method of comparing

    and contrasting these various samples were consistent with

  • 33

    the contrasted groups design (Nachmias & Nachmias, 1981).

    The distribution of the samples from Fort Worth, Michigan,

    and NORC are presented in Table 3.

    Table 3

    Composition of Samples

    *Fort Worth **Michigan ***NORC Age

    Group White Blask Qthgr White Black Other White Black Other

    Males

    18-29 0 0 0 118 87 3 30 16 1

    30-39 6 0 0 47 49 6 17 14 1

    40 + 10 0 0 25 29 1 9 11 1

    Total 16 0 0 190 165 10 56 41 3

    Females

    18-29 8 2 0 29 25 1 29 21 1

    30-39 20 2 0 9 21 1 9 20 0

    40 + 14 0 0 9 9 1 0 10 1

    Total 42 4 0 47 55 3 38 51 2

    *Half of the inmates in the Fort Worth Sample for this study

    are convicted of embezzlement and the other half are convicted

    of miscellaneous crimes.

    **This Michigan category represents Cochrane's (1971) prison

    sample.

    ***This NORC sample was drawn by Cochrane (1971) to serve as a

    matched control group.

  • 34

    The sample of embezzlers included 23 female embezzlers

    and eight male embezzlers. Because this sample consisted of

    mostly females, some possible explanations for this gender

    imbalance were in order. First, females may rely on the

    most easily detectable methods of embezzlement; hence, they

    are caught more often. Male embezzlers, by virtue of their

    position m financial institutions, may have lower

    accountability and can conceal their activities more

    readily. Second, women may be more likely to confess to

    their crimes to avoid a lengthy criminal investigation, as

    informally reported by one female embezzler to this

    investigator. Unfortunately, these explanations are more

    amenable to investigation through interviews and case

    studies, which were beyond the scope of this study.

    Procedure

    The samples of inmates at Fort Worth were tested using

    the Rokeach Value Survey (RVS: Form G) in group settings.

    The inmates participated in this study on a voluntary basis

    and provided informed consent. Out of 62 inmates, only one

    person refused to complete the survey. This person was

    replaced by another inmate who possessed the appropriate

    matched characteristics.

    Data Analysis

    Consistent with the data analyses of Rokeach (1968,

    1973) and Cochrane (1971), the responses to the RVS items

    were tabulated into aggregate medians and composite

  • 35

    rankings. For the purposes of this contrasted groups design

    the medians served as the criteria for comparing and

    contrasting these items among the various samples.

    Two procedures were used to compare and contrast these

    groups. First, the median test was applied to determine the

    statistically significant differences between the embezzlers

    and their matched control group. The median test assessed

    these differences at the .05 level of significance (one-

    tailed test). This statistical analysis also involved the

    Bonferonni technique which was used to reduce the likelihood

    of obtaining significant differences by chance (hence,

    reducing false positives) (Glass & Stanley, 1970).

    Secondly, simple comparisons of the medians between the

    inmate group (N = 62) (embezzlers and their matched control

    group) and the NORC sample (N = 1409) were employed to

    identify the similarities and differences.

    The Testing of the Hypotheses

    As mentioned in the previous chapter, the hypotheses

    were presented in four sets. The first set of hypotheses

    contrasted the male embezzlers with their matched offender

    control group. Secondly, the female embezzlers were

    contrasted with their matched offender control group. These

    two sets of hypotheses employed the median test to appraise

    statistically significant differences. Third, male

    embezzlers and their matched offender control group

    (together as one group) were contrasted with the NORC data

  • 36

    (Rokeach, 1973). Finally, the female embezzlers and their

    matched offender control group (together as one group) were

    contrasted with the NORC data. These last two sets of

    hypotheses were tested according to the simple comparison

    procedure. The following null hypotheses are presented in

    their respective categories indicating whether they have been

    accepted or rejected.

    Mal£ Qpbegslers vsrsug their matched control group.

    The embezzler group will place more importance on the

    following items when contrasted with their matched control

    group:

    H 1: Pleasure

    H 2: Wisdom

    H 3: Self-controlled

    H 4: Capable

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    The embezzler group will place less importance on the

    following items when contrasted with their matched control

    group:

    H 5: A world at peace accept the null hypothesis

    H 6: Equality accept the null hypothesis

    H 7: National security accept the null hypothesis

    H 8: Salvation accept the null hypothesis

    H 9: Honest accept the null hypothesis

  • 37

    Female QmfrezglQrs versus their matched control group.

    The embezzler group will place more importance on the

    following items when contrasted with their matched control

    group:

    H 10: An exciting life accept the null hypothesis

    H 11: A sense of accomp. accept the null hypothesis

    H 12: Freedom accept the null hypothesis

    H 13: Inner harmony accept the null hypothesis

    H 14: Wisdom accept the null hypothesis

    H 15: Broadminded accept the null hypothesis

    H 16: Capable accept the null hypothesis

    H 17: Independent accept the null hypothesis

    H 18: Intellectual accept the null hypothesis

    H 19: Logical accept the null hypothesis

    H 20: Self-controlled accept the null hypothesis

    The embezzler group will place less importance on the

    following items when contrasted with their matched control

    group:

    H 21: A world at peace

    H 22: National security

    H 23: Salvation

    H 24: Helpful

    H 25: Honest

    Fort Worth male inmates versus the NQRC samnlA The

    inmate group will place more importance on the following

    items when contrasted with the NORC sample:

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

  • 38

    H 26

    H 27

    H 28

    H 29

    Pleasure

    Wisdom

    Self-controlled

    Capable

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    The inmate group will place less importance on the following

    items when contrasted with the NORC sample:

    H 30

    H 31

    H 32

    H 33

    H 34

    A world at peace

    Equality

    National security

    Salvation

    Honest

    reject the null hypothesis

    reject the null hypothesis

    reject the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    Fort Worth female inmates versus the NORC sample. The

    inmate group will place more importance on the following

    items when contrasted with the NORC sample:

    H 35

    H 36

    H 37

    H 38

    H 39

    H 40

    H 41

    H 42

    H 43

    H 44

    H 45

    An exciting life

    A sense of accomp.

    Freedom

    Inner harmony

    Wisdom

    Broadminded

    Capable

    Independent

    Intellectual

    Logical

    Self-controlled

    reject the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    reject the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    reject the null hypothesis

    reject the null hypothesis

    reject the null hypothesis

    accept the null hypothesis

  • 39

    The inmate group will place less importance on the following

    items when compared with the NORC sample:

    H 46

    H 47

    H 48

    H 49

    H 50

    A world at peace

    National security

    Salvation

    Helpful

    Honest

    reject the null hypothesis

    reject the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    accept the null hypothesis

    Interpretati nn

    One could argue that comparing responses obtained in 1987

    with responses obtained in 1968 (NORC) may be problematic in

    the case of the value items. For example, in 1968 this nation

    was experiencing the Viet Nam War and civil rights

    demonstrations and so one may assume that citizens today would

    place less importance on some of these items than they did in

    1968. While this assumption may be true, it is imperative to

    keep in mind that Cochrane found statistically significant

    differences between a prison population and the NORC data in

    1971 for these same items ("a world at peace," "equality," and

    "national security"). An additional caution related to

    interpretation was the difference in sample size between the

    inmate group (N = 62) and the NORC sample (N = 1409).

    The hypotheses were created to replicate the research of

    Cochrane (1971) who found statistically significant

    differences between a Michigan prison sample and a matched

    control group drawn from the NORC sample (Rokeach, 1968). The

  • 40

    results of the four sets of hypotheses are presented and

    interpreted in the following section.

    lisle and female embezzlers versus their matched controls

    The research hypotheses in these sections were based on the

    theoretical assumption that the embezzler's value system has

    far more influence on his or her committing a crime than the

    value system of a criminal convicted of a crime other than

    embezzlement (for example, drug or fire arms violations).

    This was reasoned because the embezzlers, unlike other

    offenders, are educated, employable, and middle-class. Hence,

    it was proposed that those values reflecting self-centeredness

    would be endorsed by the embezzlers. These hypotheses were

    assessed by the median test (.05 level of significance, one-

    tailed test). All the null hypotheses in this section were

    accepted suggesting no differences between these groups. That

    is, the embezzlers possess those value systems which are

    characteristic of those inmates convicted of other crimes.

    £ori Wprtfr sals jrmateg versus tii£ NORC sample. As

    displayed in the list of hypotheses, three null hypotheses in

    this category were rejected. "A world at peace," "equality,"

    and "national security" are those items that demonstrated

    differences between the male inmate group as contrasted with

    the NORC data. Although only three of these items yielded

    differences between the groups these three items did fit the

    theoretical framework of Cochrane, who found that members of

    the prison population do not favor those values which do not

  • 41

    have immediate and personal relevance. Concerning the item

    equality, it must be noted that this item was favored more

    by blacks than by whites (Cochrane, 1971).

    Fort Worth female inmates versus the NORC sample. It was

    evident in the list of the hypotheses that the following items

    met the terms for rejecting the null hypothesis: "an exciting

    life," "wisdom," "independent," "intellectual," "logical," "a

    world at peace," and "national security." These findings

    partially supported the results of Cochrane (1971). Keep in

    mind that these hypotheses take into consideration the basic

    differences between male and female ranking of values

    (Rokeach, 1968, 1973). Attending to the differences between

    the inmates and the general population, female prisoners

    constitute a much more rare group as compared to their male

    counterparts (Cochrane, 1971). Cochrane noted that these

    females must violate the social norms far more often than

    males in order to be prosecuted and incarcerated.

    Cochrane revealed that female convicts differ from their

    matched control group on 14 of 36 values. Cochrane observed

    that the value systems of the female prisoners resembled -

    "masculine" preferences more than their female control group.

    These results partially supported this notion by demonstrating

    that the female inmates at Fort Worth endorsed such

    masculine value systems. In this investigation female

    inmates preferred "an exciting life," "independent,"

    "intellectual," and logical."

  • 42

    Further Analyses

    Consistent with the contrasted groups design, additional

    conclusions about these various groups were drawn by situating

    their rankings of items side by side and interpreting the

    results horizontally. Tables 4, 5, 6, and 7 illustrate the

    contrasts and comparisons of the Fort Worth inmates with

    Cochrane's (1971) Michigan sample as well as with the matched

    control group he had drawn from the NORC data set. These

    tables were stratified according to gender, offense category

    (only for Fort Worth: embezzlers versus the matched offender

    control group), and type of value (terminal and instrumental).

    While these lengthy tables can overwhelm the reader with data,

    these groups can be contrasted more readily if the most

    preferred values were listed independently to accentuate the

    differences. Six contrast groups, stratified by gender and

    type of value (terminal and instrumental), have been

    identified in order to assess their respective value systems.

    The contrasted groups for this design were: (1) Embezzlers

    (males and females as separate groups), (2) Matched Offender

    Control Groups (males and females as separate groups), (3)

    Fort Worth inmate group (embezzlers and matched offender

    control group: males and females as separate groups), (4)

    Michigan prison group (Cochrane, 1971), (5) Cochrane's (1971)

    matched control group drawn from the NORC, and (6) the NORC

    sample (Rokeach, 1968, 1973) (see Appendix C).

  • 43

    One advantage of the contrasted groups design was its

    ability to identify group differences and similarities. Of

    all the terminal values in each of the male contrasted groups,

    only one value was listed as among the five most preferred

    values: "family security." Therefore, this value was

    important to the prison groups and the general population.

    Freedom was listed in three of the four prison groups

    and the NORC sample, which demonstrated its importance to both

    populations. Cochrane (1971) mentioned in his study that

    "freedom" was valued more by the matched control group than by

    the prison group. Cochrane explained that, although prisoners

    were aware of their lack of freedom, this lower ranking may be

    an attempt to reduce the dissonance aroused by imprisonment"

    (1971, p. 76). Among other common values, "self-respect" and

    a comfortable life" were also endorsed by the prison groups

    and the general population sample.

    One value that was strongly preferred by only the general

    population sample was "a world at peace." The lack of this

    preference clearly showed that the prison group placed a low

    priority on large-scale social situations. This finding

    supported Cochrane who stated: "In general it appears that

    those values which do not have immediate and personal

    relevance are considered less important by the prison group

    than the control group" (1971, p. 76).

  • 44

    Table 4

    Value Rankings of Male Prisoners and Matched Control Groups

    Survey Item Emb.

    H = 8

    Med/Rnk

    Terminal Values

    Non-Emb. Emb. & Non ^Michigan ++NORC

    8

    Med/Rnk

    16

    Med/Rnk

    363

    Med/Rnk

    100

    A comfortable life 6 .5/5 4 .0/3 5.5/5 7 .43/6 7 .67/6 An exciting life 11 .0/13 10 .0/10 11.0/12 13 .13/15 13 .86/17 A sense of accomp. 7 .0/6 6 .0/5 6.0/6 7 .68/7 9 .00/9

    A world of peace 10 .5/12 14 .0/15.5 12.5/13 *9 .00/9 *4 .27/3

    A world of beauty 14 .0/14.5 13 .0/13 14.0/15.5 14 .51/17 13 .77/16 Equality 16 .0/18 14 .0/15.5 14.5/17 *8 .52/8 *6 .50/4 Family security 1 .0/1 4 .5/4 2.0/1 4 .61/1 3 .61/1 Freedom 7 .5/8 2 • 5/1 4.0/2.5 5 .33/3 3 .92/2 Happiness (Form D)

    *6, .49/4 Happiness (Form D)

    *6, .49/4 *8 . 10/7 Health (Form G) 5 .0/4 3, .5/2 4.0/2.5

    Inner harmony 10, .0/11 10. .5/11 10.5/11 9. 88/12 11. 30/12 Mature love 7. .5/8 8. ,0/7 7.5/7 9. 72/11 11. ,60/13 National security 15. 5/17 15. 5/18 15.5/18 *13. 88/16 *11. 63/14 Pleasure 14. 0/14.5 13. 5/14 14.0/15.5 *12. 92/14 *14. 40/18 Salvation 3. 5/3 15. 5/17 8.5/9 *15. 05/18 *11. 21/11 Self-respect 3. 0/2 7. 5/6 5.0/4 6. 73/5 7. 64/5 Social recogition 14. 5/16 12. 5/12 13.0/14 12. 03/13 13. 30/15 True friendship 7. 5/8 9. 0/8.5 8.0/8 9. 52/10 10. 00/10 Wisdom 9. 5/10 9. 0/8.5 9.0/10 *8. 17/2 *5. 16/8

    Indicates a statistically significant difference between groups (median

    test). These are the results of Cochrane's (1974) study.

    +This Michigan category represents Cochrane1s (1971) prison sample.

    +*This NORC sample was drawn by Cochrane (1971) to serve as a matched

    control group.

    Med: refers to composite median.

    Rnk: refers to composite rank order by ranking the medians.

  • 45

    Table 5

    Value Rankings of Male Prisoners and Matched Control Groups

    Survey Item

    Instrumental Values

    Emb. Non-Emb. Emb. & Hon +Michigan

    H = 8 8 16 363

    Med/Rnk Med/Rnk Med/Rnk Med/Rnk

    ++ NORC

    100

    Med/Rnk

    Ambitious 5 .0/2.5 4 .5/2 5 .0/2.5 6 .11/1 6 .00/2

    Broadminded 8 .0/6 10 .5/11 8 .5/6.5 7 .16/4 7 .79/5

    Capable 9 .5/10 11 .0/14 10 .5/11.5 *8 . 11/7 *9 .72/11

    Cheerful (Form D) 11 .97/15 10 .50/12

    Loyal (Form G) 9 .0/8 11 .0/14 9 .5/10

    Clean 9 .5/10 5 .5/4 8 .5/6.5 7 .69/6 9 .36/8

    Courageous 12, .5/14.5 10 .0/9.5 11 .5/14 8. .68/9 7 .00/4

    Forgiving 5, .0/2.5 8 .5/6 7 .0/4 10 .10/11 8 .07/6

    Helpful 11. .0/13 11, .0/14 11, .0/13 10, .32/12 9 .50/9

    Honest 3, .0/1 4 .0/1 4 .0/1 *6, .16/2 *3 .77/1

    Imaginative 12. .5/14.5 11, .0/14 12. . 5/17 14. .71/18 13, .50/17

    Independent 10. .0/12 11, .0/14 10. .5/11.5 8. .67/8 8, .50/7

    Intellectual 13. ,0/16 8. .5/6 12. 0/15.5 10. .09/10 12. ,30/15

    Logical 14. .0/17 10. ,0/9.5 12. 0/15.5 13. ,35/16 13. .37/16

    Loving 6. 0/4 8. 5/6 7. 5/5 11. 13/14 11. 50/14

    Obedient 15. 5/18 16. 0/18 15. 5/18 13. 42/17 13. 88/18

    Polite 8. 5/7 11. 5/17 9. 0/8.5 10. 63/13 11. 00/13

    Responsible 6. 5/5 5. 0/3 5. 0/2.5 7. 23/5 6. 79/3

    Self-controlled 9. 5/10 9. 0/8 9. 0/8.5 *6. 90/3 *9. 68/10

    Indicates a statistically significant difference between groups (median

    test). These are the results of Cochrane's (1974) study.

    +This Michigan category represents Cochrane's (1971) prison sample.

    ++This NORC sample was drawn by Cochrane (1971) to serve as a matched

    control group.

    Med: refers to composite median.

    Rnk: refers to composite rank order by ranking the medians.

  • 46

    Table 6

    Value Rankings of Female Prisoners and Matched Control Groups

    Terminal Values

    Survey Item Emb. Non-Emb. Emb. & Hon "^Michigan ++NORC

    & = 23 23 46 98 91

    Med/Rnk Med/Rnk Med/Rnk Med/Rnk Med/Rnk

    A comfortable life 12.0/13 10.0/11.5 11 .0/12.5 8 .50/9 7 .69/7

    An exciting life 14.0/15.5 • • 16.0/18 14 .0/15 *14 .79/18 *16 .43/18

    A sense of accomp. 9.0/8.5 9.0/9 9 .0/8.5 *8 . 10/8 *9 .94/9 A world of peace 10.0/10.5 9.0/9 9 .0/8.5 *7 .36/7 *3 .36/1

    A world of beauty 13.0/14 14.0/15.5 14 .0/15 13 .38/17 14 .45/16 Equality 11.0/12 11.0/13 11 .0/12.5 6 .17/4 6 .25/4 Family security 3.0/1 4.0/2 3 • 5/1 4 .50/2 3 .75/2 Freedom 6.0/5.5 6.0/4.5 6 .0/4.5 *3 .32/1 *4 .88/3 Happiness (Form D)

    .30/8 .86/5 Happiness (Form D)

    7 .30/8 6 . .86/5

    Health (Form G) 5.0/3.5 3.0/1 4 .5/3

    Inner harmony 8.0/7 9.0/9 8. .5/7 *8, .83/10 *10. 71/12

    Mature love 9.0/8.5 10.0/11.5 10. .0/10.5 10. ,50/11 11. 42/13

    National security 15.0/17 14.0/15.5 14. 5/17 *13. 50/14 11. 64/14 Pleasure 14.0/15.5 14.0/15.5 14. 0/15 13. 80/16 14. 69/17 Salvation 5.0/3.5 7.0/6 6 . 5/6 13. 72/15 10. 60/11 Self-respect 4.0/2 5.0/3 4. 0/2 6. 59/5 7. 39/6

    Social recognition 16.0.18 14.0/15.5 15. 0/18 13. 36/13 14. 00/15 True friendship 10.0/10.5 8.0/7 10. 0/10.5 11. 25/12 10. 10/10 Wisdom 6.0/5.5 6.0/4.5 6. 0/4.5 *5. 00/3 *9. 13/8

    "indicates a statistically significant difference between groups (median

    test). These are the results of Cochrane's (1974) study.

    +This Michigan category represents Cochrane's (1971) p r i 8 e n s a m p l e .

    ++This NORC sample was drawn by Cochrane (1971) to serve as a matched

    control group.

    Med: refers to composite median.

    Rnk: refers to composite rank order by ranking the medians.

  • 47

    Table 7

    Value Rankings of Female Prisoners and Matched Control Groups

    Survey Item

    Instrumental Values

    Non-Emb FmVi A. +-K

    fi = 23

    Med/Rnk Med/Rnk Med/Rnk Med/Rnk Med/Rnk

    Ambitious 10.0/9.5 7.0/5 8.5/6 6.33/4 6.33/3 Broadminded 11.0/12 9.0/8.5 11.0/11 .5 *5.70/1 *8.94/9 Capable 12.0/14.5 13.0/15 12.0/14 .5 *9.17/9 *10.38/13 Cheerful (Form D)

    11.50/15 Cheerful (Form D)

    11.50/15 11.58/14 Loyal (Form G) 8.0/5 9.0/8.5 9.0/8

    Clean 9.0/7 9.0/8.5 9.0/8 5.83/2 5.27/2 Courageous 12.0/14.5 12.0/13.5 12.0/14 .5 9.21/10 8.80/8 Forgiving 7.0/4 6.0/3.5 6.5/4 8.17/7 7.67/5 Helpful 9.0/7 9.0/8.5 9.0/8 *10.64/12 *8.60/7 Honest 2.0/1 3.0/1 2.5/1 *6.00/3 *3.85/1 Imaginative 15.0/17.5 14.0/16.5 14.5/17 15.90/18 16.09/18 Independent 9.0/7 6.0/3.5 7.0/5 *8.50/8 *10.33/12 Intellectual 11.0/12 14.0/16.5 11.0/11. 5 *11.00/13 *12.86/16 Logical 14.0/16 10.0/11 12.0/14. 5 *12.67/16 14.82/17 Loving 4.0/2 9.0/8.5 5.5/3 11.00/14 8.40/6 Obedient 15.0/17.5 15.0/18 15.0/18 12.92/17 12.13/15 Polite 11.0/12 12.0/13.5 12.0/14. 5 10.32/11 9.42/11 Responsible 5.0/3 5.0/2 5.0/2 6.64/5 6.58/4 Self-controlled 10.0/9.5 11.0/12 10.0/10 *7.27/6 *9.40/10

    •indict.. . s i g n i f i c a n t d i « . , . „ c . ( m . d i w

    test). These are the results of Cochrane's (1974) study.

    +This Michigan category represents Cochrane's (1971) prison sample.

    This NORC sample vas drawn by Cochrane (1971) to serve as a matched

    control group.

    Med: refers to composite median.

    Rnk: refers to composite rank order by ranking the medians.

  • 48

    Taking into consideration the alteration in RVS forms

    (D versus G), it was assumed that this instrument was

    improved by replacing "happiness" (a global and somewhat

    vague item) with "health" (a more specific term). In this

    design, "health" was strongly endorsed by those groups

    tested with Form G, and "happiness" was strongly endorsed by

    those groups tested with Form D.

    The terminal values preferred by the female groups were

    similar to those of their male counterparts. The only

    salient difference between the male and female groups was

    that "wisdom" was strongly preferred by all of the female

    prison groups while this item was strongly valued by only

    one male prison group. Consistent with the male value

    systems, the female prison groups also discounted "a world

    at peace."

    Among the instrumental value systems of the male

    groups, there were three common values listed by all of the

    groups: "honest," "ambitious" and "responsible." These

    values were preferred by both the prison groups as well as

    those members of the general population. Cochrane (1971)

    believed that the endorsement of "honest" could be

    indicative of the conflict between the prisoners' values and

    behavior, which could be used as additional material for

    dissonance reduction studies.

    It was also worth noting that the items "forgiving" and

    "loving" were preferred by three of the four prison groups

  • 49

    but not by the general population sample. This result may

    be interpreted as a function of the punishment and

    incarceration of the prison groups. Finally, "courageous"

    was favored by the general population sample, but not

    preferred by any of the prison groups. The RVS defined

    "courageous" as standing up for your beliefs, and it

    appeared that the prison groups did not consider this value

    to be important.

    The female instrumental values also appeared to be

    similar to those of the male groups which suggested that

    different patterns between genders did not exist. However,

    one exception was the item "independent," which was endorsed

    by two of the prison groups but not by either of the general

    population samples. It could be hypothesized that an

    extreme sense of independence (self-reliant and self-

    sufficient) may leave a woman vulnerable to criminal, or

    unconventional, paths toward achieving self-reliance.

    The results in this study partially supported some of

    Cochrane's findings that inmate value systems differ from

    the general population. He enhanced his interpretations by

    noting that male and female prison groups "resemble each

    other much more closely than they resemble their respective

    controls' value system" (1971, p. 78). (For additional

    comparisons based on this assumption, refer to Tables 8 and

    9, which illustrate the terminal and instrumental values as

    reported by males and females.)

  • 50

    Survey Item

    Table 8

    Fort Worth and NORC Medians for Terminal Values

    Comparison of Gender

    Fort Worth NORC

    Median Median

    Males Females Males Females (N = 16) (N = 46) (N = 665) (N = 744)

    A comfortable life 5.5 11 .0 7 .77 10.02

    An exciting life 11.0 14 .0* 14 .62 15.57*

    A sense of accomp. 6.0 9 .0 8 .29 9.40

    A world at peace 12.5* 9 . 0* 3 .75* 3.00*

    A world of beauty 14.0 14 .0 13 .61 13.51

    Equality 14.5* 11 .0 8 .87* 8.29

    Family security 2.0 3 .5 3 .86 3.78

    Freedo 4.0 6 .0 4, .91 6.05

    Happiness (Form D) — — 7. .94 7.34

    Health (Form G) 4.0 4. .5 — - -

    Inner harmony 10.5 8. .5 11. .08 9.83

    Mature love 7.5 10. 0 12. 57 12.32

    National security 15.5* 14. 5* 9. 21* 9.81*

    Pleasure 14.0 14. 0 14. 14 14.97

    Salvation 8.5 6. 5 9. 88 7.33

    Self-respect 5.0 4. 0 8. 16 7.40

    Social recognition 13.0 15. 0 13. 79 15.01

    True friendship 8.0 10. 0 9. 63 9.06

    Wisdom 9.0 6. 0* 8. 49 7.71*

    •Supports the hypothesized difference between groups.

  • 51

    Survey Item

    Table 9

    Fort Worth and NORC Medians for Instrumental Values

    Comparison of Gender

    Fort Worth NORC

    Median Median

    Males Females Males Females (M = 16) (N = 46) (fi = 665) (N = 744)

    Ambitious 5.0 8.5 5.61 7.33

    Broadminded 8.5 11.0 7.20 7.64

    Capable 10.5 12.0 8.86 10.10

    Cheerful (Form D) 10.41 9.43

    Loyal (Form G) 9.5 9.0

    Clean 8.5 9.0 9.43 8.13

    Courageous 11.5 12.0 7.49 8.06

    Forgiving 7.0 6.5 8.23 6.43

    Helpful 11.0 9.0 8.35 8.07

    Honest 4.0 2.5 3.43 3.21

    Imaginative 12.5 14.5 14.28 16.10

    Independent 10.5 7.0* 10.17 10.72*

    Intellectual 12.0 11.0* 12.77 13.22*

    Logical 12.0 12.0* 13.51 14.65*

    Loving 7.5 5.5 10.90 8.64

    Obedient 15.5 15.0 13.51 13.08

    Polite 9.0 12.0 10.85 10.71

    Responsible 5.0 5.0 6.58 6.82

    Self-controlled 9.0 10.0 9.65 9.55

  • 52

    Although Cochrane revealed more differences between the

    Michigan prison groups and their matched control groups than

    were obtained in this study, this may be attributed to the

    qualitative differences between the prison population at

    Fort Worth as compared to other prisons. For example, the

    inmates at Fort Worth are generally low risk, nonviolent,

    and minimum security. However, while not all of the inmates

    at Fort Worth are middle-class, white-collar, and educated,

    many of the embezzlers did fit this description.

    Considering these characteristics, inmates at Fort Worth

    (particularly the embezzlers) resemble the general

    population far more than they resemble other prison

    populations.

    Additional Analyses

    Because gender was only one of the demographic

    variables used to stratify these samples, it was appropriate

    to introduce those meaningful findings based on the other

    characteristics of these samples: age, race, education,

    religion, and income. The selection of these variables was

    based on those listed by Rokeach (1968, 1973). This section

    concentrates on the salient features of the additional

    analyses.

    Age. Rokeach (1973) presented a comprehensive analysis

    of values as they related to a person's individual

    development. He proposed a series of developmental patterns

    in which a person's chronological age corresponded with a

  • 53

    particular value system. This author included age as a

    variable to take into account the influence a person's age

    has on his or her value system.

    Four major age categories (and their respective N) were

    delineated for the Fort Worth samples: 20-29 (N = 10), 30-

    39 (N = 28), 40-49 (fi = 12), and 50-59 (JtJ = 12) (see Table

    3). Responses to the RVS items for these age categories

    were reported in Appendix D as they compared and contrasted

    with the data from the NORC sample. As was the case with

    many of these findings, one must exercise caution when

    comparing the sample from Fort Worth with the NORC sample

    because of the difference in size.

    Consistent with the previous findings, when the Fort

    Worth group and the NORC sample were divided into age

    groups, drastic differences in the ranking of the items "a

    world at peace," "equality" and "national security"

    appeared. Each of the inmate age categories clearly placed

    a lower emphasis on these items than the NORC sample, which

    added additional support to the conclusion that these

    inmates were more self-centered and less concerned for

    society.

    The item "self-respect" also differentiated between

    these age groups. The inmates placed more importance on

    this item, which may be indicative of their current struggle

    with dissonance. Once a person has endured the process of

    criminalization through the criminal justice system he or

  • 54

    she may seek self-respect because nobody else appears to

    respect them. This experience with the criminal justice

    system may also explain the differences on the items

    broadminded (defined as open—minded) and "courageous"

    (defined as standing up for one's beliefs) in which the Fort

    Worth group placed much less importance than the NORC

    sample. Upon experiencing punishment and degradation from

    the system, one may not be expected to retain an open mind

    or feel that it pays to stand up for one's beliefs.

    Considering this, one may also understand why the Fort Worth

    group favored "forgiving" more than the NORC sample.

    Finally, the item "loving" (defined as affectionate and

    tender) was ranked higher, which may be a function of

    gender, particularly in a co-correctional institution where

    members of the opposite sex are constant reminders of

    affection and "tenderness.

    Race. In one of his studies, Rokeach (1973) found that

    race was an important factor in determining a person's value

    system. For example, a person who has experienced racial

    discrimination may form a value system different from that

    of a person who has not endured such degradation.

    Taking this into consideration, race was selected as an

    additional variable to analyze value systems. For the

    purposes of this investigation, blacks were not sufficiently

    represented in this investigation. This is not to say that

    blacks were not sufficiently represented at F.C.I., Fort

  • 55

    Worth, but the target population was the embezzler, and

    there were only two black embezzlers (both female) in this

    sample (four blacks total, including their matched controls,

    see Table 3). Therefore, no inferences about the responses

  • 56

    and comparisons with the NORC sample in this section focused

    on these two educational categories.

    Table 10

    Frequency Distributions of Education

    Years of Education

    Frequency Percent

    4 3 4.8

    5 2 3.2

    7 2 3.2

    8 3 4.8

    9 1 1.6

    10 1 1.6

    11 1 1.6

    12 23 37.1

    13 7 11.3

    14 7 11.3

    15 6 9.7

    16 4 6.5

    17 1 1.6

    18 1 1.6

    Total 62 100.0

    Note. M - 12.00; Median = 12.00; SD = 3.178

  • 57

    As expected, both of these categories ranked the

    following items lower than the NORC sample: "a world at

    peace, "equality," "national security," "broadminded,"

    courageous," and "loving." The inmates in these categories

    Placed more importance on "salvation," and "self-respect."

    The differences in the item "self-controlled" (defined as

    restrained and self-disciplined) appeared to be masked by

    other variables until the sample was stratified by

    education. In this case "self-controlled" was ranked lower

    by those with "some college" than by the same educational

    category in the NORC sample.

    Religion. Religion was also included as an additional

    variable because of the basic relationship between religion

    and values. Religion is assumed to be comprised of an

    identifiable set of values and a person's religious

    background is assumed to influence his or her value system.

    Similar to the variable education, the inmates in this

    sample represented two major categories: Catholic (N = 19)

    and Baptist (N = 19) (see Table 11 and Appendix G).

    Considering this breakdown, the interpretations were limited

    to these religious groups as they compared to the same

    religions reported by the NORC.

    The Catholics and Baptists in the Fort Worth sample

    also placed a low priority on "a world at peace,"

    "equality," "national security," "ambitious," "broadminded,"

    "courageous," and "obedient." On the other hand, both of

  • 58

    Table 11

    Frequency Distributions of Religion

    Value Label Frequency Percent

    Catholic 19 30.6

    Episcopalian 2 3.2

    Lutheran 2 3.2

    Presbyterian 3 4.8

    Methodist 6 9.7

    Baptist 19 30.6

    None 1 1.6

    Other 10 16.1

    these religious groups placed more importance on "mature

    love," "self-respect," "independence," and "logical." The

    Baptists, as opposed to the Catholics, assigned a higher

    value on "loving," and "responsible," and both groups ranked

    "wisdom" higher than their respective NORC categories.

    An additional finding from this analysis was that the

    Catholics appeared to fit Cochrane's model that "pleasure"

    is valued more by Michigan prisoners than their matched

    control group. In this investigation, the incarcerated

    Catholics placed a higher premium on "pleasure" than those

    Catholics in the NORC sample.

  • 59

    Admittedly there was a major drawback by stratifying a

    prison sample by religion. The most obvious problem was

    that it was difficult to infer their degree of religiosity

    from their stated preference. One may assume that

    criminality is an indication of a low degree of religiosity.

    However, imprisonment often rekindles a concern for religion

    as seen by inmates who embrace Bibles and attend services.

    Consequently, it was difficult to determine if such inmates

    were sincerely religious or trying to project a "reformed-

    character (or both).

    Income. An additional variable used to stratify the

    inmate sample at Fort Worth was income. Rokeach (1973)

    described the basic differences between individuals of

    different income brackets. He argued that, when appraising

    a person's value system, income should be one of the

    variables taken into consideration. As is the case with the

    other variables, income may influence life chances and

    experiences which, in turn, affect the formation of a value

    system.

    Table 12 and Appendix H list the 16 income categories

    and their respective frequencies. In order to avoid any

    negative (or threatened) responses to this income variable

    by the inmates, broad categories were used instead of exact

    income disclosures. It was assumed that income was a

    sensitive issue with these inmates, particularly the

    embezzlers. The first six income categories were included

  • 60

    so that these responses could be compared and contrasted

    with those NORC responses published by Rokeach (1973).

    Table 12

    Frequency Distributions for Income

    Value Label Frequency Percent

    0-2,000 4 6.5

    2,000-3,999 1 1.6

    4,000-5,999 3 4.8

    8,000-9,999 1 1.6

    10-15K 12 19.4

    16-20K 6 9.7

    21-25K 10 16.1

    26-30K 12 19.4

    31-40K 4 6.5

    41-50K 3 4.8

    51-60K 1 1.6

    61-70K 1 1.6

    91K + 3 4.8

    Total 61 100.0

    Note. M = 7.672; Median = 8.000; SD = 3.295.

    A serious limitation in usin