MENTAL HEALTH OF THE UNEMPLOYED:
AN ANALYSIS OF A CETA PROGRAM

By

JOHN ANDERSON SANFORD

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
1981

This study is dedicated to the memory of

RICHARD H, JOHNSON

His challenges to the old way;
His joy in learning new things and

thinking new thoughts;
His delight in living
Touched all who knew him.

ACKNOWLEDGEMENTS

Many people have contributed time, energy, and support during the
course of this project. First, I would like to thank Harold Riker,
Rod McDavis, and James Wattenbarger for their unrelenting encouragement.
Regardless of how long each stage of this study required, they never
lost faith that the effort would reach a satisfactory conclusion.
Second, I would like to express my appreciation for the cooperation of
the staff and participants of the Northeast Florida Employment and
Training Consortium. Their tolerance of repeated testing sessions
and active participation in the collection of data made this study
possible. I would also like to acknowledge the staff and participants
of the Alachua County Comprehensive Employment and Training Program,
past and present, whose very existence provided me with the idea for
this effort. I would particularly like to recognized Jess Brannen from
the program for his guidance and encouragement in understanding the CETA
experience.

I would like to thank my parents, Richard and Jo Ann. It was from
these two students of life that I first learned the importance of
learning and clear thought in a happy life.

Finally, and most importantly, I would like to express ray gratitude
to my wife, Robin Runyan. She had the thankless task of supporting the
creation of this study while trying not to interfere with her pursuit
of her own education. During the smooth and the rough, she hung in there,
Thank you.

iii

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS iii

LIST OF TABLES vi

ABSTRACT viii

CHAPTER

I STATEMENT OF THE PROBLEM 1

Need for the Study 1

Purpose 5

Research Questions 5

Value of the Study 6

Rationale 6

Definition of Terms 11

II REVIEW OF LITERATURE 14

Unemployment 14

Mental Health 46

Summary 75

III PROCEDURES 77

Overview 77

Hypotheses 78

Subjects 80

Research Design 85

Data Collection 89

Instrumentation 91

iv

Analysis of Data 99

Limitations of the Study IO3

IV RESULTS 104

Evaluation of Comparison Groups lOA

Evaluation of Hypotheses 110

Comparison of CETA Subjects with Subjects From

Selected Validation Studies 138

Summary 1^1

V DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS ... 143

Overview I44

Discussion I45

Conclusions I57

Mental Health of the CETA Subjects: A Composite

Picture I50

Recommendations for Counseling 163

Recommendations for Further Research .... 166

Summary Igg

APPENDICES

A CETA ELIGIBILITY CRITERIA 169

B CETA PROGRAM SIGNIFICANT SEGMENTS 171

C TEST PACKAGE INTRODUCTION AND PERSONAL DATA

QUESTIONNAIRE 172

D CETA APPLICATION I75

E INFORMED CONSENT FORM I77

BIBLIOGRAPHY jyg

BIOGRAPHICAL SKETCH 201

LIST OF TABLES
Table Page

1 Cell Size of Comparison Groups 81

2 Recoded Values of Demographic Variables of Age and
Education 102

3 Comparison of Intake Group Subjects and Training Group
Subjects on 12 Selected Demographic Variables: Matched

and Unmatched for Education 106

4 Comparison of Unmatched and Matched Groups on the
Independent Variables and on the Dependent Mental

Health Variables 107

5 Comparison of Adult Work Experience Subjects and
Public Service Employment Subjects on the 12 Demo-
graphic Variables 109

6 Comparison of Intake (Group 1) Subjects and Training
(Group 2) Subjects on the Personal Orientation
Inventory Ill

7 Comparison of Intake (Group 1) Subjects and Training
(Group 2) Subjects on the Bern Sex-Role Inventory . . 113

8 Comparison of Program 1 (Adult Work Experience)
Subjects and Program 2 (Public Service Employment)
Subjects on the Personal Orientation Inventory . . . 115

9 Comparison of Program 1 (Adult Work Experience)
Subjects and Program 2 (Public Service Employment)
Subjects on the Bern Sex-Role Inventory ... 117

10 Distribution of Bem Sex-Role Inventory Sex-Type
Classifications by the Demographic Variable of Sex
for the Combined Group, Group 2 (Training), and

Program 2 (Public Service Employment) Subjects . . . 132

11 Distribution of Bem Sex-Role Inventory Sex-Type
Classifications by the Demographic Variable of Race

for Group 1 (Intake) Subjects 133

vx

12 Comparison of CETA Subjects with Normal Adult Sub-
jects Reported by Shostrom (1974) on the Personal
Orientation Inventory 139

13 Comparison of CETA Subjects with Two Samples of
College Students Reported by Rotter (1966) on the
Internal-External Scale 139

14 Comparison of CETA Subjects with Two Samples of
Stanford University Students Reported by Bern (1974,

1977) on the Bern Sex-Role Inventory 140

vii

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

MENTAL HEALTH OF THE UNEMPLOYED:
AN ANALYSIS OF A CETA PROGRAM

BY

John Anderson Sanford

March, 1981

Chairman: Dr. Harold C. Riker

Major Department: Counselor Education

The purpose of this study was to examine the mental health charac-
teristics of unemployed persons and the impact of participation in
employment training activities funded under the Comprehensive Employment
and Training Act of 1978 (CETA) on those characteristics. This study
endeavored to obtain mental health information not currently available
from previous research for use in the development of counseling and
training strategies used with unemployed persons.

The subjects for this study were 191 participants (61 males, 130
females) in a north Florida CETA program during April and May, 1980.
All subjects participated in either the Title II-B Adult Work Experience
(AWE) program (N=69) or the Titles II-D and VI Public Service Employment
(PSE) program (N=122). Each subject had either just begun training
(Intake Group, N=96) or had just completed four months of training
(Training Group, N=95) .

viii

The dependent mental health variables were self-actualization,
psychological androgyny, and locus of control. These variables were
measured using the Personal Orientation Inventory (POI) , Bem Sex-Role
Inventory (BSRI) , and Internal-External Scale (lES) , respectively.
The test package was pre-recorded and presented to the subjects in a
written and recorded format to control for the effects of limited reading
skills. The Intake and Training groups were matched on 12 independent
demographic variables to establish between group comparability.

Of seven null hypotheses tested four were rejected as a result of
finding significant differences between groups on the dependent mental
health measures. The Training Group subjects perceived human nature in
a more positive manner and perceived themselves as less feminine than
Intake Group subjects. AWE subjects were more self-actualizing than PSE
subjects in the Awareness dimension of the POI. Subjects who were older,
unmarried, with 12 or more years of education, or affiliated with liberal
religious denominations were more self-actualizing and more internally
oriented than subjects who were younger, married, less education, or
affiliated with fundamentalist denominations. Non-white subjects were
more external in locus of control than white subjects.

An examination of differences between AWE and PSE subjects on the
demographic characteristics revealed the AWE subjects more likely to be
female, older, unmarried, and unemployed longer, with less education and
more children than PSE subjects. It was concluded that the AWE program
and PSE program serve different elements of the unemployed population.

Significant differences were found between the mental health charac-
teristics of CETA subjects in this study and subjects reported in
selected validation studies of the dependent mental health measures.

ix

With the exception of the POI Self-regard scale, the CETA subjects were
less self-actualizing than a normal adult sample and were more externally
oriented than two college samples. In contrast, the CETA subjects scored
higher on the BSRI scales and had a higher percentage of androgynous
subjects than two college samples. It was concluded that the CETA sub-
jects experienced difficulty in objectively completing sophisticated
self-evaluative tasks.

These findings suggest three important implications for counselors.
Unemployed persons may be more responsive to directive or structured
counseling and training strategies as opposed to introspective approaches.
The findings also suggest that unemployed persons have difficulty per-
ceiving themselves as needing "help" or identifying problems areas,
implying that the counselor may wish to re-examine how his/her role is
presented to the unemployed client. Finally, the utility of the demo-
graphic characteristics in predicting willingness and ability to engage
in mental health growth related activities was discussed. Additionally,
recommendations were made for further research.

CHAPTER I
STATEMENT OF THE PROBLEM
Need for the Study

Work and mental health have been repeatedly related (Maslow, 1970a;
Tiffany, D., Cowan, & Tiffany, P., 1970; Work In America, 1973).
Extended unemployment has been strongly associated with increased evi-
dence of mental illness (Cassel, 1966; Work In America, 1973). Tiffany,
D., et al. (1970) found that work inhibition, one of the causes of long-
term unemployment, is directly related to poor mental health. Maslow
(1970a), in describing self-actualization, identified a direct link
between work and mental health. Maslow found that one of the identi-
fying characteristics of self-actualizing persons was a desire to have
a line of work external to their own lives without regard to the eco-
nomic benefits of working. Furthermore, food, water, shelter, and
security, the basic needs which serve as a foundation for mental health
(Maslow, 1970a), have been shown to be among the basic reasons people
work (Kaplan & Tausky, 1972; Morse & Weiss, 1955; Weiss & Kahn, 1960).

Work has long been viewed as basic to the smooth functioning of the
national socio-economic system (Lebergott, 1964). Most members of
society are expected to work; with work defined as a means of making
one's contribution to society, as a source of identity, and as a means
of supporting one's self and family (Super, 1976; Wilensky, 1966; Work
In America, 1973) .

Prior to the depression in the 1930' s and World War II, national
interest in work was primarily focused on vocational choice rather than
unemployment. Related federal legislation, the Smith-Hughes Act of 1917,
was limited to vocational education. Efforts to help the unemployed
occurred largely through agencies supported by private donations (State
of Florida Office of Manpower Planning, 1978).

Beginning in the depression and continuing after World War II,
national interest in assisting those without work increased rapidly
(Ginzberg, 1968; Levitan, 1967). This interest was reflected in the
passage of such legislation as the Wagner-Peyser Act of 1932, the Social
Security Act of 1935, the Employment Act of 1946, the Manpower Develop-
ment and Training Act of 1962 (as amended in 1963, 1965, 1966, and
1968), and the Comprehensive Employment and Training Acts of 1973 and
1978.

These legislative actions established programs designed to insure
full employment, help the unemployed, and meet the nation's manpower
needs. The goal of these programs was the development of a manpower pool
of adequate size and skill level through the provision of educational
and financial assistance to persons experiencing difficulty in finding
work. The Manpower Development and Training Act and the Comprehensive
Employment and Training Acts in particular were designed to provide
training, counseling, placement, and supportive services to the long-
term unemployed (State of Florida Office of Manpower Planning, 1978).

Interest in assisting the unemployed was stimulated by several
factors. Some, such as how to respond to the technological threat
represented by the Russian Sputnik, were of international origin.
Others, however, were of intranational origin. These included economic

3

demands for an adequate manpower pool, social demands that employment be
available for returning servicemen, and altruistic demands to help unem-
ployed persons improve their lives (Ginzberg, 1968; Levitan, 1964;
Levitan, 1967). Underlying these social factors was the belief that it
was better for the individual to be working than to be unemployed (Cassel,
1966; Work In America, 1973).

During the period following the depression and World War II, inter-
est was also increasing among psychologists in the development of the
healthy personality. Several theoretical descriptions of how and why
people develop a healthy, well-rounded, well-adjusted personality have
been presented (Jourard, 1963; Maslow, 1970a; Rogers, C, 1951; Rogers, C,
& Dymond, 1954). These descriptions represented a shift from earlier,
problem-oriented theories of personality to a growth or health orienta-
tion. Rather than describe human functioning and personality development
in degrees of pathology, the health-oriented theorists focused on descrip-
tions of the personality in degrees of strengths (Wilson, 1972).

The psychology of mental health results from the blending of exis-
tential philosophy and psychology (Jourard, 1963; Misiak & Sexton, 1973).
A subjective philosophy of the individual, existentialism begins with
descriptions of the individual and his/her role in the development of
his/her life. This philosophy holds that only through the study of the
individual can the human condition be understood:

There is no such thing as truth or reality for a living human
being except as he participates in it, is conscious of it, has
some relationship to it. (May, 1960, p. 14)

The individual is seen as uniquely responsible for his/her existence.

To exist is to engage in the process of making choices, thereby defining

4
one's essence by becoming what one is. Existence is a state of freedom
to act and choose one's life and the responsibility to accept the conse-
quences of those choices and actions (Sartre, 1957).

Implicit in existentialism is a definition of the mentally healthy
person which is signif iciantly different from earlier definitions. The
existential definition of mental health is based on the strengths and
actions of the individual rather than problems or weaknesses. The men-
tally healthy person has the courage to be what he/she can be, to be
him/herself, to act and accept the results of that action as his/her
responsibility, and the will and courage to act on that responsibility.
He/she is able to actualize his/her potential to the fullest, to tran-
scend the limiting effects of environmental and social roles, and to be
free to accept the self and others as individuals rather than as objects
or a collection of socially defined roles (Jourard, 1963).

Humanistic psychology, therefore, "is a multifaceted approach to
human experience which focuses on Man's uniqueness and his self -actu-
alization" (Misiak & Sexton, 1973, p. 127). The focus is on the mental
health characteristics of the individual, not his/her problems. An
example of this orientation is the concept of self-actualization
(Maslow, 1970a). The concept was derived, in part, through the exami-
nation of a population identified as being healthy. Maslow found that
a well defined body of characteristics served to separate the healthy
group from other, less healthy, groups. The definition of mental health
derived from this approach is one of the presence of strengths within
the individual rather than the presence of absence or problems (Wilson,
1972).

There is a well established relationship between work and mental
health (Maslow, 1970a; Tiffany, D., et al., 1970; Work In America; 1973) and

the goals of manpower programs to help the unemployed persons in society
obtain work and improve the quality of their lives (Ginzberg, 1968).
However, despite the strength of this relationship there has been no
mental health research examining the unemployed population or the impact
of going to work upon their mental health (Sheppard, 1972; Smith, E.,
1977).

Purpose
This study was designed to examine the mental health characteristics
of unemployed persons and to measure the effects of employment training
upon these mental health characteristics. This was accomplished through
observing unemployed persons as they began a Comprehensive Employment and
Training Act program and comparing these observations with similar obser-
vations of unemployed persons who had completed four months of training
in the program.

Research Questions
The specific research questions to be explored were:

1. What are the mental health characteristics of long-term
unemployed persons as thev enter a Comprehensive Employ-
ment and Training Act funded employment training program

as measured by the Personal Orientation Inventory (Shostrom,
1974), the Bem Sex-Role Inventory (Bern, 1974), and the
Internal-External Scale (Rotter, 1966)?

2. What are the mental health characteristics of long-term
unemployed persons who have completed four months of
Comprehensive Employment and Training Act funded employ-
ment training as measured by the Personal Orientation
Inventory, the Bem Sex-Role Inventory, and the Internal-
External Scale?

3. How do the mental health characteristics of long-term
unemployed persons entering training differ from the
mental health characteristics of long-term unemployed
persons completing four months of training as measured
by the Personal Orientation Inventory, the Bem Sex-Role
Inventory, and the Internal-External Scale?

6

4. IVhat is the relationship between the demographic factors
of age, race, sex, education, religious affiliation,
family history, and length of unemployment prior to the
employment training program experience and the mental
health characteristics of long-term unemployed persons
participating in the employment training program?

Value of the Studv

— ■ - - ■ -■

There are several benefits to be derived from this study. Plans
for emplojonent training programs will benefit from the identification
of mental health characteristics of their target population. Such a
description will enable the employment trainers and counselors to plan
training activities which take into consideration the mental health
characteristics of the clientele. In the process, therefore, it is
hoped to establish a foundation for programs designed to facilitate the
development of mental health in conjunction with the development of good
work habits. Furthermore, the description of mental health among long-
term unemployed persons will expand the population base of mental health
research beyond the traditional study group of college students enrolled
in introductory college psychology courses. The systematic definition
of the effects of employment training upon the mental health character-
istics of long-term unemployed persons will provide a fuller understanding
of mental health concepts.

Rationale

The basic premise of this study is that existing research in mental
health and research examining unemployed persons has failed to investi-
gate the relationship between mental health and unemployment. The need
for an examination of this relationship has long been acknowledged
(Ginzberg, 1968; Sheppard, 1972; Smith, E., 1977; Tiffany, D., et al.,
1970).

Current manpower programs were initiated and developed to aid the
unemployed persons in society and to provide for the manpower needs of
the nation (Ginzberg, 1968; Levitan, 1967). These goals were to be met,
in part, by providing unemployed persons with education, training, coun-
seling, job placement, and supportive services designed to assist them
in overcoming barriers to employment and in obtaining and retaining
fulltime emplojmient .

Despite these stated goals, evaluative studies of manpower program
results have failed to examine the psychological impact of employment
training upon the client. Research in related areas has been confined
to examining the consequences of extended unemployment, the meaning of
work, the motivation to work, job satisfaction, and the consequences of
job dissatisfaction. There have been studies examining the outcomes of
employment training programs but these have been limited to changes in
economic status and length of job retention.

This study focused on the mental health charcteristlcs of unem-
ployed persons and on the effect of employment training upon these
characteristics. Three major mental health concepts were used: self-
actualization (Maslow, 1970a), psychological androgyny (Bern, 1974), and
locus of control (Rotter, 1966). Each of these concepts was defined as
representing a different, yet related, aspect of mental health for the
purposes of this study.

Self-actualization, as developed by Maslow (1970a), is one of the
basic elements of humanistic psychology. Maslow held that the individ-
ual not only could, but must become self-reliant, self-determining, and
self-actualizing as a natural part of continuing growth and development.

8

Self-actualization is the highest level of Maslow's need hierarchy.
He identified five need levels which serve progressively as the basic
motivators for human behavior: physical, security, belongingness and
love, esteem, and self-actualization. Follov>7ing the initial definition
of self-actualization, Maslow developed and expanded the concept into an
extensive theoretical system of mental health for the individual, soci-
ety, science, religion, and values (Maslow, 1959a, 1959b, 1961, 1963a,
1963b, 1964a, 1964b, 1965, 1967, 1968, 1970b, 1971, Maslow & Mittleman,
1951).

Psychological androgyny was described by Maslow (1971) as a central

element of mental health:

The man who thinks you can be either a man, all man, or a
woman, and nothing but a woman, is doomed to struggle with
himself, and to eternal estrangement from women. To the
extent that he learns the facts of psychological "bisexuality,"
and becomes aware of the arbitrariness of either/or defi-
nitions and pathogenic nature of the process of dichoto-
mizing, to the degree that he discovers that differences
can fuse and be structured with each other, and need not be
exclusive and mutually antagonistic, to that extent will
he be a more integrated person, able to accept and enjoy
the "feminine" within himself (the "Anima" as Jung calls
it). (pp. 161-162)

The development of the concept of psychological androgyny was the
result of research on differences between men and women. The research
has found that there exists a clear and distinct difference between what
society views as appropriate for men and for women. Furthermore, these
sex-roles have been found to be valued differently by society, with the
masculine sex-role being valued more than the feminine sex-role. The
research has also found that sex-roles are flexible, varying among indi-
viduals and cultures (Angrist, 1969; Mead, 1950; Staples, 1973).

9

Work is traditionally defined as a masculine activity (Chafetz,
1974). Sex-roles are, therefore, of interest to the employment coun-
selor and trainer since the majority of unemployed persons served by
employment training programs are female (see Appendix B) . Successful
employment training may be found to necessarily include elements of
training in sex-role flexibility.

Original conceptualizations of sex-roles placed masculinity and
femininity at opposite ends of a single bi-polar dimension (Chafetz,
197A). However, subsequent research has shown that sex-roles represent
two distinct dimensions (Bern, 1974; Spence, Helmreich, & Stapp, 1975).
Psychological androgyny refers to the potential within the individual
to endorse as personally appropriate behaviors from either sex-role,
determining the appropriateness on the basis of situational determinants
and personal guidelines rather than traditional societal sex-role defi-
nitions of the behavior (Bern, 1974). Psychological androgyny may also
serve, therefore, as a measure of resistance to enculturation, a charac-
teristic of the self-actualizing person (Bern, 1974; Kelly, Caudill,
Hathorm, & O'Brien, 1977; Maslow. 1970a).

The third concept used in this study to assess mental health is that
of internal-external locus of control (Rotter, 1966). Maslow (1970a)
argued that self-actualizing persons would rely primarily upon their own
values and perceptions of the world rather than the values and judge-
ments of society. Locus of control is a related concept that is defined
as the expectancy or perceived probability that one's actions can influ-
ence the reinforcements of a given situation (Rotter, 1966). An inter-
nal orientation is defined as the expectancy that the outcome of the
situation can be influenced by one's behavior while an external locus of

10
control is defined as the expectancy that the outcome of the situation
is subject to chance and that no control can be exerted over the final
outcome.

Research has shown that locus of control is related to, though
different from, the inner-directed dimension of self-actualization
(Lambert, DeJulio, & Cole, 1976). Unlike other concepts utilized in
this study, locus of control has been used on a limited basis in studies
of unemployed persons. Lefcourt (1966), for example, found results that
suggest unemployed persons fail to take definitive action intended to
improve their situation out of the belief that such action would be to
no avail. In short, they remain unemployed as a result, in part, of an
external locus of control.

These three mental health concepts have been shown to identify
healthy, well-adjusted subjects consistently and accurately and to be
able to discriminate mentally healthy behavior from behavior that has
been clinically judged pathological or maladjusted (Bem, 1979; Rotter,
1966; Shostrom, 1974). However, none of these concepts has been used
to examine the mental health of unemployed persons or the effects of
employment training upon the mental health of the clientele.

In summary, it is argued that the CETA funded employment training
programs would be strengthened by the inclusion of mental health concepts
in the evaluation of unemployed persons and in the evaluation of employ-
ment training effects. Furthermore, it is argued that the concept of
mental health would be strengthened through the examination of a popu-
lation significantly different from traditional mental health research
populations and by examining the impact of employment training upon
mental health characteristics.

11

Definition of Terms
Several terms used in this study are defined in this section for the
purpose of establishing a common terminology.

Work. Super (1976) provides a broad, flexible definition of work
which will be used in this study. Work is:

The systematic pursuit of an objective valued by oneself
(even if only for survival) and desired by others; directed
and consecutive, it requires expenditure of effort. It may
be compensated (paid work) or uncompensated (volunteer work
or an avocation). The objective may be intrinsic enjoyment
of work itself, the structure given to life by the work role,
the economic support which work makes possible, or the type
of leisure which work facilitates. (p. 20)

Employment. The terms "work" and "employment" are often used
interchangeably resulting in confusion during discussions of related
concepts. To avoid this confusion "employment" will be given the more
limited definition of "paid work," regardless of the motivation for
working. Employment, therefore, is distinct from volunteer work.

Unemployment. This term refers to being without employment, i.e.,
without paid work.

Unemployed person. Within the definition provided for unemployment,
there is a broad range of persons defined as unemployed. College
students, for example, may be working hard towards a degree but they are
unemployed while engaged in the process. Therefore, for the purposes
of this study an individual will be considered as an unemployed person
if he/she meets the eligibility requirements for participation in a CETA
funded employment training program. Basically, these requirements are
that the individual must have been unemployed for an extended period of
time and have a sufficiently low income to be classified as economically
disadvantaged (see Appendix A). This definition will be applicable as
long as the individual is participating in an employment training program.

12

Employability. This terra Is used to refer to the individual's
ability to identify, obtain, and retain fulltime employment. This
ability includes such factors as being trained in the occupation,
good work attitudes, the ability to complete an application, and the
ability to interview for the position.

Employment training. Generally this terra refers to any structured
activity designed to enhance an individual's employability . A college
education, for example, may be considered as a form of employment
training. However, for the purposes of this study, the term will refer
to activities funded under the Comprehensive Employment and Training Act
of 1978 (CETA) . The goal of employment training is to develop sufficient
employability skills within the individual to enable them to find full-
time emplojmient outside of the CETA program.

Mental Health. This is a state of integration of needs, drives,
and values. It involves the acceptance of self and others, self-reli-
ance, non- judgemental interpersonal skills, freedom from defensive
interaction patterns, and self-actualization (Jourard, 1963). For the
purposes of this study the concepts of self-actualization, psychological
androgyny, and internal-external locus of control will be defined as
mental health concepts.

Self-actualization. The self-actualizing person has successfully
developed the ability to satisfy the basic needs (physical, security,
belongingness and love, and esteem) and is free to develop his/her
potential to the fullest (Maslow, 1970a). The ability to integrate
needs and potential represented by self-actualization is defined as a
mental health concept for the purposes of this study.

Psychological androgyny. Bern (1974) introduced the concept of psy-
chological androgyny to describe sex-role flexibility and the integration

13
of masculine and feminine sex-roles into a unified personal system.
The androgynous individual is able to engage in an activity on the
basis of its personal appropriateness in that situation rather than on
the basis of its traditional socially defined sex-role alignment. This
ability is defined as representative of mental health for the purposes
of this study.

Locus of Control. Internal-external locus of control was intro-
duced by Rotter (1966) to describe the degree of control the individual
expected to be able to exert over the rewards and outcomes of a given
situation. An internal locus of control refers to the expectancy that
such control is possible while an external locus of control refers to the
expectancy that the outcomes of the situation are, at best, randomly
determined. As with self-actualization and psychological androgyny,
an internal locus of control will be defined as indicative of mental
health for the purposes of this study.

CHAPTER II
REVIEW OF LITERATURE

Literature relevant to the current study will be discussed in this
chapter. The review will be organized into two sections: Unemployment
and Mental Health.

Unemployment

The literature discussed in this section will cover three elements
required for understanding unemployment and efforts to help the unem-
ployed person: Work, Manpower Programs, and the Comprehensive Employ-
ment and Training Act of 1978.
Work

The importance of work. Work is widely recognized as basic to the

well-being of the individual and essential to the smooth functioning of

society. Economists argue that without enough work for those who wish

to work or enough workers for the work to be done the national economy

would collapse (Heller, 196A; Heron, 1948; Killingsworth, 1964; Pierson,

1972). Work is also seen as important by the government, as evidenced

by the establishment of national manpower programs and employment goals

for the nation (Bakke, 1969; Levitan, 1964). Wilensky (1966) identified

work as an important problem for the individual and the social system

because:

(1) Every society defines it as a central obligation for most
of the population, but in modern society some men cannot obtain
enough of it. The existing skills and talents of the popula-
tion never perfectly fit the demands of the technology and the
economy. (2) Many men are discontented with the work they do.
(3) Employers and officials often feel their subordinates are

14

15

not doing enough work or good enough work. Hardworking
portions of the population sometimes complain that others
are "f eatherbedding" or "goldbricking. " (4) Lack of work
or alienation from it can place a heavy hand on the qual-
ity of life. (p. 118)

This study, however, focused on the perspective of the individual.

Rather than examine the importance of work from a social, economic, or

political point of view, the focus was on the impact of unemployment,

work, and employment training on the mental health of the individual.

In a report prepared for the Secretary of HEW (Work In America, 1973)

a task force of engineers, sociologists, social workers, economists,

anthropologists, and political scientists examined the role and function

of work in contemporary America and concluded:

Work is central "in the lives of most adults," it contributes
to identity and self-esteem, and it is useful "in bringing
order and meaning to life." Work offers economic self-suffi-
ciency, status, family stability, and an opportunity to inter-
act with others in one of the most basic activities of society.
Consequently, if the opportunity to work is absent, or if the
nature of work is dissatisfying (or worse), severe repercus-
sions are likely to be experienced in other parts of the
social system. (p. xv)

The meaning of work. One approach used in the study of the meaning
of work has been to ask the question: What is the difference between
working and not-working? Weiss and Kahn (1960) asked this question of
371 workers from several different occupations. They found that "over
three-fourths of respondents defined work either as activity which was
necessary though not enjoyed, or as activity which was scheduled or
paid" (Weiss & Kahn, 1960, p. 150). Further, they found that the defi-
nition of work as a necessary activity was associated with occupations
which accorded the worker some degree of autonomy and required advanced
education. The definition of work as scheduled or paid activity was
associated with occupations which had emotional separation from the job

16

and allowed the workers no social status or freedom on the job. In
response to the questions of how workers would feel different if they
were not working, Weiss and Kahn found that all workers except those
in low status jobs would feel emotionally worse. The low status workers
reported that they would feel the same or better if not working.

A second approach to the importance of work has been to examine the
significance of work in the lives of the workers. Dubin (1956) surveyed
A91 workers and found that only 2A% identified work as central to their
life interests although 61% identified the workplace as their central
formal organizational attachment. He concluded that the "worker has a
well developed sense of attachment to his work and workplace without a
sense of total commitment to it" (Dubin, 1956, p. 140). In a similar
study, Williams, R. , Morea, and Ives (1975) asked a sample of students
and managers to rank several items in terms of personal importance. The
items included recreation, school, church, occupation, and family. They
found that approximately two-thirds (20 out of 33) of the students
ranked their future occupation as most important, more frequently than the
sample of managers (3 out of 17). Managers were found to rank their
families in first position most frequently (14 of 17). However, 80% of
the total sample (41 out of 51) ranked occupation in first or second
place. Furthermore, the top two functions of work identified were self-
realization and social contact. Economic security ranked third for both
groups as a function of work.

A third approach to understanding the importance of work has been
to ask the workers why they work and whether they would work if the eco-
nomic necessity were removed. Morse and Weiss (1955) found that 80% of
401 workers sampled would continue to work even in the absence of

17
economic need. Reasons for continuing to work included: to keep occu-
pied or interested (31%), would go crazy if not working (14%), wouldn't
know what else to do (10%), enjoyed work (9%), work justified existence
(5%), habit (6%), and to maintain self-respect (5%). This response
pattern was consistent across all age groups ranging from just entering
work (90% would continue to work) to retired workers (65% would continue
to work) .

Kaplan and Tausky (1972) asked questions similar to those used by
Morse and Weiss (1955) of a sample of hard-core unemployed subjects
(N=275). They found that the most important reason given for having a
job was to make a living and support the self and family (52%). Other
reasons given included: liking the work (11%), security of a steady job
and income (10%), learning good work habits (4%), and having a chance to
use abilities and accomplish something (4%). They also found that most
of the respondents would rather wash cars than accept welfare (71%),
would work even if they did not need the money (84%), would work at a
lower paying respectable job rather than take a higher paying job which
was looked down on (57%), felt that the most important part of a promo-
tion was the increased respect they would get (54%), and would be con-
cerned about not being promoted even though their pay increased steadily
(54%).

Working appears to serve two basic functions in the lives of the
worker and in the lives of those who would like to work. The first
function is that of income and the support for self or family provided
by the income. The second function is equally important though less
tangible than income; the provision of self-respect, self-fulfillment,
respect from others, social standing, and a structure to life. A further

18
finding is that workers in higher level jobs with more education, gener-
ally management and white-collar positions, tend to rate the intangible
factors as more important than do the blue-collar or lower status workers
and unemployed individuals. Finally, groups at all levels of employment
have been found to express a desire to continue working independent of
economic needs.

Job satisfaction. The area of job satisfaction has been the subject
of extensive research and has been recognized as an important element
in the well-being of the working population:

Significant numbers of American workers are dissatisfied with
the quality of their working lives. Dull, repetitive, seemingly
meaningless tasks, offering little challenge or autonomy, are
causing discontent among workers at all occupational levels.
This is not so much because work itself has greatly changed;
indeed one of the main problems is that work has not changed
fast enough to keep up with the rapid and widespread changes
in worker attitudes, aspirations and values. A general increase
in their educational and economic status has placed many Americans
in a position where having an interesting job is now as important
as having a job that pays well. ... As a result, the produc-
tivity of the worker is low — as measured by absenteeism, turn-
over rates, wildcat strikes, sabotage, poor-quality products,
and a reluctance by workers to commit themselves to their tasks.
Moreover, an increasing body of research indicates that, as work
problems increase there may be a consequent decline in physical
and mental health, family stability, community participation and
cohesiveness , and "balanced" socio-political attitudes, while
there is an increase in drug and alcohol addication, aggression,
and delinquency. (Work In America, 1973, p. xvi)

Lyman (1955) identified job satisfied workers with three questions:

If you could start over, would you do the same work; Given the same pay,

what would you do; Do you want your son to do the same work as you do?

She found that only 77 of 249 workers were satisfied, i.e., answered all

three questions in the affirmative. She further found that among the

satisfied workers the white-collar subjects preferred the interesting

quality of the job and the freedom on the job while the blue-collar

19

workers preferred the economic rewards and the physical conditions of
the work. Among the dissatisfied workers the white-collar workers
found their work uninteresting and wanted to change the nature of the
job while blue-collar subjects disliked the physical demands of the job
and wanted to make the work easier.

The elements of work have been divided into intrinsic and extrinsic
categories (Herzberg, 1966). Intrinsic factors include characteristics
such as promotions, personal growth, recognition, responsibility, the
nature of the work, and an opportunity for advancement. Extrinsic
factors include working conditions, security, status, salary, super-
vision, and relationships with fellow employees.

Saleh and Singh (1973) studied the instrinsic and extrinsic elements
of the job valued by 3000 white-collar workers as moderated by the sub-
jects educational level, salary, and father's occupation. They found
that among the low salaried subjects (annual income less than $10,000)
intrinsic factors increased in importance as the father's occupational
level (unskilled versus skilled or professional) increased and as the
community size increased. Among the high salary group this relation-
ship was significant only for the university educated subjects. Over-
all, the higher salary group was significantly more intrinsically
oriented than the low salary group.

Herzberg, Mausner, and Snyderraan (1959) suggested that jobs which
allowed for increased opportunities for achievement, recognition,
responsibility, advancement, and growth would be more satisfying than
jobs which did not offer these elements. Sheppard and Herrick (1972)
found that satisfaction was shown to increase as job variety, autonomy,
meaningfulness, and responsibility increased. Kahn (1972), following

20

an extensive literature review, concluded that improved job satisfaction
resulted from: increases in job rewards and autonomy; decreases in repe-
tition, overspecialization, and fractionalism; and improvement in super-
vision, wages, promotions, peer relationships, and working conditions.

Chisholm (1975) interviewed 100 college graduates to assess the rela-
tionship between alienation on and off the job. He found that lack of
good supervisory evaluations, lack of upward mobility, and being young,
rather than the quality of work or type of job activities, were asso-
ciated with the highest levels of on or off the job alienation. Heron
(1948), in an earlier study of why individuals choose to work or are
satisfied with their work, concluded that workers needed to be involved
with their work. Without work involvement, workers were found to choose
no work at all, even when the work was available for them.

Slocum and Strawser (1972) compared assessments of job settings by
black and white Certified Public Accountants. They asked the subjects
to identify which elements of the job were satisfying and which personal
needs were not being satisfied on the job. They found that the black
subjects expressed a significantly higher unmet need for an opportunity:
to help people, for friendship, for a feeling of self-esteem, for indepen-
dent thought and action, for growth and development, and for compensation
than was expressed by the white subjects. After classifying these needs
according to Maslow's need heirarchy (1970a), they concluded that the
black subjects were expressing a pattern of unmet needs which was much
less self-actualizing than the white subjects.

Blood (1969) developed an instrument which measured endorsement of
the Protestant Ethic. He found that 306 United States Air Force airmen
instructors exhibited a positive relationship between the endorsement

21
of the Protestant Ethic and job satisfaction. Another 114 student
airmen did not exhibit the relationship between the Protestant Ethic
and job satisfaction.

Wanous (1974) surveyed 80 female telephone operators using the
Blood (1969) Protestant Ethic Scale and a questionnaire which gathered
data on childhood background, preferences for job characteristics, and
job satisfaction. Results indicated that the subjects' need for job
satisfiers, such as variety, autonomy, task identity, and task feedback
was a better predictor of job satisfaction than endorsement of the
Protestant Ethic or childhood background.

Although these studies make a strong case for the importance of
assisting workers in achieving job satisfaction, theorists have cau-
tioned that satisfaction may not represent the total explanation of
work related behavior. Kahn (1972) noted that:

For most workers it is a choice between no work connection
(usually with severe attendant economic penalties and con-
spicuous lack of meaningful alternative activities) and a
work connection which is burdened with negative qualities
(routine, compulsory scheduling, dependency). In these
circumstances the individual has no difficulty with the
choice; he chooses work, pronounces himself moderately
satisfied, and tells us more only if the questions become
more searching. (p. 200)

Rottenburg (1960) argued that economic factors will serve as a strong

influence in job choice in addition to job satisfaction and should not

be discounted. Blaumer (1960) noted that other factors influence job

choice including the need for structure in life, endorsement of the

Protestant Ethic, and the fact that the individual may never be fully

satisfied but will assign meaning to work which theorists would find

meaningless.

22

In summary, the research has shown that many workers are dissatis-
fied with their work and this dissatisfaction affects the overall quality
of their lives. Several factors have been found to affect the degree of
job satisfaction. Job characteristics such as autonomy, variety, feed-
back on performance, involvement in work, work conditions, pay, and
status as well as factors within the individual worker such as endorse-
ment of the Protestant Ethic, education level, desire for job satisfiers,
and family background all interact to determine the degree of job satis-
faction. Finally, research has shown that regardless of the degree of
satisfaction felt by the workers, they would continue to work even when
the monetary necessity had been removed suggesting that work is basic to
the self-concept of the workers independent of other functions.

The unemployed person. The motivations of unemployed persons
become increasingly important in view of the findings discussed in the
preceding section. It is important to understand why some people remain
unemployed while others will continue to work in unsatisfying jobs, even
if the economic motivations were absent.

Surveys have examined the demographic characteristics of long-term
unemployed persons and found four major groupings (Hargrett, 1965;
Killingsworth, 1968; Zeisel, 1964). First, blacks have a higher unem-
ployment rate than whites. Killingsworth reported an unemployment rate
for blacks at 7.3% as compared to a 3.3% rate for whites. Second, age
is a factor. Younger workers have difficulty finding employment due to
a lack of skills and experience while older workers have difficulty due
to the obsolescence over time of their skills and experience. Third,
those with less education and limited educational opportunities have
higher rates of unemployment. Finally, residency is a factor. Many

23

areas have limited employment opportunities due to changes in economic
conditions which force local companies to move to another area or to
cease operation.

Tiffany, D., et al. (1970) warn, however, that it is important to
avoid the misconception that the unemployed are a homogeneous group with
a common set of needs and barriers to employment. Lack of homogeneity
among unemployed persons was first demonstrated by Bakke (1933) in a
study of the impact of unemployment insurance on the persons' willingness
to return to work. Bakke's general finding was that if the subjects
were willing to support themselves before entering a period of unemploy-
ment then use of unemployment insurance would not decrease that desire
to work. Likewise, unemployment insurance was not found to create a
willingness to work, or a desire for self-sufficiency, if they were not
present before the period of unemployment began.

Further studies by Bakke (1940a, 19A0b) found that unemployed
workers would use all available methods and skills to maintain themselves
and their self-esteem as the period of unemployment was extended. He
found that if the unemployed workers had a wide range of job seeking and
personal maintenance skills before being unemployed, they would use those
skills during the period of unemployment for survival and to fill the role
normally held by work. If no prior skills were present, then the workers
were found to rely on the unemployment insurance sooner and to a greater
extent.

Davis (1946) studied individual cases of persons unemployed for an
extended period of time and found several adjustments being made as the
period of unemployed grew longer. She found a gradual lowering of
aspirations, hopes, and educational and life goals. As control over
working decreased, ambition and a hope for the future were gradually

24
relinquished as luxuries too painful to maintain. Alfano (1973) found
that unemployed persons experienced few changes in their attitudes
towards work during the first few months of unemployment. However, as
the period of unemployment grew longer their attitudes towards work
deteriorated as did the level of their savings, the strength of their
family unit, their social standing, and their self-confidence.

Feldman (1973a, 1973b, 1974) conducted a series of studies designed
to examine assessments of work related outcomes among black and white
subjects who were employed and unemployed. He concluded that the intent
to work and job seeking behavior were not completely tied to race, socio-
economic status, employment status, or the expectation that work would
result in positive outcomes but are also moderated by local factors such
as available job markets, availability of support outside of work, and
individual differences.

Long-term unemployed persons should not, however, be construed to
have no desire to work. Searls, Braucht , and Miskimins (1974) compared
work values of supervisors, employees, and long-term unemployed subjects
on five work-related variables: personal habits, supervisory relations,
employee relations, company relations, and work habits. I'Jhile they found
significant differences between work values endorsed by the supervisors
and the unemployed subjects, there were no differences between the
employed and unemployed subjects.

Cook (1971), in a comparison of motivation to work between 100
employed and 100 unemployed blacks found comparable levels of motivation
to work between the groups. Likewise, Kaplan and Tausky (1974) surveyed
275 long-terra unemployed subjects with a questionnaire similar to those
used by Morse and Weiss (1955), Tausky (1969), and Kaplan and Tausky

25

(1972). They concluded that:

While the economic meaning and function of work weighed
heavily in many of our subjects' responses, there was also,
overall, a commitment to work not only similar to that of
employed blue-collar workers, but at times resembling the
values of people of higher socio-economic status. These
findings contradict the prevalent stereotype in the United
States that the poor have little desire or commitment to
work. (Kaplan & Tausky, 1974, p. 195)

Unemployment and mental illness. Several studies have demonstrated
that extended unemployment can be detrimental to the individual.
Zawailski and Lazarsfeld (1935) identified a sequence of mental health
deterioration among long-term unemployed persons through a review of their
autobiographies. The initial reaction to being unemployed was feeling
hurt by the dismissal, followed by, in order: numbness or apathy, hope
or calmness, a weakening of hope, hopelessness and suicidal ideas, and
dumb apathy. Shanthamani (1973) found that as unemployment lasted longer
among engineering graduates their level of neuroticism increased.

Triandis, Feldman, Weldon, and Harvey (1975) examined the trust
levels of long-term unemployed persons and found that they exhibited
little trust for others, even those from their own in-group. Further-
more, the subjects did not trust or believe that hard work would result
in any desired benefit and felt unimportant and, as a result, rejected
the establishment.

A strong relationship was demonstrated between social class and the
frequency and types of mental illness (Hollingshead and Redlich, 1958).
This relationship was confirmed in a follow-up study by Myers and Bean
(1968). Together these two studies demonstrated that the lower social
classes had higher Incidences of mental illness and that the lower class
subjects were more likely to be diagnosed as having a mental disorder
and to receive custodial care.

26

Warheit, Schwab, Holzer, and Nadreau (1973) identified a similar
pattern of mental illness in an in-depth study of Alachua County,
Florida. They found a strong increase in psychopathological symptoma-
tology as the socio-economic status of the subjects decreased. Scott,
W., (1958) in a review of relevant literature, identified a pattern of
socio-psychological factors associated with mental illness that included:
the presence of a stressful event, being black, being in the lower socio-
economic classes, living in an urban area as opposed to rural areas, and
living in the inner-city as opposed to suburban areas.

Cassel (1966) concluded that "with one important exception every
study over the last 45 years has found the highest concentration of
mental disorders occurring in the lowest social class" (p. A2) . The one
exception was a study by Pasamanic , Roberts, Lemkau, and Kreuger (1964)
which discovered that the greatest incidence of psychiatric diagnosis
was not among the lowest income group or among blacks. They concluded
that these findings resulted from a diagnostic focus on the more severe
disorders such as psychosis. Freedman (1969) similarly concluded from a
study of psychiatric records that the lower classes are more likely to
exhibit neurotic disorders expressed somatically and behaviorally than
they were to exhibit psychotic disorders of the thought processes.

Kornhauser (1965) studied the mental health of auto workers and
summarized his findings regarding unemployment and poor working condi-
tions stating that:

Poorer mental health occurs whenever conditions of work and
life lead to continuing frustration by failing to offer means
for perceived progress toward attainment of strongly desired
goals which have become indispensable elements of the indivi-
dual's self-identity as a worthwhile person. Persistent failure
and frustration bring lowered self-esteem and dissatisfaction
with life, often accompanied by anxieties, social alienation
and withdrav\7al , a narrowing of goals and curtailing of aspira-
tions— in short — poor mental health. (p. 269)

27

Unemployment and mental health. It is not surprising that evidence
has been found to indicate that work is conducive to the development of
mental health. Massimo and Shore (1964), for example, conducted a voca-
tionally oriented treatment program for anti-social youth and concluded
that not only was work therapeutic in itself but it also served as a focus
for other interventions, facilitated identity formation, provided an
avenue for channeling aggressive and sexual energies, and served to alle-
viate material needs. In a follow-up study, however, they found that
these benefits were lost if the work performed was demeaning or meaningless
(Shore & Massimo, 1969). Similarly, Tiffany, D. , Cowan, Eddy, Glad, and
Woll (1967), in a study of hospitalized psychiatric patients, found that
patients involved in their work situations exhibited higher levels of
self-identity and self-satisfaction than patients not involved in their
work.

Clearly, working is conducive to the development of mental health.
However, a question must be raised, since unemployment has also been
associated with increased evidence of mental illness, as to whether
employment is a necessary condition for the development of mental health?
Bakke (1933, 1940a, 1940b), for example, stated that unemployed persons
continued to maintain their esteem following the loss of employment
using the same skills they had used while employed.

Tiffany, D., et al. (1970) warned against assuming that all long-
term unemployed persons were similar. As an example, they identified an
assumption that unemployed persons communicate in the same manner as
middle-class persons while being assumed to be cognitively deficient.
Reissraan (1964) studied the cognitive and verbal patterns of lower-class
ghetto children and middle-class children and found no evidence to

28

suggest cognitive or verbal deficiency among the ghetto children. He
found the ghetto children to be as expressive in their verbal patterns
as middle-class children. The ghetto children excelled in fantasy and
spontaneous, unstructured self-expression instead of the more tradi-
tional patterns of self-expression exhibited by the middle-class
children.

Schesta (1976) studied the relationship between job satisfaction,
work as a central life interest, age, and an affinity for leisure
activities vis-a-vis working. He found that there was a stronger rela-
tionship between leisure and age when the worker was younger, dissatis-
fied with the work setting, or did not perceive work as central to his/
her life. Similarly, Hausknecht (1964) found that blue-collar workers
facing dry, uninvolving, routine work demands will use leisure and
voluntary group associations outside of the work setting to achieve and
maintain satisfaction. He found that by participating in social groups
the blue-collar workers had a medium for expressing themselves and
establishing social contact. Price and Levinson, H., (1964) developed
a description of the mentally healthy worker which might be applied to
mentally healthy persons regardless of their employment status. The
mentally healthy workers possessed a wide variety of sources of satis-
faction and psychological rewards, were flexible under stress, treated
others as individuals, recognized and accepted their own capacities and
limitations, and were active and productive.

Based upon these observations it may be possible to apply these
descriptions to appropriate groups of unemployed persons and identify
or anticipate the development of mental health among unemployed people.
As Tiffany, D., et al. (1970) stated:

29

If we were to indicate what human beings need, whether they
are an advantaged or disadvantaged group, we would probably
list the following: Human beings need to be self-determining,
to have control over their own fate, to be a person rather
than a thing, to be able to make decisions, to be able to
plan, to be loved rather than counted, and to be recognized
for one's accomplishments. (p. 115)

Summary . Research has found that work is important to society and
the individual. For most people work provides not only income but also
structure and meaning to life as well as personal identity. Without
satisfying work, or without any work at all, the overall quality of life
has been seen to deteriorate. Studies of the healthy worker have identi-
fied a pattern of life and needs similar to Maslow's (1970a) description
of the self-actualizing person. Some of the characteristics identified as
applicable to mental health among workers may also be found to apply to
mental health among unemployed persons. However, job dissatisfaction and
unemployment have been generally found to exert a strong negative effect
on the mental health and overall quality of life of the individual.
Manpower Programs

Manpower program goals. The current manpower training program for
the United States is emcompassed in the Comprehensive Employment and
Training Act (CETA) of 1978. The CETA, along with its predecessors, CETA
of 1973 and the Manpower Development and Training Act (MDTA) of 1962, was
designed to fund a coordinated and comprehensive approach to the solution
of unemployment related problems. Included in CETA are provisions for
training youth through classroom training, work experience, and summer
jobs, and for adults by providing classroom training and on-the-job
training along with special programs designed to meet the needs unique
to the local community.

30

It is not within the scope of this study to provide a comprehensive

history of American manpower policies. Such a history may be obtained

from other sources (Levitan, 1967; Mirengoff & Rindler, 1976; State of

Florida Office of Manpower Planning, 1978). It is however, useful to

identify some of the key factors which have influenced the development

of manpower goals. Levitan (1964) identified six forces affecting the

development of national manpower policies: Russian success in space

technology, high levels of unemployment following the 1957 recession,

changes in defense technology resulting in changes in regional funding,

increased international competition for local markets, demographic

changes following the "baby boom," and long-term problems such as racial

and sexual discrimination in hiring and training practices. Ginzberg

(1968) stated:

World War I necessitated a balancing of military and civilian
needs. Next, the Immigration Laws of 1920 and 1924 put an end
to Europe's serving as our reserve pool of unskilled and skilled
workers. The Great Depression of the 1930 's revealed that the
adjustments between supply and demand for labor were not auto-
matic. The World War II and the cold war which followed illumi-
nated the critical role of scientists and development engineers
in assuring national survival and technical progress. The
passage of the Employment Act of 1946 committed the federal
government to establishing and maintaining a high level of
employment. The civil rights movement . . . [and] a growing
demand for services, particularly health and education . . .
led to further shifts. (p. 4)

Several different goals and tasks have been suggested for manpower
programs. Pierson (1972) argued that full employment is basic to the
stability of our economy and that employment programs should be indepen-
dent of the political structure. Ferman (1969) stated that an additional
goal for manpower programs should be expanded job development and
placement for unemployed persons. Killingsworth (1964) identified
as a major manpower goal the retraining of persons unemployed as a
result of automation. Goodman (1969) defined the restructuring of the

31

social system within which the trainee is trained and eventually

employed as a major task of manpower programs. Bakke (1969) argued

that the federal manpower programs should include efforts to expand

the total number of jobs available in addition to helping those without

a job find work. Ginzberg (1968) identified ten manpower program goals:

(1) The rapid growth of employment. (2) The matching of men
and skills. (3) The expansion of the supply of trained man-
power. (4) The diffusion of the benefits of automation. (5)
The reduction and elimination of poverty. (6) The development
of human potential. (7) The rehabilitation of depressed areas.
(8) The elimination of discrimination. (9) The Improvement of
labor market institutions. (10) The broadening and deepening
of manpower research. (p. 225)

It should be noted that the discussions cited above were concerned
with the MDTA. These discussions served as the basis for the creation
of CETA. There were three basic goals written into CETA: decentrali-
zation, structural training, and countercyclical employment. The first
goal was intended to place more control of manpower programs in the hands
of the local governments. This was designed to correct the lack of
responsiveness in meeting local needs common to many of the MDTA programs.
Structural training was common to both MDTA and CETA and involved training
for persons unemployed as a result of the structure of society, i.e..
Inability to find employment through lack of skills, education, or
opportunity resulting from some structural element such as racial or
sexual prejudice. The major difference in this goal is that CETA combined
all structural training under one Title for efficient administration.
Countercyclical employment was established to provide training and
employment opportunities to those persons unemployed as a consequence of
the cyclical changes in the national economy. Funding for this program
was based upon the current unemployment rate (Mirengoff & Rindler, 1978).

32

Evaluation of manpower programs. Efforts to evaluate the effec-
tiveness of manpower programs have met repeatedly with frustration
resulting first from the poor validity and quantity of evaluative data
and second from the limited scope of the evaluative studies. Mangum
(1967, 1969) has repeatedly noted the lack of valid data for the evalu-
ation of manpower programs:

For no program are there adequate valid data for evaluation of
strengths and weaknesses and ... no program has a reporting
system capable of producing such data. Data on services are
weak and follow-up data on program results are grossly inade-
quate and undependable. Ad hoc internal evaluations have been
limited, their data weak, and their investigations less than
probing. (Mangum, 1969, p. 131)

Another part of the problem lias been the lack of adequate samples
for testing. Several researchers have noted that long-term unemployed
persons are singularly transient and difficult to use in controlled
research (Barnes, 1972; Hardin, 1972; Lewis, 1972). Their recommenda-
tions have included the paying of unemployed subjects to serve as a
stable research population; although they did not address the question
of whether such subjects continue to be unemployed. Scanlon, Nay, and
Wholey (1972) identified three problems basic to evaluative research
conducted on manpower programs. First evaluations have been one-shot,
one-time efforts rather than continuous, longitudinal studies. Second,
they have been done in terms of program categories and are weak on
process data. Finally, the sample populations used in the evaluations
have been too small when regional samples are needed.

The problems with program evaluation identified with MDTA did not
improve with the conversion to CETA even though the criticism made of
MDTA evaluation resulted in the inclusion of requirements for improved
data collection and reporting procedures in CETA:

33

Quantitative measurement of program results is seriously
hampered by limitations in the CETA reporting system. The
Employment and Training Administration restructured and
unified the system to streamline and reconcile the separate
reporting systems of a number of individual programs. As
a result, it is impossible to isolate CETA data with suffi-
cient detail to make comparisons with pre-CETA programs or
even to fully analyze CETA outcomes. (Mirengoff & Rindler,
1978, p. 221)

Several studies designed to examine aspects of manpower training
programs warrant review. Although there were several changes in MDTA
and CETA, the primary goal and basic methodology of training programs
remained the same. Furthermore, with limited exceptions, there has
been no research on CETA applicable to this study. Studies being done
on CETA are concerned with other areas of CETA not relevant to this
study (Klein & Ghozeil, 1979). For these reasons evaluations of both
programs will be reviewed together.

Several studies have been conducted to identify those factors which
contribute to the individual's completion of training and subsequent
success on the job. Beatty, R., (1975) examined the use of social self-
esteem, scholastic achievement, and job skills as predictors of job suc-
cess following training. He found that self-esteem did not improve
during training and was unrelated to success on the job. Both scholastic
achievement and job skills improved during the course of training and
were related to job success. Improved scholastic achievement was
directly related to increased salary while increased job skills were
directly related to improved supervisory evaluations.

Beatty, R., and Beatty, J., (1975) found that during the first six
months of employment following placement absenteeism was related to lower
salary, low supervifriry evaluations, supervisors perceived as unsup-
portive, jobs which required little ability, and a short prior work

34
history. Absenteeism after 24 months on the job was related to percep-
tions of the supervisor, self-esteem in the work setting, skill levels
required by the job, and duties assigned. Goodman and Salipante (1976)
found that hard-core unemployed trainees were more likely to remain on
the job if there were frequent raises during the early part of employ-
ment, regardless of the size of the raise, and if there was frequent
job performance feedback. They concluded that, while telling the unem-
ployed person about the benefits of employment will improve job retention
on a limited scale, greater success will be achieved with frequent and
iiranediate reinforcement of good job behavior.

Personal characteristics of the trainees have also been found to
contribute to the success of training. Henke (1976) examined the rela-
tionship between completion of training and five groups of personal
attributes: Personal — marital status, age, and race; Academic — literacy,
level of education, and math and word knowledge; Economic — possession of
material goods, prior hourly wages, and length of prior unemployment;
Vocational — prior training, prior employment history, and veteran's
status; and Program Status — referral source and length of pre-vocational
training. He found that only the subjects' academic achievement and prior
vocational history were related to program completion. Similarly, in a
study of the characteristics of the successful trainee. Cole (1974)
found that completers were more likely to be female, older, with a better
academic background, and no prior arrest record.

The relationship between training success and expectancy theory has
been examined. Gurapper (1971) found that attitudes towards work and a
generalized expectation of success were not related to program comple-
tion while prior work history was predictive of success. On the other

35

hand, Goodman, Salipante, and Paransky (1973) found that job retention
was not the result of employment training but of the expectation that
remaining on the job would bring rewards.

Counseling has been found to facilitate completion of training.
Bryant, F.,and Showalter (1971) found that participation in a pre-
training orientation and assessment program which included strategies
to improve test taking, job seeking, and vocational exploration served
to significantly increase training completion. Baron and Bass, A.,
(1968) found that verbal reinforcement was more effective in improving
trainee self-esteem than material rewards while material rewards served
to increase task performance. Both types of rewards were found to be
more effective when the trainee perceived them as situation appropriate.
A Philadelphia study (Philadelphia Manpower Utilization Commission,
1969) found that participation in pre-training work adjustment training,
family services counseling, and social work counseling improved train-
ing completion. Salipante and Goodman (1976) found that counseling
during training in conjunction with role playing of job related skills
served to extend the effective length of the training process.

Goursslin and Roach (1964) concluded that while early MDTA programs
did indeed help the upper level trainee they were not significantly
effective with the disadvantaged or hard-core unemployed person. Solie
(1968) examined completers and non-completers of early training pro-
grams and found that completers had a better job experience following
training than did rejects and non-completers, but that the evidence
suggested a selection bias which screened out the less qualified, low
potential applicants. This problem has continued in later programs:

CETA programs have been least successful in finding unsubsi-
dized jobs for the hard-core unemployed — minority partici-
pants, persons with less than a high school education, younger
workers, and the poor. (Mirengoff & Rindler, 1978, p. 239)

36

Roskind (1975) compared retention records of hard-core unemployed
individuals who received special consideration in being selected for
employment and were oriented to the job with the retention records of
individuals hired through the regular process. He found that, as a
group, the hard-core unemployed subjects had a poorer track record in
the job situation than the regular employees. Qulnn, Fine, and Levitan
(cited in Goodman et al . , 1973), however, found that the hard-core
unemployed individuals completing a regular training program prior to
entry into work did as well on the job as regular hires. They also
found that the trained subjects sought greater on-the-job autonomy and
evaluated their supervisors lower than regular employees. Several other
studies have shown that completers of training programs generally tend
to have better job retention and a higher wage rate than hard-core
unemployed persons who do not complete training (Harris & Associates,
1967; Mangum, 1967; Mangum & Walsh, 1973; Thorpe, 1973). Levitan
(1969) suggested that training per se may not enrich the trainees so
much as it serves as an "aging vat" in which the trainees receive a job
and wages until they are ready to cope with the world of work on their
own.

Gordon and Scott, R., (1972) used a structured interview and alien-
ation scale to identify areas of the trainee's lives which were signi-
ficantly affected by employement training. Graduates of the training
program had a higher spending rate and better housing, read more of the
local news, were more likely to be members of unions, community action
groups, or civil rights groups than they were to be members of social
or recreational groups, attended church more frequently, and were less
alienated from society than were non-graduates.

37

Christensen (1974) matched parolees completing MDTA training with
parolees not completing or applying to MDTA training on several cri-
teria Including age, type of crime for which the subject was convicted,
educational level, achievement test scores, and parental occupation.
He found that the program completers had significantly lower recidivism
rates, committed fewer felonies, and had a better employment rate than
non-completers .

The conclusions drawn from the evaluation studies of manpower pro-
grams are less than clear cut. In all studies examined there were no
reports of unqualified success in meeting training goals, although most
argued that the data suggested at least partial success. However, the
evaluators of the programs consistently support them as being benefi-
cial :

The conclusions reached by a number of critics of retraining
are disheartening. . . . But the expectations must both be seen
in perspective. Retraining has three major objectives: (1)
short run increases in employment; (2) long-run economic
growth; (3) improvement in the welfare and general well-being
of the trainees and society. Regardless of the short run
deficiencies, there can be little doubt that retraining the
unemployed, like the educational process as a whole, is a
worthwhile enterprise from the standpoints of the long-run
economic growth of the American economy and the general well-
being of its citizens. (Somers, 1964, pp. 152-153)

While the record is mixed, there are significant benefits
that should not be overlooked. CETA has provided access to
public service jobs to large numbers of minority group mem-
bers and other disadvantaged persons who might not otherwise
have had an opportunity for employment. There are also non-
economic advantages for participants, in terms of improved
morale, health, and ability to function in the labor market,
that cannot be evaluated by statistics. (Mirengoff & Rindler,
1978, p. 240)

Both of the above statements, one from the early years of manpower

programs and one from the most recent evaluation of the programs, refer

to the intangible, health oriented benefits In an effort to take some

38
of the sting out of criticisms of poor program performance. However,
there have been no studies designed to support these assertions even
though several writers and researchers have argued that the planning and
effectiveness of training could be significantly improved by research
into mental health and other factors associated with employment training
(Baron & Bass, A., 1968; Borus & Tash, 1970; Ginzberg, 1966; Sheppard,
1972; Smith, E. , 1977; Tiffany, D., et al., 1970; Work in America, 1973).
Comprehensive Employment and Training Act of 1978

Several of the evaluative studies cited above appeared to examine
similar variables in employment training but had differing results.
These contradictory studies Included examinations of job retention (Quinn
et al., cited in Goodman et al., 1973; Roskind, 1975), expectancy theory
(Goodman et al., 1973; Gumpper, 1971), and self-esteem (Baron & Bass, A.,
1968; Beatty, R. , 1975). One possible explanation for this variation in
findings is variation in training programs. Programs funded under CETA
and MDTA are frequently different from each other in specific services
provided trainees. CETA programs in particular are Intended to vary from
program to program to accommodate local labor market requirements.
Therefore, the characteristics of the local CETA program participating
in this study will be included in the discussion of the overall
Comprehensive Employment and Training Act of 1978.

CETA is designed to address the needs of several different groups
of unemployed persons. The Act was written to allow a degree of
responsiveness to local needs and environmental conditions by encour-
aging local control in the development of administrative procedures and
operational structure. The Act does establish general guidelines for
all programs and is divided into eight titles. The discussion will

39

follow the general format of the Act with additional information drawn
from the Master Plan of the participating Prime Sponsor (Northeast
Florida Employment and Training Consortium, 1979).

Title I — Administrative provisions. This Title contains the
administrative provisions for the overall operation of employment training
programs under the Act. CETA programs are administered by the local
governmental agency or agencies which represent a population group of
100,000 persons or more. The administrative body(ies) is(are) called
the Prime Sponsor. l,nien the Prime Sponsor consists of two or more
governmental agencies working together to represent a population group
of 100,000 or more persons they are referred to as a consortium. In the
event a region of the state cannot be formed into a Prime Sponsor area
(such as counties with less than 100,000 population) the state government
serves as the Balance of State Prime Sponsor. The employment training
program participating in this study is a consortium composed of three
north Florida counties.

Also included in this Title are limitations on the time a person
may participate in CETA funded activities. The maximum participation
period is 30 months with individual programs having special limits:
Public Service Employment— 78 weeks. Work Experience--1 ,000 hours, and
classroom training-104 weeks of allowances. Limitations are also placed
on annual salaries earned by CETA participants while in training. The
Title also includes extensive additions and exceptions to these limitations
and administration of the program which are not relevant to this discussion
of the Act.

Title II— Comprehensive employment and training services. This
Title is divided into four parts. Part A is a general introduction

40
and statement of purpose for the Title. Parts B and C cover programs
for training, education, upgrading and retraining, and supportive ser-
vices to prepare economically disadvantaged persons for unsubsidized
employment. Part D provides for long term employment training for
economically disadvantaged, structurally unemployed persons.

Part B of Title II provides for supportive services and training
in two activities: classroom training and work experience. Classroom
training is intended to assist economically disadvantaged persons who
do not have employable skills by providing them with the opportunity
to attend relevant skill training classes. The major forms of assis-
tance are the payment of an hourly allowance for class attendance and
the payment of tuition, fees, and book costs. The employment training
program in this study subcontracts classroom training activities to a
local junior college and an Opportunity Industrial Center in the area.
Classes are offered in programs designed to match the participants'
interests and potentials. Typical classes include welding, auto
mechanics, auto body work, cashiering, clerical skills, bookkeeping,
health care, and the General Education Degree (GED) .

There are two types of work experience activities funded under
Title II-B: On-the-job training (OJT) and Adult Work Experience (AWE).
On-the-job training consists of placing a participant with a private
business which provides the participant with job experience and train-
ing in job skills under a contractual agreement with the employment
training program. The employment training program pays up to 50% of
the expense of training the participant and the business provides the
balance. The business is expected to hire the participant on a fulltime
basis following the contracted training period provided the participant

41
has demonstrated adequate progress. Adult Work Experience is quite
similar to On-the-job training. The major differences are that the
participant is placed with a public or private, non-profit agency, the
training program provides 100% of the participant's wages (usually the
federally established minimum wage), and the training agency is not
expected, though encouraged, to hire the participant at the end of
training. Adult Work Experience participants usually have fewer skills
prior to training than the On-the-job training participants and often
require further training following completion of the Adult Work Experi-
ence program activities.

Title II-C provides for upgrading and retraining programs. The
goal of these programs is to help workers who are underemployed and
economically disadvantaged or who are about to be laid off and have
no expectation of future employment with their current skills. The
employment training program participating in this study does not
currently offer training under Title II-C.

Title II-D provides for Public Service Employment for the economi-
cally disadvantaged, structurally unemployed person. Structurally
unemployed persons are those who are unable to obtain fulltime employ-
ment due to factors inherent in the very fabric and structure of soci-
ety over which they have no control. These factors include racial or
sexual discrimination in hiring practices, limited access to full or
quality education, or lack of mobility to the available jobs. Partici-
pants in this part often possess the basic skills necessary for the
performance of the positions for which they are training. Public Service
Employment training consists of placing the participant with a private,
non-profit or public agency to perform a normal job function in the
agency. The job slot must have a corresponding position in the regular

42
employment structure of the training agency and the participant is paid
a wage comparable to that received by persons in the regular position.
Participants under this part are expected to have been unemployed for a
longer period of time than participants under part II-B (15 of 20 weeks
prior to the application versus seven days) and may train longer (78
weeks versus 1000 hours — 25 weeks). Title II-D also provides for a
limited educational program to expand the skill level of the partici-
pants similar to the classroom training program.

Title II includes provisions for counseling activities as a part of
the training process. Within the participating employment program
participants under Title II receive regular contact with the employment
training program counseling staff. On-the-job training participants do
not receive counseling as a part of the training process. The Adult
Work Experience participants receive two two-hour counseling-job reten-
tion skills training group sessions and two job site visits per month.
The Public Service Employment participants receive one two-hour group
session and one job site visit per month. The classroom training parti-
cipants receive one to two hours of group counseling per week from the
counselors at the training facility. These contacts are designed to
provide training in skills needed to identify, obtain, and retain full-
time unsubsidized employment. These skills include how to relate to a
supervisor, what does it feel like to work, what does it mean to be a
good employee, keeping regular attendance, personal hygiene, how to cope
with stress and anger, and communication skills. The counseling staff
also provides crisis intervention in crises on the job site and in the
participants' personal life as necessary. All contacts are structured
to assist in the completion of training and entry into the world of work.

43

Title III — Special federal responsibilities. This Title provides
for the meeting of special federal responsibilities such as assistance
for minorities, Native Americans, displaced homeraakers, and women. The
participating Prime Sponsor in this study does not currently provide
or subcontract training under this Title.

Title IV — Youth programs. This Title provides for a wide range of
programs designed to meet the needs of economically disadvantaged youth.
Included in these programs are part-time employment opportunities to
encourage youth to remain in school or return to school, summer employ-
ment, and on-the-job training positions in private businesses. Youth
programs are quite similar to the adult programs outlined in Title II-B
in general design. There is less pressure on the training agency to
employ the youth following training, there is more leeway in inappro-
priate behavior, fewer of the youth positions are fulltime, and there
is more counseling contact with the youth.

Title V — National commission for employment policy. This Title
creates the National Commission for Employment Policy, formerly the
National Commission for Manpower Policy. The Commission acts as an
advisory group to the President of the United States in the development
of relevant national policy.

Title VI — Countercyclical public service employment. This Title
provides for a public service employment program to assist persons who
have been unemployed for an extended period of time as a result of the
cyclical variations in the national economy. The program is quite
similar to the Title II-D Public Service Employment program and the
employment training program participating in this study has combined
Title II-D and Title VI training activities in a common department for
administrative purposes. Overall funding for Title VI activities is

44
tied to the national economy and unemployment rates. There are two
types of public service employment training positions funded under Title
VI. One type is identical in structure and function to those provided
in Title II-D. The other type of positions is a part of a special public
works "Project" program. A Title VI Project is an activity sponsored
by local community based organizations (CBOs) to accomplish a specific
task beneficial to the public. Projects have a maximum operational
period of 18 months with an option to extend 18 months. Sponsoring
agencies must be public or private, non-profit and assume complete
operational and administrative responsibility for the project. The
employment training program provides the sponsoring agency with the
wages for the participants, screens participants for CETA eligibility,
provides counseling and some skill training, and monitors the progress
of the project. Projects in the employment training program participating
in this study have included area and school beautif ication, development
of a minority affairs library, rehabilitation of low-income housing, a
Teen theatre program, and general public works.

Title VII — Private sector opportunities for the economically
disadvantaged. This Title authorizes the creation of the Private Sector
Initiative Program (PSIP) and the Private Industry Council (PIC). The
PIC is intended to oversee the PSIP and to develop opportunities within
the private business community for training and employment activities for
the economically disadvantaged population. The PIC and PSIP are mandated
for all Prime Sponsors and serve to greatly expand the On-the-job
training program and related activities.

Title VIII — Young adult conservation corps. This Title provides
for training and employment activities for young people from all social
and economic backgrounds through work on conservation and other projects

45
on federal and public lands. As such this is the only program funded
under the Act which does not require the participants to be economically
disadvantaged or poverty level. The employment training program partici-
pating in this study does not currently offer a YACC program.

All participants follow the same basic track through the employ-
ment training program. The first stage consists of applying to the
program and being selected for training. The second stage is the pre-
training orientation during which the participants complete relevant
paper work and have their Employability Development Plans (EDP) prepared.
The EDP is a statement of what training the participants need to achieve
an adequate level of employability and identifies how that training will
be accomplished. The third stage is the training phase during which the
EDP is implemented. While the EDP often specif ies one training program
it is possible to include more than one program. As an example, a
participant with absolutely no skills may be placed in classroom training
followed by a transfer to a Public Service Employment program position
to complete their training. The final stage of the training process is
job development during which the program assists the participants in
finding fulltime unsubsidized employment.
Summary

This review of research in the area of unemplojrment and work has
shown that the role of work and efforts to help the unemployed are basic
to the well-being of the individual and society. Without sufficient
work opportunities the individual has been shown to experience both eco-
nomic difficulties and a deterioration of health, family, and psycho-
logical well-being. The federal government has recognized this impor-
tance and has assumed increased responsiblity for controlling the
impact of unemployment upon the individual and society. The strength of

46
this responsibility Is reflected in the size and scope of the social
security, employment service, and manpower programs provided by the
government .

The primary goal of manpower programs is to insure an adequate man-
power pool for the economic structure of the nation and to assist the
long-term, economically disadvantaged persons in overcoming blocks to
employment and to improve the quality of their lives. Evaluation of
manpower programs, however, has not indicated an overwhelming success
in achieving the stated goals. In response to criticism generated by
these evaluations supporters of manpower programs have consistently
referred to the intangible psychological and social benefits of training
and employment. Researchers have also indicated a need for the inclusion
of psychological factors in planning and evaluation of manpower programs.
To date, however, there have been no studies which utilized mental
health concepts in an evaluative or planning capacity.

Mental Health

Three concepts of mental health are reviewed in this section for
use in evaluating manpower programs and examining unemployed persons:
self-actualization, psychological androgyny, and locus of control.
Self -Actualization

General research. The wide acceptance of Maslow's (1970a) concept
of self-actualization as a description of psychological health was fol-
lowed by a demand for an objective measure of the concept. Shostrom
(1974) introduced the Personal Orientation Inventory (POI) to meet this
demand. The POI was designed and tested as a measure of the values and
behaviors related to the development of self-actualization as described
by Maslow (Shostrom, 1974). Shostrom reports that the initial normative

47
studies were done on 2607 college freshmen. The initial validation
studies were comparisons of responses from a wide range of samples
including: supervisors, student nurses, service organization volunteers,
college students, high school students, psychiatric in-patients, delin-
quent males, alcoholic males, and psychopathic felons. The POI was found
to consistently discriminate between the groups in the predicted
direction.

King (1974) compared the level of self-actualization within 115
married couples. He found no sex differences in self-actualization as
measured by the POI, suggesting that self-actualizing subjects may
experience a stronger interpersonal attraction for other self-actualizing
persons. Olczak and Goldman (1975) examined attitude similarity and
interpersonal attraction as moderated by self-actualization among 59
college students. The self-actualizing subjects were more attracted to
persons with similar attitudes than were non-self-actualizing subjects.
Furthermore, non-self-actualizing subjects were more attracted to sub-
jects with dissimilar attitudes than were self-actualizing subjects.

Several studies have examined the impact of the college experience
upon levels of self-actualization. Schroeder (1973) compared responses
of 278 male and 290 female college freshmen on the POI at the beginning
and completion of a school year. He found that the female students had
higher levels of self-actualization than the male students and that
male and female subjects increased in self-actualization over the course
of the school year. Scott, S., (1975) reports that residence hall resi-
dents experienced increases in self-actualization more frequently than
non-residence hall students. Furthermore, student assistants in the
residence halls had higher initial levels of self-actualization and

48
exhibited greater increases than regular residents. Female student
assistants were more self-actualizing than male student assistants.

Schroeder and LeMay (1973) compared scores on the POI of students
living in coeducational and single-sex residence halls. Students in
the coeducational residence halls were found to have higher levels of
self-actualization than single-sex halls. Similarly, Brown, R. ,
Winkworth, and Brascamp (1973) reported that coeducational residence
halls were more conducive to increases in self-actualization than single-
sex halls. McCleod (1973) studied college students' perceptions of self-
actualizing characteristics and behavior as moderated by residential
environment. Students in a restrictive environment (fraternity house)
viewed the self-actualizing characteristics as being deviant and rejected
them more frequently than students in a more open environment (a female
residence hall) .

Participation in group experiences has been found to have an impact
on levels of self-actualization. Gilligan (1973) compared responses on
the POI and Omnibus Personality Inventory of selectors and non-selectors
of sensitivity training. His comparisons revealed that those who selected
the sensitivity training scored higher on the POI scales of Inner-
directed support, Existentiality, Spontaneity, and Capacity for Intimate
Contact than non-selectors. Similarly, selectors scored higher on the
Omnibus Personality Inventory scales of Introversion, Estheticism, Com-
plexity, Autonomy, Religious Orientation, and Impulse expression than
non-selectors and lower on Omnibus Personality Inventory scales of
Practical Outlook and Masculinity/femininity than non-selectors.
Montgomery (1975) found that growth group experiences increased levels of
self-actualization among college students without affecting college
retention.

49

Kimball and Gelso (1974) compared shifts in self-actualization and
scores on an ego strength scale (Barron, 1953) following participation
in a marathon growth group. They found that self-actualization increased
following the experience and that ego strength levels were unchanged
after the group experience. Furthermore, ego strength was not related
to changes in self-actualization even though high ego strength and high
self-actualization scores were directly correlated prior to the group
experience. Jones, D., and Medvene (1975) clarified the results obtained
by Kimball and Gelso with the finding that the low ego strength subjects
did not exhibit changes in self-actualization following participation in
a marathon group while high ego strength subjects increased in self-actu-
alization following the group experience.

Olim (1968) urged parents to avoid using enforced conformity, triv-
ialization, and dehumanizing child-raising practices which would inhibit
the child's later development of self-actualization. Hjelle and Smith, G.,
(1975) compared retrospective reports of child-rearing practices and self-
actualization levels among college women. They found subjects reporting
parental practices of encouragement, acceptance, psychological autonomy,
and lax controls had higher levels of self-actualization than subjects
reporting more restrictive child-rearing practices. Rogers, M.,(1970)
compared reports of nuclear family interaction patterns and found that
greater amounts of family participation during childhood were related
to higher levels of self-actualization later in life. Gibb (1967)
reported that a retrospective analysis of the characteristics of self-
actualizing students revealed that the self-actualizing student most
likely was female, from a home where the parents finished high school,
from families with one to three children, had a mother who worked on a

50
fulltime basis, had little or no formal religious training or was not
currently involved in an active religious group, graduated from a high
school with a graduating class between 101 and 500 students, had spent
the first two years of college at a large university, was enrolled in
the College of Liberal Arts, had prior work experience but was not
currently working, and was engaged in extracurricular activities nine
or more hours per week.

Maslow (1970b) stated that present day religion was becoming formu-
listic, restrictive, and hampered the development of self-actualization.
Graff and Ladd (1971) compared self-actualization levels and degree of
religious commitment among 152 protestant males. They reported a strong
inverse relationship between strength of religious commitment and self-
actualization levels. Subjects reporting less religious commitment were
found to be significantly more self-accepting, spontaneous, accepting
of their own aggression, more inner-directed, and less dependent than
subjects with a strong religious commitment. Similar studies using a
Catholic population (Burke, 1973) and a Methodist population (Anderson,
1973) also indicated that endorsement of the more conservative doctrines
of the denomination was related to lower levels of self-actualization.
Reynolds (1970) reported similar results in a study of unrelated and
progressively more conservative religious groups. Hjelle (1975) and Lee
and Piercy (1974) found that among college students the frequency and
regularity of church attendance were inversely related to self-actu-
alization levels. Lee and Piercy also identified a sex difference in
the results with male subjects attending church regularly having lower
levels of self-actualization than females attending church regularly or
subjects of either sex who did not attend church on a regular basis.

51

There have been several studies comparing self-actualization to
responses on psychological measures. Vance (1967) found only limited
agreement between the POI and the Mental Health Analysis Test and
concluded that the two instruments were measuring different dimensions
of mental health. Wills (1974) compared the responses of a sample of
college males and females on the POI and the Tennessee Self-Concept
Scale. He reported the finding that high self-actualizing males were
more open to self-criticism, had a less positive self-concept from a
moral-ethical perspective, and generally felt better about themselves
than low self-actualizing males. High self-actualizing females were
more open to self-criticism, liked themselves better physically, and felt
better about their social interactions than low self-actualizing females.
Overall, the female subjects were found to be more inner-directed than
the male subjects.

Wexler (1974) found that self-actualization was positively related
to an increased desire for new experiences and creativity. Self-actu-
alizing subjects have also exhibited a more holistic and integrated time
sense than non-self-actualizing subjects (Gestinger, 1975). Likewise,
Yonge (1975) reported that increased self-actualization was positively
related to improved time perspectives and creativity.

Foulds and Warehime (1971b) explored the relationship between the
concept of emotional repression and sensitization (Byrne, 1961) and self-
actualization. They found, contrary to expectations, that there existed
a strong, direct relationship between repression of emotions and self-
actualization. Ginn (1974) replicated the Foulds and Warehime study
and found that defensiveness served to moderate the observed pattern
with non-defensive repressors having higher levels of self-actualization
than defensive repressors.

52

Wesch (1971) studied the relationship between self-actualization
and anxiety about death and reported an inverse relationship between the
two dimensions. deGrace (1974) compared self-actualization and anxiety
level. He found that self-actualization was not Incompatible with mod-
erate levels of anxiety.

McClain and Andrews (1969) compared reports of peak experiences
from 139 college juniors and seniors with their responses on the Rokeach
Dogmatism Scale, the California F (Authoritarianism) Scale, and the
16 Pf Test. They found that those students reporting peak experiences
were significantly more open-minded, assertive, intelligent, tender-
minded, forthright, experimenting, self-sufficient, expedient, imagi-
native, and anti-authoritarian than students reporting no peak experiences,
Orraand (1973) compared 140 college students on self-actualization,
purpose in life, and dogmatism. He reported a significant inverse rela-
tionship between dogmatism and self-actualization or a positive purpose
in life.

Specific research. While the studies reviewed in the preceding
section do not represent an exhaustive review of the research on self-
actualization, they are representative of the field. It is important
to note that in all but five of these studies the research sample was
composed of college students. Four of the five exceptions used religious
membership to identify study subjects and one used married couples.
Similarly, samples reported in the POl test manual (Shostrom, 1974) were
composed of college students, persons functioning well in society, or
persons with severe adjustment problems. None of the samples were drawn
from populations with identified employment problems.

53
Lessner and Knapp (1974) compared 40 owners of small businesses
on self-actualization and business orientation. The owners were class-
ified on the basis of definitions developed by Smith, N.,(1967). A
"craft" orientation referred to persons perceiving and reacting to a
limited range of cultural input, narrowness of education, low social
awareness, lack of flexibility in dealing with the social and economic
environment, and a time sense focused on the past and present. The
"merchandising" orientation was related to persons responding to broad
social and cultural stimuli, with a broader education, increased social
awareness, flexibility, and a present and future time perspective. The
results of the study were that merchandising owners were signficantly
higher than craft oriented owners on the Inner-directed, Spontaneity,
Self-acceptance, and Acceptance of aggression scales of the POI.
Ohlbaum (1971) compared self-concepts, values, and self-actualization
levels of professional, career, and non-professional women. Professional
women were those working in occupations requiring advanced education and
career women were those working in occupations not requiring advanced
education but which were self-supporting. Non-professional women were
those without a self-supporting occupation. She found that, in general,
professional and career women had higher levels of self-actualization
than the non-professional women.

Goldstein (1967) examined differences between 81 mental health
clinic outpatients and 77 non-patients on self-actualization, socio-eco-
nomic status, intelligence, and number of ungratified needs. She con-
cluded that socio-economic status was not a key determinant of mental
health while the number of ungratified needs served to predict the degree
of mental health. The quality of family life, mental health and self-

54
actualization were inversely related to the number of ungratified needs.
Socio-economic status was related to identification with parents, mood,
intelligence, and quality of interpersonal relationships, but not mental
health.

Reeves and Shearer (1973) examined the relationship between per-
ceived conformity, race, self-concept, and self-actualization. Five
groups were compared: members of the Black Student Union, persons who
identified themselves as average and conforming, persons identifying
themselves as being non-conforming in private while being publicly
conforming, persons identifying themselves as being non-conforming both
in private and in public, and a general group. All subjects were white
with the exception of the Black Student Union subjects. Subjects identi-
fying themselves as being non-conforming, either in public or private,
had higher levels of self-actualization than the other three groups.
The Black Student Union subjects had lower self-concepts than the other
subjects.

White (1971) compared 100 college sophomore females on several
scales intended to measure self-actualization, locus of control, and
alienation from self and society. Subjects with higher levels of self-
actualization tended to be more internal in their locus of control and
less alienated from self and society than low self-actualizing subjects.

Summary. The concept of self-actualization has been widely
researched and accepted as a measure of mental health. Research has
shown that self-actualization is affected by such factors as sex, family
history, size of family, education, and ungratified needs. It is related
to interpersonal attraction, liberal value systems, college experiences,
lower religious commitment, and sensitivity training. Self-actualization

55
has also been associated with a liberal (merchandising) business orien-
tation, non-conformance, an internal locus of control, and lower levels
of alientation from self and society. Research in the area of self-actu-
alization has been found to have a primary focus on populations which
do not have an identified history of employment problems.
Psychological Androgyny

General research. Early sex-role theory was primarily based upon
gender and conceptualized masculine and feminine sex-roles as being
inversely related on a single, bi-polar continuum (Maccoby & Jacklin,
1974; Roczak, B.,& Roczak, T., 1969). Other theorists have argued that
sex-roles are culturally determined rather than biologically determined
and were not necessarily inversely related (Holter, 1971; Komorovsky,
1950; Mead, 1950; Staples, 1973). Furthermore, sex-role theorists have
argued that the mentally healthy individual is able to engage in both
masculine and feminine behaviors regardless of gender and that it is
possible to integrate masculinity and femininity (Angrist, 1969; Bem,
1974; Block, 1973; Chafetz, 1974; Maslow, 1970a; Singer, 1976; Spence
et al. , 1975).

The focus of sex-role research in recent years has been the examina-
tion of sex-role development and implications. Sternglanz and Serbin
(1974), after eliminating all children's television programs which did
not offer male and female characters, found twice as many male as female
characters and found that the males were consistently more aggressive,
more constructive, and more likely to be rewarded for their behavior than
were female characters. Freuh and McGee (1975) identified a strong rela-
tionship between strength of traditional sex-role endorsement among
children and time spent watching television and that the strength of the
relationship increased as the children grew older. McArthur and Eisen

56
(1976) found that when storybook characters achieved well, children
of the same sex were more likely to increase their own achievement
efforts. Chafetz (1974) reported that characters and performers in adult
movies, books, television shows, and popular songs and song groups
tended to be male or male oriented and that the presentation of roles
for men and women consistently showed the male role as more competent,
more performance oriented, and stronger.

There have been numerous studies assessing differences between male
and female perceptions of masculinity and femininity. Lunneborg, P., and
Lunneborg, C.,(1970) factor analyzed the responses of male and female
subjects on items drawn from the Minnesota Multiphasic Personality Inven-
tory M/F Scale, the Guilford-Martin M Scale, the Gough Fe Scale, and the
California Personality Inventory Fe Scale. Eleven factors were found to
distinguish women from men: Feminine interests. Rejection of adventure.
Emotional sensitivity, Philistine versus artistic interests. Masculine
interests. Neurotic symptoms. Self-confidence, Indifference, Social
adequacy, Extraversion, and Unsociable non-conformity.

Engel (1966) factor analyzed responses on the Strong Vocational
Interest Blank, the Minnesota Multiphasic Personality Inventory, the
Terman-Miles Attitude and Interest Analysis Test, the Gough Fe Scale,
and the Frank Drawing Completeness Test. Women were found to score
higher than men on the factors: Aversion to mechanical and scientific
occupations; Interest in domestic, nurturant, social, and journalistic
endeavors; Cultural, aesthetic, and verbal interests; Aversion to
detailed work; and Aversion to finance and business math.

Stewart and Winter (1974) compared Thematic Apperception Test
results of self-defined women planning self-supporting careers and

57
socially-defined women planning only marriage and family related activity.
Self-defined women were signficantly more likely to perceive causality
than feelings, see themselves as their own agent, have fathers who were
self-employed, have no older brothers, have mothers who worked, and have
majors in masculine fields such as law, medicine, or engineering. Women
who were socially-defined were more likely to perceive feelings than
causality, see themselves as passive, have a father who worked in a
bureaucracy and a mother who did not work, have older brothers, and have
a feminine major.

Helland (1973) found that adolescent boys were more interested in
job aspirations and expectations while adolescent girls were more inter-
ested in numbers of dates and dating patterns. Banikotes, P., and
Banikotes, F., (1972) found that both male and female college students
considered liberated women to be more intelligent, less moral, and more
aware of current events than other women.

Shemberg and Leventhal (1968) found that, on the Edwards Social
Desirability Scale, male subjects exhibited a stronger desire to avoid
appearing weak than female subjects. Derlega and Chaikin (1976) reported
that expressive males and quiet females were seen as less adjusted than
quiet males and expressive females, regardless of the sex of the rater.
Okman (1973) found that, while there were high levels of agreement
between male and female students on the basic content of sex-role
descriptions, the males were more likely to use physical and sexual words
to describe sex-roles while females were more likely to use emotional
and interpersonal words. Garske (1975) found that undergraduates tended
to perceive female graduate students as more masculine, than non-college
or undergraduate females. McGovern, Ditzian, and Taylor (1975) found

58
that persons In a helping role perceived male help seekers as strong,
brave, leader, dominant, competitive, and happy while female help seekers
were perceived as weak, cowardly, attractive, honest, and sad.

There have been several studies which asked subjects to complete
sex-role instruments for themselves, for their own-sex in general, for
the opposite-sex in general, for their own ideal-sex, for the opposite-
sex ideal-sex, and how the opposite-sex would describe their own ideal-
sex (Lunneborg, P., 1968; McKee & Sherriffs, 1959; O'Leary & Depner, 1975;
Rosenkrantz, Vogel, Bee, Broverman, I., & Broverman, D., 1968; Steinmann,
1958; Steinmann, Fox, & Farkas, 1968). While there are some minor vari-
ations in the instructions and results of these studies, the general
conclusions have been consistent: the masculine sex-role related
characteristics are significantly different from the feminine sex-role
characteristics. A study typical of this body of research (Rosenkrantz
et al., 1968) found that the sex-role traits valued in males were active,
aggressive, independent, unemotional, dominant, competitive, logical,
worldly, direct, adventuresome, and self-confident. Female valued traits
were described as talkative, tactful, gentle, aware of others' feelings,
religious, quiet, and neat. Furthermore, all of these studies have found
that the masculine traits are more socially desirable than the feminine
traits .

Several studies have examined attitudes associated with traditional
sex-roles. Ellis and Bentler (1973) analyzed responses on the Sex-Role
Stereotype Questionnaire developed by Rosenkrantz et al. (1968) and the
Bentler Psychological Inventory. They found that endorsement of tradi-
tional sex-role stereotypes increased as the self-perceptions of male
and female subjects deviated from their perceptions of the opposite sex.
They also found that disapproval of the traditional sex-roles among both

59
male and female subjects was associated with liberal, extra-legal, non-
religious, and low status-seeking values. Among the female subjects the
rejection of traditional sex-roles was also associated with increased
intelligence.

Dunbar, Brown, M. , and Amoroso (1973) found that strong endorsement
of conservative sex-roles among male subjects was directly correlated
with strong anti-homosexual attitudes, feelings of personal sex-guilt,
rejection of many heterosexual sex-practices, and rejection of any male
who did not adhere to their more constricted definition of male appro-
priate behavior. Langston (1975) found that females generally had
stronger feelings of sex-guilt and that more of their behavior was moti-
vated by sex-guilt. Furthermore, conservative behavior patterns, e.g.,
assertiveness in males and passivity in females, were associated with
elevated levels of sex-guilt. Morado (1973) and Woudenberg (1973) both
found that sexist attitudes and conservative sexual behavior were
strongly associated with racism, high valuation of material goods, sex-
guilt, depersonalization of sex, and an emphasis on personal and social
control of sexual expression.

There have been several studies which assessed perceptions of sex-
roles and mental health (Broverman, I., Vogel, Broverman, D., Clarkson, &
Rosenkrantz, 1972; Ginn, 1975; Heilbrun, A., 1968; Kravetz, 1976; Maslin
& Davis, 1975; Zeldow, 1976). These studies have consistently demonstrated
the same conclusions: that masculine sex-role traits are considered more
desirable and healthier by society than feminine sex-role characteristics.
As an example, Broverman, I.,et al. (1972) asked male and female mental
health professionals to describe the healthy male, the healthy female,
and the healthy adult. The responses were extremely consistent with

60
male and female professionals agreeing that the characteristics of the
healthy male and the healthy adult were quite similar and that the healthy
female was quite different. In general, the research has indicated that
as long as the bi-polar, single dimension definition of masculine and
feminine sex-roles is used it is not possible for females to behave in
a healthy female and healthy adult manner at the same time.

The concept of psychological androgyny was introduced in response
to these findings. Bem (1974) based the development of the Bem Sex-
Role Inventory on two assumptions:

(A) Largely as a result of historical accident, the culture
has clustered a quite heterogeneous collection of attributes
into two mutually exclusive categories, each category consid-
ered more characteristic and more desirable for one or the
other of the two sexes. These cultural expectations and
prescriptions are well known by virtually all members of the
culture. (B) Individuals differ from one another in the
extent to which they utilize these cultural definitions as
idealized standards of femininity and masculinity against
which their own personality and behavior are to be evaluated.
(Bem, 1979, p. 1048)

The Bem Sex-Role Inventory (BSRI) was designed to measure the
strength with which the individual endorsed socially desirable masculine
and feminine traits. These traits have been judged to be significantly
more desirable for one or the other sex. An individual endorsing an
equal number of masculine and feminine items was defined as psycho-
logically androgynous. In response to comments from Spence et al. (1975),
this definition was later modified to define an androgynous individual
as one endorsing an equal and high number of masculine and feminine
items (Bem, 1977). Other instruments introduced to measure psychological
androgyny are the Personal Attributes Questionnaire (Spence et al.,
1975), a scale adapted for use with the Personal Research Form (Berzins,
Welling, & Wetter, 1978), and a subscale for the Adjective Checklist
(Heilbrun, A. , 1976).

61
General research on psychological androgyny has found that the
androgynous individual is not influenced as strongly as non-androgynous
individuals by traditional sex-role stereotypes and expectations. Bern
(1975) compared masculine and feminine sex-typed subjects with androgynous
subjects in tests of willingness to engage in acts of independence and
nurturance. She found that masculine sex-typed and androgynous subjects
freely engaged in the independence activity more frequently than the
feminine sex-typed subjects. Feminine sex-typed and androgynous subjects
engaged more freely in the nurturance activity than the masculine sex-
typed subjects. She also found that feminine females hesitated in
engaging in the nurturance activity of playing with a kitten. This
hesitation was found to disappear when the nurturant activity was helping
a human (Bem, Martyna, & Watson, 1976). The relationship between sex-
typing and willingness to engage in same-sex or cross-sex behavior has
been consistently replicated (Bem & Lenney, 1976; Kelly et al., 1977;
Spence et al., 1975; Wiggins & Holzmuller, 1978).

Kelly and Worrell, L., (1976), in a study similar to that done by
Hjelle and Smith, G., (1975), identified a strong relationship between
parental childrearing practices and psychological androgyny. They found
that subjects reporting parental practices encouraging non-traditional
behavior in the child, e.g., emotional sensitivity among boys and
achievement among girls, exhibited a greater incidence of androgyny
than subjects reporting parental encouragement of the more traditional
patterns.

Specific research. Several studies have found that there exists
a strong positive relationship between freedom from restrictive sex-role
stereotypes and psychological health (Bem, 1974; Broverman, I., Broverman,

62
D., Clarkson, Rosenkrantz, & Vogel, 1970; Chafetz, 1974; Heilbrun, C,
1973; Rosenkrantz et al., 1968). Spence et al. (1975) and O'Conner,
Mann, and Bardwick (1978) compared responses on the Personal Attributes
Questionnaire and a social self-esteem instrument. Androgynous subjects
exhibited higher levels of self-esteem than sex-typed or undifferen-
tiated subjects. Gump (1972) reported a strong inverse relationship
between strength of traditional sex-role attitudes and ego strength.

The relationship between psychological androgyny and psychological
adjustment is not, however, clear cut. In recent literature reviews
Kelly and Worrell, J., (1977) and Worrell, J., (1978) identified patterns
in the research suggesting that the development of androgyny may have
dramatically different implications for males and females. The process
of developing sex-role flexibility may mean making qualitatively differ-
ent decisions and adjustments for men and women. The social pressures
and consequences against women engaging in masculine behavior, even when
situatlonally appropriate, appear to be stronger than those associated
with males engaging in feminine behavior. Bern (1977), for example, in
an exploration of the median split sex-type classification method pro-
posed by Spence et al. (1975), found that self-esteem among female sub-
jects was related to high scores in both the masculine and feminine
dimensions while high self-esteem among the male subjects was related
only to high scores in the masculine dimension. She also found a direct
correlation between masculine males and conservative scores on the Spence
and Helmreich (1972) Attitudes Toward Women Scale. No significant corre-
lations were obtained between the BSRI and the Internal-External Scale
(Rotter, 1966), the Attitudes Towards Problem-Solving Scale (Carey, 1958),

63
the Mack IV Scale (Christie & Geis, 1970), or the Self-disclosure Scale
developed by Jourard (1971) .

Deutsch and Gilbert (1976) compared scores on the Bell Adjustment
Inventory with descriptions of perceived-self sex-role, ideal-self sex-
role, ideal-self sex-role for the opposite sex, and the opposite sex's
description of own-sex ideal-self sex-role. They found that masculine
males scored higher on the adjustment instrument than androgynous males
while feminine females scored lower on the adjustment inventory than the
androgynous females. Furthermore, while the male descriptions of the
various sex-roles tended to agree with each other regardless of the sex-
role being described, the female descriptions showed a wide divergence
between perceived-self sex-role, ideal-self sex-role, and perceptions of
how males perceived the female ideal-self sex-role. They concluded that,
at least in terms of social utility and effectiveness, the masculine sex-
type was more desirable than the feminine sex-type, with androgyny some-
where between the two.

Wiggins and Holzmuller (1978) compared male and female subjects on
responses to an extensive interpersonal interaction questionnaire which
included an androgyny scale. They identified eight basic dimensions of
interpersonal behavior: dominant-ambitious, arrogant-calculating, lazy-
submissive, cold-quarrlesome, aloof -introverted, unassuming-igenious,
warm-agreeable, and gregarious-extraverted. Each subject had a profile
based upon their composite scores in each dimension. It was found that
each sex-type group had unique profile patterns, suggesting that the
different sex-types were associated with different interpersonal styles.
Furthermore, the androgynous males had a relatively flat profile, while
the androgynous females exhibited a profile which was a mirror image of the

64
female feminine sex-type subjects' profile. Wakefield, Sasek, Friedman,
and Bowden (1976) found that high masculinity scores on the MMPI M/F
Scale and the Omnibus Personality Inventory were positively correlated
with androgyny while high femininity scores were not correlated with
androgyny.

Jones, W. , Chernovetz, and Hansson (1979) conducted a study in
which they compared responses of 1404 college students on the BSRI and
instruments measuring adjustment, including the Women's Liberation and
Attitudes Scale, the Internal-External Scale, tests of neuroses, alco-
holism, political awareness, and a self-report history. They concluded:

The most succinct description of the present findings is that
the more adaptive and flexible, unconventional, and competent
patterns of responding occurred among the more masculine sub-
jects independent of gender.... In one regard this is not
surprising, since the items that comprise the masculinity sub-
scale have the underlying commonality of being instrumental in
nature, that is, the ability to effectively and efficiently
accomplish objectives. Similarly, most of the vairables
examined involved stereotypically masculine endeavors. (p. 311)

The instrumental-expressive dimension noted by Jones, W., et al.

(1979) has been further established in several factor analytic studies

of the BSRI. These studies have consistently identified three, and

occasionally four, factors using a wide range of populations: 253 male

amd female industrial workers, 36 police officers, and 36 unemployed

housewives (Gaudreau, 1977); 419 undergraduate and 121 graduate students

(Gross, Batlis, Small, & Erdwins, 1979); and 1464 graduate students

(Pedhazur & Tetenbaura, 1979). These factors were most commonly labelled

as: Emotional and Interpersonal Sensitivity, consisting of items drawn

from the Femininity scale and the Social Desirability scale of the BSRI;

Assertiveness, consisting of items drawn from the BSRI Masculinity

65
scale; Self-sufficiency, consisting of items from the Masculinity scale;
and Gender, consisting of the two items — "masculine" and "feminine." Bem
(1979) notes that these factors are not inconsistent with the theoretical
foundations of the concept of androgyny and that the only factor not
predicted by the theory was the Gender factor.

Studies examining the relationship between psychological androgyny
and self-actualization, however, have had unambiguous results. Nevill
(1977) compared responses of a socially stratified sample on the POI,
the BSRI, and the Tennessee Self-Concept Scale. The androgynous subjects
scored higher on the POI and the Tennessee Self-Concept Scale than did
the sex-typed subjects. Similarly, Cristall and Dean (1976), Ott (1976),
and Pettus (1976) compared responses of college students on the BSRI,
the POI, and measures of anxiety, sex-role attitudes, and intelligence.
All three studies demonstrated a direct relationship between self-actu-
alization and androgyny. Other results included a positive relationship
between androgyny and an internal locus of control (Ott, 1976) and
between androgyny and low anxiety and non-traditional sex-role attitudes
(Pettus, 1976). No significant relationships were identified between
intelligence or social desirability and androgyny (Pettus, 1976).

In a study of liberal and conservative female responses on the Per-
sonal Attributes Questionnaire and the POI, Hjelle and Butterfield
(1974) concluded:

The results of this study convincingly indicate that college
age females holding pro-feminist (liberal) attitudes vis-a-vis
their social and sexual roles exhibit a markedly higher level
of personal growth than their peer counterparts expressing
tradition bound attitudes. Specifically, these findings
suggest that pro-feminist subjects perceive themselves as
relying more confidently upon their own internal norms without
seeking constant support from others for self-validation. (p. 229)

66

Sex-role definitions have been found to associate masculine sex-
role behavior with an internal locus of control and feminine sex-role
behavior with an external orientation. House (1974) found that female
subjects tended to avoid an appearance of competition and were externally
oriented while male subjects activitely engaged in competition and had
an internal orientation. Hochreich (1975) found that, although the male
ideal-self descriptions among college students were internally oriented
and female ideal-self descriptions were externally oriented, perceived-
self descriptions were not significantly different on the locus of con-
trol dimension. She concluded that, while there existed clear stereo-
types which were markedly different in locus of control, the subjects
did not uniformly adhere to them on a day-to-day basis.

Research on relationships between sex-roles and vocational choices
has focused on the impact of traditional sex-role endorsement on female
vocational choices. Altman (1975) found that the mother's occupational
history and satisfaction with the work significantly influenced the
subject's adult life choices. If the mother did not work then her satis-
faction with the housewife role influenced the subject's willingness to
assume that role during her adult life. Darley (1976) argued that the
vocational choices open to females are usually limited by societal
pressures. As noted above, Stewart and Winter (1974) found that self-
defined women were more likely to engage in masculine occupations while
socially-defined women were more likely to pursue feminine occupations.
Vogel, Rosenkrantz, Broverman, I., Broverman, D. , and Clarkson (1975)
found that non-traditional sex-role orientations were directly related
to a desire for more education, fewer children (if any), and pursuit of
a vocation, even if there were children.

67

Moreland, Harren, Krimsky-Montague, and Tinsley (1979) compared
male and female college student responses on the BSRI and the Harren
Assessment of Career Decision Making. They found that among men and
women, the androgynous subjects were more developed in their college
major and occupational choices and were more rational in their career
decision making processes than were the sex-typed subjects.

Summary . Several observations may be drawn from the literature
on sex-roles and psychological androgyny discussed in this section.
First, there exist culturally accepted stereotypes of masculinity and
femininity. Second, the masculine stereotype is identified as the most
socially desirable and mentally healthy of the two sex-roles. Third,
strict adherence to these sex-roles has been found to be associated with
increased incidence of sex-guilt, racism, authoritarianism, and dog-
matism. Fourth, work and vocational choices are influenced by sex-role
orientation and working has traditionally been considered a masculine
activity. Fifth, psychological androgyny appears to describe a blending
of the more desirable characteristics of both sex-roles. Sixth, androgyny
has been strongly associated with self-actualization while masculinity
has been associated with social adjustment. Seventh, the Bem Sex-Role
Inventory has been found to consist of three major factors which
serve to consistently discriminate between persons who have developed
a level of sex-role flexibility and those endorsing the more traditional
sex-roles. Finally, as with research on self-actualization, androgyny
has yet to be applied on an in-depth basis to persons with employment
problems, or persons outside of the college environment (Bem, 1979;
Locksley and Colton, 1979).

68
Locus of Control

General research. Rotter, Chance, and Phares (1972) identified as
their principal unit of investigation in the development of the Social
Learning Theory the interaction between the individual and the environment.
They focused on learned behavior, postulating that the person's prior
experiences influence what is learned from subsequent experiences. One
of the key concepts identified as predicting behavior in any given situ-
ation was the expectancy, based upon prior experiences, that one's behav-
ior might have an impact on the outcome of the situation. If the indi-
vidual believed that he/she could have an impact on his/her environment
and acted as if this were the case, then the person was defined as having
an internal locus of control. Conversely, an external locus of control
referred to the belief that one is subject to external forces and is
unable to exert control over the environment.

The Internal-External Scale was introduced by Rotter (1966) as an
instrument to assess the locus of control orientation of the individual.
This scale has been used to study the relationship between locus of
control and many different elements in the personality. Adams-Weber
(1969) found that subjects with an internal orientation were more likely
to view punishment as an appropriate consequence of an "immoral act"
than external subjects. Phares (1965) found that internally oriented
students were able to exert a greater amount of influence on the opin-
ions of mid-range (neither internal or external) students than were
externally oriented students.

Several studies have examined differences in degrees of willingness
to work at a difficult task. They have found that when the possibility
of success at a difficult task was presented as contingent upon the

69
skill and effort applied by the subject, internally oriented subjects
attempted the task with significantly higher levels of energy and dili-
gence than externally oriented subjects (Cohen, Rothbart, & Phillips,
1976; Newmanm 1977; Phares, 1957; Rotter & Mulray, 1965). Similarly,
internally oriented subjects were found to prefer a greater number of
internally controlled social activities than the externally oriented
subjects (Clark, 1976).

An internal locus of control has been found to be a characteristic
of individuals engaging in, or committing to engage in, voluntary action.
Gore and Rotter (1963) found that internally oriented black students were
more likely to commit to anti-racism activities than were externally
oriented black students. Strickland (1965), in a follow-up study to
the Gore and Rotter study, found that members of civil rights groups
were more internally oriented than members of non-activist groups. Kinder
(1976), while finding no differences in orientation between students
who volunteered to participate in psychological experiments, found that
the internally oriented subjects were more likely to actually complete
the experiment than were the externally oriented volunteers.

Bryant, B., and Trockel (1976), in a study similar to the Hjelle and
Smith, G., (1975) and Kelly and Worrell, L., (1976) studies, examined the
effects of early childhood history upon the locus of control orientations
of college women. Externally oriented women were found to recall a higher
frequency of dramatic life changes during their early childhood years
than recalled by internally oriented women. Conversely, internally ori-
ented women were found to recall the greatest frequency of their stressful
life changes during their high school years.

70
Specific research. Several studies have examined the relationship
between locus of control and psychological adjustment. A series of
studies used the concept of locus of control as a measure of alienation
or perceived powerlessness (Seeman, 1959, 1963, 1972; Seeman & Evans,
1962). Seeman (1959) defined alienation as "the expectancy or proba-
bility held by the individual that his own behavior cannot determine the
outcomes or reinforcements he seeks" (p. 784). He foind that subjects
who were institutionalized for health reasons (Seeman & Evans, 1962) or
for legal reasons (Seeman, 1963) were less likely to seek and use accurate
information pertaining to their status in the institution if they were
alienated or perceived themselves as having no control. Seeman (1972)
concluded that the powerless individual would be less likely to engage
in organizations which increase control over the environment (e.g., labor
unions), less likely to engage in planned action, more likely to engage
in short-term, unplanned protests (e.g., walk off the job when angry
with the supervisor), less likely to learn control relevant information
about the environment , and more likely to feel alienated if a member of
a minority group.

Efran (cited in Rotter, 1966) found that internal high school
students were more likely to repress failure experiences. Butterfield
(1964) found that internal subjects were less likely to experience
debilitating anxiety in test situations. Shybut (1968) reported that
the more disturbed psychiatric patients in a hospital setting were more
external and tended to have a shortened time sense. Externally oriented
students exhibited a higher suicide potential than internally oriented
students, especially among the female subjects (Williams, C.,& Nickels,
1969). Wolk (1976) reported that residents of low constraint (open-

71
entry, open-exit) retirement communities were more internally oriented
than residents of high constraint (nursing home) environments. Finally,
Chandler (1976) found that externally oriented students had lower self-
concepts, lower self-acceptance, and higher perceived-self ideal-self
discrepancies than internally oriented students.

Studies comparing locus of control and self-actualization are few
in number. Wall (1970) failed to find a significant relationship
between the POX Inner-directed scale and the Internal-External Scale,
although positive correlations were obtained between an internal locus
of control and the POI Self-regard, Self-acceptance, and Nature of Man
scales. Bass, B.,and Stek (1972), however, failed to find any signifi-
cant correlations between the POI and locus of control measures.
Lambert et al. (1976) factor analyzed college student responses on the
Internal-External Scale, the POI, the Tennessee Self-Concept Scale, and
Kohlberg's Scale of Social Problems. They identified three major
factors: Mental health and effective living consisting of items from
the POI, Internalized values and principles of justice consisting of
items from the POI and Kohlberg's Scale of Social Problems, and Locus
of control consisting of Internal-External Scale items. They concluded
that while there were several strong correlations among the different
instruments, they appeared to measure different elements of mental
health and the substitution of one instrument for another would result
in a loss of information.

In addition to the studies of the relationship between locus of
control and sex-roles discussed in the preceding section, two other
studies warrant comment. Fox, L.,(1976) compared responses of black and
white freshmen women on a sex-role scale and the Internal-External Scale.
She found no significant differences in locus of control by race or sex

72
and suggested that a wider age range would reveal sex-role and racial
differences in locus of control. Foster (1975) compared responses
of subjects ranging in age from 20 to 40 years or more on the Internal-
External Scale, the Manifest Anxiety Scale, and a sex-role attitude
inventory. She found that an internal orientation was significantly
related to a lower anxiety level and a strong endorsement of masculine
attitudes.

Research has demonstrated a strong relationship between race and
locus of control. Battle and Rotter (1963) compared black and white
students in the sixth and eighth grades from upper and lower socio-eco-
nomic brackets of both sexes. They found that the black students were
more externally oriented than the white students and that the lower-
class blacks with high intelligence were more externally oriented than
middle-class students of either race. Lefcourt and Ladwig (1965) using
the Powerlessness Scale (Dean, 1961) compared locus of control among
black and white prisoners matched for age, socio-economic background,
intelligence, and reason for Incarceration and found that the black pris-
oners were more external than the white prisoners. Lefton (1968) found
that among auto workers from high and low pay categories increased pay
was associated with decreased feelings of anomie and increases in inter-
nal orientation. He also found that among the high pay group the black
workers were more external and had higher anomie scores than the white
workers. Hall, Joesting, and Woods (1977) studied students from predomi-
nantly black and white universities and found evidence to suggest that
there existed a sense of powerlessness against a hostile environment
among the black and female subjects.

73

Gurin, P., Gurin, G., Lao, and Beattie (1969) factor analyzed

responses on an extended locus of control instrument administered to a

black population. They found that subjects with higher job aspirations

or achievements were more external in orientation than the subjects with

lower aspirations. Further analysis of the data identified a "blame

the system" pattern among the more achievement oriented subjects:

To summarize, our results indicate that when internal-external
control refers to Negroes' conceptions of causes for their
condition as Negroes, and these conceptions are related to
more innovative coping criteria, it is the external rather
than the internal orientation which is associated with the
more effective behaviors. When an internal orientation
implies self-blame as a Negro, it also seems to involve a
readiness to accept traditional restraints on Negro behavior,
(p. ^7).

Locus of control has been studied with relation to work more exten-
sively than the other mental health concepts being considered in this
study. Hartley (1976) reported that among college officials an internal
locus of control was significantly related to increases in reported job
satisfaction. Andrisani and Nestel (1976), in a longitudinal study,
found that internally oriented subjects were more likely to have had
higher occupational achievement, higher hourlv earnings, higher levels
of job satisfaction, higher achievement of occupational status, and
higher increases in annual earnings than subjects with an external locus
of control. Internally oriented subjects were also more likely to have
re-entered the labor force if they had left it than were externally
oriented subjects. Kimmons and Greenhaus (1976) reported trends towards
positive correlations between an internal locus of control and the
likelihood of reporting job satisfaction when autonomy, job feedback,
performance rewards, and work involvement were present in the work situ-
ation regardless of gender. Oldham (1976) found a direct relationship

74

between internality and an increased expenditure of effort in work,

a desire for growth in work, and supervisory and co-worker satisfaction

with the quality of the subjects' work performance.

Locus of control has also been related to socio-economic status.
Levinson, P., (1964) found that welfare recipients expressed evidence of
alienation, feelings of resignation, helplessness, pessimism, passivity,
and an inability to mobilize themselves to take corrective action.
Battle and Rotter (1963), as noted earlier, identified a pattern of
lower-class blacks with high intelligence being more externally oriented
than middle-class subjects of either race. Franklin (cited in Rotter,
1966), using a nationally stratified high school sample, found a direct
relationship between more internal scores and higher socio-economic
class. Crandall, V. C, Katkovsky, and Crandall, V. J., (1965) reported
that age, size of family, and ordinal position in the family served to
predict internality better than socio-economic status. They found that
older subjects had higher internality scores, with females scoring higher
than males, and that children of smaller families, or the first born
child, had higher internality scores.

Summary . In general, an internal locus of control has been shown
to be associated with more effective functioning. Individuals with an
internal orientation have been found to be more adaptive to unusual and
threatening situations and better able to obtain and utilize control
relevant information about the environment. Internal locus of control
has been associated with improved psychological functioning, a more open-
mind, non-traditional sex-role values and perceptions, childhood stress
experiences during high school years rather than earlier, smaller family
size, being first born, sex, and age. Blacks tend to be more external
than whites though among achievement oriented blacks this indicates an

75
improved ability to succeed while an internal orientation has been
associated with a willingness to accept the status quo of racial
inequality.

Summary

The literature reviewed in this chapter permits several observations.
As regards the concept of work, working is clearly significant to those
who work, providing them not only with an income for survival but also
with self-definition and meaning to life. Lack of satisfaction on the
job has been shown to result in lower self-esteem and a general dissatis-
faction with life. Job satisfaction has been seen to be based not only
on monetary factors but more importantly upon the worker's sense of work
involvement and the psychological rewards and social acceptance derived
from the work setting.

Persons without work for extended periods of time have been found
to have an increased incidence of mental illness. Evidence was found to
suggest that it may be possible to develop other sources of self-esteem
and psychological well-being outside of the work setting. Manpower
programs have been designed to address the problems of unemployment in
an effort to both help the unemployed persons and meet the nation's
manpower needs. National manpower programs have not, however, been
adequately evaluated in terms of the psychological impact of training.

The unemployed population has not been evaluated in an in-depth
manner on the basis of mental health measures of self-actualization,
psychological androgyny, and locus of control. Self-actualization is a
basic indicator of mental health, providing a measure of overall psycho-
logical adjustment and growth towards mental health. Psychological
androgyny is a concept of sex-role flexibility resulting from the

76
integration of the masculine and feminine sex-roles. Locus of control
is a measure of the individual's beliefs about his/her ability to
affect the outcomes and reinforcements of a situation. Overall these
concepts have been seen to be affected by such factors as age, race,
family position, sex, religious affiliation, socio-economic status, and
criminal or psychiatric records. Together these three concepts serve
to describe mental health for the purposes of evaluating the mental
health characteristics of unemployed persons.

CHAPTER III
PROCEDURES
Overview

This study was designed to examine the mental health character-
istics of long-term unemployed persons and to investigate the effects of
employment training on these mental health characteristics. This study
was accomplished through the performance of four comparisons of subjects
drawn from the clientele of an emplojrment training program in north
Florida funded under the Comprehensive Employment and Training Act of
1978.

The dependent variables on which these comparisons were made were
three measures of mental health: the Personal Orientation Inventory
(POI), the Bern Sex-Role Inventory (BSRI), and the Internal-External Scale
(lES). Also used in the comparison of subjects and subject groups were
seven independent variables consisting of the demographic characteristics:
age, sex, race, education, religious affiliation, family history,, and
length of unemployment prior to beginning the employment training
program.

The first comparison was on the basis of training status of the
subjects. Group 1 (Intake) was composed of all clients of the employment
training program beginning training during the months of April and May,
1980, in either the Adult Work Experience or Public Service Employment
programs. Group 2 (Training) was composed of all clients of the
employment training program who had completed four months of training

77

78
during the months of April and May, 1980, in either the Adult Work
Experience or the Public Service Employment programs.

The second comparison consisted of comparing subjects on the basis
of the component or program in which they were participating. Program 1
(Adult Work Experience) was composed of all subjects, regardless of
training status, participating in this study and receiving training in
activities funded under the Adult Work Experience section of Title II-B
of the Comprehensive Employment and Training Act of 1978. Program 2
(Public Service Employment) was composed of all subjects participating
in this study and receiving training in activities funded under Title
II-D or Title VI of the Comprehensive Employment and Training Act of
1978 regardless of the subjects' training status.

The third comparison of this study was an examination of the vari-
ance in the scores of the subjects in this study on each of the dependent
mental health measures. The scores of the Combined Group on each of the
mental health measures was tested for a significant relationship with the
independent demographic variables.

The final comparison consisted of comparing the mean scores of the
Combined Group on each of the mental health measures with the mean scores
of subjects reported in selected validation studies for each of the mental
health measures.

Hypotheses

Based upon the research questions and the literature the following
hypotheses were developed for examination in this study:

Hypothesis 1. No differences exist in self-actualization character-
istics as a result of training in the CETA Adult Work Experience program
and the CETA Public Service Employment program between Intake subjects

79
just beginning training (Group 1) and Training subjects who have
completed four months of training (Group 2) as measured by the Personal
Orientation Inventory.

Hypothesis 2. No differences exist in psychological androgyny
characteristics as a result of training in the CETA Adult Work Experience
program and in the CETA Public Service Employment program between
Group 1 subjects and Group 2 subjects as measured by the Bem Sex-Role
Inventory.

Hypothesis 3. No differences exist in locus of control as a result
of training in the CETA Adult Work Experience program and the CETA
Public Service Employment program between Group 1 subjects and Group 2
subjects as measured by the Internal-External Scale.

Hypothesis 4. No differences exist between subjects in Program 1
(Adult Work Experience) and subjects in Program 2 (Public Service
Employment) in self-actualization as measured by the Personal Orientation
Inventory.

Hypothesis 5. No differences exist between Program 1 subjects and
Program 2 subjects in psychological androgyny as measured by the Bem
Sex-Role Inventory.

Hypothesis 6. No differences exist between Program 1 subjects and
Program 2 subjects in locus of control as measured by the Internal-
External Scale.

Hypothesis 7. No relationships exist between variance in self-
actualization, psychological androgyny, and locus of control character-
istics of the combined subject groups and the independent demographic
characteristics of age, sex, race, education, religious affiliation,
family history, and length of unemployment prior to participation in
employment training.

80
Subjects

Subjects for this study were drawn from the clientele of an
employment training program in northern Florida funded under the Compre-
hensive Employment and Training Act of 1978. All subjects beginning
training in either the Adult Work Experience or Public Service Employment
programs during the months of April and May, 1980, were invited to
participate in this study as a part of the Intake group (Group 1).
None of the subjects declined to participate. Of the 113 possible
subjects tested for inclusion in Group 1, six were eliminated due to
improperly completed test packages and eleven were eliminated as a
result of matching the Intake and Training groups on the independent
variable of education. The final Group 1 was composed of 96 subjects
(31 males, 65 females). All subjects were tested during the second
half of their pre-training orientation.

All employment training program clients beginning training during
the months of December, 1979, and January, 1980, and therefore com-
pleting four months of training during the months of April and May,
1980, were invited to participate in the study as a part of the Training
group (Group 2) . Approximately 160 clients were invited to attend the
testing sessions. Of the 135 possible subjects attending the testing
sessions for inclusion in Group 2, six declined to participate and
returned to work, ten were subsequently eliminated due to improperly
completed test packages, and Ik were eliminated as a result of matching
the Intake and Training groups on the independent variable of education.
The final Group 2 was composed of 95 subjects (30 males, 65 females).
All subjects were tested during a special testing session organized by
the employment training program staff for this purpose during the two
weeks following the completion of the subject's fourth month of training.

81
Of the total 191 subjects in this study, 69 were participating in
Program 1 (Adult Work Experience) , eight of whom were males and 61 of
whom were females. Program 2 (Public Service Employment) contained the
remaining 122 subjects (53 males, 69 females). The final cell size
distribution of the subjects is presented in Table 1 below.

Table 1
Cell Size of Comparison Groups

Program 1 (Adult
Work Experience)

Program 2 (Public
Service Employment)

Total

Group 1
(Intake)

35

61
96

Group 2
(Training)

34

61
95

Total

69

122
191

Group 1 (Intake)

All CETA participants complete the application process prior to
beginning the pre-training orientation. This process includes filling
out the CETA application, assessment of basic skill levels, participating
in a vocational assessment to identify vocational interests and goals,
and selection to a CETA employment training program. The person moves
from applicant status to CETA participant status and begins to receive
a wage for time spent in training activities at the beginning of pre-
training orientation. Pre-training orientation consists of completion
of all personnel and payroll paperwork necessary for participation in
training, explanation of the rules and regulations governing partici-
pation in CETA employment training, assignment to a training job site,
development of the clients' Employability Development Plan (EDP) , and

82
discussion of job relevant attitudes such as punctuality, personal
hygiene and appearance, and employer-employee relations. The EDP is a
plan detailing the training goals of the client based upon his/her
interests, potentials, and needs. The EDP identifies the training
program(s) in which the client will participate to achieve these training
goals. Pre-training orientation requires six to eight hours to complete.
Group 2 (Training)

CETA training follows completion of the pre-training orientation.
The training provided in both Program 1 and Program 2 is quite similar
consisting of placing the participant in a job slot at a public or
private, non-prof it agency. The job slot must be identical to regular
positions in the agency in all respects except that the salary is paid
by CETA. Training is provided in job performance skills and job reten-
tion skills by the job site supervisor and the CETA counselor.

Job performance training is provided, for the most part, by the
job site supervisor. I'Jhile in training the CETA participant performs a
regular job function at the job site. The job site supervisor provides
the CETA participant with instructions in the proper performance of job
duties and corrects the participant's work until mastery of the job
skills is achieved. The skills trained will vary across training slots.
Typical training slots will include clerical work, playground supervision,
traffic management and parking meter monitoring, carpentry, grounds
maintenance, Sunland Cottage Parent, child care aide, or supply clerk.
The counselor is involved in job skills training only as required to
mediate in differences of opinion between the participant and the job
site supervisor.

Job retention training is provided for the most part by the CETA
counselor. Job retention skills include all factors involved in the

83
identification, acquisition, and retention of full-time employment.
Some of these skills are knowing where to look for an employment position,
how to complete an application, good Interview techniques, proper dress
and personal appearance for a job, proper employer-employee relations,
punctuality and attendance, and peer relations. The CETA counselor also
provides training in life skills such as preparation and management of a
budget, planning a healthy diet, personal medical attention, personal
hygiene, and how to resolve family problems. Training in these skill
areas is provided in group sessions and during job site visits by the
counselor. Program 1 participants receive group training sessions twice
a month and are visited on the job site by the counselor twice a month.
Program 2 participants receive group sessions and job site visits once
a month. Group sessions are usually two hours in duration and job site
visits vary from 15 to 60 minutes. The job site supervisor is involved
in the training of these skills only through the identification of the
specific job retention skills required by the training agency.

CETA counselors in both Program 1 and Program 2 provide the CETA
participants with limited personal-social adjustment counseling. Most
of this counseling is situation specific, designed to correct a personal
problem which is causing difficulties on the job site. This type of
counseling is usually done in conjunction with the job site supervisor
in a three-way conference to identify the problem as it pertains to the
work setting. Individual counseling topics will include resolution of
personal problems which are being carried onto the worksite and problems
between members of the worksite. The counselor also engages in a
continual review and update of the client's EDP to monitor client
progress towards training completion and to adjust the EDP to changes in
the client's interests and needs. The CETA counselor will also provide

84
group counseling as needed to address problems cominon to all members of
the training group as a part of the job retention group training sessions.
Personal-social problems requiring extended counseling are referred to a
fulltime counseling staff subcontracted by the employment training
program.
Program 1 versus Program 2

The intake and training phases of Program 1 (Adult Work Experience)
and Program 2 (Public Service Employment) are quite similar. The basic
structure of both programs was described in the preceding chapter.
There are, however, a few between program differences which can be
identified. First, examination of the eligibility criteria for partici-
pation in the two programs (Appendix A) reveals that Program 1 partici-
pants are required to be unemployed for a much shorter time period than
Program 2 participants. Program I requires a one week unemployment
period while Program 2 requires 10 of 12 weeks (Title VI) or 15 of 20
weeks (Title II-B) of unemployment. Second, Program 1 is designed for
persons unable to obtain employment due to a lack of most basic employ-
ability skills. Program 2 is designed for persons who possess the
basic skills required for employment but who are unable to obtain employ-
ment due to conditions beyond their control. Third, participants in
Program 1 receive a basic training wage while Program 2 participants
receive a salary equivalent to persons performing similar job functions.
Finally, participants in Program 1 may train for a maximum of 25 weeks
at 40 hours per week while participants in Program 2 may participate
in training activities for a maximum of 78 weeks.

85
Taken together, these differences between Program 1 and Program 2
serve to identify the basic difference in the intent of the two programs.
Both programs are designed to assist participants in the acquisition of
skills and attitudes necessary for the performance of fulltime employ-
ment. This is accomplished by placing the participants in a job slot at
a training agency for training in the performance of that job. The
basic difference is that Program 1 is intended to be a basic skills
training program while Program 2 is intended to more nearly approximate
a typical job.

Research Design
The primary comparison in this study between Group 1 (Intake) and
Group 2 (Training) was performed utilizing a modified form of the
quasi-experimental separate-sample, pretest-posttest design described
by Campbell and Stanley (1966) :

Pretest R 0 (X)
Posttest R - X 0

R = Random assignment of subjects to the
pre- and posttest conditions.

0 = Observation/data collection.

X = Treatment received by subjects.
(X) = Treatment received by subjects but

not affecting the data collected
since it occurs after the data is
collected for that group, (pp. 53-54)

Modifications In the basic design were necessitated by the projected

low average monthly intake rate for both programs. The intake rate for

Program 1 was estimated at 10-15 new participants per month and estimated

at 15-20 new participants per month in Program 2. In order to achieve a

reasonable sample size within a manageable time frame it was necessary

to modify the design to increase the efficiency with which the low

number of new participants could be utilized in the study.

86

The first modification was the elimination of the random group
assignment and simultaneous treatment elements of the design. The
original design calls for the assignment of all incoming subjects to
either the pretest or posttest condition on a random basis, testing the
pretest group, permitting both groups to receive the same treatment, and
testing the posttest group at the conclusion of the treatment phase.
This modification permitted the assignment of all new participants to
the pretest group, Group 1, and the use of participants beginning train-
ing four months prior to the data collection period for the posttest
group. Group 2. This modification permitted the completion of the data
collection process in two months as opposed to a minimum of six to eight
months without the modification.

The second modification was the collection of data over an extended
time frame. The original design used a simultaneous data collection
process, testing all subjects at the same time. However, CETA partici-
pants are oriented and enter training as training sites become available.
Consequently, not all subjects were available for testing at the same
time. It was therefore necessary to collect data over an extended time
period. Subjects were assigned to a monthly subgroup classification as
they were tested to facilitate a test of homogeneity (described below) to
control for variance in the data as a result of this modification.

87

The modified design is as follows for this study:
Group

Group 1 (Intake)

Monthly
Subgroup

(1)

Data
Collection Month

1
0

2

(X)

(2)

0

Group 2 (Training)

(1)

X

(X)

(2)

X

X

Monthly Subgroup = The subgroup assignment on the
basis of the month in which the data was
collected for use in the homogeneity test.
Data Collection Month = The month in which the data

was collected.
0 = Observation/data collection.
X = Training received by participants/subjects.
(X) = Training received by participants/subjects
but not affecting the data collected.
(1), (2) = Monthly subgroup classification for the
purposes of the homogeneity test.

These modifications raised two problems. The first was the possi-
bility of uncontrolled selection biases. In the basic design, as
described by Campbell and Stanley (1966), the comparability of comparison
groups was assured by the random assignment of subjects into the pretest
and posttest conditions. In order to reestablish comparability a test
of comparability was included in the design modification to control for
possible selection biases. The test of comparability consisted of the
comparison of Group 1 and Group 2 on the independent demographic
variables. It was found that Group 1 and Group 2 differed on the
independent variable of education. Group 1 was found to have a lower
average education level (M=11.A6 years) than Group 2 (M=11.96 years,
F=4.12,df=l/224,p<.05) . The two groups were matched on this variable

88
by the elimination of 11 cases with extremely low levels of education and
24 cases with extremely high levels of education. Upon completion of
matching Group 1 with Group 2 on the education variable the matched
Group 1 and Group 2 were compared with the unmatched Group 1 and 2 on
each of the dependent variables. No significant differences were
identified between the matched and unmatched groups and it was concluded
that comparability had been achieved as a result of matching the groups
without altering the information contained in the data.

A second problem was posed by the modification of the testing
period. The sampling of subjects over an extended period of time could
result in a loss of homogeneity within Group 1 and within Group 2. To
determine whether variance occurred between the first and second month
of data collection, a test of homogeneity was performed within the
Combined Group (Group 1 and Group 2), within Group 1, within Group 2,
and within both Program 1 and Program 2. The test of homogeneity con-
sisted of comparing the average scores of the subjects tested during the
first month of data collection (Monthly Subgroup 1) with the average
scores of subjects tested during the second month of data collection
(Monthly Subgroup 2) within each of the comparison groups on the
dependent mental health measures. No differences between monthly
subgroups were identified within the combined group and within Program 1.
The differences identified in the tests of homogeneity are discussed in
the following chapter. In general, however, it was concluded that there
was very limited variance in the mental health scales between the first
and second month of the data collection process and the data was
considered to be homogeneous.

89

Thus, the modified design retains the integrity of the original
design. Each threat to internal and external validity is controlled
for or identified within the modified design. Historical, maturational,
and mortality effects were controlled for by the homogeneity test. The
comparability procedure served to control for selection biases as well
as interactions between selection and maturational processes or treatment
effects. There were no testing effects since each subject was tested
only once and the instrumentation effects were controlled through the use
of a uniform presentation of test instruments and by all subjects being
administered the same test package. Statistical regression was not a
factor as the subjects were not selected for participation on the basis
of extreme test scores. Finally, multiple treatment effects and reactive
effects were not problems since the training experience was a single,
continuous, non-laboratory process.

Data Collection

Data for this study was collected during April and May, 1980. Sub-
jects for Group 1 (Intake) were tested during the pre-training orientation.
Intake subjects participating in Program 1 were tested at the conclusion
of normal orientation activities. Intake subjects in Program 2 were
tested just prior to the last pahse of the orientation activities. All
participants in the employment training program who had completed four
months of training during the months of April and May, 1980, in either
Program 1 or Program 2 were invited to attend the testing sessions for
inclusion in the Training Group (Group 2). All Group 2 subjects were
tested during group sessions organized by the training program to assist
in the collection of data for this research project. Each of the testing

90
sessions were scheduled during the two weeks immediately following the
completion of the fourth month of training.

Prior to administering the test package all subjects received and
had read to them the Informed Consent Form which they were then asked to
sign (See Appendix F) . The test package consisted of an Introduction and
Personal Data Questionnaire (See Appendix C) , the Personal Orientation
Inventory (Shostrora, 1974), the Bern Sex-Role Inventory (Bern, 1974), and
the Internal-External Scale (Rotter, 1966). In an effort to eliminate
potential testing effects due to a limited reading ability among some of
the subjects the test package was pre-recorded on cassette tape and played
back to the subjects as they read along. All instructions and questions
of each test instrument were presented in this manner. An answer period
of five to six seconds (six to seven on the BSRI) was allowed between
items for the subjects' response. The playback length of the combined
test package was 72:08 minutes with each section having playback times of
6:22 minutes (Introduction and Personal Data Questionnaire), 10:00 minutes
(Internal-External Scale), 9:50 minutes (Bem Sex-Role Inventory), and
42:56 minutes (Personal Orientation Inventory). The actual test admin-
istration time was more nearly 90 minutes allowing for questions, seating
time, and the distribution and collection of materials. To avoid possible
serial effects the order of presentation of the segments of the test
package was randomly varied.

Each instrument was scored according to the published instructions.
If an instrument contained more than 10% of the items incorrectly answered
the instrument was eliminated from the subject's data set and referred
to as a missing case during the data analysis. If two or more of the
instruments in a subject's data set were eliminated the subject was
eliminated from the combined data set.

91
Instrumentation

The mental health variables were measured through the use of the
following instruments: the Personal Orientation Inventory (Shostrom,
1974), the Bem-Sex Role Inventory (Bem, 1974), and the Internal-External
Scale (Rotter, 1966).
Personal Orientation Inventory

The Personal Orientation Inventory (POI) assesses levels of self-
actualization in this study. The POI consists of 150 paired statements.
These forced-choice items require the respondent to make a comparative
value judgement on which statement of each pair is more true for him/her.
The items are scored twice: once for the two basic personal orientation
scales, the Inner-directed support scale (127 items) and the Time-competent
scale (23 items); and once for the ten subscales, each of which measures
a conceptually important element of self-actualization.

The two personal orientation scales. Inner-directed support and
Time-competent, have a reflexive scale, Other-directed support and Time-
incompetent, respectively. Together, these two pairs of scales form
ratio scales identifying the personal orientation of the respondent.
The support ratio is comprised of the Other-directed/lnner-directed (O/I)
ratio and represents the general orientation of the respondent's mode of
interpersonal interaction. Inner, or self-directed, respondents are
guided primarily by internalized values while other-directed subjects
are motivated by peer group influences and other external forces.
Similarly, the time-competent respondent, as measured by the ratio of the
Time-incompetent /Time-competent scales (Ti/Tc) , lives in the present while
the time-incompetent subject dwells upon events in the past and future.
The POI Test Manual suggests that the self-actualizing Ti/Tc ratio score

92
ranges between 1:6 and 1:22, while the self-actualizing O/l ratio ranges
between 1:3.2 and 1:6. A (Shostrom, 1974, pp. 14-16).

In addition to these basic personal orientation scales the POI
contains ten subscales which serve to measure five conceptually
important factors of self-actualization.

Valuing. This factor is assessed by using the Self-Actualizing
Values (SAV) scale and the Existentiality (Ex) scale. The SAV scale
contains 26 items and measures the degree of affirmation of the primary
values of self-actualizing persons. The Ex scale, containing 32 items,
assesses the person's degree of flexibility in the application of
values to living.

Feeling. This factor is composed of the Feeling Reactivity (Fr)
and Spontaneity (S) scales. The Fr scale measures the degree of sensi-
tivity and responsiveness to one's own needs while the S scale measures
the freedom with which the individual expresses these needs. The
scales contain 23 and 18 items respectively.

Self-Perception. This factor, considered equivalent to a measure
of self-concept, is measured through the use of the Self-Regard (Sr)
scale and the Self-Acceptance (Sa) scale. The Sr scale (16 items)
measures degree of affirmation of self-worth while the Sa scale (26
items) measures willingness to accept both one's strengths and one's
weaknesses .

Awareness . This factor is composed of the Nature of Man,
Constructive (Nc) scale and the Synergy (Sy) scale. The Nc scale (16
items) measures the degree to which the individual sees humankind as
essentially good and the extent to which he/she can understand the
synergistic nature of humankind. The Sy scale (9 items) is a measure of

93
the extent to which the individual can synergistically resolve such
dichotomies of life as work and play or selfishness and unselfishness.

Interpersonal Sensitivity. The final factor consists of the
Acceptance of Aggression (A) and Capacity for Intimate Contact (C)
scales. The A scale (25 items) measures the ability to accept one's own
natural aggressive feelings and express them constructively rather than
repress them. The C scale measures the ability to establish and maintain
intimate relationships with others, unencumbered by destructive or
limiting expectations and obligations (28 items).

The validity of the POI has been repeatedly demonstrated using
nominated groups. This procedure involved identifying groups as healthy
or unhealthy on several different criteria including: other established
measures (e. g., hospitalized psychiatric patients versus fully function-
ing businessmen) and expert diagnosis. The POI has consistently
discriminated between subjects clinically judged to be self-actualizing
and subjects judged to be non-self-actualizing (Shostrom, 1964),
psychiatric inpatients (Fox, J., Knapp, & Michael, 1968), psychopathic
felons (Fisher, 1968), and neurotic groups as measured by the Eysenck
Personality Inventory (Knapp & Comrey, 1973), and the MI>IPI (Shostrom,
1974).

Test-retest reliability coefficients for the major scales (I and Tc)
are reported at 0.71 and 0.84 respectively (Klavetter & Morgar, 1967).
Test-retest reliability coefficients have been reported as significant
for all scales but the Fr scale (Ilardi & May, 1968). Reliability has
also been tested by instructing subjects to attempt to "fake good"
responses on the POI. Even subjects provided with information on self-
actualization were unable to "fake good" responses successfully (Foulds &
Warehime, 1971a ; Ckildman & Olczak, 1976; Warehime, Routh, & Foulds, 1974).

94

Test reviews of the POI have generally agreed that it should not be
used as a diagnostic tool, though it can be used profitably as a thera-
peutic tool (Ranaan, 1973; Tosi & Lindamood, 1975). Ranaan (1973)
further suggests that the POI would be of limited use as a research
instrument due to questions regarding instrument validity and reliability.
However, Tosi & Lindamood (1975), following a review of later studies,
concluded that the POI should be regarded as a valid research instrument.
Use of the POI in this study provides a psychological health measure that
has been widely used and accepted (Shostrom, 1973).
Bem Sex-Role Inventory

The Bem Sex-Role Inventory (BSRI) was introduced by Bern (1974) as a
measure of psychological androgyny. The BSRI contains 60 adjectives and
adjective phrases which the respondents are asked to rate as describing
themselves on a scale of 1-7. The scale values are: (1) never or almost
never true for me, (2) often not true for me, (3) sometimes not true for
me, (4) neither true or not true for me, (5) sometimes true for me, (6)
often true for me, or (7) always or nearly always true for me. The BSRI
has three subscales: Masculinity, Femininity, and Social Desirability.

The items were selected by judges rating 200 personality character-
istics on their desirability for males and females. If an item was
judged by both male and female judges as being significantly more
desirable for males than females then the item was included in the
Masculinity subscale. Likewise, items judged more desirable for females
were placed in the Femininity subscale. Both the Masculinity and
Femininity scales contain 20 positive items. The Social Desirability
scale contains 10 items judged to be negative for both males and females
and 10 items judged to be positive for both males and females.

95
The Masculinity and Femininity scales are scored by computing the
average ranking of the 20 items in each scale. The Social Desirability
scale is scored by reversing the rankings of the 10 negative items and
computing the average for all 20 items. The Masculinity and Femininity
averages are used in the determination of the subject's sex-type classi-
fication. The original sex-type classification method was developed by
Bem (1974) and consisted of applying the Student's _t-test of differences
between means to the Masculinity and Femininity averages. If the t-test
value was significant then the subject was classified as sex-typed in the
direction of the higher average. Therefore, if the ^-test value was
significant and the Masculinity average was higher than the Femininity
average then the subject was classified as masculine sex-typed, regard-
less of the subject's gender. If the _t-test value was not significant
then the subject was classified as androgynous, since the level of
endorsement of masculine and feminine items was approximately the same.
An alternative scoring method was proposed by Bem (1974) as an approxi-
mation of the _^-test value. This method consists of multiplying the
absolute value of the difference between the masculine and feminine means
by a constant. The approximate _t-test value is used to classify the
subject in the same manner as the exact t-test value.

Spence et al. (1975) questioned the ;t-test classification method.
They suggested that the classification of all subjects with no signifi-
cant difference between the masculine and feminine averages as andro-
gynous may be inappropriate. They argued that two low levels of endorse-
ment may be indicative of a condition different from psychological andro-
gyny. They proposed the median-split classification method as an alter-
native. This method consists of calculating the median score for the

96
the sample on each of the two sex-role scales. If an individual subject
in the sample had masculinity and femininity averages above the sample
median scores, then the subject would be classified as androgynous. If
only one of the subject's average scores was above the median, then the
subject would be classified as sex-typed in the direction of the higher
average. If both of the subject's scores were below the median values
for the two scales, then the subject would be classified "undifferentiated.'
Following an examination of the median-split classification method, Bern
(1977) agreed that, although a limited amount of additional information
was obtained, the median-split method of classifying sex-types should be
used as the more appropriate.

Regardless of the scoring and classification method used, the BSRI
has been shown to demonstrate validity consistently by predicting andro-
gynous versus sex-typical behavior. Several studies have used the BSRI
and found that persons classified as sex-typed are less willing to engage
in cross-sex behavior; i. e., behavior normally associated with the other
sex-type sex-role. Androgynous subjects do not exhibit this unwillingness
(Bern, 1974, 1977; Bern & Lenney, 1976; Bern et al., 1976). The BSRI has
also been found to correlate strongly with the POI as a measure of mental
health (Cristall & Dean, 1976; Nevill, 1977; Ott; 1976; Pettus, 1976).
Test-retest reliability values over a four week test period were quite
high: Masculinity scale, r=.90; Femininity scale, r=.90; Social Desira-
bility, r=.89; and Androgyny, r=.93 (Bem, 1974).

The Bem Sex-Role Inventory has been widely accepted and used as
a measure of psychological androgyny (Bem, 1979) and is used in this
study as a measure of androgyny.

97

Internal-External Scale

Rotter (1966) introduced the Internal-External Scale (lES) as a
measure of locus of control. The lES contains 29 paired statements,
six of which are fillers. The 23 scoring pairs contain an external and
an internal statement. The respondents are asked to select the statement
in each pair which is more true for them. The lES is scored by totaling
the number of external items endorsed. The resultant score is a measure
of external locus of control. Since locus of control is conceptualized
as a single bi-polar dimension, a low score is considered indicative of
an internal locus of control.

The validity of the lES has been demonstrated by discriminating
between persons willing to engage in external or internal actions. These
studies have examined subjects' willingness to take risks or attempt a
difficult problem (Cohen et al., 1967; Phares, 1957; Rotter & Mulray,
1965) as well as willingness to engage in social action (Gore & Rotter,
1963; Kinder, 1976; Strickland, 1965). Rotter (1966) reports test-retest
reliability correlations of r=.49 over a two month interval and r=.83
over a one month interval on the lES. Zerlga, Tseng, and Greever (1976)
reported an eight month test-retest correlation of r=.55 (p <.001). As
with the POI and the BSRI, the lES is selected as the most widely used
measure of locus of control for use in this study.
Demographic Variables

This study used seven demographic factors in the analysis of the
dependent measures and in the comparability test. The seven factors
were age, sex, race, education, religious affiliation, family history,
and length of unemployment prior to participation in CETA training.
The variables of age and education were measured in years and the length

98

of unemployinent was measured in months. Three variables were used to
assess degree of religious affiliation: church membership, church atten-
dance ( frequency --never or rarely, one to six times a year, twice a
month, at least once a week), and denomination of religious affiliation
(not affiliated; Baptist; Fundamentalist-including Church of Christ,
Church of God, Church of the Nazarene, Holiness, and Pentacostal; and
Liberal-including Methodist, Presbyterian, Catholic, Episcopal, and
Unitarian). Four variables were used to measure family history.
Two of these variables represented past family history: Parent (raised
by both parents, by one parent and one step-parent, by one parent only,
by other relatives, or by legal guardians) and Ordinal position in the
family (only child, oldest child, second oldest child, third oldest child,
fourth or younger child, youngest child). Two of the variables were
measures of the current family history: Marital Status (single, married,
divorced, widowed, or separated) and Number of children.

The demographic variables were drawn from two sources. The subjects'
age, race, sex, education, length of unemployment, marital status, and
number of children were drawn from their CETA applications (see Appendix
D) . The subjects' church membership, church attendance, denomination,
parents, and family position were drawn from the Personal Data Question-
naire (Appendix C) .

A total of 12 demographic variables were assessed: age, race, sex,
education, church membership, church attendance, denomination, parents,
family position, marital status, number of children, length of
unemployment .

99

Analysis of Data

The primary purpose of this study was to determine whether persons
beginning CETA training differed significantly from persons completing
four months of CETA training on three dependent mental health measures.
The secondary purpose of this study was to determine whether partici-
pants in the CETA Adult Work Experience program differed significantly
from participants in the CETA Public Service Employment program on the
selected independent demographic variables and the dependent mental
health measures. Finally, this study had the purpose of determining
whether there existed a relationship between the mental health character-
istics of the CETA participants and their selected demographic character-
istics and whether differences existed between the mental health
characteristics of the CETA participants and subjects reported in a
selected validation study for each of the three mental health measures.

The three mental health measures used in this study were the
Personal Orientation Inventory (POI), the Bern Sex-Role Inventory (BSRI),
and the Internal-External Scale (lES) . The demographic variables used
in this study were age, race, sex, education, church membership, church
attendance, denomination of church affiliation, parents, family position,
marital status, number of children, and length of unemployment prior to
participation in CETA training.

Data collected in this study were evaluated using statistical proce-
dures described by Roscoe (1975) and by Nie, Hull, Jenkins, Steinbrenner,
and Bent (1975). The confidence level for all comparisons was set at the
p <.05 level. Differences which had significance levels of p <.10 and
p >.05 were identified as trends.

The first stage of the data analysis was to examine between group
comparability and the within group homogeneity. The between group

100
comparability was demonstrated by comparing Group 1 and Group 2 on each
of the 12 demographic variables. A one-way Analysis of Variance
(ANOVA) , described by Nie et al. (1975, pp. 2A9-261), was used to com-
pare differences between Group 1 and Group 2 on the demographic variables
of age, education, number of children, and length of unemployment. A Chi-
square test of homogeneity was used to compare differences between
Group 1 and Group 2 on the remaining demographic variables (Nie et al.,
1975, pp. 218-248). The between group tests of comparability were
also performed examining differences between Program 1 and Program 2 on
the demographic variables using the same statistical procedures.

Within group homogeneity was examined to identify differences
within each comparison group on each of the dependent mental health
variables between subjects sampled during the first month of testing
(Monthly Subgroup 1) and subjects tested during the second month of
data collection (Monthly Subgroup 2) . The one-way ANOVA was used to
compare monthly subgroups within the Combined Group (all subjects) ,
Group 1, Group 2, Program 1, and Program 2 on all scales of the POI ,
the lES, and the Masculinity, Femininity, and Social Desirability scales
of the BSRI. A Chl-square test of homogeneity was used to compare
monthly differences on the BSRI sex-type classification.

The seven null hypotheses were tested in the following manner:
Hypotheses 1-3. The hypotheses state that no differences exist
between Group 1 (Intake) and Group 2 (Training) subjects on measures of:

a. Self-actualization (Hypothesis 1)

b. Psychological Androgyny (Hypothesis 2)

c. Locus of Control (Hypothesis 3).

These null hypotheses were tested using a one-way ANOVA to compare
differences between Group 1 and Group 2 on the POI scales, the lES, and

101

the Masculinity, Femininity, and Social Desirability scales of the BSRI .
The Chi-square test of homogeneity was used to compare Group 1 and Group
2 on the BSRI sex-type classification.

Hypotheses 4-6. These hypotheses state that no differences exist
between participants in Program 1 (Adult Work Experience) and Program 2
(Public Service Employment) of the CETA program on measures of:

a. Self-actualization (Hypothesis 4).

b. Psyhcological Androgyny (Hypothesis 5).

c. Locus of Control (Hypothesis 6).

These null hypotheses were tested using the one-way ANOVA procedure
to compare differences between Program 1 and Program 2 on the POI scales,
the lES, and the Masculinity, Femininity, and Social Desirability scales
of the BSRI. The Chi-square test of homogeneity was used to compare
differences between Program 1 and Program 2 on the BSRI sex-type
classification.

Hypothesis 7. This hypothesis states no relationships exist between
the variance in the dependent mental health measures and the selected
independent demographic varaibles. This hypothesis was tested using the
three-way ANOVA procedure (Nie et al., 1975, pp. 398-422). The three
factors used in the three-way ANOVA were: Factor 1 — Group 1 versus Group
2, Factor 2 — Program 1 versus Program 2, Factor 3 — the twelve demographic
variables. The three-way ANOVA procedure was used to examine variance
in the POI scales, the lES, and the Masculinity, Femininity, and Social
Desirability scales of the BSRI. A series of Chi-square tests of homo-
geneity were used to examine variation in the distribution of the BSRI
sex-type classification. The age and education variables were recoded
for ease of analysis (see Table 2 for grouped values) .

102
Table 2
Receded Values of Demographic Variables of Age and Education

Age

Education

Range

Code

Range

Code

18 years or less.

1

0-5 years.

1

19-20 years.

2

6-8 years.

2

21-22 years.

3

9-10 years.

3

23-25 years.

4

11 years.

4

26-30 years.

5

12 years.

5

31-35 years.

6

13 years.

6

36-45 years

7

14 years.

7

45 years or more.

8

15 years or more.

9

The final analysis of the study was the examination of differences
between the mental health characteristics of the CETA subjects and sub-
jects reported in selected validation studies of the three mental health
measures. The POI scores of the CETA subjects were compared with a
"normal" adult sample reported by Shostrom (1974, p. 24). The lES scores
of the CETA subjects were compared with the lES scores of students at
the Ohio State University and the Florida State University reported by
Rotter (1966, pp. 279-280). The Masculinity, Femininity, and Social
Desirability scores on the BSRI of the CETA subjects were compared with
the scores of Stanford University students on these scales reported by
Bem (1974, p. 161). All of the above comparisons were performed using
the one-tailed _t-test of differences between Independent means (Roscoe,
1975, p. 217). The one-tailed test was used, as opposed to the two-tailed
test, since the CETA subjects were expected to have lower mental health
scores than the subjects reported in the literature based upon the
discussion of results reported in prior studies. The distribution of
BSRI sex-type classifications among the CETA subjects was compared with

103
the distribution of sex-type classifications among Stanford University
students reported by Bern (1977, p. 198) using the Chi-square test of
homogeneity procedure (Roscoe, 1975, pp. 254-262).

Limitations of the Study

The overall purpose of this study was to examine the mental health
characteristics of unemployed persons through the comparison of samples
of CETA participants. Although the definition of the unemployed person
for the purpose of this study was to be eligible for participation in a
CETA funded employment training program, the question may be raised
whether all eligible persons actually apply to CETA for training. If
they do not apply for training, then it may be possible that the sample
of subjects used in this study and the findings obtained are not fully
generalizable to all unemployed persons. In view of the multitude of
reasons a person may be unemployed for extended periods of time, the
results of this study may be most applicable to persons who are eligible
for and apply to CETA employment training programs. While these results
certainly have meaning for non-participant unemployed persons, fullscale
application of these results to those persons may be speculative and
should be done with caution.

Another limitation of this study is the locale from which the sub-
jects were drawn. This study examined subjects participating in a north
Florida metropolitan based CETA program. There may exist differences
between unemployed persons applying for and receiving training in rural
versus metropolitan areas, as well as between different regions of the
nation. Likewise, all subjects were tested during the Spring of the
year. Seasonal differences in subject characteristics may exist which
should be considered in understanding and using the results of this study.

CHAPTER IV
RESULTS

The purpose of this study was the investigation of mental health
characteristics of unemployed persons participating in CETA funded employ-
ment training activities. There were four major comparisons performed
in this investigation. The demographic and mental health characteristics
of persons beginning training were compared with persons completing four
months of CETA training. The demographic and mental health character-
istics of participants in the CETA Adult Work Experience program were
compared with the characteristics of participants in the CETA Public
Service Employment program. The relationships between the mental health
characteristics and the demographic characteristics of the CETA partici-
pants were examined. Finally, the mental health characteristics of the
CETA participants were compared with the mental health characteristics
of subjects reported in the initial validation studies of the mental
health measures used in the study. Three dependent measures of mental
health were used: the Personal Orientation Inventory, the Bern Sex-Role
Inventory, and the Internal-External Scale. The subjects were partici-
pants in a northern Florida employment training program and were identi-
fied and tested during the Spring of 1980.

Evaluation of Comparison Groups
Comparability of Intake and Training Groups

The first stage of the data analysis was the establishment of
between group comparability. A test of comparability comparing the

104

105

demographic characteristics of the subjects in the Intake Group (Group 1)
with those of the subjects in the Training Group (Group 2) was performed.
The results of these comparisons, presented in Table 3, reveal that the
Group 1 subjects did not differ significantly from the Group 2 subjects
on 11 of the 12 demographic variables examined. The average level of
education among the Group 1 subjects (11.46 years) was found to be
significantly lower than the average level of education among the
Group 2 subjects (11.96 years, F=4.123, df=l/224, p <.05).

In view of the difference in education levels Group 1 and Group 2
were matched on the demographic variable of education. The matching
process involved the elimination of 24 subjects with extremely high
educational levels and the elimination of 11 subjects with extremely
low levels of education. This process reduced the overall sample size
from 226 subjects to 191 subjects. Following completion of the matching
process, the between group test of comparability was repeated. The
results of these comparisons, also presented in Table 3, support the
conclusion that the matched groups may be judged comparable on the basis
of the 12 demographic variables examined. The matched data set consisting
of 191 subjects was compared with the original unmatched data set (N=226) on
each of the dependent mental health measures as well as on each of the
demographic variables. The absence of significant differences between
the matched and unmatched data sets on the dependent mental health vari-
bles (See Table 4) supported the conclusion that the mental health
and demographic characteristics of the overall CETA sample had not been
significantly altered by the matching process.

106

Table 3

Comparison of Intake Group Subjects and Training Group Subjects
on 12 Selected Demographic Variables: Unmatched and Matched for
Education

Unmatched Data

Matched Data

Categorical
Variables

Sex

Race

Church Membership

Church Attendance

Denomination of

Church
Parent

Family Position
Marital Status

Continuous
Variables

Age

Education

Number of Children

Length of

Unemployment

df

0.324

1

5.410

2

1.025

1

3.975

3

4.957

3

4.787

5

5.728

5

0.880

4

x'

df

0.011

1

3.914

2

0.170

1

3.158

3

5.136

3

5.458

5

3.974

5

0.415

4

MSB

(a)

MSW

(b)

43,

,755

103,

.349

0.423

5.904

14,

.087

3,

.417

4.123 *

0.099

1.

,410

2.

.382

0.592

0.002

10,

,816

4,

.455

2.428

0.874

108.003 0.055

2.686 0.037

2.368 O.OOl

4.619 0.189

(a) Mean Square Between, df=l

(b) Mean Square Within, df=224

(c) Mean Square Between, df=l

(d) Mean Square Within, df=189
* p <.05

Homogeneity Within Comparison Groups

The second stage of the data analysis was the within group tests of

homogeneity. This test consisted of comparing the Monthly Subgroups

within the Combined Group (Group 1 plus Group 2), the Intake Group

(Group 1), the Training Group (Group 2), the Adult Work Experience program

(Program 1), and the Public Service Employment program (Program 2) on

each of the dependent mental health measures. These tests were performed

to identify variations within the comparison groups during data collection.

107
Table 4

Comparison of Unmatched and Matched Groups on the Independent
Demographic Variables and on the Dependent Mental Health Variables.

Categorical
Variables

df

Sex

Race

Church Membership

Church Attendance

Denomination of Church

Parent

Family Position

Marital Status

Bern Sex-Role Classication

Continuous
Variables

Age

Education

Number of Children

Length of Unemployment

Internal-External Scale

Personal Orientation Inventory

I

0

Tc

Ti

SAV

Ex

Fr

S

Sr

Sa

No

Sy

A

C

Inner-directed
Other-directed
Time-competent
Time- incompetent
Self-Actualizing Values
Existentiality
Feeling Reactivity
Spontaneity
Self-regard
Self -acceptance
Nature of Man
Synergy

Acceptance of Aggression
Capacity for Intimate
Contact
Bern Sex-Role Inventory
Masculinity
Femininity
Social Desirability

M
F
SD

0.134
0.238
0.105
1.024
0.368
0.584
0.490
0.176

0.181

0.016

4.452
5.589
0.019
0.007
1.069
1.927
0.139
0.329
0.004
0.013
0.818
0.533
0.472

0.472

0.000
0.001
0.000

1
2
1
3
3
5
5
4

Mean

Mean

Square

Square

Between

df

Within

df
415

F

5.136

1

105.090

0.048

0.618

1

3.102

415

0.199

0.000

1

2.368

415

0.000

0.234

1

4.536

415

0.052

12.758 404

2.403
2.268
2.370

415
415
415

0.001

97.587

396

0.046

95.141

396

0.059

8.197

396

0.002

7.913

396

0.001

9.809

396

0.109

17.844

396

0.108

7.902

396

0.018

4.739

396

0.069

5.942

396

0.001

8.633

396

0.002

4.316

396

0.190

2.334

396

0.237

9.239

396

0.051

14.519 396 0.052

0.000
0.000
0.000

p <.05

108
The Combined Group subjects and Program 1 subjects exhibited no
significant differences between subjects tested during the first month
of data collection (Monthly Subgroup 1) and subjects tested during the
second month of data collection (Monthly Subgroup 2). Within Group 1
(Intake), Monthly Subgroup 1 subjects were found to score higher on the
POI Spontantiety scale (F=4.699, df=l/90, p <.05) and higher on the POI
Capacity for Intimate Contact scale (F=4.398, df=l/90, p <.05) than the
Monthly Subgroup 2 subjects. Within Group 2 (Training), Monthly Sub-
group 1 subjects scored lower on the POI Acceptance of Aggression scale
than the Monthly Subgroup 2 subjects (F=6.859, df=l/90, p <.05). Monthly
Subgroup 1 subjects within Program 2 (Public Service Eraplojmient) scored
lower on the POI Acceptance of Aggression scale than the Monthly Subgroup 2
subjects (F=4.297, df=l/114, p <.05). These differences were examined in
the context of other differences discussed in this study. In view of
the absence of significant within group differences within the Combined
Group and within Program 1, and the limited number and pattern of within
group differences observed within Group 1, Group 2, and Program 2, it was
concluded that the comparison groups were sufficiently homogeneous for
use as comparison groups in this study.

Comparability of Adult Work Experience Program and Public Service
Employment Program

The subjects in this study were also compared on the basis of the
CETA training program in which they were participating. Of the 191 sub-
jects in the final matched data set, 69 subjects (8 males, 61 females)
were participants in the Adult Work Experience program (Program 1) and
122 subjects (53 males, 69 females) were participants in the Public Service
Employment Program (Program 2). Program 1 subjects were compared with
subjects in Program 2 on each of the 12 independent demographic variables

109

for the purpose of obtaining a clear demographic description of the two

programs. The results of these comparisons are presented in Table 5.

Table 5

Comparison of Adult Work Experience Subjects and Public Service
Employment Subjects on the 12 Demographic Variables.

Categorical

X^

df

Variables

Sex

19.127

**A 1

Race

5.395

2

Church Membership

0.459

I

Church Attendance

3.730

3

Denomination of Chur

ch

1.985

3

Parent

5.031

5

Family Position

3.170

5

Marital Status

27.145

*** 4

Mean Mean

Program Program Square Square

Continuous 1 Mean 2 Mean Between Within df F

Variables

Age 33.8 yr. 25.6 yr. 2971.69 92.31 1/189 32.19 ***

Education 10.9 yr. 12.1 yr. 60.03 2.37 1/189 25.34 ***

Number of Children 1.9 1.0 39.95 2.16 1/189 18.53 ***
Length of

Unemployment 5.9 mo. 4.4 mo. 103.13 4.08 1/189 42.31 ***

*** p <.001 """

Examination of the results in Table 5 permits two observations.

First, unlike the Group 1 versus Group 2 comparison of the subjects'

demographic characteristics, Program 1 and Program 2 subjects are not

comparable in their demographic characteristics. Significant differences

were found between subjects in Program 1 and subjects in Program 2 on six

of the 12 demographic variables examined. Subjects in Program 1 were

more likely to be female (x^=19.127, df=l, p <.001) and more likely to

be unmarried (x^=27.145, df=4, p <.001) than subjects in Program 2.

110
Furthermore, Program 1 subjects were found to be older (F=32.19,
df=l/189, p <.001), have less education (F=25.34, df=l/189, p <.001),
have more children (F=18.53, df = l/189, p -^.001), and have been unem-
ployed for a longer period of time prior to entering CETA (F=42.31,
df=l/189, p <.001) than subjects in Program 2. Based upon these findings,
the second observation is that Program 1 and Program 2 may be serving
different segments of the unemployed population. Program 1 appears to
be serving the single, older mother who has limited education and has
been unemployed for an extended period of time. Program 2, on the other
hand, is serving younger unemployed persons of either sex, who have
completed high school, are married with few, if any, children, and have
not been unemployed for as long as the Program 1 subjects. Thus, Pro-
gram 1 appears to be serving the welfare mother while Program 2 is serving
the unskilled high school graduate.

Evaluation of Hypotheses

Seven null hypotheses were tested on the basis of the data collected
in this study. The results of these analyses are presented in this
section.
Hypothesis 1

This hypothesis states: No differences exist in self-actualization
characteristics as a result of training in the CETA Adult Work Experience
program and in the CETA Public Service Employment program between Intake
(Group 1) subjects just beginning training and Training (Group 2) subjects
who have completed four months of training as measured by the Personal
Orientation Inventory. This hypothesis was evaluated using a one-way
analysis of variance (ANOVA) to compare differences between the mean
scores of Group 1 subjects and Group 2 subjects on each of the POI scales.
The results of these comparisons are presented in Table 6.

Ill

Table 6

Comparison of Intake (Group 1) Subjects and Training (Group 2)
Subjects on the Personal Orientation Inventory.

Mean

Mean

Group 1

Group 2

Square

Square

POI

Mean

Mean

Between

Within

df

F

Scales

Inner-directed

74.97

75.31

10.49

97.49

1/177

0.743

Other-directed

50.24

50.14

4.99

94.26

1/177

0.053

Time-competent

14.27

13.98

4.97

8.10

1/177

0.614

Time-incompetent

8.40

8.74

5.11

7.86

1/181

0.650

Self -Actualizing

Values

18.09

18.57

8.83

9.55

1/181

0.925

Existentiality

15.65

15.66

0.00

18.53

1/181

0.000

Feeling Reactivity

14.08

14.24

1.26

7.93

1/181

0.158

Spontaneity

10.97

10.95

0.02

4.73

1/181

0.005

Self-regard

12.20

12.33

0.68

6.08

1/181

0.112

Self -acceptance

13.41

13.24

1.34

8.87

1/181

0.151

Nature of Man

10.02

10.57

13.82

4.29

1/181

3.223

Synergy

5.46

5.54

0.31

2.30

1/181

0.134

Acceptance of

Aggression

14.60

14.77

1.34

9.24

1/181

0.145

Capacity for

Intimate Contact

17.14

17.12

0.02

14.92

1/181

0.001

* p <.05

Examination of the results presented in Table 6 clearly supports the
acceptance of the null hypothesis. None of the differences on the POI
scales between Group 1 subjects and Group 2 subjects were significant at
the p <.05 level. The greatest difference between the groups was on the
POI Nature of Man scale which indicated a trend for Group 2 subjects to
score higher than the Group 1 subjects (F=3.223, df=l/181, p <.10).
However, significant results identified in the evaluation of Hypothesis
7 were found to strengthen this trend and subsequently the null hypothesis
was rejected.

112
Hypothesis 2

Hypothesis 2 states: No differences exist in psychological andro-
gyny characteristics as a result of training in the CETA Adult Work
Experience program and in the CETA Public Service Employment program
between Group 1 subjects and Group 2 subjects as measured by the Bem
Sex-Role Inventory, This hypothesis was evaluated using the one-way
ANOVA procedure and the Chi-square test of homogeneity.

During the evaluation of the data obtained from the Bern Sex-Role
Inventory (BSRI) two observations were made regarding the character of
the data. First, it was observed that a higher percentage of the respon-
dents failed to complete the BSRI appropriately (8%0 than failed to
apporpriately complete the POI (4%) or the Internal-External Scale (3%).
As discussed in the preceding chapter, appropriate completion of a test
instrument consists of correctly responding to at least 90% of the test
items. The second observation was that 30% of the subjects correctly
completing the BSRI exhibited a pattern of responses which is, at best,
questionable. The BSRI requires the respondent to rank 60 adjectives
on a scale of 1-7 ranging from "1" (never or almost never true for me)
to "7" (always or almost always true for me). The questionable response
patterns consisted of a majority of the 60 items being ranked at one or
two levels, e. g. , most of the items were ranked mostly "7's," mostly "6's'
and "7's," mostly "I's" and "7's," mostly "6's" and "7's," or mostly "2's"
and "6's." It seems questionable that the items would realistically be
ranked in such a regular fashion. These two observations, in conjunction
with the results of later analyses of the data in this chapter (Comparison
of CETA Subjects with Subjects in Selected Validation Studies), lead to
serious questions regarding the utility of the BSRI in the study of sex-

113

roles and sex-role flexibility among the CETA subjects in this study.

With these reservations in mind the results of the evaluation of

Hypothesis 2 are presented in Table 7.

Table 7

Comparison of Intake (Group I) Subjects and Training (Group 2)
Subjects on the Bern Sex-Role Inventory.

Mean

Mean

Group 1

Group 2

Square

Square

BSRI

Mean

Mean

Between

Within

df

F

Scales

Masculinity

5.68

5.58

1.82

2.34

1/181

0.777

Femininity

5.87

5.45

12.64

2.16

1/181

5.845 *

Social

Desirabili

■ty

4.90

4.89

3.72

2.56

1/181

1.648

BSRI Sex-Type Classification
Distribution Table and Chi-square test

Sex-Type Classification

Group 1

Group 2

Total

Masculine Sex-Typed

6

13

19

Feminine Sex-Typed

20

10

30

Androgynous

54

55

109

Undifferentiated

6

11

17

Total

86

89

175

X^=7.402
df=3

* p <.05

The results presented in Table 7 support the rejection of the null
hypothesis. Subjects in Group 1 (Intake) were found to score signifi-
cantly higher on the BSRI Femininity scale than subjects in Group 2
(Training) after four months of training (F-5.845, df=l/181, p <.05).
This difference indicates that the four months of training results in
a decrease in the degree of endorsement of feminine items among the CETA

114
subjects in this study, suggesting that working leads to a less feminine
self-perception. No differences were found between subjects in Group 1
and Group 2 on the BSRI Masculinity scale, the BSRI Social Desirability
scale, or the BSRI Sex-Type Classification.
Hypothesis 3

This hypothesis states: No differences exist in locus of control
as a result of training in the CETA Adult Work Experience program and
the CETA Public Service Employment program between Group 1 subjects and
Group 2 subjects as measured by the Internal-External Scale. This hypothe-
sis was tested using the one-way ANOVA procedure to compare the mean
scores of subjects in Group 1 and Group 2 on the Internal-External
Scale.

Group 1 subjects (Mean=10.04) were not found to differ significantly
from Group 2 subjects (Mean=9.59) on the Internal-External Scale
(F=0.767, df=l/177, p=NS) . The null hypothesis was accepted and it was
concluded that four months of training did not result in changes in the
locus of control among the CETA subjects in this study.
Hypothesis 4

Hypothesis 4 states: No differences exist between subjects in
Program 1 (Adult Work Experience) and subjects in Program 2 (Public
Service Employment) in self-actualization characteristics as measured
by the Personal Orientation Inventory. This hypothesis was evaluated
using the one-way ANOVA procedure to compare the mean scores of Program 1
subjects on the POI scales with the mean scores of the Program 2 subjects
on the POI scales. The results of these comparisons are presented in
Table 8.

115

Table 8

Comparison of Program 1 (Adult Work Experience) Subjects and
Program 2 (Public Service Employment) Subjects on the Personal
Orientation Inventory.

Mean

Mean

Program

Program

Square

Square

POI

1 Mean

2 Mean

Between

Within

df

F

Scales

Inner-directed

73.81

75.91

156.02

96.66

1/177

1.614

Other-directed

51.16

49.63

84.23

93.82

1/177

0.898

Time-competent

13.76

14.34

15.74

8.04

1/177

1.959

Time- incompetent

8.91

8.37

12.37

7.82

1/181

1.583

Self -Actualizing

Values

18.22

18.20

0.03

9.60

1/181

0.003

Existentiality

14.94

16.07

54.10

18.23

1/181

2.968

Feeling Reactivity

13.69

14.43

23.54

7.81

1/181

3.015

Spontaneity

10.61

11.16

12.53

4.66

1/181

2.691

Self-regard

11.87

12.35

9.75

6.03

1/181

1.616

Self -acceptance

13.37

13.30

0.22

8.87

1/181

0.024

Nature of Man

10.67

10.08

14.99

4.28

1/181

3.500

Synergy

5.85

5.29

13.21

2.22

1/181

5.939

Acceptance of

Aggression

14.31

14.90

14.44

9.17

1/181

1.575

Capacity for

Intimate Contact

16.40

17.55

56.04

14.61

1/181

3.835

* p <.05

Examination of the results presented in Table 8 supports the rejec-
tion of the null hypothesis on the basis of significant differences
obtained in the POI Synergy scale. All other POI scales did not exhibit
significant differences between Program 1 and Program 2 subjects. The
results indicate that Program 1 subjects were more flexible in their
perception of the dichotomies of life and able to perceive the dichotomies
as meaningfully related.

In addition to the significant difference obtained on the POI Synergy
scale, the results in Table 8 identify the existence of non-significant
trends toward differences between subjects in the two programs. Program 1
subjects tended to score lower than Program 2 subjects on the POI

116
Existentiality scale (F=2.97, df=l/181, p <.10), the POI Feeling
Reactivity scale (3.02, df=l/181, p <.10), and the POI Capacity for
Intimate Contact scale (F=3.84, df=l/181, p <. 10) . Program 1 subjects
tended to score higher than Program 2 subjects on the POI Nature of Man
scale (F=3.50, df=l/181, p <.10). These trends suggest that the subjects
in this study participating in Program 1 had higher levels of mental
health than Program 2 subjects in their view of human nature and the
dichotomies of life and lower levels of mental health than the Program 2
subjects in their sensitivity to their own feelings, their ability to
establish intimate interpersonal relationships, and the flexibility of
their values system.
Hypothesis 5

Hypothesis 5 states: No differences exist between Program 1 subjects
and Program 2 subjects in psychological androgyny characteristics as
measured by the Bern Sex-Role Inventory. This hypothesis was evaluated
using the one-way ANOVA procedure and the Chi-square test of homogeneity.
The reservations regarding the utility of the BSRI when used with the
CETA subjects in this study identified in the evaluation of Hypothesis 2
are applicable to the results discussed in relation to this hypothesis.
Examination of the results of this evaluation, presented in Table 9,
reveals no significant differences between Program 1 subjects and Pro-
gram 2 subjects on the BSRI scales and on the BSRI Sex-type classification.
Therefore, the null hypothesis was accepted.

117

Table 9

Comparison of Program 1 (Adult Work Experience) Subjects
and Program 2 (Public Service Emplojmient) Subjects on the
Bern Sex-Role Inventory

Mean

Mean

Program

Program

Square

Square

BSRI

1 Mean

2 Mean

Between

Within

df

F

Scales

Masculinity

5.7A

5.60

0.26

2.35

1/181

0.112

Femininity

5.96

5.52

5.71

2.20

1/181

2.597

Social

Desirabili

-ty

5.51

5.22

1.75

2.27

1/181

0.770

BSRI Sex-Type Classification
Distribution Table and Chi-Square Test

Sex-Type Classification

Program
1

Program
2

Total

Masculine Sex-Typed

8

11

19

Feminine Sex-Typed

14

16

30

Androgynous

36

73

109

Undifferentiated

3

lA

17

Total

61

114

175

X^=4.595
df=3

* p <.05
Hypothesis 6

This hypothesis states: No differences exist between Program 1
subjects and Program 2 subjects in locus of control as measured by the
Internal-External Scale. This hypothesis was tested using the one-way
ANOVA procedure to compare the mean score of the Program 1 subjects
with the mean score of the Program 2 subjects on the Internal-External
Scale.

Subjects in Program 1 tended to have a slightly higher mean Internal-
External Scale score than subjects in Program 2 (Program 1 Mean=10.29,

118
Program 2 Mean=9.56) indicating a slightly more external orientation
among Program 1 subjects than among Program 2 subjects. The difference
between the subjects in the two programs was a non-significant trend
(F=3.002, df=l/177, p <.10).
Hypothesis 7

Hypothesis 7 states: No relationships exist between variance in
self-actualization, psychological androgyny, and locus of control
characteristics of the subjects in this study and the independent demo-
graphic characteristics of age, sex, race, education, religious affili-
ation, family history, and length of unemployment prior to participation
in employment training. This hypothesis was evaluated using a three-way
ANOVA procedure to identify the amount of variance within the dependent
mental health scores of the subjects in this study explained by the
selected demographic characterisitcs, the training status of the subjects,
and the program status of the subjects.

Two of the factors in the three-way ANOVA procedure, the subjects'
training status (Factor 1: Group 1 versus Group 2) and the subjects'
program status (Factor 2: Program 1 versus Program 2), were the same
for all ANOVAs. Since only the mental health characteristics (the
dependent variable) and the demographic characteristics of the subjects
(Factor 3) changed across ANOVAs, each ANOVA will be identified by the
dependent variable and the demographic characteristic of that ANOVA.
Therefore, the Inner-directed/denoraination ANOVA refers to the three-
way analysis of the variance in the POI Inner-directed scale by Group
by Program by Denomination. The term "variance" is used to refer to the
measure of deviation of individual subject scores on the scale from the
Combined Group (Group 1 plus Group 2) Mean for that scale.

119

There were 18 dependent mental health scales analyzed in the evalu-
ation of the null hypothesis. These were the 14 POI scales, the three
subscales of the BSRI, and the Internal-External Scale. A total of 12
independent demographic variables were used in the evaluation of the null
hypothesis: age, sex, race, education, church membership, church atten-
dance, denomination (of church affiliation), parents, family position,
marital status, number of children (in subjects' current family), and
length of unemplojrment (prior to participation in employment training).
In all a total of 216 three-way ANOVAs were performed. Of these 216
only 30 ANOVAs were found to identify a significant relationship between
the variance in mental health scores of the subjects and one or more of
the three factors in the ANOVA. These results are presented in the order
of the mental health variable analyzed rather than by factors.

POI Inner-directed scale. The Inner-directed /denomination ANOVA
explained a significant amount of the variance in the POI Inner-directed
scores of the subjects in this study. The ANOVA obtained a significant
main effect on the Denomination Factor (F=3.004, df=3/163, p <.05) and a
significant interaction between the Program Factor and the Denomination
Factor (F=3.029, df=3/163, p <.05). The Scheffe' Test for Multiple Com-
parison of Group Means, referred to as the Scheffe' test, was used to
further examine the results of the ANOVA (Roscoe, 1975, pp. 313-315).
Application of the Scheffe' test revealed that the subjects in the
Fundamentalist subgroup scored lower on the POI Inner-directed scale
than did subjects in the Non-denomination, Baptist, and Liberal subgroups
combined (F=3.047, df=l/163, p <.05).

Application of the Scheffe' test to the Program by Denomination
interaction revealed the source of interaction to be a significant

120
difference between the subjects in the Program 1-Liberal subgroup and
the Program 2-Liberal subgroup. Program 1-Liberal subgroup subjects were
found to have POI Inner-directed scores at the same level as the Funda-
mentalist subgroup subjects of both programs. The Program 2-Liberal
subgroup subjects scored significantly higher than the Fundamentalist
and Program 1-Liberal subgroup subjects combined (F=2.04, df=7/163,
p <.05).

The interaction pattern between the Program Factor and the Denomi-
nation Factor on the POI Inner-directed scores of the CETA subjects will
be shown to repeat several times during the following discussion. In
subsequent analyses this interaction pattern will be referred to as simply
the Program-Denomination Interaction. Specific subgroups will be identi-
fied only when they deviate from the pattern identified above.

POI Other-directed scale. The Other-directed/denomination ANOVA
was found to account for a significant amount of the variance in the POI
Other-directed scores of the subjects in this study. The POI Other-
directed scale is conceptually and functionally the inverse of the POI
Inner-directed scale, decreasing as the Inner-directed scale increases.
As a result, the Denomination Factor main effect was significant in the
ANOVA (F=3.203, df=3/163, p <.05) as was the Program-Denomination inter-
action (F=3.841, df=3/163, p <.05). Application of the Scheffe'test
revealed that the Fundamentalist subgroup subjects scored significantly
higher on the POI Other-directed scale than the combined subgroup of the
subjects in the Non-denomination, Baptist, and Liberal subgroups (F=3.299,
df=3/163, p <.05). The Scheffe' test examination of the data revealed
the Program-Denomination interaction to be the exact inverse of the
POI Inner-directed Program-Denomination interaction (F=2.136, df=7/184,
p <.05).

121

POI Time-competent scale. The Time-competent/education ANOVA was
found to account for a significant amount of the variance in the POI Time-
competent scale scores of the subjects in this study. A significant main
effect was obtained on the Education Factor (F=3.469, df=7/169, p <.05).
The Scheffe' test revealed that subjects with an education level of at
least 11 years completed or more scored higher on the POI Time-competent
scale, were more time-competent, than subjects with less education
(F=2.477, df=7/169, p <.05).

POI Time-incompetent scale. The Time-incompetent/education ANOVA
was found to account for a significant amount of the variance in the POI
Time-incompetent scale scores of the subjects in this study. The POI Time-
incompetent and POI Time-competent scales are inversely related. The
ANOVA obtained a significant main effect on the Education Factor (F=3.382,
df=7/173, p <.01). The application of the Scheffe' test revealed that
subjects with at least 12 years of education scored lower, were less
time-incompetent, than subjects with less education (F=3.106, df=7/173,
p <.01).

POI Self-Actualizing Values scale. None of the three-way ANOVAs
was found to account for a significant amount of the variance in the
POI Self-Actualizing Values scale scores of the subjects in this study.

POI Existentiality scale. The Existentialiy/denomination ANOVA was
found to account for a significant amount of the variance in the POI
Existentiality scale scores of the subjects in this study. Two signifi-
cant interactions were obtained: the Program by Denomination interaction
(F=2.831, df=3/167, p <.05) and the Group by Program interaction (F=5.096,
df=l/167, p <.05). The Program by Denomination interaction was found to
follow the Program-Denomination interaction pattern identified in the

122
Inner-directed/denomination ANOVA with the exception that subjects in the
Program 2-Non-denomination subgroup scored as high as the Program 2-Lib-
eral denomination subgroup on the POI Existentiality scale. The applica-
tion of the Scheffe' test failed to identify significant differences in
the interaction pattern although it was quite close: F=1.99, df=3/180,
p <.10; where p=.05 at 3/180 degrees of freedom when F=2.06. Furthermore,
the Scheffe' test application did not yield significant results in the
Group by Program interaction effect (F=2.12, df=l/180, p <,10). A more
detailed analysis of the two interaction patterns identified the source
of both interactions to be situated among the scores of the Group 1 sub-
jects where the Program-Denomination interaction was repeated (F=6.353,
df=7/83, p <.01). The strength of this interaction among the Group 1
subjects' scores was sufficient to result in the significant interaction
effects identified in the overall ANOVA procedure.

POI Feeling Reactivity scale, POI Spontaneity scale. POI Self-Regard
scale. None of the three-way ANOVAs was found to account for a signifi-
cant amount of the variance in the scores of the CETA subjects in this
study on these three scales.

POI Self-acceptance scale. The Self-acceptance/denomination ANOVA
was found to account for a significant amount of the variance in the POI
Self-acceptance scale scores of the subjects in this study. The ANOVA
obtained a significant main effect on the Denomination Factor (F=4.412,
df=3/167, p <.01) and a significant interaction between the Denomination
and Program factors (F=3.774, df=3/167, p <.01). The pattern of responses
on both the main effect and the interaction pattern was similar to the
patterns identified in the Inner-directed/denomination ANOVA. The Scheffe'

123
test revealed that the subjects in the Fundaraentallst subgroup were
significantly less self-accepting than subjects in the other three sub-
groups of the denomination variable (F=3.588, df=3/180, p <.05). The
Scheffe' test also revealed that the Program-Denomination interaction
pattern identified above was repeated in the Self-acceptance/denomination
ANOVA (F=2.685, df=7/183, p <.05).

POI Nature of Man, Constructive scale. A significant amount of the
variance in the scores of the subjects in this study on the POI Nature
of Man scale was accounted for in six of the ANOVAs . The demographic
variables associated with these ANOVAs were: race, family position, mari-
tal status, number of children, age, and education.

Although no significant main effects or interaction effects were
obtained in the Nature of man/race ANOVA, the analysis was found to
account for a significant amount of the variance in the scores on the
POI Nature of Man scale. Two of the main effects were found to approach
significance: the Group Factor main effect (F=3.227, df=l/178, p <.10)
and the Program Factor main effect (F=3.124, df=l/178, p <.10). The
trends were for the subjects in Group 1 and Program 2 to score lower on
the POI Nature of Man scale than subjects in Group 2 and Program 1,
respectively, without either difference achieving significance.

The Nature of man/family position ANOVA obtained a significant main
effect on the Program Factor (F=4.424, df=l/158, p <.05) and a significant
interaction effect between the Group Factor and the Family Position Factor
(F=2.482, df=5/158, p <.05). Application of the Scheffe' test revealed
that Program 1 subjects scored higher on the POI Nature of Man scale than
Program 2 subjects (F=4.501, df=l/158, p <.05). Application of the
Scheffe' test to the interaction effect revealed that subjects in the

124
Group 2-Only Child subgroup scored higher on the POT Nature of Man
scale than subjects in the combined subgroup of Group 1-Only Child and
Group 1-Third Oldest Child while the other subjects scored in the midrange
(F=2.209, df=ll/171, p <.05).

The Nature of man/marital status ANOVA was found to obtain two
significant main effects. The Group Factor main effect was found to
reveal Group 2 subjects scoring higher on the POI Nature of Man scale
than Group 1 subjects (F=4.195, df=l/164, p <.05). The Marital Status
main effect was found to reveal the married subjects scoring lower on
the POI Nature of Man scale than the combined subgroups of single, divorced,
widowed, and separated subjects (4.100, df=4/183, p <.01).

The Nature of man/number of children ANOVA obtained a significant
main effect on the Group Factor. Group 2 subjects were found to score
higher on the POI Nature of Man scale than the Group 1 subjects (F=4.893,
df=l/173, p <.05).

The Nature of man/age ANOVA obtained a significant main effect on
the Age Factor (F=2,620, df=7/173, p <.05). Examination of the data with
the Scheffe' test revealed that subjects who were 31 years of age or older
scored higher on the POI Nature of Man scale than subjects 30 years of
age or younger (F=2.228, df=7/173, p <.05).

Finally, the Nature of man/education ANOVA obtained a significant
main effect on the Education Factor (F=3.167, df=7/173, p <.01). The
application of the Scheffe' test revealed that subjects with 13 years of
education or more scored higher on the POI Nature of Man scale than the
subjects with a high school education or less (F=3.013, df=7/173, p <.01).

In general the following observations may be made with regard to the
variance observed in the scores on the POI Nature of Man scale. Group 2

125
subjects were found to score higher on the POI Nature of Man scale than
Group 1 subjects. These findings support the rejection of Hypothesis 1
since the indication is that training does affect subjects' responses on
this scale when those responses are moderated by the subjects' marital
status and number of children. Program I subjects were found to score
higher on the POI Nature of Man scale than subjects in Program 2, further
supporting the findings discussed in the rejection of Hypothesis 4.
Furthermore, subjects who scored higher on this scale were found to be
older, unmarried, have more education, and more likely to be affiliated
with a liberal religious denomination, or no affiliation at all, than
subjects who scored lower on the scale. Higher scores on the POI Nature
of Man scale indicate a more positive perception of human nature while
lower scores indicate a negative perception of human nature.

POI Synergy scale. Variance in the POI Synergy scale scores of
subjects in this study was associated with the demographic variables of
race, denomination, marital status, age, and education. Two of the three-
way ANOVAs obtained Program Factor main effects indicating that Program 1
subjects scored significantly higher on the POI Synergy scale than the
Program 2 subjects. These two ANOVAs were the Synergy /race ANOVA
(F=5.506, df=l/178, p <,05) and the Synergy/denomination ANOVA (F=7.368,
df=l/178, p <.01).

The Synergy /marital status ANOVA obtained a significant main effect
on the Marital Status Factor (F=4.637, df=4/183, p <.001). The Scheffe'
test revealed the married subjects to have scored significantly lower on
the POI Synergy scale than the combined single, divorced, widowed, and
separated subjects (F=2.933, df=4/I83, p <,05).

126

The Synergy/age ANOVA obtained a significant main effect on the Age
Factor (F=2.858, df=7/173, p <.01). Subsequent application of the Scheffe'
test revealed a continuous increase in the POI Synergy scale scores of
subjects in this study as the subjects' age increased. Subjects 23 years
of age or older scored higher on the POI Synergy scale than subjects 22
years of age or younger (F=2.142, df=7/173, p <.05) and subjects who were
31 years of age or older scored higher on the POI Synergy scale than sub-
jects who were 30 years of age or younger (F=2.327, df=7/173, p <.05).

Finally, the Synergy/education ANOVA obtained two significant main
effects. The Program Factor main effect was significant with Program I
subjects scoring higher on the POI Synergy scale than Program 2 subjects
(F=6.745, df=l/173, p <.01). The Education Factor main effect was also
significant (F=2.078, df=7/173, p <.05). However, subsequent application
of the Scheffe' test to differences between subgroup means did not identify
significant differences between individual subgroups or combinations
thereof. The pattern of scores was generally a repeat of earlier
patterns discussed which identified a relationship between increased
levels of education and increased levels of self-actualization. Upon
examination of the results it was found that the main effect was mar-
ginally significant since the p <.05 significance value of F at 7/173
degree of freedom is F=2.06. The marginal nature of the significance
value in conjunction with the extremely conservative nature of the Scheffe'
test (Roscoe, 1975, pp. 313-315) resulted in the failure to identify the
specific source of the significant main effect.

In general, POI Synergy scale scores among the subjects in this study
were found to increase as the subjects' age or educational levels
increased. Furthermore, Program 1 subjects scored higher on the scale

127
than Program 2 subjects and unmarried subjects scored higher on the
scale than married subjects. The pattern of results identified on the
POI Synergy scale is very similar to the pattern of results discussed
in the preceding section on the POI Nature of Man scale. Together
these two scales form the Awareness dimension of the POI.

POI Acceptance of Aggression scale. The Acceptance of aggression/
denomination ANOVA obtained a significant Program-Denomination inter-
action effect (F=4.009, df=3/167, p <.01). Application of the Scheffe'
test revealed that this interaction pattern was a repeat of the Program-
Denomination interaction discussed earlier in the Inner-directed/denomi-
nation ANOVA (F=2.135, df=3/167, p <.05).

The Acceptance of aggression/education ANOVA obtained a significant
main effect on the Education Factor (F=2.741, df=7/173, p <.01). The
Scheffe' test analysis of the data revealed subjects who had 11 years of
education or more were more accepting of their own aggression as a natural
process than subjects who had completed less than II years of education
(F=2.287, df=7/173, p <.05).

POI Capacity for Intimate Contact scale. The Capacity for intimate
contact/denomination ANOVA was found to account for a significant amount
of the variance in the POI Capacity for Intimate Contact scale scores of
of the subjects in this study. The ANOVA obtained a significant inter-
action effect between the Program Factor and the Denomination Factor
(F=5.114, df=3/167, p <.01) and a significant three-way interaction
between the Group Factor, the Program Factor, and the Denomination Factor
(F=3.072, df=3/167, p <.05). Application of the Scheffe' test revealed
the Program-Denomination interaction to be a repeat of the interaction
pattern discussed earlier (F=2.483, df=7/167, p <.05). Examination of

128
the three-way interaction pattern with the Scheffe' test revealed the
source of the interaction to be a Group by Denomination interaction among
the Program 1 subjects. Program 1-Group 1-Fundamentalist subjects were
found to score higher on the POX Capacity for Intimate Contact scale than
Program 1-Group 2-Fundamentalist subjects (F=2.847, df=7/59, p <.05).

The Capacity for intimate contact/age ANOVA obtained a significant
main effect on the Age Factor (F=2.712, df=7/173, p <.05). Application
of the Scheffe' test revealed subjects who were 35 years of age or older
scored higher on the POI Capacity for Intimate Contact scale than subjects
who were 34 years of age or younger (F=2.326, df=7/173, p <.05).

The Capacity for intimate contact/education ANOVA obtained a signifi-
cant main effect on the Education Factor (F=2.896, df=7/173, p <.01).
Application of the Scheffe' test revealed that subjects who had completed
12 or more years of education scored higher on the POI Capacity for Inti-
mate Contact scale than subjects with less than a high school education
(F=2.259, df=7/173, p <.05).

Bern Sex-Role Inventory Masculinity scale. The Masculinity/family
position ANOVA was found to account for a significant amount of the variance
in the BSRI Masculinity scale scores of the subjects in this study. A
three-way interaction pattern between the Group Factor, the Program Factor,
and the Family Position Factor was obtained (F=3.219, df=5/158, p <.01).
Application of the Scheffe' test revealed the source of the interaction
to be among the scores of the Program 1 subjects. Among the Program 1
subjects an interaction pattern was identified where subjects who were
either the Oldest-Child or the Third-Oldest Child scored higher on the
BSRI Masculinity scale than the other Program 1 subjects (F=2.790,
df=5/69, p <.05) while Program 2 subjects did not exhibit this pattern.

129
Bern Sex-Role Inventory Femininity scale. The Femlnlnity/famlly
position ANOVA was found to account for a significant amount of the
variance in the BSRI Femininity scale scores of subjects in this study.
The ANOVA obtained a three-way interaction between the Group Factor, the
Program Factor, and the Family Position Factor (F=3.219, df=5/158.
p <.05). Examination of the results Identified the source of the
three-way interaction to be located in a Program by Family Position inter-
action. Subjects in the combined subgroups of Program 1-Only-Chlld,
Oldest-Child, and Third-Oldest Child were found to score higher on the
BSRI Femininity scale than the other Program 1 and Program 2 subjects
(F=1.870, df=ll/158, p <.05). The Femininity/family position ANOVA also
obtained a significant main effect on the Group Factor with Group 1
subjects scoring higher on the BSRI Femininity scale than Group 2 subjects
(F=6.66A, df=l/158, p <.05).

The Femininity/number of children ANOVA was also found to account
for a significant amount of the variance in the BSRI Femininity scores
of the subjects in this study. A significant main effect was obtained
on the Group Factor with Group 1 subjects scoring higher on the BSRI
Femininity scale than Group 2 subjects (F=6.335, df=l/173, p <.05).
Similarly, the Femininity/ length of unemployment ANOVA obtained a signifi-
cant main effect on the Group Factor with Group 1 subjects scoring higher
than the Group 2 subjects on the BSRI Femininity scale (F=7.145, df=l/173,
p <.01).

Bem Sex-Role Inventory Social Desirability scale. The Social desira-
bility/family position ANOVA obtained a significant three-way interaction
between the Group Factor, the Program Factor, and the Family Position
Factor to account for a significant amount of the variance in the BSRI
Social Desirability scale scores of the subjects in this study (F=3.972,

130
df=5/158, p <.01). As was the case in the other two BSRI scales, the
interaction was found to have its source located in the Program by
Family Position interaction. The Program 1-Only Child, Oldest-Child
and Third-Oldest-Child subjects were found to score higher on the BSRI
Social Desirability scale than the combined group of the remaining
Program 1 subjects and the Program 2 subjects (F=1.783, df=5/158, p <.05).

The Program by Family Position interaction pattern identified in
the three BSRI scales is unusual and resists ready explanation. The
confusing point about the pattern is the elevated scores among the
Program 1-Third-Oldest-Child subjects. This researcher is not familiar
with any theory of family position dynamics which would account for this
elevation of scores on all three sex-role scales. Further study of this
interaction pattern is needed before an adequate explanation may be
attempted.

Internal-External Scale. The Internal-external scale/race ANOVA was
found to account for a significant amount of the variance among the
Internal-External Scale scores of the subjects in this study. A signifi-
cant main effect v/as obtained on the Race Factor (F=6.160, df=2/17A,
p <.01). Application of the Scheffe' test revealed that non-white sub-
jects scored higher on the Internal-External Scale than the white subjects
(F=6.152, df=2/174, p <.01) indicating a more external locus of control
among the non-white subjects than among the white subjects.

The Internal-external scale/age ANOVA obtained a significant main
effect on the Age Factor (F=2.111, df=7/169, p <.05) to account for a
significant amount of the Internal-External Scale score variance of the
subjects in this study. The Scheffe' test revealed subjects who were 45
years of age or older scored in the more internal locus of control

131
direction than subjects who were 20 years of age or younger (F=2.058,
df=7/169, p <.05). This ANOVA also obtained a significant main effect
on the Program Factor with Program 1 subjects having a more external
locus of control orientation than Program 2 subjects (F=7.604, df=l/169,
p <.01).

The Internal-external scale/education ANOVA explained a significant
amount of the variance in the Internal-External Scale scores of subjects
in this study with a significant main effect in the Education Factor
(F=2.028, df=7/169, p=.05). Application of the Scheffe' test revealed
that subjects with 11 years or less of education had higher Internal-
External Scale scores, i.e., more external, than subjects who had 12 or
more years of education (F=2.223, df=7/169, p <.05).

Bem Sex-Role Inventory sex-type classification. The final mental
health measure to be examined was the BSRI Sex-type classification of the
subjects on the basis of their BSRI sex-role scale scores using the median-
split classification method. The classification is a categorical variable
and required the use of a series of Chi-square tests of homogeneity to
test Hypothesis 7. The results of these tests indicated that the indepen-
dent demographic variables of sex and race were related to the distribu-
tion of sex-type classifications. The significant Chi-square distribu-
tions are presented in Table 10 and Table 11.

Examination of the results in Table 10 reveals that the distribution
of sex-type classifications by sex is not uniform. Male subjects were
classified as Masculine sex-typed and Undifferentiated more frequently
than expected in the Chi-square test of homogeneity. Conversely, female
subjects in this study were classified as Feminine sex-typed and Andro-
gynous more frequently than the Chi-square test predicted. The finding

132
associating males with the masculine sex-type and females with the femi-
nine sex-type is not surprising. However, the noteworthy finding is that
female subjects were more likely to be classified as Androgynous while
male subjects were more frequently classified as Undifferentiated. The
indication is that the female subjects in this study reported a higher
level of sex-role flexibility than the male subjects in this study.

Table 10

Distribution of Bem Sex-Role Inventory Sex-Type Classifications
by the Demographic Variable of Sex for the Combined Group, Group
2 (Training), and Program 2 (Public Service Employment) Subjects.

Combined Group

Male Female Total

Masculine Sex-Typed Observed 10 9 13

Expected 6 13

Feminine Sex-Typed Observed 4 26 30

Expected 10 20

Androgynous Observed 32 77 109

Expected 34 74

Undifferentiated Observed 10 7 17

Expected 5 12

Total 56 119 175 x^=l^-5 **

df=3

Group 2 (Training)

Male Female Total

Masculine Sex-Typed Observed 7 6 13

Expected

Feminine Sex- Typed Observed 1 9 10

Expected

Androgynous Observed 13 42 55

Expected 17

Undifferentiated Observed 7 4 11

Expected

7

6

4

9

1

9

3

7

3

42

7

38

7

4

3

8

X^=12.0 **

Total 28 55 89 df=3

133
Table 10-continued

Program 2 (Public Service Employment)

Male

Masculine Sex-typed Observed 9

Expected 4

Feminine Sex-typed Observed 3

Expected 7

Androgynous

Undifferentiated

Total

Observed 27
Expected 31

Observed 10
Expected 6

A9

Female

2
6

13
9

46
42

8

12

65

Total

11

16

73

18

118

X^=16.3 ***
df=3

* p <.05
** p <.01
•** p <,001

Table 11

Distribution of Bern Sex-Role Inventory Sex-Type Classifications
by the Demographic Variable of Race for Group 1 (Intake) Subjects.

Group 1 (Intake)

Black

White

Total

72

11

Other Total

Masculine Sex-typed

Observed

5

0

1

6

Expected

5

1

0

Feminine Sex-Typed

Observed

12

7

1

20

Expected

17

3

1

Androgynous

Observed

49

4

1

54

Expected

45

7

2

Undifferentiated

Observed

6

0

0

6

Expected

5

1

0

86

X^=16.2 *
df=6

* p <.05

134

Examination of the results presented in Table 11 reveals that the
racial distribution of sex-type classifications is not uniform among the
subjects in Group 1 (Intake). The black subjects were found to be classi-
fied as feminine less frequently than the white subjects and more fre-
quently as androgynous than the white subjects. The results suggest that
among the Intake subjects in this study the black subjects reported a
greater amount of sex-role flexibility than the white subjects.

The most striking aspect of the distributions presented in Table 10
and Table 11, however, is not the relationship between the sex-type
classifications and demographic variables but the overall percentage of
subjects classified as androgynous. Of the 175 subjects in this study
appropriately completing the BSRI, 109 were classified as androgynous,
66% of the total. This is an unusually high percentage of subjects
classified as androgynous, suggesting that the BSRI may not be accurately
measuring sex-role flexibility among the CETA subjects in this study.

In general, the results discussed in the evaluation of Hypothesis 7
support the rejection of the null hypothesis and the acceptance of the
conclusion that a relationship does exist between the mental health
characteristics of the CETA subjects in this study and selected demographic
characteristics of those subjects. Of the 19 dependent measures of mental
health examined, 15 were found to have a significant amount of the vari-
ance in the scores of the subjects in this study accounted for by one or
more of the demographic variables. Of the 12 independent demographic
variables used in the study, eight were found to be related to the
variance in the scores on at least one of the dependent mental health
measures .

135

The independent demographic variable of Denomination of Church
Affiliation was found to contribute significantly to the explanation of
variance in the mental health scores of the subjects in this study on
six of the POI scales: Inner-directed support. Other-directed support,
Existentiality, Self-acceptance, Acceptance of Aggression, and Capacity
for Intimate Contact. An interaction effect between the Program Factor
and the Denomination Factor was obtained on all six scales. The inter-
action pattern consisted of subjects in the Program 2-Liberal denomination
subgroup scoring significantly higher, i.e., in the self-actualizing
direction, than subjects in the Program 1-Liberal denomination subgroup
and subjects in the Fundamentalist subgroup from either program. Further-
more, a significant main effect was obtained on the Denomination Factor
in the POI Inner-directed scale, the POI Other-directed scale, and the
POI Self-acceptance scale. The subjects in the Fundamentalist subgroup
were found to score significantly lower, in the less self-actualizing
direction, than the subjects in the combined group of Non-denominational,
Baptist, and Liberal subgroup. The overall pattern of responses on these
six POI scales indicated that those subjects affiliated with fundamental
denominations or with fundamental teachings within liberal denominations
were generally less self-actualizing than subjects affiliated with more
liberal denominations or teachings.

The independent demographic variable of Education was found to
contribute significantly to the explanation of variance in the mental
health scores of the subjects in this study on five scales: the POI Nature
of Man scale, the POI Synergy scale, the POI Acceptance of Aggression
scale, the POI Capacity for Intimate Contact scale, and the Internal-
External Scale. Subjects with more than a high school education were

136
found to score in the more mentally healthy direction than subjects
with less than a high school education.

Similarly, the independent demographic variable of Age was found to
contribute significantly to the variance in mental health scores among
subjects in this study on the POI Nature of Man scale, the POI Synergy
scale, the POI Capacity for Intimate Contact scale, and the Internal-
External Scale. It was found that subjects who were at least 30 years
of age or older generally scored in the more mentally healthy direction
on these scales than subjects who were younger.

The independent demographic variable of Race was found to contribute
significantly to the explanation of variance in the Internal-External
Scale scores among the subjects in this study. The non-white subjects
were found to score in the less mentally healthy, external locus of
control direction than white subjects.

The independent demographic variable of Marital Status was found to
contribute significantly to the explanation of variance in the mental
health scores of subjects in this study on the POI Nature of Man scale
and the POI Synergy scale. Married subjects were found to score in the
less self-actualizing direction on both of the scales to a greater extent
than the subjects who were single, divorced, widowed, or separated.

The independent variable of Family Position was found to contribute
significantly to the explanation of variance in the mental health scores
of subjects in this study on four scales: the POI Nature of Man scale,
the BSRI Masculinity scale, the BSRI Femininity scale, and the BSRI Social
Desirability scale. Group 1-Only-Child and Group 1-Third Child subjects
were found to score lower on the POI Nature of Man scale than Group 2-
Only-Child subjects. The Program 1-Oldest-Child and Program 1-Third-

137
Oldest-Child scored higher on the BSRI Masculinity scale, the BSRI
Femininity scale, and the BSRI Social Desirability scale than the
other subjects in either program.

Seven different independent demographic variables were found to
contribute indirectly to the explanation of variance in the mental
health scores of the subjects in this study. An indirect contribution
to the explanation of variance was identified when an ANOVA obtained a
significant main effect in either the Program Factor or the Group Factor
without obtaining a significant main effect in the Demographic Factor of
the ANOVA. As an example, the Femininity/number of children ANOVA and
the Femininity/length of unemployment ANOVA both obtained significant
Group Factor main effects in the explanation of the variance in BSRI
Femininity scale scores. The results of the ANOVAs indicated that neither
the Number of Children or Length of Unemployment variables contributed
directly to the significant explanation of the variance in the scores
while both demographic variables served to indirectly contribute to the
identification of a significant difference between subjects in Group 1
and subjects in Group 2. Similarly, the demographic variables of Marital
Status and Number of Children were found to contribute indirectly to the
identification of a significant difference between Group 1 subjects and
Group 2 subjects on the POI Nature of Man scale. The Family Position
variable was found to contribute indirectly to the identification of
a difference between subjects in Program 1 and Program 2 on the POI Nature
of Man scale. The demographic variables of Race, Education, and the
Denomination of Church Affiliation were found to contribute to the
identification of significant differences between subjects in Program 1
and Program 2 on the POI Synergy scale.

138
Finally, the Bern Sex-Role Inventory sex-type classifications were
found to be significantly related to the demographic variables of sex
and race. Male subjects were found to be classified more frequently as
masculine sex-typed and undifferentiated than females while female sub-
jects were classified more frequently as feminine sex-typed or andro-
gynous than the male subjects in the Combined Group, Group 2 (Training),
and Program 2 (Public Service Employment) . The finding was that the
female subjects in this study appear to report higher degrees of sex-role
flexibility than the male subjects. The other finding was that black
subjects were more likely to be classified as androgynous than white sub-
jects in Group 1 (Intake).

Comparison of CETA Subjects with Subjects
From Selected Validation Studies

The final analysis of the data collected in this study was the com-
parison of the Combined Group mean score on each of the dependent mental
health measures with the mean scores of subjects reported in selected
validation studies of each of the mental health measures. The results of
these comparisons are presented in Table 12, Table 13, and Table 14.

Examination of the results presented in Table 12 reveals that the
CETA sample in this study scored significantly lower on the POI than the
"normal" adult sample reported by Shostrom (1974, p. 24). The normal
adult sample was composed of subjects nominated as being neither self-
actualizing or non-self-actualizing by expert judges as a part of the
validation process. The results permit the observation that the CETA
subjects were less self-actualizing on 11 of the 12 POI scales than the
normal adult subjects. The CETA subjects were found to hold themselves
in as high a self-regard as the normal adult subjects.

139
Table 12

Comparison of CETA Subjects with Normal Adult Subjects Reported
by Shostrom (1974) on the Personal Orientation Inventory.

POI
Scale

Inner-directed
Time- competent
Self-Actualizing Values
Existentiality
Feeling Reactivity
Spontaneity
Self-regard
Self -acceptance
Nature of Man
Synergy

Acceptance of Aggression
Capacity for Intimate
Contact

Shostrom
Sample

(a)

Mean SD

87.2
17.7
20.2
21.8
15.7
12.7
12.0
17.1
12.4
7.6
16.6

18.8

13.6
2.8
3.0
5.1
3.3
2.9

4.6

CETA
Sample

Mean SD

75.1

14
18
15
14
11.0
12.2
13.3
10.3
5.5
14.7

17.1

9.9

2.9

3.1

4.3

2.8

2.2

2.5

3.0

2.1

1.5

3.0

3.9

* p <.05
** p <.01
*** p <.001

(a) Shostrom, 1974, p. 24

df

339
339
339
339
339
339
339
339
339
339
339

339

t-value

9.44 ***
11.55 ***

5.93 ***
12.15 ***

4.64 ***

2.27 **
.61
11.53 ***

9.64 ***
12.02 ***

5.25 ***

3.63 ***

Table 13

Comparison of CETA Subjects with Two Samples of College Students
Reported by Rotter (1966) on the Internal-External Scale.

Rotter Sample

CETA
Mean

Sample
SD

df

College

Mean

SD

t-value
5.19 ***

Ohio State University^^^

8.15

3.88

9.82

3.59

744

Florida State
University ^^^

9.05

3.66

9.82

3.59

300

1.79 *

* p <.05
** p <.01
*** p <.001

(a)
(b)

Rotter,
Rotter,

1966, p
1966, p

. 279
. 280

Examination of the results presented in Table 13 reveals a signifi-
cant difference in the locus of control orientation between the CETA sub-
jects and both samples of college students reported by Rotter (1966).

140

The CETA subjects were found to be more externally oriented in their

locus of control than either the racially mixed Ohio State University

students or the all black Florida State University students.

Table 14

Comparison of CETA Subjects with Two Samples of Stanford Univ-
versity Students Reported by Bem (1974, 1977) on the Bern Sex-
Role Inventory.

Sex/BSRI Scales

Bem
Sample

(a)

Mean

Male Subjects

Masculinity 4.97
Femininity 4.44

Social Desirability 4.91

Female Subjects

Masculinity 4.57
Femininity 5.01

Social Desirability 5.08

CETA
Sample

SD

Mean

SD

df

t-value

.67

5.77

1.57

633

8.80 ***

.55

5.42

1.43

633

12.19 ***

.50

5.24

1.54

633

3.96 ***

.69

5.60

1.52

468

9.91 ***

.52

5.70

1.44

468

7.33 ***

.50

5.37

1.51

468

2.98 **

BSRI Sex-Type Classification
Distribution Tables and Chi-Square Test

Sex-Type Classification

Androgynous

Undifferentiated

Bem^ -* CETA

Sample Sample Total

204

Observed

185

19

Expected

163

41

Observed

159

29

Expected

150

38

Observed

162

103

Expected

212

53

Observed

159

17

Expected

140

35

188

265

176

Total

665 168 833 x =88.24 ***

df=3

* p <.05
** p <.01
*** p <.001

(a) Bem, 1974, p. 161

(b) Bem, 1977, p. 198

141

Examination of the results in Table 14 reveals an entirely different
set of findings than the results in Table 12 and Table 13. The CETA sub-
jects were found to have higher mean scores on the three BSRI subscales
than the Stanford Unviersity subjects reported by Bem (1974). Furthermore,
the CETA subjects were found to have a higher frequency of subjects
classified as Androgynous than the Stanford University students reported
by Bem (1977). The results in Table 14 indicate that the CETA subjects
are more mentally healthy than the college students as measured by self-
reported levels of sex-role flexibility. This finding is discrepant with
the findings reported in Table 12 and Table 13 indicating that the CETA
subjects were less mentally healthy than the comparison subjects on
measures of self-actualization and locus of control. In view of the high
degree of intercorrelation identified between these three mental health
measures in the literature, the discrepancy identified in these findings
is indicative of an effect not previously obtained. The implications of
this discrepancy will be discussed in the following chapter.

Summary
In this chapter the data collected on the CETA subjects were analyzed.

The comparability of Group 1 (Intake) and Group 2 (Training) subjects was
established and the homogeneity of the comparison groups examined.
Differences between the subjects in Program 1 (Adult Work Experience) and
subjects in Program 2 (Public Service Employment) on selected demographic
characteristics were also examined. Seven null hypotheses were evaluated
on the basis of the data collected resulting in the rejection of Hypotheses
1, 2, 4, and 7 on the basis of identifying significant relationships
between the mental health characteristics of the CETA subjects and their
training status, their program status, and their demographic variables.

142
Finally, the CETA subjects were compared with "normal" adults and college
samples in selected validation studies for each of the mental health
measures used in this study. These comparisons identified a discrepancy
between the relative mental health characteristics of the CETA subjects
and the validation study subjects as measured by the three mental health
measures. The CETA subjects were found to be less mentally healthy than
the validation study subjects reported for the Personal Orientation
Inventory and the Internal-External Scale and more mentally healthy than
the validation study subjects on the Bern Sex-Role Inventory. These
results are discussed in the following chapter.

CHAPTER V
DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS

This study was designed to identify and investigate the mental
health characteristics of long-term unemployed persons as represented by
participants in a Comprehensive Employment and Training Act funded
employment training program. Previous research examining the psycho-
logical characteristics of long-term unemployed persons has been limited
to investigations of elements of mental illness. Research into the
effects of employment training has been limited to examinations of
changes in job retention, income levels, and social activities. Research
into mental health has been generally limited to subjects without an
extended history of employment difficulties. Relevant literature reviews
have repeatedly suggested an increase in the use of psychological
research on the impact of employment training upon long-term unemployed
persons.

This study was prompted by an interest in examining the impact of
CETA training on participants in mental health terms rather than degrees
of mental illness or changes in economic status. Furthermore, this
study was prompted by the desire to expand mental health research to
include persons with extended histories of employment difficulties.
Hopefully, the results of this study may serve as a basis for improving
the effectiveness of employment training and counseling strategies
intended for use with long-term unemployed persons.

143

144
Overview

During the Spring of 1980, 226 CETA participants were studied. The
mental health characteristics of these participants were compared on the
basis of training status, the program to which they were assigned, and
the demographic characteristics of age, race, sex, education, marital
status, number of children, parents, family position, church membership,
church attendance, denomination of religious affiliation, and length of
unemployment prior to beginning training. They were also compared with
subjects described in studies used in the validation of the Personal
Orientation Inventory, the Bern Sex-Role Inventory, and the Internal-
External Scale. All subjects in this study were participants in the
Adult Work Experience program (Program 1) or the Public Service Employ-
ment program (Program 2) of the CETA program examined in this study.

Subjects who were just beginning training were assigned to Group 1
(Intake) and subjects who had completed four months of training were
assigned to Group 2 (Training) . Group 1 and Group 2 were compared on
the basis of the independent demographic variables to establish between
group comparability prior to the evaluation of Hypotheses 1-3. A signifi-
cant (p <.05) between group difference was identified on the demographic
variable of education. This difference was eliminated by removing 35
subjects with extreme levels of education from the combined subject group.
Of the 191 subjects in the final combined subject group 96 were assigned
to Group 1 (Intake) and 95 were assigned to Group 2 (Training) . A total
of 69 subjects was receiving training in Program 1 (Adult Work Experience)
and 122 subjects were in Program 2 (Public Service Employment).

A mental health test package was administered to all subjects. The
test package contained an Introduction, a Personal Data Questionnaire,

145

the Personal Orientation Inventory, the Bern Sex-Role Inventory, and the
Internal-External Scale. Subjects with limited reading skills were
assisted in understanding the test items by pre-recording the complete
test package and playing it back to all subjects during the testing
procedure. Potential serial presentation effects were controlled by
random variation of the order in which the three mental health measures
were presented. The testing procedure was the final activity of the
pre-training orientation program for subjects in Group 1-Program 1.
Subjects in Group 1-Program 2 were tested just prior to the final activ-
ity of their pre-training orientation. All subjects in Group 2 were
tested during the two weeks following completion of four months of
training in special group sessions organized by the employment training
program staff.

The data collection process required two months to complete. A test
of within group homogeneity was performed to identify possible threats to
the validity of this study which may have resulted from the extended data
collection period. The test consisted of comparing subjects tested during
the first month of data collection with subjects tested during the second
month of data collection. Examination of the results of these comparisons
failed to identify systematic or pervasive significant differences within
comparison groups. It was concluded that the sample of data obtained
was sufficiently homogeneous for use in the evaluation of the hypotheses
examined in this study.

Discussion

Four major comparisons were made in the evaluation of the null
hypotheses and data analysis. The first comparison examined differences
between Group 1 (Intake) subjects and Group 2 (Training) subjects on

146

measures of self-actualization, psychological androgyny, and locus of
control (Hypotheses 1-3) . The second comparison consisted of examining
differences between participants in Program 1 (Adult Work Experience)
and participants in Program 2 (Public Service Employment) on the measures
of self-actualization, psychological androgyny, and locus of control
(Hypotheses 4-6) . The third comparison examined the relationships
between the mental health and demographic characteristics of the CETA
subjects in this study (Hypothesis 7). The final comparison examined
differences between the mental health characteristics of subjects in
this study and the mental health characteristics of subjects utilized
in the validation of the mental health instruments in this study.
Comparison 1; Group 1 (Intake) versus Group 2 (Training)

The first major comparison of this study was an examination of
changes in mental health among CETA subjects resulting from completion
of four months of training. The greatest difference obtained on the
POI was a non-significant (p <.10) trend which indicated that subjects
completing four months of training (group 2) tended to score higher on
the POI Nature of Man scale than subjects who were just beginning
training (Group 1). This difference was significant (p <.05) when the
independent demographic variables of marital status or number of children
were included in the analysis. This finding supported the rejection of
Hypothesis 1.

Shostrom (1974) defined the POI Nature of Man scale as a measure
of the degree to which "one sees man as essentially good ... [and] can
resolve the goodness-evil, masculine-feminine, selfishness-unselfishness,
and spirituality-sensuality dichotomies in the nature of man" (p. 18).
The data suggest that the CETA subjects in this study were more likely

147
to view human nature as basically good after four months of training
than before they began training. Since this difference in the perception
of human nature was most noticable when the subjects' marital status
and number of children were included in the data analysis it would
appear that some element of the training experience interacts with these
variables to alter the subjects' perception of human nature toward
improved mental health. Findings discussed in a later section of this
chapter suggest that this element is in part a function of economic
changes which result from entering training

The BSRI Femininity scale measures perceived-self femininity by
assessing the degree to which subjects will describe themselves with
feminine sex-role related attitudes and behaviors. Subjects in this
study who had completed four months of training (Group 2) ranked the
Femininity scale items of the BSRI as being significantly (p <.05) less
self-descriptive than did subjects who were just beginning training
(Group 1). Group 2 subjects were concluded to have significantly less
feminine self-perceptions than Group 1 subjects and Hypothesis 2 was
rejected.

Sex-role theory identifies work as a masculine sex-role related
behavior (Chafetz, 1974). CETA subjects, as a result of not working for
extended periods of time prior to beginning CETA training, have had
limited exposure to this masculine activity. It is reasonable to con-
clude that CETA subjects will begin to shift toward a more masculine
self-perception as they progress in training and become more acclimatized
to work. Based upon the results obtained in this study it would appear
that one manifestation of this shift is a decrease in the level of
perceived-self femininity after four months of employment training.

148
This finding is not surprising, in view of the high percentage of females
among the CETA subjects in this study (68%). However, based upon other
findings in this study, there is some question as to the validity
of instruments which contain sophisticated self-evaluative items, such
as the BSRI, when used with CETA subjects. Therefore, these findings
must be considered as tentative pending further research on sex-roles
and the utility of the BSRI for use with CETA subjects.

No differences were identified between subjects in Group 1 and
Group 2 on the Internal-External Scale. It was concluded that four
months of CETA training did not significantly affect the locus of control
orientation of the CETA subjects examined in this study. The acceptance
of Hypothesis 3 was supported.

The evaluation of Hypotheses 1-3 supports the conclusion that CETA
employment training had a limited effect upon the mental health charac-
teristics of the CETA participants examined in this study. This conclu-
sion suggests that either the development and enhancement of mental
health among the CETA participants is not one of the primary goals of
the CETA employment training staff or that the mental health enhancement
stategies employed by the training staff are not effective. There is
evidence to support both alternatives. It is the personal experience
of this researcher that the primary statistic used by the U.S. Department
of Labor to assess program effectiveness is the percentage of partici-
pants completing training and entering fulltime, unsubsldized employment.
As a consequence, although employment training programs include coun-
seling services as provided for by the Act, the primary goal of the
programs is more often concerned with completion of training and job
placement rather than with development of mental health. Improvement in

149

the mental health of the participant is apparently a byproduct of training

rather than a targeted goal. At the same time, findings from this study

which are discussed later suggest that certain counseling strategies may

not be effective with long-term unemployed persons.

Comparison 2; Program 1 (Adult Work Experience) versus Program 2
(Public Service Employment)

This comparison examined demographic and mental health differences
between participants in Program 1 and Program 2. Program 1 participants
differed significantly (p <.001) from Program 2 participants on six of the
twelve independent demographic variables. Program 1 participants were
more likely to be female, unmarried, and older with more children,
less education, and a longer history of unemployment prior to entering
training than Program 2 participants. These differences support the
conclusion that the two programs serve different segments of the unemployed
population in this study. Program 1 is apparently used by the CETA
program participating in this study to serve the unskilled, unmarried,
welfare mother with an extended unemployment history while Program 2 is
used to serve married high school graduates with limited skills and a
relatively short history of unemployment.

The evaluation of Hypotheses 4-6 comparing the mental health differ-
ences of participants in Program 1 and Program 2 identified only one
significant difference. Participants in Program 2 were found to be
significantly (p <.05) more self-actualizing in their ability to perceive
life's dichotomies as being meaningfully related (POI Synergy scale).
The data supported the rejection of Hypothesis 4 and the acceptance of
Hypotheses 5 and 6.

150
However, there was a contradictory pattern of trends which suggested
the existence of mental health differences between participants in the
two programs which warrant discussion. As noted above. Program 1 par-
ticipants were more self-actualizing than Program 2 participants in
their synergistic perception of life's dichotomies as being related.
A non-significant trend (p <.10) also indicated a tendency for Program 1
subjects to be more self -actualizing in their perception of human nature.
In contrast. Program 1 participants exhibited non-significant (p <,10)
trends to be less self-actualizing than Program 2 participants in their
poorer sensitivity to their own emotions, in their less flexible values,
and their limited capacity to establish meaningful interpersonal rela-
tionships. Program 1 participants also tended (p <,10) to be more
external in their locus of control orientation than Program 2 partici-
pants. The pattern of these trends is inconsistent with previous
research which found that increased mental health in one area would be
associated with increased mental health in other areas, Shostrom (1974)
has reported a high degree of positive intercorrelation among the
scales of the POI. Likewise, a strong positive correlation has been
reported between high scores on the POI and an internal locus of control
as measured by the Internal-External Scale (White, 1971), It is reason-
able to suggest, therefore, that there exist some characteristics of the
participants in the different programs which result in the discrepant
mental health differences identified. Findings of this study, discussed
in a later section, indicate that this discrepancy is associated with
the independent demographic variables of age, sex, race, and religious
affiliation.

151

Comparison 3: Mental Health Variables versus Demographic Variables

The third major comparison of this study was an examination of the
relationship between the mental health characteristics and selected
demographic characteristics of the CETA subjects. Several significant
relationships were identified between mental health and the demographic
characteristics of the subjects,

CETA participants affiliated with liberal religious denominations
were generally more self-actualizing than those participants affiliated
with fundamentalist denominations. However, a significant interaction
effect indicated that Program 1 participants affiliated with liberal
denominations exhibited levels of self-actualization similar to partici-
pants affiliated with fundamentalist denominations. This interaction
suggests that fundamentalist teachings exist within the liberal denomi-
nations which Program 1 participants are more likely to endorse than
Program 2 participants. Overall, CETA participants affiliated with
liberal religious teachings were more inner-directed, more accepting of
themselves, more accepting of aggression as a natural process, more
flexible in the application of their values, and better able to establish
meaningful interpersonal relationships than participants affiliated with
fundamentalist religious teachings. These findings are consistent with
previous research which has identified a strong inverse relationship
between self-actualization and association with a fundamentalist religion
(Anderson, 1973; Burke, 1973; Reynolds, 1970).

The independent demographic variables of age and education were
found to be associated with the mental health levels of CETA subjects
examined in this study. Those subjects with approximately 12 years of
education or more, or who were 30 years of age or older, were generally

152
more mentally healthy than subjects who had less education or who were
younger. Specifically, the more educated or older subjects were more
internal in their locus of control, more oriented in the here and now,
more likely to perceive the dichotomies of life as being meaningfully
related, more positive in their perception of human nature, better able
to accept their aggression as being natural, and better able to establish
meaningful interpersonal relationships.

A strong relationship was found to exist between race and locus of
control orientation. As suggested by previous research (Battle & Rotter,
1963; Hall et al., 1977; Lefton, 1968) the white subjects in this study
were more internal in their locus of control than the non-white subjects.
These findings are consistent with the findings of Gurin, P. et al.
(1968), since participation in CETA employment training is voluntarily
initiated and maintained by the participants.

Unmarried CETA subjects were more self-actualizing than married
CETA subjects. Unmarried subjects were more likely to perceive the
dichotomies of life as being meaningfully related and were more likely
to have a positive view of human nature than married subjects. Some
characteristic of the marriage relationship or of married persons in
this study would appear to limit the development of these self-actual-
ization characteristics among the married subjects. The exact source
of this limiting effect must be the subject of future research.

Sex-role flexibility was related to the sex and race of the CETA
subjects. In three of the comparison groups (the Combined Group, Group
2, and Program 2) the male subjects were less willing to describe them-
selves with feminine sex-role related items than were female subjects
to describe themselves with masculine sex-role related items. As a

153

result, a higher percentage of females were classified as androgynous
than were males and a higher percentage of males were classified as
undifferentiated or sex-typed than were females. Among the subjects in
Group 1, the black subjects were more willing to describe themselves
with opposite-sex sex-role related items, indicating a higher level of
sex-role flexibility, than were white subjects. In general, the data
suggest that the blacks and females tended to be more flexible in their
sex-role self-descriptions than were males and whites among the subjects
in this study. This finding may reflect the predominantly female (68%)
and black (82%) composition of the CETA sample since, as discussed pre-
viously, work is primarily a masculine sex-role related activity.
Working, therefore, would require the development of greater flexibility
in sex-role self-perceptions among females than males. However, in view
of the tentative nature of the data obtained with the BSRI as discussed
earlier, these findings must await further clarification.

Finally, a relationship was identified between mental health and
the family position of CETA subjects. A significant interaction in the
data analysis indicated that subjects who were the oldest or only child
in their families were more self-actualizing in their view of human
nature as positive than were other CETA subjects. Furthermore, this
same group had higher levels of perceived-self masculinity, feminity,
and social desirability, i.e., scored higher on these three subscales
of the BSRI, than the other CETA subjects. While evidence in the litera-
ture suggests that higher levels of mental health might be expected
among subjects who are the oldest or only child in their families
(Crandall, V. C.,et al., 1965; Gibb, 1967), no evidence has been found
to suggest increased levels of mental health among subjects who are the

154
third oldest member of their families. This pattern is unusual and
further research is needed in this area.

Hypothesis 7 was rejected because these findings identified strong
relationships between mental health and demographic characteristics of
the CETA subjects in this study. Of the 12 demographic variables
selected for study, denomination, age, education, race, marital status,
sex, number of children, and family position were found to be related
to mental health differences.
Comparison 4: CETA subjects versus Subjects in the Validation Studies

The final comparison was an examination of mental health differences
between CETA subjects and subjects reported in studies validating the
three mental health measures used in this study. Previous research has
identified a strong positive relationship between the concepts of self-
actualization, psychological androgyny, and locus of control as measured
by the Personal Orientation Inventory, the Bern Sex-Role Inventory, and
the Internal-External Scale (Bass, B., & Stek, 1972: Hochreich, 1975;
House, 1974; Lambert et al., 1976; Nevill, 1977; Ott, 1976; Pettus,
1976; Wall, 1970; White, 1971).

The relative levels of mental health found by comparing the CETA
subjects with the subjects in the validation studies are inconsistent
with previous research. CETA subjects were significantly (p <.01) less
self-actualizing than the sample of normal adults (defined as being
neither self-actualizing nor non-self-actualizing) reported by Shostrom
(1974, p. 24), Similarly, CETA subjects were significantly (p <.05)
less internal in their locus of control than samples of students from
Ohio State University and Florida State University (Rotter, 1966,
pp. 279-280). In contrast, however, CETA subjects exhibited a

155

significantly (p <.05) higher level of mental health as measured by sex-
role flexibility than subjects reported by Bern (1977, p. 198). The CETA
subjects also had higher scores on the three subscales of the Bern
Sex-Role Inventory than did Stanford University students used by Bern
(1974, p. 161) in the development of the BSRI. These contrasts suggest
that the data measured the existence of factors other than relative
differences in mental health among the subjects in this study.

Perhaps one factor reflected in these findings is the ability to
engage in accurate self-evaluative behavior. The only mental health
scale which did not differ significantly between the CETA subjects and
the subjects in the validation studies was the POI Self-regard scale.
The CETA subjects were found to hold themselves in as high a self-regard
as the "normal" adults reported by Shostrom (1974, p. 24). The high
self-regard level is noteworthy since the CETA subjects were less men-
tally healthy than the normal adult subjects on every other POI scale
and than college students on the Internal-External Scale. The solitary
elevation of the POI Self-regard scale, in association with the relative
depression of the other POI scales and Internal-External Scale, serves
to suggest that the CETA subjects hold themselves in comparatively high
regard without the relative strengths to support this high regard. They
do not, therefore, appear to be engaging in accurate, critical self-
analysis.

An examination of the item structure in the three mental health
measures used in this study and the pattern of responses to these items
also serves to indicate that the CETA subjects experienced difficulty
in self-evaluative activities. The items used in the POI and the lES
are constructed of pairs of statements. Subjects are required to select

156
the statement in each pair which is more true for them. This is a
relatively structured task permitting, at the most, three choices:
either statement 1 is true, statement 2 is true, or both statements are
true and the item, is left blank. The items in the BSRI, however, require
much more sophisticated responses. Each item on the BSRI requires the
subjects to decide, on a scale of 1-7, how accurately the items describe
themselves. The BSRI items clearly require a more complex self-evalu-
ative response than either the POI or the lES.

The unusual response pattern observed on the BSRI and reported in
Chapter IV (page 112) becomes more significant in view of the relative
complexity of the items. It was observed that 30% of the subjects cor-
rectly responding to the BSRI did so in a systematic fashion. These
subjects ranked a majority of the 60 items at one or two levels: either
mostly "7's," mostly "I's" and "7's," mostly "6's" and "7's," or mostly
"2's" and "6's." It seems unlikely that a majority of the items would
realistically be ranked in such a regular pattern if the subjects were
not experiencing difficulty in responding to the items. Furthermore,
these responses resulted in 66% of the CETA subjects being classified
androgynous as compared to the 24% so classified among the Stanford
University students (Bem, 1977, p. 198). It seems reasonable to conclude
that CETA subjects experienced difficulty in accurately evaluating the
personal applicability of each BSRI item because of the complexity of
the response required.

These findings raise serious questions about the applicability of
the BSRI to populations which might be expected to have difficulty with
sophisticated self-evaluation. As a consequence, the findings obtained
with the BSRI on the mental health characteristics of CETA participants

157
are not discussed further. There is a need for research to develop a
more appropriate sex-role inventory for use with the CETA-type subjects.

Conclusions

The findings of this study support the following conclusions.
Conclusion 1

The CETA employment training program investigated in this study
does not appear to improve the mental health of the participants. With
the exception of an improved perception of human nature as being inher-
ently good, no significant mental health differences were identified as
a result of four months of training among the CETA subjects examined.
The improved perception of human nature was significant only when the
subject's marital status and number of children were included in the
comparison.
Conclusion 2

The CETA Adult Work Experience program and the CETA Public Service
Employment program seem to be serving different elements of the unem-
ployed population.

Mental health differences. The mental health profile of partici-
pants in the CETA Adult Work Experience program and the CETA Public
Service Employment program appeared to differ. The Adult Work Experience
participants tended to be more mentally healthy than the Public Service
Employment participants in their ability to perceive human nature as
being basically good and in their ability to experience the dichotomies
of life as being meaningfully related. The Public Service Employment
participants tended to be more mentally healthy than Adult Work Experi-
ence participants in their greater sensitivity to their own emotions,
more flexible values, greater ability to establish intimate interpersonal
relationships, and more internal locus of control.

158

Demographic differences. The demographic characteristics of those
in the two programs were significantly different. The Adult Work Experi-
ence participants were more likely to be female, older, and unmarried
with more children, less education, and a longer history of unemployment
than the Public Service Employment participants.
Conclusion 3

Mental health differences among the CETA subjects in this study were
related to differences in their demographic characteristics.

Religious affiliation. CETA subjects affiliated with liberal reli-
gious teachings seemed to be more mentally healthy than CETA subjects
affiliated with fundamentalist religious teachings by virtue of their
more inner-directed support, increased interpersonal sensitivity, greater
self-acceptance, and greater flexibility of values.

Education. CETA subjects with at least a high school education
appeared to be more mentally healthy than CETA subjects with less than a
high school education. CETA subjects with more education were more ori-
ented in the here and now, interpersonally sensitive, able to experience
the dichotomies of life as being meaningfully related, likely to perceive
human nature as being good, and more internally oriented.

Age. CETA subjects who were approximately 30 years of age or older
indicated that they were more mentally healthy than CETA subjects who
were younger. The older subjects were better able to experience the
dichotomies of life as being meaningfully related, to perceive human
nature as being good, to establish meaningful interpersonal relation-
ships, and exhibited a more internal locus of control.

Marital status. CETA subjects who were not married appeared to be
more mentally healthy than the married subjects on the basis of their

159

greater ability to perceive human nature as basically good and to
experience the dichotomies of life as meaningfully related.

Race. White subjects indicated greater mental health than the non-
white subjects by virtue of being more Internally oriented.
Conclusion 4

CETA subjects in this study were apparently less mentally healthy
than subjects reported in validation studies of the Internal-External
Scale and the Personal Orientation Inventory. CETA subjects were more
external in their locus of control than college students and less self-
actualizing than normal adults.
Conclusion 5

CETA subjects in this study seemed to have difficulty with accurate
self -evaluation. Comparisons of CETA subjects with samples of college
students and normal adults included in the validation studies of the
Internal-External Scale and the Personal Orientation Inventory indicated
that the CETA subjects held themselves in relatively high regard without
substantial support for that opinion. Furthermore, the CETA subjects'
response patterns on the Bex Sex-Role Inventory suggested difficulty in
responding to the relatively more sophisticated questionnaire items.
Conclusion 6

The usefulness of the Bern Sex-Role Inventory for use in assessing
the sex-roles among the CETA subjects is questionable. CETA subjects
exhibited difficulty in responding to the more sophisticated BSRI items
in an accurate and consistent manner. The findings obtained through the
use of the Bex Sex-Role Inventory are identified as tentative pending
further research.

160
Mental Health of the CETA Subjects: A Composite Picture

Based upon these conclusions, it is possible to generate a composite
mental health picture of the CETA subjects in this study. Although this
picture is in part speculative, the intent is to propose an overall
picture of the mental health of the CETA subjects. The composite picture
is composed of elements drawn from the combined Group, Program 1 versus
Program 2, and Group 1 versus Group 2 comparisons.
Element 1 : The Combined Group

In comparison with normal adults and with college students, CETA
subjects were found to be less mentally healthy in their locus of con-
trol and levels of self-actualization. CETA subjects exhibited an unusu-
ally elevated level of self-regard without the supporting strengths to
justify the opinion. These findings are consistent with research identi-
fying increased levels of mental illness among persons with extended
histories of unemployment (Tiffany, D. et al., 1970; Work in America,
1973). Jourard (1963) notes that one of the most common forms of poor
mental health is an inaccurate or incomplete self-concept resulting
from a failure to acknowledge unpleasant or undesirable elements of
the self.

Mental health differences among the CETA subjects in this study
were found to be related to several demographic characteristics.
Subjects who were older, unmarried, or had completed high school were
found to have higher levels of mental health than those who were younger,
married or less educated. Subjects affiliated with liberal religious
teachings were found to have higher levels of mental health than subjects
affiliated with more fundamentalist teachings. Finally, white subjects
were found to have a more internal locus of control than non-white
subjects.

161

Based upon these findings a composite mental health picture of the

demographic characteristics of the typically most mentally healthy subject

would a unmarried, white, female over the age of 30 with a high school

education and affiliated with liberal religious teachings. This rather

general mental health picture of the subjects in this study may be further

refined by examining mental health differences between different groups

of subjects.

Element 2; Program 2 (Adult Work Experience) versus Program 2 (Public
Service Employment)

Mental health differences and demographic differences among the CETA
subjects were found to be associated with differences in program partici-
pation. Program 1 participants were found to have higher mental health
levels than Program 2 participants in their ability to experience the
dichotomies of life as being meaningfully related and to perceive human
nature as being inherently good. However, Program 1 participants exhibited
a trend toward lower mental health levels by virtue of having a more
external locus of control, poorer sensitivity to their own emotions, less
flexibility of values, and a poorer ability to establish meaningful
interpersonal relations. In addition, an interaction pattern between
religious affiliation and program participation indicated that Program 1
participants affiliated with liberal denominations were more likely to
endorse the more fundamental teachings within the liberal denominations.
Finally, Program 1 participants were found to differ markedly from Program
2 participants on the basis of demographic characteristics. Program 1
participants were more likely to be female, older, and unmarried with
more children, less education, and a longer history of unemployment
than Program 2 participants.

162

Based upon the strength of the demographic differences and the
pattern of trends in mental health differences between participants in
the two programs, it is possible to suggest the existence of two patterns
of mental health within the general composite mental health picture.
First, increased mental health as measured by perceptions of human nature
and the dichotomies of life appears to be associated with being female,
older, unmarried, and having children. Second, increased mental health
as measured by locus of control, sensitivity to one's own emotions,
flexibility of values, and capacity for intimate contact appears to be
associated with being white, being affiliated with liberal religious
teachings, having at least a high school education, and having a shorter
history of unemployment.
Element 3: Group 1 (Intake) versus Group 2 (Training)

The final element of the composite picture of the mental health of
the CETA subjects in this study is the effect of four months of training
on their mental health. The only mental health difference which was
associated with completion of four months of training was a tendency for
subjects in Group 2 to have a more positive perception of human nature.
This trend became significant when the subjects' marital status and
number of children were included in the comparison.

In view of the preceding discussion it would appear that unmarried
mothers with limited education and an extended history of unemployment
responded to training by perceiving human nature more positively. While
the specific element of training is not readily evident, it seems possi-
ble that one of the most immediate effects of training is that the par-
ticipants begin to receive a regular income, either again or for the
first time. The suggestion is that the perception of human nature as

163
being inherently good Is sensitive to economic pressures, especially
when the person Is economically responsible for others.

Recommendations for Counseling

Based upon the conclusions of this study, several recommendations
may be made for counselors and trainers working with CETA clients.
Recommendation 1

It is recommended that CETA counseling staffs develop counseling
strategies which are not dependent upon introspection or responses to
open-ended questions. The CETA subjects in this study were unable to
respond to test items which required sophisticated, self -evaluative
behavior. CETA clients may be expected, therefore, to have a noticeable
degree of difficulty in responding to interventions which require them
to engage in complex self-analysis, such as open-ended paraphrases and
questions. It has been this researcher's experience that CETA clients
are more receptive to directive or structured counseling efforts. As an
example, the question, "In this situation were you happy or sad?" was
found to elicit a more productive response than the question, "How did
you feel when this happened?". While arguments have been made that this
type of approach limits the client's response, it has been this
researcher's experience that only be limiting the range of possible
responses is it possible to obtain genuine responses. The findings of
this study support the experience that ambiguous, open-ended questions
will result in ambiguous, uninformative responses.
Reconmiendation 2

CETA counselors should reexamine the manner in which their role and
function as counselors is presented to CETA clients. Traditionally,
counselors are presented as being available to "help" the clients deal

164

with problems. However, if the difficulty in self-evaluation identified

in this study results in an inability to perceive problem areas in the

self then CETA clients may not respond to efforts to "help" them. CETA

clients encountered by this researcher were most open to help offered

within the context of current problems. As an example, CETA clients have

been more willing to admit the need for help when failure to obtain help

would clearly result in some aversive consequence, such as loss of job.

Without the identification of a problem which the clients could not aviod ,

ignore, or discount, the presentation of help from the counselor was

often spurned as insulting and unnecessary. To recall a quote from

May (1960):

There is no such thing as truth or reality for a living human
being except as he participates in it, is conscious of it,
has some relationship to it. (p. 14)

Recommendation 3

It is recommended that CETA counseling staffs carefully reexamine

CETA clients' understanding of work related statements. Previous

research has identified strong similarities between blue-collar workers

and unemployed persons in their respective endorsement of work related

values and motivation to work (Cook, 1971; Kaplan & Tausky, 1974;

Searls et al., 1974). However, in view of the findings in this study

regarding self-evaluative behavior among CETA clients it may be that

statements of desire to work or endorsing work related values and

activities with which the clients have no prior experience are made

without a full understanding of the implications of such statements.

As an example, the work related value of punctuality may be endorsed

without considering how it will feel to get out of bed early enough to

be at work on time. The resultant discomfort of having to get out of

165

bed early or face an angry boss for being late may cause the client to
withdraw from the program. CETA counselors should anticipate such
discrepancies between the clients' stated and actual values and habits
and prepare interventions to help the clients overcome the shock asso-
ciated with the discovery that there is more to working than receiving
a paycheck.
Recommendation 4

CETA counselors should be sensitive to the demographic character-
istics of their clients. Differences in mental health among CETA
subjects in this study were found to be related to differences in the
subjects' age, race, sex, education, marital status, number of children,
religious affiliation, and length of unemployment. The clients' willing-
ness and ability to engage in successful attitudinal change or mental
health counseling may well be predicted and affected by their demographic
characteristics.
Recommendation 5

CETA counselors should also be aware of the program in which the
clients are training. Different mental health patterns were identified
among CETA subjects in this study on the basis of the program in which
they were participating. Although there are several possible explana-
tions for these different patterns such as different selection criteria
for the programs, demographic differences between the clients in the
programs, or the effects of different training staffs or activities, it
is clear that these patterns exist and should be considered in the prepa-
ration of training and counseling interventions.
Recommendation 6

It is recommended that the counseling element of CETA employment
training programs be restructured to facilitate the development of

166
mental health among CETA clients. As discussed in preceding sections,
the goals of CETA training programs are currently focused upon the place-
ment of clients in fulltime employment. Other training outcomes seem to
be secondary to this goal. It is not surprising, therefore, that the
CETA subjects examined in this study were found to exhibit limited
improvement in mental health as a result of four months of training.
Previous research has demonstrated a strong relationship between good
mental health and good productivity (Tiffany, D.,et al., 1970; Work In
America, 1973). For this reason, greater attention to the improvement
of mental health seems likely to increase the chances for satsifactory
work performance.

Recommendations for Further Research

Several areas of interest for further research are indicated by
this study.
Recommendation 1

This study should be replicated in other regions of the nation.
Subjects participating in this study were enrolled in a CETA training
program located in a north Florida metropolitan area. It would be useful
to determine whether similar mental health differences exist among unem-
ployed persons in other regions of the nation.
Recommendation 2

A related question is whether mental health differences exist
between unemployed persons applying for and participating in CETA
training and unemployed persons who live in the same area but do not
apply for or participate in the training program. Research in this area
may prove difficult, however, in view of reported difficulty in obtain-
ing a stable research sample from this extremely transient population
(Barnes, 1972; Hardin, 1972; Lewis, 1972).

167
Recommendation 3

Research is needed to determine the length of CETA training required
to effect a significant improvement in the overall mental health of CETA
clients. CETA subjects in this study exhibited little change in mental
health following four months of training. A longitudinal study of
changes in the mental health of unemployed persons as they enter and
complete training would be beneficial.
Recommendation 4

Further research should be conducted to investigate the effective-
ness of various counseling and training strategies for facilitating the
development of mental health among CETA clients. In view of the diffi-
culties experienced with accurate self-analysis by CETA subjects in this
study, counseling strategies other than one-on-one, client-centered
therapy should be studied.
Recommendation 5

Research is needed to investigate sex-roles and sex-role flexibility
of CETA clients as well as to identify an appropriate sex-role instrument
for use with CETA clients. The CETA subjects in this study were unable
to respond with sufficient accuracy and consistency to the items on the
Bern Sex-Role Inventory to produce reliable information regarding sex-
roles and sex-role flexibility.
Recommendation 6

Research would be helpful to examine further the relationships
between mental health differences and demographic characteristics of CETA
clients. This study identified several significant relationships between
the mental health levels of the CETA subjects and their demographic
characteristics of family position, marital status, number of children.

168
length of unemployment, religious affiliation, age, sex, race, and
education. Determining how these characteristics are associated with
mental health differences would be beneficial.
Recommendation 7

Finally, further research would be helpful in identifying the
sources and implications of the self-analysis difficulties identified
among the subjects in this study. The difficulty experienced by CETA
subjects in accurately and consistently completing self-evaluative tasks
has significant implications for future planning of counseling strategies
for CETA clients. Improved understanding of this mental health dimension
of CETA clients would greatly enhance the planning and development of
counseling and training activities designed for the CETA eligible, unem-
ployed population.

Summary

This study sought to examine the mental health of unemployed persons
from the perspective of humanistically oriented mental health theory.
This study grew out of the need to understand and help the long-term
unemployed and particularly, the need to integrate employment training
and mental health. Hopefully, through identification of mental health
characteristics among unemployed persons participating in a CETA funded
employment training program, this study will provide a foundation for
improved understanding of the unemployed population of our society.
Such an understanding can contribute to the development of more effective
measures for improving the mental health of unemployed persons and
assisting them to live as more fully functioning members of the American
society.

APPENDIX A
CETA ELIGIBILITY CRITERIA
Participation in employment training activities funded under the
Comprehensive Employment and Training Act of 1978 is contingent upon
meeting federally established eligibility criteria. These criteria for
the programs examined in this study are presented as follows.
Title II-B— An eligible participant must:

1. Be a resident of the geographical area served by the CETA

program from which training is to be received.

2. Be economically disadvantaged.* (Annualize the last six months

income to determine annual income.)

3. Be unemployed for seven (7) days prior to submitting the

application, underemployed, or in school (youth only).

Title II-D— An eligible participant must:

1. Be a resident of the geographical area served by the CETA

program from which training is to be received.

2. Be economically disadvantaged.*

3. Be unemployed at the time of application.

A. Have been unemployed at 15 of the 20 weeks prior to the

date of application.
5. Not have voluntarily terminated without good cause in the last

six months, his/her last fulltlme position paying at least

minimum wage.

Title VI— An eligible participant must:

1. Be a resident of the geographical area served by the CETA

program from which training is to be received.

2. Be unemployed at the time of application.

3. Be unemployed for at least 10 of the 12 weeks prior to the

date of application.
A. Have a family income which does not exceed 100 per cent of the

Lower Living Standard Income Level,** annualized on the

basis of the three months prior to the application date.
^. Not have voluntarily terminated without good cause in the last

six months, his/her last fulltime position paying at least

minimum wage.

169

170
Income Eligibility guidelines for CETA participation are as follows.

Size of Family
Unit

1

2

3

4

5

6

*Economical

ly

**100% Lower

Disadvanta

iged

Living Standard

$3,400

$3,950

4,350

6,470

6,210

8,880

7,670

10,960

9,050

12,930

10,580

15,120

(Add $1,530 to
annual income
for each family
member over 6)

(Add $2,190 to
annual income
for each family
member over 6)

APPENDIX B
CETA PROGRAM SIGNIFICANT SEGMENTS
The federal government requires each Prime Sponsor to serve the most
needy segments of the population. In order to insure this goal is accom-
plished the Department of Labor Bureau of Labor Statistics defines for
each Prime Sponsor the demographic characteristics of the local unemployed
population. The Prime Sponsor's clientele must reflect these character-
istics. The significant percentages for the Prime Sponsor participating
in this study are presented as follows.

Title II-B

Total Participants
for Fiscal Year
1979-1980: 2040

Title II-D

1519

Title VI

1318

Percentages

of Total:

Sex

Male

37.7

Female

62.3

44.0
56.0

48.0
52.0

Race

Age

Black 80.0

White 19.0

Hispanic 0.4
Native American 0.0
Asian 0.6

19 or

less

23.9

20-21

13.7

22-44

55.4

45-54

4.3

55 or

older

2.7

77.0

21.0

0.6

0.2

0.2

15.7

15.2

62.6

4.2

2.3

58.4

39.8

1.8

0.0

0.0

11.0

11.9

69.0

5.7

2.4

171

APPENDIX C
TEST PACKAGE INTRODUCTION AND PERSONAL DATA QUESTIONNAIRE
(Title on Subject's copy was "Introduction and History")

You have in front of you a research packet. The packet contains
questions which ask you to tell us about who you are and what you believe
to be true for you. We are asking these questions so we can learn how
to make the CETA program better help you prepare for work. Before you
begin it is important for you to understand four Important points.

First, this study cannot work unless you are completely honest in
your answers to the questions. If at any point you feel that you cannot
honestly answer the questions in the packet please stop and return the
packet to me. There is no penalty for stopping and it would harm the
study more if you answered falsely.

Second, it is important that you understand that your answers will
be kept completely confidential. I am the only person who will know how
you answered the questions. Before anyone else sees the results of this
study I will have removed your name from the answer sheets and I will
have assigned your answers a code number. The only reason I need to
know which answers are yours is so I can tell you the results of your
answers and so I can follow you through your CETA experience.

Third, as I mentioned above, I plan to tell you what your answers
were. If you want, after I have completed the study I will be happy to
sit down with each of you individually and explain the results of the
study and what they mean.

172

173

Finally, it is important to point out that while we have tried to
design this packet so it is easy to complete with clear instructions,
we are not perfect. There will be areas where some of you will not
understand the questions or instructions. If this happens, please feel
free to ask me for help. It is not your fault, it just means that we
did not do our job right.

Thank you for your help in this study. Before you start the first
group of questions please answer the History questions below.

History
In this group of questions you are asked to answer questions about
yourself. These questions are about your history. They will help us
understand your answers to the other questions better. Please read with
me and circle the letter on the answer sheet which best describes you.
Please answer each question only once.

1. Are you a member of a church?

A. Yes.

B. No.

2. If you answered YES to question #1 above please write the
name of your church on the answer sheet and the denomination
(Baptist, Methodist, Catholic, or others) of your church in
the space on the answer sheet.

3. How often do you attend church?

A. Do not attend church, not currently a member, or an
inactive member.

B. Attend church only on special occasions, such as
Christmas, Easter, or others, or between one (1)
and six (6) times a year.

C. Attend church about once or twice a month,

D. Attend church at least once a week.

174

4. Who raised you? (For this question please circle all answers

that are true for you)

A. Both of your parents.

B. One of your parents and a step-parent.

C. One parent or one step-parent.

D. Other relatives such as aunt, uncle, brother, sister,
or grandparents.

E. Foster parents or legal gaurdians.

5. How many older brothers and sisters do you have? If none
please write a zero (0).

6. How many younger brothers and sisters do you have? If none
please write a zero (0) .

THIS COMPLETES THE HISTORY SECTION, PLEASE STOP UNTIL TOLD TO PROCEED.

THANK YOU.

APPENDIX D
CETA APPLICATION

•.?«1.1CATICM I»TU_

rnA mm PFrnm owin htscr

I. ^A^£

,

7 <Jfl 3. TEUPK}N€ NO. i «REA CCOE

':. BIRlHMre

5. tcf

^6. SEX

7. MWinAL STAinstciROE a.T) n S D H SEP

3. RESiDoa:

MJCOI

Sn!£ET

CIfY COUNTY STATt HF COOE

;. CITIZENSHIP; (circle OriE) UNITED STArtS ffXVUiHT BESIDEJ/T ALIEN

IF A NONCITIZm, IS A^PLICAST PERMITTED | K>/?(C1BCI-E QUE) 1-551

TO WRK IN Trt UlITtD STATES

? /ZTJ YES ^_J KJ

ffcllLY STATJS - CIRCLE ONE t FILL IN SPAC£S
SIICLE PAdtKT tCAO OF KXEEHXD

PARENT IN TVO PWIENT FWILY
mrSER Of DEPt-OEXTS

OT>€R FA-ILT ri?2ER NrTOE?E.-v]e<T INDIVIDLW.
HLTQER IN FAIILY

l-^ «-3

J. DO YDU HAVE A VALID driver's LICENSS?

12. PRIOR CETA PARTICIPATION
WICH PRDGHAM

NO IF SO, »€RE_

KTi. I LOCATION Of PBIrE SPONSOR

DATES Of P«RTICIPATIC»1_

jots TOTAL PRIOR CETA PARTICIPATION EXCEED 30 POmS

3 RECTIVINC ue«jCWEKT / 7 YES / 7 M) PEC U. 1 . 15 WEEKS CR MORE i / YES I / NO

°'*'°*"^'°' CCT^INEDHEEKS UNEM>/hKS REC U.I. EXHAUSTEE / 7 YES / 7 NO aiGIBU NEVER «J>PLIE3 k/a

ill. I HAVE BEEN INFOWH) THAT I MY BE ELIGIBLE FOR l«ErPL£m«ffI COMPENSATION

AH) HAVE BEEX ENCiJWGED TO APPLY FOR SAME WITH TVE fU)RID« STATE EfPLOYhtNT SERVICE.

IS. ECOHCr^lCALLY DISADVAMTACED: APPLICANT OR FAMILY RECEIVES AFK / / YES / / >0

If fONE, aiGlBLE NOT RECEIVING PA / 7 YES / / NO

OT>tR PA

YES
YES

NO

OntR TYPE PA RECEIVING.

16. HWiDICAPPED

YES

NO IF YES, STATE TYPE (CIRCLE Of) "EMAL iMrSICAL RECE'.vlrjG FOiTtR CARE ZZ7 YES ZZ7 <0

x7. REGULAR OUTPATIENT Of (CIRCLE CUE ) RECTAL HD5P I TAL RE/«B FACILITY CTXER ( S?EC I FY)
Wfl Of INSTITUTION

13. INSTITUTICNALIZED IN A (CIRCLE ONE) PBISO* HOSPITAL OIXER (SPECIfTf)

NA^E Of INSTITUTION

OFFENDER / 7 YES / 7 NO A alENT Of A SHELTERED WORKSK)? ZZ7 YES C / NO DISPUICCD tOtfAXER

EX-CONVICT (circle one) YES

DATE Of release/conviction

'.9. education stat\)S (circle Che)

HiGtcsT grade OCrnETED

ZO. race/etwic group (circle one)

BLA£K(nOT HISPANIC)

hHITE(K)T HISPANIC)

21. MIGSJWT OR SEASONAL FAfM FAMILY f&KH I I YES

23. ENGLISH ABILITY (CIRCU ONE) FLUENT LIMITED

25.

aiGlBLE FOR PROGRAM (clBCLE CNE) YES

NO

AMERICAN INDIAN OR
ALASKAN NATIVE

ASIAN OR PACIFIC ISLAMEH

22. FAW RESIDENCE

I 7 YES /Zr7 NO

t^. PROGRAM COf*ClNEIfT/llTLE NO.

26. EtBOULMENT STATUS (ciRCU) «M ENROtLEE PEM)ING INTERTITLE TTVl-iSfER

27. aiGlBLE FOR ALLCHAICE (CIRCLE) BASIC DEPENDENTS INCENTIVE OTTSi

28. BATE ENROLLED IN PWXRAM

29. lr««JJ»rENT DATt_

_fM«£H OF DAYS UNErt>LDYED_

PREVIOUSLY ETPUDYED BY $PCIN90«/rE>I1RE (cIHCLE) YES NO

50. RECEIVED lAYDfF NOTICE HITMIN LAST 6 fOfms (CIRCLE ONE) YES K)

t*€H (mTEI.

31. R£E>*LDYtefT OPTCHTLMTY (circle 0«) fair COCO POOR

33.LAa» force status - employed fUI. TI^E PART Tit

Lr«}IE>FUJtED

IN-SOCOL YOUTH) (Wt Of SCHOO

OTVER (specify)

KXR.Y HACE

72. atGiainTY STATUS (circle ore OR fWE) IIB lie IID IV VI VII
VIII T.ITC (^mm OMK (SreiIFY

IHEMPLOYED - LAST UAY HDRKED (fcKm, DAY, YEAH)_

WEEKS l>Irt»J3YED OmlNG LAST 12 WEEKS

WEEXS WEfPLmED DIKING LAST 20 WEEKS

_KXRS WORKED PER WEEH_

PAGE 1 OF 2 PAGES

175

176

Jl. lODNQMIC Xt/inS (CIROE at) AT OR BaUM 0« PRDVBinf l£Va OR TCI oil L
35. fwiiLT INCcre:! Ktv€b< 86-100 peromt ujil

KTXEN 71-85 PERCOff LLSIL
ABWt lOO rtHCtNT IXSIL

KAfE OF F/«iLr feeoa

liSAIIONSHlP
A/TUCAW

itcrre tnxtn

tA.Tr 5 fty. [AST 6 Tn^.
-Xit -O-

-Jfc_

<?-

_X<fc_

..iZi.

TOIM- FW1ILY wan FCR 5 rtwms »_

. RDR 6 MWnS $_

35. WRK HISTOn OllNI«Ji LAST 25 kCEKs) - fRQEHT EMUTTER (POR E^*^Dr^) APPLICAHTS): H1M_

Aim£ij_

HDLRS PER »€«_
STARTItC MTH

_K3U«.r H«ce_

REASCH FOR LEAVING

COCO CAUSE (clROf ore) YES lO XB TITLE .

KA«

HXRS PQ) I
STARTING DATt.

J.O.T.OXE .

REASCN FCR LEAVIN5_

aXO CAUSE (CIRCU ore) TES N} JOB TITU .

_J).O.T.CCDe_

KXRLY »•«_

STARTING n»IE_

AWANCBtNI OPtOmHITT (CIWIE 0r€)

_T0 PR£SEKT_

VETERAN SIATUS (CIROE OTC)

IF SO, n»res OF service

YES

SPECIAL disables vtierm (circle oc) yes k>

EXTEMJeD ACTIVE B/IY (CIRCLE Ot) YES lO

DISCHIRGE On€R THW DISKXMOLE (CIRCLE Oc) YES n
3(K DISABILITY C« DISABIUTY 01 SCH»RC£ (circle C»«) YES K>
VIETMAn-EKA (35 TEARS OD CR YIXMGBl) (CIRCLE Oc) YU K]
NOTE: rCR H«VE HAD ANY EXTErCED ACTIVE tUTY tN T>C U. S. KveO
FORCES IN T>€ 60 DAY PERICD PRIOR TO E>nj01MJ<r (tJTC APfUCANTs)

38i I tn «*«£ T»T IF AW rB^Bt OF m IrtlDIATE FAMILY IS E>*1JJ»ED IN Afl AIMINISTMTIVE CAPACITY BY A RECIPIEKT, SUBRXIPIEKT OR E>*UJYING AGENCr OF A
CETA FunS) «£MCY I AM FOffllCDEII BY LAW TD ACCEPT A POSITim IN AN AIHINISTJWTIVE CAPACITY, STAFF POSITIO*, HJBL1C SERVICE ErHjOwe*T POSITIcn, CB
Ott-T>«-JDa TRAINING PaSITlm. T>f TEHn 'PERSCW IN AN AmiNISIRATIVE CAPACITY* INCULCES THKE PERSCNS >«) mVE OVTRILL AtMINlSTRAIlVE RESPOISIBIHTY
fa> A PRCCRArt, INCUCIW. AIL EUKTO) Alt) AFTOINIED OFFICIALS HO H4VE AW RESPOISIBILITY FCR T>C OBTAINING OF AMVOH APPflWAL OF ANY GRANT FLtCED
l»«R T>C tCX, AS WELL AS OTHER OFFICIALS HO HAVE IHFUtNCE CR CDHTROL EVER T>E ACrlNlSTRATlOtt OF T>C PROGRAM, SICH AS TVt PROJECT DIRECTCR, DEPUTY
DIRFCTOR tm U1II CHIEFS, AX) PERSCrC HO WVE SELECTION, HIRING, PLACEfENI CR SUPERVISCRY RiSPOISIBILlTltS FOR PUBLIC SERVICE BVUmfKl CR C»-TX£-
X8 TRAINING PARTICIPANTS. ( (ED H»^) CdO IOT HAVE) RELATIVES RELATED TO It ST BUX]0,rW«<IAG£, OK ADOPTION HO ARE EMPUTrtD BY T>C GCM»»efIS Of
T>€ CITY OF JACKSOft'ILLt. BAttR OH NASSAU CO/<TIES, OR A«Y OF TIC ABOVE ItMTIOiD AGBCIES.

relative's NAr<_

J«ELATIOr<SHIP_

_EffUJTER_

_fasiTicre

CERTIFICATICW - I tCREBf tUmFY TWT 1>€ It«)l»VlTICH PRCr/IDED IS TRUE TD Tl« BEST OF Kl COA-EEGE. I AM ALSO AWARE TmT T>€ INFOimTION I HAVE
PWJVIOED IS SUBJECT TO REVIBf HO VERIFICATIOI Alf> I «Y HAVE TO PfiCMDC DOCLTOITS TD Sl^TOST THIS APPVICATIOr<. I AM ALSO AMARE THVT I AM SUS-IECT
TD IrrtDlAIE TEmlHATiat IF I AM FCAM) INELIGIBLE AFTER EMKUTENT A>4) fttV BE FHOSEO/TED FOR FRAU) AMVOR PE/LAITf. I ALLCW RELEASE OF THIS INKR-
mnCM FCR VERIFICATICW PIJIPOSES Arc IMJERSTAfC THAT IT HIU Bf USED TO DETEWINE ELIGIBILITY. I AM FIRHIWING MY SOCIAL SECVRIDf HTBER 01 A
VOULMARY BASIS WITH Tl€ IMJEHSTAfCING SUH IS NOT REQUIRED BY FEDERAL STATUE CR REOOLATICrt. I HAVE BEEN ACWISED TmT CEIA Hia UTILIZE TMIS Ht^ER
0«.Y 10 FACILIIAIE TIC LDCATIOt OF E«\JCm«NT, MILITARY, CREDIT, HO EEUCATIOUL RECORDS COCERNIIC « IN COrMECTION WITH THIS APPUCATlOh.

applicant's siciutuie_

_ft»TI (rO<T>l,nAY,YEAR)_

PARSIT OR CUWSIAN SIG>WTUC_

NOT REOJIRai IF HEAD OF HUSEHXJD OR ABOVE Hi)

IlfTERVIEVOt S SIO«TVI<E_

IKIAXE OFFICER $ SIG»iATmE_

_PATE (MDrmi,BAY, year)_
_BATE(rOrH,DAY,YEAR)_
J»TE(refnH,DAY,YEAR)_

VERIFICATION officer's sio«»njie_

REVIEWING officer's SIGNAn«E__

_I»TE(M>nH,nAY,YEAR)_

_»te(mchtm,!»y,year)_

PAGE 2 Of ? PACES

APPENDIX E

INFORMED CONSENT FORM

(Title was omitted on the subjects' copies)

You are about to participate in a research project. This project
is designed to increase our understanding of CETA employment training
programs. This project is being conducted in conjunction with your CETA
program and your participation is greatly appreciated. You will receive
your usual pay for group participation for participating in this study.

This study consists of several questions which you are asked to
answer. The questions have no right or wrong answers but are designed
to have you describe what is true for you. If you do not understand any
of the questions you are asked to seek clarification from the researcher
present .

You may withdraw from this study at any time if you wish and return
to your normal CETA group activities.

If you wish you may receive a complete interpretation of your
responses from the researchers at the end of the research project. If
you wish this interpretation please indicate this at the bottom of this
page. If you change your mind please contact the researchers at the
address or telephone number listed below.

Please read the following statements carefully and sign your name
if you chose to participate in the project:

I have read and understand the procedures described above. I agree
to participate in the procedures and I have received a copy of this
description.

I do/do not (circle one) wish to receive an interpretation of my
responses to the questions in this project.

I do/do not (circle one) give the research staff permission to share
my responses by name with others, including my CETA counselor.

Your Name Counselor

John A. Sanford (9C4) 376-0210
P. 0. Box 14253
Gainesville, FL 32604

177

BIBLIOGRAPHY

Adams-Webber, J. Generalized expectancies concerning locus of control
of reinforcements and the perception of moral sanctions. British
Journal of Social and Clinical Psychology, 1969, 8(4), 340-343.

Alfano, A. M. A scale to measure attitudes toward working. Journal of
Vocational Behavior, 1973, 3(3), 329-333.

Altman, S. L. Women's career plans and maternal employment (Doctoral
dissertation, Boston University Graduate School, 1975).
Dissertation Abstracts International, 1975, 3_5, 3569B.
(University Microfilms No. 75-12)

Anderson, P. G. An analysis of the scales and items of the Personal

Orientation Inventory as it relates to adherents of two religious
groups (Doctoral dissertation. University of Northern Colorado,
1972). Dissertation Abstracts International, 1973, 33, 3364A.
(University Microfilms No. 73-254)

Andrisani, P. J., & Nestel, G. Internal-external control as a con-
tributor to and outcome of work experience. Journal of Applied
Psychology, 1976, 6^(2), 156-165.

Angrist, S. S. The study of sex-roles. Journal of Social Issues, 1969,
25(1), 215-232.

Bakke, E. W. The unemployed man. London: Nisbet & Co., 1933.

Bakke, E. W. Citizens without work. New Haven: The Yale Press, 1940.
(a)

Bakke, E. W. The unemployed worker. New Haven: The Yale Press, 1940.
(b)

Bakke, E. W. The mission of manpower policy. Kalamazoo, Mi.: The W.
E. Upjohn Institute for Employment Research, 1969.

Banikotes, P., & Banikotes, F. Male and female perceptions of lib-
erated versus conventional sex-roles. Psychonomic Science, 1972,
29(2), 111-112.

Barnes, H. N. Finding and interviewing the hard-to-locate : The DMI
experience. In M. E. Borus (Ed.), Evaluating the impact of man-
power programs. Lexington f4a . : D. C. Heath & Co., 1972.

Baron, R. M. , & Bass, A. R. The role of social reinforcement parameters
in improving trainee task performance and self-image. Detroit,
Mi.: Department of Psychology, Wayne State University, 1968.

178

179

Barron, F. An ego strength scale which predicts response to psycho-
therapy. Journal of Consulting Psychology, 1953, j^(5), 327-333.

Bass, B. A., & Stek, R. J. Perceived locus of control and self-actu-
alization: Failure to replicate. Perceptual and Motor Skills,
1972, 35(2), 646.

Battle, E. S., & Rotter, J. B. Children's feelings of personal control
as related tn social class and ethnic group. Journal of Person-
ality, 1963, 11(4), 482-490.

Beatty, R. W. A two year study of hard-core unemployed clerical

workers: Effects of scholastic achievement, clerical skill, and
self-esteem on job success. Personnel Psychology, 1975, 28(2),
165-173.

Beatty, R. W. , & Beatty, J. R. Longitudinal study of absenteeism of

hard-core unemployed. Psychological Reports, 1975, 26(2), 395-406.

Bem, S. L. The measurement of psychological androgyny. Journal of
Consulting and Clinical Psychology, 1974, 42(2), 155-162.

Bem, S. L. Sex-role adaptability: One consequence of psychological
androgyny. Journal of Personality and Social Psychology, 1975,
31(4), 634-643.

Bem, S. L. On the utility of alternative procedures for assessing
psychological androgyny. Journal of Consulting and Clinical
Psychology, 1977, 45(2), 196-205.

Bem, S. L. Theory and measurement of androgyny: A reply to the

Pedhazur-Tetenbaum and Locksley-Colton critiques. Journal of
Personality and Social Psychology, 1979, 17(6), 1047-1054.

Bem, S. L., & Lenney, E. Sex typing and avoidance of cross sex behav-
ior. Journal of Personality and Social Psychology, 1976, 33(1),
48-54. —

Bem, S. L., Martyna, W. , & Watson, C. Sex typing and androgyny: Fur-
ther explorations of the expressive domain. Journal of Personality
and Social Psychology, 1976, 14(5), 1016-1023.

Berzins, J., Welling, M. , & Wetter, R. A new measure of psychological
androgyny based on the Personality Research Form. Journal of
Consulting and Clinical Psychology, 1978, 46(1), 126-138.

Blaumer, R. Work satisfaction and industrial trends. In W. Galenson
& S. M. Upset (Eds.), Labor and trade unionism. New York: Wiley
& Sons, 1960.

Block, J. H. Conceptions of sex roles. American Psychologist, 1973,
28(5), 512-526.

180

Blood, M. R. Work values and job satisfaction. Journal of Applied
Psychology, 1969, 53(6), 456-459,

Borus, M. E., & Tash, W. R. Measuring the impact of manpower programs:
A primer. Policy Papers in Human Resources and Industrial Rela-
tions, 1970, j_7.

Broverman, I. K. , Broverman, D. M. , Clarkson, F. E., Rosenkrantz, P. S.,
& Vogel , S. R. Sex role stereotypes and clinical judgements of
mental health. Journal of Consulting and Clinical Psychology,
1970, 34(1), 1-7.

Broverman, I. K., Vogel, S. R., Broverman, D. M., Clarkson, F. E., &
Rosenkrantz, P. S. Sex role stereotypes: A current appraisal.
Journal of Social Issues, 1972, ^(2), 59-78.

Brown, R. , Winkworth, J., & Brascamp, L. Student development in a coed
residence hall: Promiscuity, prophylactic, or panacea? Journal
of College Student Personnel, 1973, i4(2), 98-104.

Bryant, B. K. , & Trockel, J. F. Personal history of psychological

stress related to locus of control orientation among college women.
Journal of Consulting and Clinical Psychology, 1976, 44^(2), 266-
271.

Bryant, F. L., & Showalter, J. M. An evaluation of orientation and

assessment effectiveness in the Work Incentive Program. Journal
of Employment Counseling, 1971, 8^(2), 59-64.

Burke, J. F. The relationship between religious orientation and self-
actualization among selected Catholic groups (Doctoral disser-
tation. United States International University, 1973). Disser-
tation Abstracts International, 1973, 34_, 1721B-I722B. (University
Microfilms No. 73-22, 655)

Butterfield, E. C. Locus of control, test anxiety, reactions to

frustration, and achievement attitudes. Journal of Personality,
1964, 32(3), 355-370.

Byrne, D. The Repression-Sensitization Scale: Rationale, reliability,
and validity. Journal of Personality, 1961, ^(3), 334-349.

Campbell, D. T. , & Stanley, J. C. Experimental and quasi-experimental
designs for research. Chicago: Rand-McNally , 1966.

Carey, G. L. Sex differences in problem solving performance as a func-
tion of attitude differences. Journal of Abnormal and Social Psy-
chology, 1958, 56, 256-260.

Cassel, J. Social class and mental disorders: An analysis of the
limitations and potentialities of current epidemiological
approaches. In K. S. Miller & C. M. Grigg (Eds.), Mental health
and the lower social classes. Tallahassee, Fl . : Florida State
University, 1966.

181

Chafetz, J. S. Masculine/ feminine or human? An overview of the soci-
ology of sex-roles. Itasca, II.: F. E. Peacock Publishers, 1974.

Chandler, T. A. A note on the relationship of internality-externality ,
self-acceptance, and self-ideal discrepancies. Journal of Psy-
chology, 1976, _9i(n, 145-146.

Chisholm. R. F. Alienation and activities in on-the-job and off-the-job
life spheres (Doctoral dissertation, Case-Western Reserve Univer-
sity, 1975). Dissertation Abstracts International, 1975, 36,
1962B. (University Microfilms No. 75-19, 193)

Christensen, D. T. Relationship of Manpower Development and Training
Act education/training and parole success of one-hundred-eleven
matched pairs of parolees from Utah State Prison (Doctoral disser-
tation. University of Utah, 1974). Dissertation Abstracts Inter-
national, 1974, 35, 2044A. (University Microfilms No. 74-23, 098)

Christie, R. , & Geis, F. L. Studies in machiavelianism. New York:
Academic Press, 1970.

Clark, R. A. The Internal-External Scale: Control of what? Journal
of Consulting and Clinical Psychology, 1976, ^(1), 154.

Cohen, S. , Rothbart, M. , & Phillips, S. Locus of control and the

generality of learned helplessness in humans. Journal of Person-
ality and Social Psychology, 1976, 34(6), 1049-1056.

Cole, A. Mooney Problem Checklist differences among unemployed indi-
viduals in a work evaluation setting (Doctoral dissertation. State
University of New York at Buffalo, 1974). Dissertation Abstracts
International, 1974, 35., 1520A-1521A. (University Microfilms
No. 74-19, 971)

Cook, V. S. A comparison of work values of disadvantaged black males

with work values of advantaged black males (Doctoral dissertation.
The Catholic University of America, 1971). Dissertation Abstracts
International, 1971, 32, 1848A. (University Microfilms No. 71-23,
401)

Crandall, V. C, Katkovsky, W. , & Crandall, V. J. Children's beliefs
in their own control of reinforcements in intellectual-academic
situations. Child Development, 1965, 36.(1), 91-109.

Cristall, L. , & Dean, R. S. Relationship of sex-role stereotypes and
self-actualization. Psychological Reports, 1976, _32(3, Pt. 1),
842.

Darley, S. A. Big-time careers for the little women. Journal of
Social Issues, 1976, ^(3), 85-98.

Davis, A. The motivations of the underpriveleged worker. In W. F.

Whyte (Ed.), Industry and society. New York: McGraw-Hill, 1946.

182

Dean, D. G. Alienation: Its meaning and measurement. American Soci-
ological Review, 1961, 26(5), 753-758.

deGrace, G. The compatibility of self-actualization and anxiety.
Journal of Clinical Psychology, 1974, 30(4), 566-568.

Derlega, V. J. , & Chaikin, A. L. Norms affecting self-actualization in
men and women. Journal of Consulting and Clinical Psychology,
1976, 44(3), 376-380.

Deutsch, C. J., & Gilbert, L. A. Sex-role stereotypes: Effects on

perceptions of self and others and on personal adjustment. Journal
of Counseling Psychology, 1976, 23(4), 373-379.

Dubin, R. Industrial workers' worlds: A study of the "Central Life
Interests" of industrial workers. Social Problems, 1956, 2(3),
131-142.

Dunbar, J., Brown, M. , & Amoroso, D. Some correlates of attitudes

toward homosexuality. Journal of Social Psychology, 1973, 89(2) ,
271-279.

Ellis, L. J., & Bentler, P. M. Traditional sex determined role stan-
dards and sex-stereotypes. Journal of Personality and Social
Psychology, 1973, 25(1), 28-34.

Engel, I. M. A factor analytic study of items from five masculinity-
femininity tests. Journal of Consulting Psychology, 1966, 30(6) ,
565.

Feldman, J. Race, economic class, and evaluation of work. Journal of
Applied Psychology, 1973, 58(1), 10-15. (a)

Feldman, J. Race, economic status, and perceived outcomes of work and
unemployment. Journal of Applied Psychology, 1973, 58(1), 16-22.
(b)

Feldman, J. Race, economic class, and the Intention to work: Some
normative and attitudinal correltatlons. Journal of Applied
Psychology. 1974, 59(2), 179-186.

Ferman, L. A. Job development for the hard to employ. Policy Papers
in Human Resources and Industrial Relations, 1969, 11 .

Fisher, G. Performance of psychopathic felons on a measure of self-
actualization. Educational and Psychological Measurement, 1968,
28(2), 561-563.

Foster, N. J. Women: Locus of control and attitudes toward femininity
and masculinity (Doctoral dissertation. Northwestern University,
1975). Dissertation Abstracts International, 1975, 35.. 5109B.
(University Microfilms No. 75-7912)

ion

183

Fox, J., Knapp, R. , & Michael, W. An assessment of self-actualization
of psychiatric patients: The validity of the Personal Orientat
Inventory. Educational and Psychological Measurement, 1968
28(2), 565-5697^ "

Fox, L. L. A comparative analysis of internal-external locus of control
and sex role concepts in black and white freshmen women (Doctoral
dissertation. East Texas State University, 1975). Dissertation

Abstracts International. 1976, 36, 5143A-5144A. (University

Microfilms No. 76-4631)

Foulds, M. L., & Warehime, R. G. Effects of a "Fake Good" response set
on a measure of self-actualization. Journal of Counseling Psy-
chology, 1971, 18(3), 279-280. (a) ""

Foulds, M. L., & Warehime, R. G. The relationship between repression-
sensitization and a measure of self-actualization. Journal of
Consulting and Clinical Psychology. 1971, 36(2), 257-259. (h)

Freedman, L. Z. Psychopathology and poverty. In A. B. Shostak & W
Gomberg (Eds.), The blue collar world: Studies of the American
worker. Englewood Cliffs, N.J.: Prentice-Hall, 1969.

Freuh, T., & McGhee, P. Traditional sex role development and the amount
of time spent watching television. Developmental Psychology 1975
11(1), 109. ■ ^ ^ '

Garske, J. P. Role variation as a determinant of attributed masculinity
and femininity. Journal of Psychology, 1975, 91(1), 31-37.

Gaudreau, P. Factor analysis of the Bem Sex-Role Inventory. Journal of
Consulting and Clinical Psychology, 1977, 45(2), 299-302.

Gestinger, S. H. Temporal relatedness: Personality and behavioral

correlates. Journal of Personality Assessment, 1975, 39(4), 405-
408 .

Gibb, L. L. A study of differences in sex, home background, educational
background, work experience, extra-curricular participation, and
self-actualization in college juniors (Doctoral dissertation.
Northern Illinois University, 1966). Dissertation Abstracts. 1967,
£7, 2358A-2359A. (University Microfilms No. 66-11, 538)

Gilligan, J. F. Personality characteristics of selectors and non-selec-
tors of sensitivity training. Journal of Counseling Psychology,
1973, 20(3), 265-268. " ^ ^ ^'

Ginn, R. 0. Defensive and non-defensive repressors and sensitizers and
self-actualization. Journal of Clinical Psychology. 1974, 30(1)
82-83. — '

184

Ginn, R. 0. Male and female estimates of personal problems of men and
women. Journal of Counseling Psychology, 1975, ^(6), 518-522.

Ginzberg, E, The development of human resources. New York: McGraw-
Hill, 1966.

Ginzberg, E. Manpower agenda for America. New York: McGraw-Hill,
1968.

Goldman, J. A., & Olczak, P. V. Effects of knowledge about self-actu-
alization on faking the Personal Orientation Inventory. Journal
of Consulting and Clinical Psychology, 1976, 44_(4), 680.

Goldstein, M. W. Mental health, social status, and Maslow's need system
(Doctoral dissertation, Yeshiva University, 1967). Dissertation
Abstracts, 1967, ^S, 2123B-2124B. (University Microfilms No. 67-
14, 571)

Goodman, P. S. Hiring and training the hard-core unemployed: A problem
in system definition. Human Organization, 1969, 2^(4), 259-269.

Goodman, P. S., & Salipante, P., Jr. Organizational rewards and reten-
tion of the hard-core unemployed. Journal of Applied Psychology,
1976, 6j^(l), 12-21.

Goodman, P. S., Salipante, P., Jr., & Paransky, H. Hiring, training,

and retraining hard-core unemployed: A selected review. Journal
of Applied Psychology, 1973, 58(1). 23-33.

Gordon, M. E. , & Scott, R. P. Evaluation of a manpower development
project in terms of its effects on the personal lives of its
graduates. Journal of Vocational Behavior, 1972, 2^(4), 467-478.

Gore, P. M. , & Rotter, J. B. A personality correlate of social action.
Journal of Personality, 1963, 2i(U, 58-64.

Goursslin, 0. R. , & Roach, J. C. Some issues in the training of the
unemployed. Social Problems, 1964, 21(1), 86-98.

Graff, R. W. , & Ladd, C. E. Personal Orientation Inventory correlates
of a religious committment inventory. Journal of Clinical Psy-
chology, 1971, 22(4), 502-504.

Gross, R. , Batlis, N. , Small, A., & Erdwins, C. Factor structure of the
Bem Sex-Role Inventory and the Personal Attributes Questionnaire.
Journal of Consulting and Clinical Psychology, 1979, 47^(6), 1122-
1123.

Gump, J. P. Sex-role attitudes and psychological well-being. Journal
of Social Issues, 1972, 28(2), 79-92.

185

Gurapper, D. C. Implications of economic need, expectancy, and attitudes
towards work (Doctoral dissertation. The Pennsylvania State Univer-
sity, 1970). Dissertation Abstracts International, 1971, 32, 605B.
(University Microfilms No. 71-16, 605)

Gurin, P., Gurin, G., Lao, R. C., & Beattie, M. Internal-external
control in the motivation dynamics of Negro youth. Journal of
Social Issues, 1969, 25(3), 29-53.

Hall, E. R. , Joesting, J., & Woods, M. J. Relationships among measures
of locus of control for black and white students. Psychological
Reports, 1977, 40(1), 59-62.

Hardin, E. On the choice of control groups. In M. E. Borus (Ed.),
Evaluating the impact of manpower programs. Lexington, Ma.:
D. C. Heath and Co., 1972.

Harford, T. C., Willis, C. H., & Deabler, H. L. Personality correlates
of masculinity and femininity. Psychological Reports, 1967, 21(3) ,
881-884.

Hargrett, A. J. The education-unemployment relationship in Chicago as
revealed in the 1960 Census. Journal of Negro Education, 1965,
34(2), 121-129.

Harris, L., & Associates. A study of Job Corps non-graduate termina-
tions. In Hearings on the Economic Opportunity Amendments of
1967: 90th Congress, First Session. Washington, D. C: U. S.
Government Printing Office, 1967 (Part 1).

Hartley, M. P. The relationship of locus of control and need achieve-
ment to job satisfaction (Doctoral dissertation, Rutgers Univer-
sity-The State University of New Jersey, 1975). Dissertation
Abstracts International, 1976, 3^, 6553A. (University Microfilms
No. 76-8689)

Hausknecht, M. The blue collar joiner. In A. B. Shostak & W. Gomberg
(Eds.), The blue collar world: Studies of the American worker.
Englewood Cliffs, N.J.: Prentice-Hall, 1964.

Heilbrun, A. B. Sex-role, instrumental-expressive behavior, and psycho-
pathology in females. Journal of Abnormal Psychology, 1968, 73(1) ,
131-136.

Heilbrun, A. B. Measurement of masculine and feminine sex-role iden-
tities as independent dimensions. Journal of Consulting and Clin-
ical Psychology, 1976, 44(2), 183-190.

Heilbrun, C. G. Toward a recognition of androgyny. New York: Harper
& Row, 1973.

186

Helland, D. J. Sex-role correlates of adolescent self-esteem (Doc-
toral dissertation, The University of Michigan, 1973). Disser-
tation Abstracts International, 1973, 34. 2026A-2027A. (Univer-
sity Microfilms No. 73-24, 584)

Heller, W. W. Employment and manpower. In S. Lebergott (Ed.), Men

without work: The economics of unemployment. Englewood Cliffs,
N.J.: Prentice-Hall, 1964.

Henke, R. 0. Factors in successful job training and job performance
of the disadvantaged. Journal of Experimental Education, 1976
45(1), 61-68.

Heron, A. R. Why men work. Stanford: Stanford University Press, 1948.

Herzberg, F. Work and the nature of man. Cleveland: World Publishing,
1966.

Herzberg, F. , Mausner, B., & Snyderman, B. The motivation to work.
New York: Wiley, 1959.

Hjelle, L. A. Relationship of a measure of self-actualization to reli-
gious participation. Journal of Psychology, 1975, ^(2), 179-182.

Hjelle, L. A., & Butterfield, R. Self-actualization and women's atti-
tudes toward their roles in contemporary society. Journal of
Psychology, 1974, 87(2), 225-230.

Hjelle, L. A., & Smith, G. Self-actualization and retrospective reports
of parent-child relationships among college females. Psychological
Reports, 1975, 36(3), 755-761.

Hochreich, D. J. Sex-role stereotypes for internal-external control
and interpersonal trust. Journal of Consulting and Clinical
Psychology, 1975, 43(2), 273.

Hollingshead, A. B. , & Redlich, F. C. Social class and mental illness.
New York: Wiley & Sons, 1958.

Holter, H. Sex-roles and social change. Acta Sociologica, 1971, 14(1) ,
2-12.

House, W. C. Actual and perceived differences in male and female expec-
tancies and minimum goal levels as a function of competition.
Journal of Personality, 1974, 42(3), 493-509.

Ilardi, R. , & May, W. T. A reliability study of Shostrum's Personal
Orientation Inventory. Journal of Humanistic Psychology, 1968,
8(1), 68-72.

Jones, D. S., & Medvene, A. M. Self-actualization effects of a marathon
growth group. Journal of Counseling Psychology, 1975, 22^(1), 39-43.

187

Jones, W. , Chernovetz, M. E., & Hansson, R. 0, The enigma of androgyny:
Differential implications for males and females. Journal of Con-
sulting and Clinical Psychology, 1979, j46(2), 298-313.

Jourard, S. M. Personal adjustment; An approach through the study of
the healthy personality (2nd ed . ) . New York: MacMlllan, 1963.

Jourard, S. M. Self-disclosure: An experimental analysis of the trans-
parent self. New York: Wiley-Interscience, 1971.

Kahn, R. L. The meaning of work: Interpretation and proposals for
measurement. In A. Campbell & P. E. Converse (Eds.), The human
meaning of social change. New York: Russell Sage Foundation, 1972.

Kaplan, R. , & Tausky, C. Work and the welfare cadillac: The function
and commitment to work among the hard-core unemployed. Social
Problems, 1972, J_9(4), 469-483.

Kaplan, R. , & Tausky, C. The meaning of work among hard-core unemployed.
Pacific Sociological Review, 1974, ]J_(2) , 185-189.

Kelly, J. A., Caudill, S. , Hathorm, S. , & O'Brien, C. G. Socially
undesirable sex correlated characteristics: Implications for
androgyny and adjustment. Journal of Consulting and Clinical
Psychology, 1977, 45(6), 1185-1186.

Kelly, J. A., & Worrell, J. New formulations of sex-roles and androgyny:
A critical review. Journal of Consulting and Clinical Psychology,
1977, 45(6), 1101-1115.

Kelly, J. A., & Worrell, L. Parent behaviors related to masculine,

feminine, and androgynous sex-role orientations. Journal of Con-
sulting and Clinical Psychology, 1976, 44(5), 843-851.

Killingsworth, C. C. Automation, jobs, and manpower. In S. Lebergott
(Ed.), Men without work: The economics of unemployment.
Englewood Cliffs, N.J.: Prentice-Hall, 1964.

Killingsworth, C. C. Jobs and income for negroes. Policy Papers in
Human Resources and Industrial Relations, 1969, 6.

Kimball, R. , & Gelso, C. T. Self-actualization in a marathon growth
group: Do the strong get stronger? Journal of Counseling Psy-
chology, 1974, 2A(l), 38-42.

Kimmons, G. , & Greenhaus, J. H. Relationships between locus of control
and reactions of employees to work characteristics. Psychological
Reports, 1976, 39(3, Pt. 1), 815-820.

Kinder, B. Locus of control and "pseudovolunteerlng. " Journal of
Applied Psychology, 1976, 6^(2), 251-252.

188

King, M. Sex differences in self-actualization. Psychological Reports,
1974, 35(1, Pt. 2), 602.

Klavetter, R. E., & Mogar, R. E. Stability and internal consistency of
a measure of self-actualization. Psychological Reports, 1967,
21^(2), 422-424.

Klein, L. R. , & Ghozeil, S. A popularized version of 21 doctoral

dissertations (Research and Development Monograph No. 70, U. S.
Department of Labor Employment and Training Administration) .
Washington, D. C. : U. S. Government Printing Office, 1979.

Knapp, R. R. Relationship of a measure of self-actualization to

neuroticism and extraversion. Journal of Consulting Psychology,
1965, 29(2), 168-172.

Knapp, R. R. , & Comrey, A. L. Further construct validation of a measure
of self-actualization. Educational and Psychological Measurement,
1973, 33(2), 419-425.

Komorovsky, M. Functional analysis of sex-roles. American Sociological
Review, 1950, ^5(4), 508-516.

Kornhauser, A. Mental health and the industrial worker. New York:
Wiley, 1965.

Kravetz, D. F. Sex-role concepts of women. Journal of Consulting and
Clinical Psychology, 1976, 44(3), 437-443.

Lambert, M. J., Dejulio, S. S., & Cole, A. M. Internality versus exter-
nality and personal adjustment: A factor analytic study.
Psychological Reports, 1976, 39(3, Pt . 1), 920-922.

Langston, R. D. Stereotyped sex-role behavior and sex-guilt. Journal
of Personality Assessment, 1975, 3^(1), 77-81.

Lebergott, S. (Ed.). Men without work: The economics of unemployment.
Englewood Cliffs, N.J.: Prentice-Hall, 1964.

Lee, R. , & Piercy, F. Church attendance and self-actualization. Journal
of College Student Personnel, 1974, j_5(5) , 400-403.

Lef court, H. M. Belief in personal control: Research and implications.
Journal of Individual Psychology, 1966, 22^(2), 185-195.

Lefcourt, H. M. , & Ladwig, G. M. The American negro: A problem in

expectancies. Journal of Personality and Social Psychology, 1965,
i(4), 377-380.

Lefton, M. Race, expectancy, and anomia. Social Forces, 1968, 46(3) ,
347-352.

189

Lessner, M., & Knapp, R. R. Self-actualization and entrepreneurial

orientation among small business owners: A validation study of
the Personal Orientation Inventory. Educational and Psychological
Measurement. 1974, 34(2), 455-460. ~

Levinson, P. Chronic dependency: A conceptual analysis. Social Service
Review, 1964, 38(4), 371-381.

Levitan, S. A. Federal manpower policies and programs to combat unem-
ployment. Kalamazoo, Mi. : The W. E. Upjohn Institute for Eraploy-
ment Research, 1964.

Levitan, S. A. The design of federal antipoverty strategies. Policy
Papers in Human Resources and Industrial Relations, 1967, 1.

Levitan, S. A. Antipoverty work and training efforts: Goals and

reality. Policy Papers in Human Resources and Industrial Relations,

1969, 3. '

Lewis, M. V. Finding the hard-to-locate: A review of best practices.
In M. E. Borus (Ed.), Evaluating the impact of manpower programs.
Lexington, Ma.: D. C. Heath, 1972.

Locksley, A., & Colton, M. E. Psychological androgyny: A case of

mistaken identity. Journal of Personality and Social Psychology,
1979, 37(6), 1017-103E ^

Lunneborg, P. Stereotypic aspects to masculinity-femininity measurement.
Proceedings of the 76th Annual Convention of the American Psy-
chological Association, 1968, 3, 163-164. (Summary)

Lunneborg, P., & Lunneborg, C. E. Factor structure of masculinity-
femininity scales and items. Journal of Clinical Psychologv.

1970, 26(3), 360-366. " ^

Lyman, E. Occupational differences in the value attached to work.
American Journal of Sociology, 1955, ^(1), 138-144.

Maccoby, E. E. , & Jacklin, C. The psychology of sex-differences.
Stanford: Stanford Univeristy Press, 1974.

Mangum, G. L. Contributions and costs of manpower development and

training. Policy Papers in Human Resources and Industrial Rela-
tions, 1967, 5. ~'

Mangum, G. L. The emergence of manpower policy. New York: Holt,
Rinehart, & Winston, 1969.

Mangum, G. L. , & Walsh, M. F. A decade of manpower development and
training. Salt Lake City: Olympus Printing, 1973.

Maslin, A., & Davis, J. L. Sex-role stereotyping as a factor in mental
health standards among counselors in training. Journal of Coun-
seling Psychology, 1975, 22^(2), 87-91.

190

Maslow, A. H. Cognition of being in the peak experience. Journal of
Genetic Psychology, 1959, 94(1), 43-66. (a)

Maslow, A. H. Critique of self-actualization I: Some dangers of being
cognition. Journal of Individual Psychology, 1959, J^(l), 24-32.
(b)

Maslow, A. H. Eupsychia-The good society. Journal of Humanistic
Psychology, 1961, ^(1), l-U.

Maslow, A. H. Fusion of facts and values. American Journal of Psycho-
analysis. 1963, 22(2). 117-131. (a)

Maslow, A. H. The need to know and the fear of knowing. Journal of
General Psychology, 1963, 68(1), 111-128. (b)

Maslow, A. H. Further notes on the psychology of being. Journal of
Humanistic Psychology, 1964, 4_(1), 45-58. (a)

Maslow, A. H. Synergy in the society and the individual. Journal of
Individual Psychology, 1964, 20(2). 153-164. (b)

Maslow, A. H. Humanistic science and transcendent experiences. Journal
of Humanistic Psychology, 1965, 5(2), 219-227.

Maslow, A. H. A theory of metamotivation: The biologic rooting of the
value life. Journal of Humanistc Psychology, 1967, _7(2), 93-127.

Maslow, A. H. Toward a psychology of being (2nd ed . ) . New York: D. Van
Nordstrand, 1968.

Maslow, A. H. Motivation and personality (2nd ed . ) . New York: Harper
& Row, 1970. (a)

Maslow, A. H. Religion, values, and peak experiences. New York: Viking
Press, 1970. (b)

Maslow, A. H. The farther reaches of hviman nature. New York: Viking
Press, 1971.

Maslow, A. H. , & Mittleman, B. Principles of abnormal psychology: The
dynamics of psychic illness. New York: Harper & Brothers, 1951.

Massimo, J. L. , & Shore, M. F. Job focused treatment for anti-social
youth. Children, 1964, n.(4) , 142-146.

May, R. Existential psychology (2nd ed . ) . New York: Random House,
1960.

191

McArthur, L. Z., & Eisen, S. V. Achievements of male and female story-
book characters as determinants of achievement behavior by boys and
girls. Journal of Personality and Social Psychology, 1976, 33(4),
467-473. ~~ ~

McClain, E. W. , & Andrews, H. B. Some personality correlates of peak
experiences: A study in self-actualization. Journal of Clinical
Psychology. 1969, 25(1), 36-38.

McGovern, L. P., Ditzian, J. L., & Taylor, S. P. Sex and perceptions of
dependency in a helping situation. Bulletin of the Psychonomic
Society. 1975, 5(4), 336-338.

McKee, J. P., & Sherriffs, A. C. Men's and women's beliefs. Ideals, and
self-concepts. American Journal of Sociology. 1959, 64(4), 356-
363. —

McLeod, A. I. The self-actualizer as a deviant (Doctoral dissertation.
University of Nebraska, 1972). Dissertation Abstracts Interna-
tional, 1973, 33, 3796A-3797A. (University Microfilms No. 73-120)

Mead, M. Sex and temperment in three primitive societies. New York:
The New American Library of World Literature, 1950. "

Mirengoff, W. , & Rindler, L. The Comprehensive Employment and Training
Act: Impact on people, places, and programs — An interim report.
Washington, D. C. : National Academy of Sciences. 1976.

Mirengoff, W. , & Rindler, L. The Comprehensive Employment and Training
Act: Manpower programs under local control. Washington, D. C.:^
National Academy of Sciences, 1978,

Misiak, H. K. , & Sexton. V. S. Phenomenological, existential, and

humanistic psychologies: A historical survey. New York: Grune &
Stratton, 1973.

Montgomery, E. F. A study of the effects of career and personal group
counseling on retention rates and self-actualization (Doctoral
dissertation. University of Southern Mississippi. 1975). Disser-
tation Abstracts International. 1975, 36, 2028A-2029A. (Univers'ity
Microfilms No. 75-22, 512)

Morado, C. Racism and sexism: An investigation of common traits (Doc-
toral dissertation, Boston University Graduate School, 1973).
Dissertation Abstracts International, 1973, 34, 1729B-1730B.
(University Microfilms No. 73-23, 501)

Moreland, J. R. , Harren. V. A., Krimsky-Montague, E., & Tinsley. H. E.
Sex-role self-concept and career decision making. Journal of
Counseling Psychology. 1979, 26(4), 329-336.

Morse, N. C, & Weiss, R. S. The function and meaning of work and the
job. American Sociological Review. 1955, 20(2), 191-198.

192

Myers, J. K. , & Bean, L. L. A decade later: A follow up of social
class and mental illness. New York: Wiley & Sons, 1968.

Nevill, D. Sex-roles and personality correlates. Human Relations.
1977, 30(8), 751-759.

Newman, J. M. Comparison of the Internal-External Scale and a specific
locus of control measure in predicting risk taking behavior under
novel task conditions. Psychological Reports, 1977, 40(3, Pt.2),
1035-1040. ~

Nie, N. H., Hull, C. H., Jenkins, J. G. , Steinbrenner , K. , & Bent, D. H.
Statistical package for the social sciences (2nd. Ed.). New York:
McGraw-Hill, 1975.

Northeast Florida Employment and Training Consortium. Northeast Florida
Emplo>Tnent and Training Consortium Master Plan. Jacksonville, Fl . :
Author, 1979.

O'Conner, K. , Mann, D. W. , & Bardwick, J. M. Androgyny and self-esteem
in the upper-middle class: A replication of Spence. Journal of
Consulting and Clinical Psychology, 1978, 46(5), 1168-1769.

Ohlbaum, J. S. Self-concepts, value characteristics, and self-actu-
alization of professional and non-professional women (Doctoral
dissertation. United States International University, 1971).
Dissertation Abstracts International, 1971, yi, 1221B-1222B.
(University Microfilms No. 71-14, 181)

Okman, G. An exploration of sex differences in the coiistruction of sex-
roles (Doctoral dissertation. University of Connecticut, 1973).
Dissertation Abstracts International, 1973, 34^, 420B. (University
Microfilms No. 73-16, 718)

Olczak, P. v., & Goldman, J. A. Self-actualization as a moderator of
the relationship between attitude similarity and attraction.
Journal of Psychology, 1975, 89(2), 195-202.

Oldham, G. R. Job characteristics and internal motivation: The mod-
erating effects of interpersonal and individual variables. Human
Relations, 1976, 29.(6), 559-569.

O'Leary, V. E., & Depner, C. E. College males' ideal female: Changes

in sex-role stereotypes. Journal of Social Psychology, 1975, 95(1),
139-140.

Glim, E. G. The self-actualizing person in the fully functioning family:
A humanistic viewpoint. Family Coordinator, 1968, ll_0) , 141-148.

193

Ormand, H. A. Relationship of measurements of dogmatism, purpose in

life, and self-actualization (Doctoral dissertation. United States
International University, 1973). Dissertation Abstracts Interna-
tional, 1973, 34, 1730B-1731B. (University Microfilms No. 73-23,
160)

Ott, T. J. Androgyny, sex-role stereotypes, sex role attitudes, and
self-actualization among college women (Doctoral dissertation.
University of Notre Dame, 1976). Dissertation Abstracts Inter-
national, 1976, 37.. 3527A. (University Microfilms No. 76-27
292)

Pasamanick, B., Roberts, D. W. , Lemkau, P. W., & Kreuger, D. B. A

survey of mental diseases in an urban population: Prevalence by
race and income. In F. Reissman, J. Cohen, & A. Pearl (Eds.),
Mental health of the poor. New York: The Free Press. 1964.

Pedhazur, E. J., & Tetenbaum, T. J. The Bern Sex-Role Inventory: A

theoretical and methodological critique. Journal of Personality
and Social Psychology. 1979, 37(6), 996-1016.

Pettus, J. P. Psychological androgyny: Construct validation and rela-
tionship to mental health and sex stereotypes (Doctoral disser-
tation. University of Montana, 1976). Dissertation Abstracts
International, 1976, 37, 2575B. (University Microfilms No. 76-
25, 703)

Phares, E. J. Expectancy changes in skill and chance situations.
Journal of Abnormal and Social Psychology, 1957, 54, 339-342.

Phares, E. J. Internal-external control as a determinant of amount of
social influence exerted. Journal of Personality and Social
Psychology, 1965, 2(5), 642-647. '

Pierson, J. H. G. Essays on full employment. Metuchen, N.J.: Scarecrow
Press, 1972.

Philadelphia Manpower Utilization Commission. A study of the effects
of pre-vocational training and family services on the long term
unemployed (Office of Manpower Research Contract //14-64, Title I,
MDTA, U. S. Department of Labor). Washington, D. C. : Government'
Printing Office, 1969.

Price, C. R., & Levinson, H. Work and mental health. In A. B. Shostak
& W. Gomberg (Eds.), Blue collar world: Studies of the American
worker. Englewood Cliffs, N.J. : Prentice-Hall , 1964.

Ranaan, S. L. Test Review. Journal of Counseling Psvcholoov 1973
20(5), 477-478. -^ ^ ^'

Reeves, T. G., & Shearer, R. A. Differences among campus groups on a

measure of self-actualization. Psychological Reports. 1973, 32(1),

194

Reissman, F. Are the deprived non-verbal? In F. Reissraan, J. Cohen,
& A. Pearl (Eds.). Mental health and the poor. New York: The
Free Press, 1964.

Reynolds, E. N. Interpersonal risk and self-actualization in four reli-
gious groups (Doctoral dissertation, Case-Western Reserve University,
1968). Dissertation Abstracts International, 1970, S^' 4019A-4029A.
(University Microfilms No. 70-5046)

Roczak, B., & Roczak, T. Masculine/feminine: Readings in sexual

mythology and the liberation of women. New York: Harper & Row,
1969.

Rogers, C. Client Centered Therapy. Boston: Houghton-Mifflin, 1951.

Rogers, C, & Dymond, R. (Eds.). Psychotherapy and personality change.
Chicago: University of Chicago Press, 1954.

Rogers, M. I. Self-actualization as process (Doctoral dissertation,
Case-Western Reserve University, 1968). Dissertation Abstracts
International. 1970, 30, 4380B-4381B. (University Microfilms No.
70-5049)

Roscoe, J. T. Fundamental research statistics for the behavioral

sciences (2nd ed.). New York: Holt, Rinehart, & Winston, 1975.

Rosenkrantz, P. S., Vogel, S. R. , Bee, H. , Broverman, I. K. , & Broverman,
D. M. Sex-role stereotypes and self-concepts in college students.
Journal of Consulting and Clinical Psychology, 1968, 32(3), 287-
295.

Roskind, W. L. A longitudinal study comparing chronically unemployed
with other employees in an industrial organization (Doctoral
dissertation, Wayne State University, 1974). Dissertation Abstracts
International, 1975, 35.. 3638B. (University Microfilms No. 74-29,
852)

Rottenburg, S. On choice in the labor markets. In W. Galenson & S. M.
Lipset (Eds.), Labor and Trade Unionism. New York: Wiley & Sons,
1960.

Rotter, J. B. Generalized expectancies for internal versus external
control of reinforcement. Psychological Monographs, 1966, 8£
(1, Whole No. 609).

Rotter, J. B., Chance, J. E. , & Phares, E. J. Applications of a social
learning theory of personality. New York: Holt, Rinehart, &
Winston, 1972.

Rotter, J. B., & Mulray, R. C. Internal versus external control of

reinforcement and decision time. Journal of Personality and Social
Psychology, 1965, 1, 598-604.

195

Saleh, S. D. , & Singh, T. Work values of white collar employees as a

function of sociological background. Journal of Applied Psvcholosv
1973, 58(1), 131-133. ~ ^ ^'

Salipante, P., Jr., & Goodman, P. S. Training, counseling, and retention
of hard-core umemployed. Journal of Applied Psychology, 1976
62_(1), 1-11. ^

Sartre, J. Existentialism and human emotions. New York: The Wisdom
Library, 1957.

Scanlon, J. W. , Nay, J. N., & Wholey, J. S. An evaluation system to

support a decentralized, comprehensive manpower program. In M. E.
Borus (Ed.), Evaluating the impact of manpower programs. Lexington,
Ma.: D, C, Heath, 1972. ~~~

Schesta, J. F. Leisure: Compensation for job dissatisfaction (Doctoral
dissertation. University of Missouri-Columbia, 1975). Dissertation
Abstracts International, 1976, 36, 6A88A. (University Microfilms
No. 76-7547) ~~

Schroeder, C. C. Sex differences and growth toward self-actualization
during the freshmen year. Psychological Reports, 1973, 32(2),
416-418.

Schroeder, C, C, , & LeMay, M. L. The impact of coed residence halls on
self-actualization. Journal of College Student Personnel. 1973
i4(2), 105-110, ~

Scott, S. H, Impact of residence hall living on college student develop-
ment. Journal of College Student Personnel, 1975, 16(3), 214-219,

Scott, W. A. Social psychological correlates of mental illness and
mental health. Psychological Bulletin, 1958, 55^(2), 65-87.

Searls, D. J., Braucht, G. N., & Miskimins, R. W. Work values of the
chronically unemployed. Journal of Applied PsycholoRy, 1974
59(1), 93-95, " ^ ^

Seeman, M, On the meaning of alienation. American Sociological Review,
1959, 24(6), 783-791,

Seeman, M. Alienation and social learning in a reformatory, American
Journal of Sociology, 1963, 69(3), 270-284,

Seeman, M. Alienation and engagement. In A. Campbell & P. Converse
(Eds,), The human meaning of social change. New York: Russell
Sage Foundation, 1972,

Seeman, M. , & Evans, J, W. Alienation and learning in a hospital situ-
ation. American Sociological Review, 1962, 27^(6), 772-783,

Shanthamani, V. S, Unemployment and neuroticism. The Indian Journal
of Social Work, 1973, 34(1), 43-45.

196

Shemberg, K. M. , & Leventhal, D. B. Masculinity-femininity and the need
for social approval. Journal of Projective Techniques and Person-
ality Assessment, 1968, 32(6), 575-577.

Sheppard , H. L. The value of attitude and opinion measures in manpower
evaluation research. In M. E. Borus (Ed.), Evaluating the impact
of manpower programs. Lexington, Ma.: D. C. Heath, 1972.

Sheppard, H. L. , & Herrick, N. Where have all the robots gone? New
York: The Free Press, 1972.

Shore, M. F., & Massimo, J. L. Five years later: A follow up study

of comprehensive vocationally oriented psychotherapy. In C. Caplan
& S. Lebovici (Eds.), Adolescence, psychosocial perspectives. New
York: Basic Books, 1969.

Shostrom, E. L. An inventory for the measurement of self-actualization.
Educational and Psychological Measurement, 1964, TAil) , 207-218.

Shostrom, E. L. Comment on a test review: The Personal Orientation

Inventory. Journal of Counseling Psychology, 1973, ^(5), 479-481.

Shostrom, E. L. Manual: Personal Orientation Inventory. San Diego:
Educational & Industrial Testing Service, 1974.

Shostrom, E. L., & Knapp , R. A. The relationship of a measure of self-
actualization (Personal Orientation Inventory) to a measure of
pathology (Minnesota Multiphasic Personality Inventory). American
Journal of Psychotherapy, 1966, 20(1), 193-202.

Shybut, J. Time perspective, internal versus external control, and

severity of psychological illness. Journal of Clinical Psychology,
1968, 24_(3), 312-315.

Singer, J. Androgyny: Toward a new theory of sexuality. Garden City,
N.J.: Anchor Press/Doubleday , 1976.

Slocum, J. W. , & Strawser, R. H. Racial differences in job attitudes.
Journal of Applied Psychology, 1972, 56(1), 28-32.

Smith, E. J. Work attitudes and the job satisfaction of black workers.
Vocational Guidance Quarterly, 1977, 25(3), 252-263.

Smith, N. R. The entrepreneur and his firm. East Lansing, Mi.: Graduate
School of Business, Michigan State University, 1967.

Solie, R. J. Employment effects of retraining the unemployed. Indus-
trial and Labor Relations Review, 1968, 2J.(2), 210-225.

Somers, G. G. Retraining the unemployed: A preliminary survey. In S.
Lebergott (Ed.), Men without work: The economics of unemployment.
Englewood Cliffs, N.J.: Prentice-Hall, 1964.

197

Spence, J. , & Helmreich, R. The Attitudes Towards Women Scale: An
objective instrument to measure attitudes toward the rights and
roles of women in contemporary society. JSAS Catalogue of Selected
Documents in Psychology, 1972, 1, 66-67. (Ms. No. 153)

Spence, J., Helmreich, R. , & Stapp, J. Ratings of self and peers on

sex-role attributes and their relation to self-esteem and concep-
tions of masculinity and femininity. Journal of Personality and
Social Psychology, 1975, 32(1), 29-39.

Staples, R. Male-female sexual variations: Function of biology or
culture? Journal of Sex Research, 1973, 9^(1), 11-20.

State of Florida Office of Manpower Planning. History and coordination
of employment and training programs: Summer, 1978. Tallahassee,
Fl.: Author, 1978.

Steinmann, A. Lack of communication between men and women. Marriage
and Family Living, 1958, 20(4), 350-352.

Steinmann, A., Fox, D., h Farkas , R. Male and female perceptions of

male sex-roles. Proceedings of the 76th Annual Convention of the
American Psychological Association, 1968, 2. 421-422. (Summary)

Sternglanz, S., & Serbin, L. Sex-role stereotyping in children's tele-
vision programs. Developmental Psychology, 1974, 10(5), 710-715.

Stewart, A., & Winter, D. Self-definition and social definition in
females. Journal of Personality, 1974, 42(2), 238-259.

Strickland, B. R. The prediction of social action from a dimension of
internal-external control. Journal of Social Psychology, 1965
66(2), 353-358. ^

Super, D. Career education and the meaning of work. Washington, D. C. :
U. S. Government Printing Office, 1976.

Tausky, C. Meaning of work among blue collar men. Pacific Sociological
Review, 1969, j_2(l), 49-54.

Thorpe, C. B. An exercise in the reclamation of human resources: A study
in manpower development. Journal of Negro Education, 1973, 42(1)
11-18. —

Tiffany, D. W, , Cowan, J. R. , Eddy, W. , Glad, D. , h Woll, S. Part I:
Work involvement and self-perceptions of ex-psychiatric patients:
An exploratory study. Kansas City, Mo.: Institute for Community
Studies, 1967.

Tiffany, D. W. , Cowan, J. R. , & Tiffany, P. M. The unemployed; A

social-psychological approach. Englewood Cliffs, N. J. : Prentice-
Hall, 1970.

198

Tosi, D. J., & Lindamood, C. A. The measurement of self-actualization:
A citical review of the Personal Orientation Inventory. Journal of
Personality Assessment. 1975, 2i(3), 215-224.

Triandis, H. C. , Feldman, J., Weldon, D. , & Harvey, W. Ecosystem

distrust and the hard to employ. Journal of Applied Psychology,
1975, 60(1), 45-56.

Vance, E. M. Relationship of self-actualization to mental health (Doc-
toral dissertation. North Texas State University, 1967). Disser-
tation Abstracts, 1967, ^, 135A. (University Microfilms No. 67-
8084)

Vogel, S., Rosenkrantz, P. S., Broverman, I. K. , Broverman, D. M. , 6.

Clarkson, F. Sex-role concepts and life style plans of young women.
Journal of Consulting and Clinical Psychology, 1975, 4^(3), 427.

Wakefield, J. A., Sasek, J., Friedman, A. F. , & Bowden, J. D. Androgyny
and other measures of masculinity-femininity. Journal of Consulting
and Clinical Psychology, 1976, 44(5), 766-770.

Wall, J. B. Relationship of locus of control and self-actualization.
Psychological Reports, 1970, 22(1), 282.

Wanous, J. P. Individual differences and reactions to job character-
istics. Journal of Applied Psychology, 1975, _59(5) , 616-622.

Warehime, R. G. , Routh, D. R. , & Foulds, M. L. Knowledge about self-
actualization and presentation of self as self-actualized. Journal
of Personality and Social Psychology, 1974, 2^(1), 155-162.

Warheit, G. J., Schwab, J. J., Holzer, C. E. , III, & Nadreau, S. E. New

data from the south on race, sex, age, and mental illness. Paper

Presented at the American Sociological Association National Conven-
tion, New York, New York, 1973.

Waters, C. W. , Waters, L. K. , & Pincus, S. Factor analysis of masculine
and feminine sex-typed items from the Bern Sex-Role Inventory.
Psychological Reports. 1977, 40(2), 567-570.

Weiss, R. S. , & Kahn, R. L. Definitions of work and occupation. Social
Problems, 1960, 8(2), 142-151.

Wesch, J. E. Self-actualization and the fear of death (Doctoral disser-
tation. University of Tennessee, 1970). Dissertation Abstracts
International, 1971, 2i» 6270B-6271B. (University Microfilms No.
71-7692)

Wexler, D. A. Self-actualization and cognitive processes. Journal of
Consulting and Clinical Psychology, 1974, j42(l)> 47-53.

199

White, H. M. An investigation of some characteristics of high and low
self-actualization and their relationship to alienation from self
and society (Doctoral dissertation. University of California,
Berkeley, 1970). Dissertation Abstracts International, 1971, 31,
5142B. (University Microfilms No. 71-9954)

Wiggins, J. S. , & Holsmuller, A. Psychological androgyny and inter-
personal behavior. Journal of Consulting and Clinical Psychology,
1978, 46(1), 40-52.

Wilensky, H. Work as a social problem. In H. S. Becker (Ed.), Social
problems: A modern approach. New York: Wiley, 1966.

Williams, C, & Nickels, J. Internal-external control dimension as

related to accident and suicide proneness. Journal of Consulting
and Clinical Psychology, 1969, 33i^) , 485-494.

Williams, R. S. , Morea, P. C. , & Ives, J. M. The significant os work:
An empirical study. Journal of Occupational Psychology, 1975,
48(1), 45-51.

Wills, B. S. Personality variables which discriminate between groups
differing in levels of self-actualization. Journal of Counseling
Psychology, 1974, 2j_(3), 222-227.

Wilson, C. New pathways in psychology: Maslow and the post-f reudian
revolution. New York: The New American Library, 1972.

Wolk, S. Situational constraint as a moderator of the locus of control-
adjustment relationship. Journal of Consulting and Clinical Psy-
chology, 1976, 44(3), 420-427.

Work in America: A special report to the Secretary of Health, Education,

and Welfare, prepared under the auspices of the W. E. Upjohn Institute
for Employment Research. Cambridge, Ma.: The MIT Press, 1973.

Worrell, J. Sex-roles and psychological well-being: Perspectives on

methodology. Journal of Consulting and Clinical Psychology, 1978,
46(4), 777-791.

Woudenberg, R. A. The relationship of sexual attitydes, sexual stereo-
types, racial-sexual stereotypes, and racial attitudes (Doctoral
dissertation, Michigan State University, 1973). Dissertation
Abstracts International, 1973, 2i» 2958B. (University Microfilms
No. 73-29, 805)

Yonge, G. D, Time experiences, self-actualizing values, and creativity.
Journal of Personality Assessment, 1975, 39_i6) , 601-606.

Zawailzki, B. , & Lazarsfeld, P. Psychological consequences of unemploy-
ment. Journal of Social Psychology, 1935, 6^, 224-251.

200

Zeisel, J. A profile of unemployment. In S. Lebergott (Ed.). Men

without work; The economics of unemployment. Englewood Cliffs,
N.J.: Prentice-Hall, 1964.

Zeldow, P. B. Effects of non-pathological sex-role stereotypes on

student evaluations of psychiatric patients. Journal of Consulting
and Clinical Psychology. 1976, 44(2), 304.

Zerlga, W. D., Tseng, M. S., & Greever, K. B. Stability and concurrent
validity of the Rotter Internal-External Locus of Control Scale.
Educational and Psychological Measurement, 1976, 36^(2), 473-475.

BIOGRAPHICAL SKETCH

John Anderson Sanford was born on May 6, 1948, in Wichita, Kansas.
He was raised and educated in Borger, Texas, and Bartlesville, Oklahoma,
and graduated from the College High School, Bartlesville, in May 1966.
He attended the University of Kansas, Lawrence, Kansas, from 1966 through
1970 and received the Bachelor of Arts degree in the double fields of
psychology and philosophy. He attended graduate school at the Kansas
State University, Manhattan, Kansas, from 1972 through 1974 and received
the Master of Science Degree in guidance and counseling. He thesis title
was Self-Esteem Levels of Homosexuals in Two University Communities.

Mr. Sanford was married to Robin Ann Runyan in 1976. He served with
the United States Army between 1970 and 1972 as a counselor for enlisted
personnel stationed in the Stuttgart, Germany, US Army Medical Service
Area. During his graduate work Mr. Sanford has been actively involved
in social serve activities. He has served as a volunteer trainer with the
Alachua County Crisis Center training suicide and crisis intervention
volunteers from 1974 to 1980. He was employed as a supervisor and
counselor by the Alachua County Comprehensive Employment and Training
Program from 1976 to 1979.

201

I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation
and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy,

Harold C. Riker, Chairman
Professor of Counselor Education

I certify that I have read this study and that in ray opinion
it conforms to acceptable standards of scholarly presentation
and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.

lid liQA )j;(^.(\iU

Roderick McDavis

Associate Professor of Counselor
Education

I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation
and is fully adequate, in scope and quality, as a dissertation
for the degree of Doctor of Philosophy.

^James L. wattenbarger
Professor of Education *^dministration

This dissertation was submitted to the Graduate Faculty of the
Department of Counselor Education in the College of Education
and to the Graduate Council, and was accepted as partial
fulfillment of the requirements for the degree of Doctor of
Philosophy.

March, 1981

Dean, Graduate School