University of Southern California Dissertations and Theses

A Biosocial Model Of Sex-Atypicality In Female Employment Aspirations

(USC Thesis Other)

A Biosocial Model Of Sex-Atypicality In Female Employment Aspirations

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. tJMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may be
from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI
A Bell & Howell Information Company
300 North Zeeb Road, Ann Arbor MI 48106-1346 USA
313/761-4700 800/521-0600

A Biosocial Model of Sex-Atypicality in Female Employment
Aspirations
by
Dawn McKenna
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
(Occupational Therapy)
December 1994
Copyright 1994 Dawn McKenna
UMI Number: 1378426
UMI Microform 1378426
Copyright 1996, by UMI Company. Ail rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
UNIVERSITY O F SOU TH ERN CALIFORNIA
TH E G R AD U A TE SC H O O L
U N IV E R SIT Y PA R K
LOS A N G ELE S. C A L IF O R N IA 9 0 0 0 7
This thesis, written by
.Dawn.. .McKenna..........................................................
under the direction of her. Thesis Comm ittee,
and approved by all its members, has been p re
sented to and accepted by the Dean of The
Graduate School, in partial fulfillment of the
requirements for the degree of
Master o f Arts
CC. . ^
Dean
Date...$szmh&i:.15.±.19.9A..
THESIS COMMITTEE
ACKNOWLEDGMENT
I wish to thank the members of my thesis committee, Dr.
Cynthia Hedricks, Dr. Ann Neville-Jan, and Dr. Diane Parham,
for their time and suggestive comments. I was fortunate to
work with Dr. Cynthia Hedricks as my thesis chair. Without
her, I would not have been able to accomplish my goals. I
would also like to thank Dr. Michael Carlson for his assistance
with the initial statistical analysis.
I am most appreciative of my family and friends for all
their advice and support, especially, Gardner Monks, for
teaching me how to use a computer and for listening.
TABLE OF CONTENTS
ACKNOWLEDGMENT ......................................................................... ii
LIST OF TABLES................................................................................ vii
ABSTRACT ......................................................................................... x
CHAPTER I. THE PROBLEM............................................................. 1
Background ............................................................................. 3
Rationale and Significance ............................................... 5
The Sam ple............................................................................. 7
Research Design .................................................................... 8
Expectations .......................................................................... 9
Assumptions .......................................................................... 9
Limitations ..............................................................................10
CHAPTER II. REVIEW OF THE LITERATURE ............................... 11
Introduction ........................................................................... 11
Relevance of Female Employment Aspirations to
Occupational Science and Occupaitonal Choice .......... 13
Social Model................................................................................. 18
Ordinal position of birth .......................................... 19
Mother's employment s t a t u s ................................. 19
Father's occupational prestige................................20
Parental educational level ...................................... 20
Childhood encouragement ........................................ 21
Summary ....................................................................... 22
Biological Model .................................................................... 23
Androgens ..................................................................... 25
Behavioral correlates of prenatal
exposure to androgen..................................... 26
Behavioral correlates of circulating
androgen in adulthood ................................... 27
Estrogens ...................................................................... 29
Behavioral correlates of prenatal
exposure to estrogen ..................................... 29
Behavioral correlates of circulating
estrogens in adulthood ................................. 30
Progestogens ................................................................30
Behavioral correlates of prenatal
exposure to progestogen................................31
Behavioral correlates of circulating
progestogens in adulthood.............................31
Oral contraceptive use...............................................32
Summary .................................................................................. 33
CHAPTER III. METHODOLOGY ..........................................................35
Subject Recruitment ............................................................35
Instrumentation ................................. 37
Female Senior Questionnaire ..................................37
Index of Sex-Role Orientation ................................38
Other inventories ....................................................... 38
Experimental Protocol ......................................................... 39
Radioimmunoassay of Blood Samples .............................43
Data Analysis ..........................................................................44
Predictor and criterion variables .........................44
Cycle to cycle differences ...................................... 45
Non oral contraceptive versus oral
contraceptive users .................................................. 45
Transformations .........................................................45
Relationships between variables: Single
correlations .................................................... 46
Biological, social, and biosocial models of
sex-atypical behavior: Multiple regression
analysis..............................................................................47
CHAPTER IV . RESULTS .................................................................... 48
Comparisons Between Non Oral Contraceptive
Users (Non OC) Versus Oral Contraceptive Users
(OC).............................................................................................48
Non Oral Contraceptive Users (Non OC
Correlation analyses ............................................................48
Employment aspirations ...............................53
College major .........................................................................54
ISRO .....................................................................54
Multiple regression analyses: Social model 54
Employment aspirations ..........................................56
College major .................................................. 56
Multiple regression analyses: Biological
m o d els........................................................................... 60
Employment aspirations ...............................60
Multiple regression analyses: Biosocial
m o d els........................................................................... 60
Employment aspirations ...............................63
Summary ............................................................63
Oral Contraceptive Users (OC) ..........................................67
Correlation analyses ................................................ 67
Employment aspirations ...............................69
College major .................................................. 69
ISRO .....................................................................69
Multiple regression analyses: Social models .. 70
Employment aspirations ...............................70
College major .................................................. 70
ISRO .....................................................................74
Multiple regression analyses: Biological
m o d els........................................................................... 76
Multiple regression analyses: Biosocial
m o d els........................................................................... 76
Summary .......................................................................76
CHAPTER V. DISCUSSION ....................... 77
Introduction .......................................................................... 77
Comparisons between non oral contraceptive
users (non OC) versus oral contraceptive users
(OC) ........................................................................................... 80
Non oral contraceptive u s e r s .............................................80
Employment aspirations ..........................................80
College major ............................................................. 84
Summary .......................................................................85
Oral contraceptive users ....................................................86
Employment aspirations ........................................86
College major ............................................................87
ISRO ...............................................................................87
Summary .................................................................... 88
Summary ............................................................................... 88
Comparisons between the non OC versus the
OC users ......................................................................88
Social predictor variables .................................... 89
Hormonal predictor variables ...............................90
Behavioral criterion variables.............................. 90
Employment aspirations .............................90
College major .................................................93
ISRO scores ................................................... 93
Conclusions ..........................................................................94
REFERENCES ......................................................................................97
APPENDIX A. Definition of term........................................................106
APPENDIX B. Operational definitions of variable.................. 108
APPENDIX C. Definition of variable labels ..............................113
APPENDIX D. Correlation analyses between predictor and
criterion variables for the non oral
contraceptive users...................................................115
APPENDIX E. Correlation analyses between predictor and
criterion variables for the oral
contraceptive users...................................................119
LIST OF TABLES
TABLES
1 Mean (+SD) hormonal differences between non oral
contraceptive (Non OC) users (n=33) and oral
contraceptive (OC) users (n=24) ......................................... 49
2 Mean (+SD) differences between non oral
contraceptive (Non OC) users (n=33) and oral
contraceptive (OC) users(n=24) on the behavior
criteria variables .................................................................... 50
3 Mean (+SD) differences between non oral
contraceptive (Non OC) users (n=33) and oral
contraceptive (OC) users (n=24) on the continuous
social predictor variables .................................................... 51
4 Differences between non oral contraceptive (Non
OC) users (n=33) and oral contraceptive (OC) users
(n=24) on the dichotomous social predictor
variables ..................................................................................... 52
5 Summary of significant correlations between social
and hormonal predictor variables and behavioral
criterion variables: Non oral contraceptive users
(n= 33)...........................................................................................55
6 Social model for sex-atypicality of employment
aspirations (RMALE80) of non oral contraceptive
(Non OC) users ...........................................................................57
7 Social model for occupational prestige of
employment aspirations (RPREST) of non oral
contraceptive (Non OC) users ...............................................58
8 Social model for sex-atypicality of college
major (UNCFIELD) of non oral contraceptive (Non OC)
users .............................................................................................59
9 Biological model for sex-atypicality of employment
aspirations (RMALE80) for non oral contraceptive
(Non OC) users ...........................................................................61
10 Biological model for occupational prestige of
employment aspirations (RPREST) for non oral
contraceptive (Non OC) users .............................................. 62
11 Biosocial model for sex-atypicality of employment
aspirations (RMALE80) for non oral contraceptive
(Non OC) users ...........................................................................64
12 Biosocial model for occupational prestige of
employment aspirations (RPREST) for non oral
contraceptive (Non OC) users ...............................................65
13 Comparison of social, biological, and biosocial
models to predict sex-atypicality of the
employment aspiration (RMALE80): Non oral
contraceptive users .................................................................68
14 Summary of significant correlations between
social and hormonal predictor variables and
behavioral criterion variables: Oral contraceptive
users (n=24) ...............................................................................71
15 Social model for occupational prestige of
employment aspirations (RPREST) of oral
contraceptive (OC) users ....................................................... 72
16 Social model for sex-atypicality of college major
(UNCFIELD) of oral contraceptive ( OC) users .................73
Social model for sex-atypicality of the ISRO scores
of oral contraceptive (OC) u s e rs ......................................... 75
ABSTRACT
Social and hormonal factors were simultaneously
evaluated for their ability to predict sex role behavior of 57
female college seniors. Sex role behavior was defined as sex-
atypicality of employment aspirations, occupational prestige,
college major, and score on the Index of Sex Role Orientation
(ISRO).
Behavioral data were gathered through self-administered
questionnaires. Levels of androgens (testosterone,
androstenedione, DHEA, DHEAS), SHBG, estradiol and
progesterone were assayed from blood samples collected
across two menstrual cycles. Due to hormonal and behavioral
differences between those who used oral contraceptives (n=24)
and those who did not (non OC users; n=33), data from these
two groups were analyzed separately.
Social and hormonal factors significantly predicted sex
role behavior. However, a biosocial regression model that
included social and hormonal factors predicted more
variability in sex role behavior than did either a social or
biological (hormonal) model, alone. Because this result was
seen only for non OC users, it was recommended that oral
contraceptive use be considered in future studies that examine
social or hormonal factors related to female sex role behavior.
Chapter I
THE PROBLEM
In the past twenty years, there has been an increase in
the number of women who have entered into employment that
has traditionally been dominated by men (Reskin & Roos,
1987), such as law, engineering or politics (Jones & Lamke,
1985). Most women, however, are concentrated in a small
number of jobs that employ predominantly women (Kaufman,
1989), such as sales and clerical or in professional jobs such
as nursing or elementary education (Jones & Lamke, 1985). For
the purposes of this paper, sex-typicality of employment was
defined as the percentage of all males who held that particular
job according to the 1980 U.S. Department of Labor. A job held
by a high percentage of males was considered sex-atypical for
women, and conversely, a job held by a low percentage of
males was considered sex-typical for women. The actual
percentage of males who held the job in 1980 that the women
in this study aspired to became the women's "score" for
examining the degree of sex-atypicality of her employment
aspiration.
This project emerged from the fact that biological and
social scientists have separately attem pted to predict sex-
atypicality of female employment. A social model proposes
l
that differences between women with respect to sex-
atypicality of employment reflect differences in their family
backgrounds and childhood experiences. A biological model
proposes that differences between women with respect to
sex-atypicality of employment reflect differences in their
reproductive hormone activity. This project attem pted to
construct a biosocial model of sex-atypicality of female
employment aspirations by using components of both the social
and the biological model.
The data have been taken from an existing database on
female college seniors, collected in 1987 at the University of
North Carolina at Chapel Hill. Senior college women were
chosen for the study because it was expected that: (1) there
would be more variability in the sex-atypicality of their
employment aspirations compared to females with a high
school degree or less education; (2) they were being faced with
what to do with their lives following graduation; and (3) the
hormonal system of these women had not yet been influenced
by the experience of engaging in employment.
Additionally, social and hormonal factors were examined
for relationships to sex-atypicality of college major, because
of the potential relationship of college major to future
employment, and for relationships to the woman's score on the
Index of Sex-Role Orientation (ISRO), because it is a general,
2
standardized index of sex-role behavior. Appendix A contains
the operational definitions of the independent and the
dependent variables examined in this study.
The women in this study were divided into two groups:
non oral contraceptive users and oral contraceptive users.
Preliminary data analysis revealed that there were significant
differences in hormone levels, sex-atypicality of employment
aspirations, and sex-atypicality of college majors between the
two groups of women.
Background
Social scientists have identified significant correlates
of sex-atypical employment choice in women. These
predictors include: (1) being first born or only children; (2)
having a mother who worked outside the home; (3) having
fathers with high prestige jobs; (4) having educated parents;
and (5) receiving academic encouragement from parents
(Broverman, 1980; Houser & Garvey, 1985; Johnson, Nagoshi,
Ahern, Wilson, DeFries, McClearn, & Vandenberg, 1983; Lemkau,
1983; Mendelson; 1985 O'Connell, Betz, & Kurth, 1989;
Williams & McCullers, 1983). Therefore, this study examined
the influences of being a first born or only child, having a
mother who worked outside the home, having a father with a
high prestige job, having educated parents, and receiving
3
academic encouragement from parents on sex-atypicality of
employment aspirations, college major and ISRO scores of
senior college females.
Biological scientists have also identified significant
correlates of sex-atypical employment choice in women.
Specifically, women in sex-atypical employment were found to
have higher androgen levels than women in sex-typical
employment (Bancroft, Sanders, Davidson, & Warner, 1983;
Dongyun & Yumin, 1990; Purifoy & Koopmans, 1980; Schindler,
1979). Additionally, the hormone estrogen often exerts
behavioral effects similar to those of androgen, while the
hormone progesterone may counteract the influences of
androgen and estrogen (Feder, 1984). This study examined the
influences of androgens, estrogen and progesterone on sex-
atypicality of employment aspirations, college major and ISRO
scores of senior college females.
Both social and biological scientists have identified
significant correlates of sex-atypical employment in females.
Neither group of investigators has examined social and
biological factors simultaneously in the same subject
population to evaluate social versus hormonal influences.
Therefore, this project examined both social and hormonal
correlates with sex-atypicality of female employment
aspirations using regression analyses. For reasons previously
4
stated, this project also examined these simultaneous effects
on sex-atypicality of college major and the womens' scores on
the Index of Sex-Role Orientation (ISRO).
Previous studies that have examined female employment
choice have not distinguished between women with respect to
oral contraceptive use. However, studies that have focused on
sexual or maternal behaviors have found hormonal and
behavioral differences between these two groups of women
(Alexander, Sherwin, Bancroft, & Davidson, 1990; Bancroft,
Sherwin, Alexander, Davidson, & Walker, 1991a, 1991b;
Harding, Vail, & Brown, 1985; Hedricks & Udry, 1990).
However, relationships between hormones and an index of sex
role behavior were not statistically significant (Bancroft et
al., 1991a). In the present study, 42% of the women were using
oral contraceptives; these women were found to have
significant hormonal and behavioral differences compared to
those women not using oral contraceptives. Therefore, the data
from the two groups of women were analyzed separately.
Rationale and Significance
Occupational science originated, and is being further
developed, at the University of Southern California (USC). It is
defined as the "systematic study of the human as an
occupational being" (Clark, Parham, Carlson, Frank, Jackson,
5
Pierce, Wolfe, & Zemke, 1991), where occupation is defined as
"chunks of culturally and personally meaningful activity in
which humans engage" (USC, Department of Occupational
Therapy, 1989, as cited in Clark et al., 1991). Employment is
one type of occupation; other occupations are carried out
during rest, play and leisure.
One purpose of occupational science is to investigate
why people choose to engage in certain occupations (Clark et
al., 1991). The faculty in the USC Department of Occupational
Therapy have developed a conceptual model consisting of six
subsystems to explain engagement in occupation. This model
was not intended to be equated with occupational science, but
rather it was developed to provide a coherent framework for
the organization of knowledge generated by occupational
scientists (Clark et al., 1991). Two of the subsystems that
this project addressed are the biological subsystem and the
sociocultural subsystem.
Within the area of occupational (i.e., employment) choice,
there are four different theoretical perspectives that have
been applied to women. These theories (Ginzberg, 1966,1972;
Holland, 1963; Roe, 1966; Super, 1966) propose that
employment choice is a function of interindividual differences
such as early parent-child interactions, personality, abilities
and interests, and environmental demands. These theories
6
were developed from limited empirical observations. Any data
collection that did occur was primarily on men. Furthermore,
although hormonal correlates of female employment choice
have been reported (Bancroft, et al., 1983; Dongyun & Yumin,
1990; Purifoy & Koopmans, 1980; Schindler, 1979), none of
the theories of occupational choice include a hormonal
contribution. It was expected that this project would expand
and improve upon current theories because this study: (1) was
on women, who are under-represented in research in this area;
and (2) it evaluated a hormonal contribution to occupational
choice.
The Sample
The data that I used were collected in 1987; thus, this
project was a secondary analysis of those data. The sample
consisted of 57 college senior females from The University of
North Carolina at Chapel Hill. Data were collected, in a series
of self-administered questionnaires, on future employment
aspirations, college major, sex-role orientation, family
background, childhood experiences, medical/reproductive
history, and personality. Blood samples were taken and
hormonal levels were obtained for total testosterone (TT), sex
hormone binding globulin (SHBG), androstenedione (A),
dehydroepiandrosterone (DHEA), dehydroepiandrosterone
sulfate (DHEAS), progesterone (P), and estradiol (E2). A free
testosterone index (FTI) was calculated for each specimen
using the following formula: FTI = TT/SHBG (Clark, Marcellus,
DeLory, & Bird, 1975) because not all of TT can influence
behavior. Only free testosterone (FT) and a considerable
fraction of testosterone (T) bound to albumin are biologically
active and thus can influence behavior (Manni, Pardridge,
Cefalu, Nisula, Bardin, Santner, & Santen, 1985). For reasons
previously stated, the sample was divided into two groups (non
oral contraceptive and oral contraceptive users) for the
purposes of data analysis.
Research Design
The research design was that of an ex post facto study.
To analyze the data, I first ran a series of descriptive
statistics and examined the distributions of the main
variables. From these descriptive statistics I looked at
means, variances and single correlations between variables.
Only those social and hormonal variables which had a
significant relationship to the dependent variables (sex-
atypicality of employment aspiration, college major and ISRO
scores), where p < 0.10, were included in the social, biological
and biosocial models. The statistical models were constructed
using multiple regression analysis (Statistical Analysis
8
System, Cary, NC). The statistical model did not test cause
and effect hypotheses but rather correlations between certain
behaviors, hormone levels and social factors.
Expectations
The following variables were expected to positively
correlate with sex-atypicality of employment aspirations,
college major and ISRO scores: (1) being the oldest or only
child; (2) having a mother who worked; (3) having a father in a
high prestige job; (4) having both parents with high levels of
education; (5) having received academic encouragement from
parents; (6) having a high FTI; (7) having a high blood level of
A, E2, DHEA, or DHEAS; and (8) having a low blood level of P.
I had no expectations regarding which subsystem
(sociocultural or biological) would correlate more
significantly with sex-atypicality of female employment
aspirations, college major and ISRO scores.
Assumptions
For the purposes of this study, it was assumed that the
data collection protocol was followed as reported in the
Methods Section, which was taken from notebooks kept by the
Principal Investigator (Dr. C. Hedricks) during subject
recruitment and data collection.
Limitations
A limitation of this study is that it was a correlational
design; therefore, I was unable to identify causal
relationships. The focus of the project was on finding
relationships between the variables. The study results are not
generalizable to: (1) men, because it examined sociocultural
and hormonal characteristics of women only; (2) women who
are not college seniors, i.e., those already in work force; (3)
women who have been pregnant because of possible hormonal
effects of pregnancy on later hormonal activity; or (4) non-
Caucasian women, as 84.2% of the sample were Caucasian.
Self-reported data based on memory were used. There was no
follow-up to determine whether the women actually engaged in
those occupations that they listed in the study. Also, the
sample studied was small (n=57), thus limiting the power of
the models obtained.
10
CHAPTER II
REVIEW OF THE LITERATURE
Introduction
The term "sex (or gender) role behaviors" has been used
as a term to describe behaviors that are thought to be
appropriate according to one's sex (or gender). Examples of
such behaviors may include employment choice, play rehearsal
of parenting behavior, physical energy expenditure or social
aggression (Longino, 1990). Two independent models have been
proposed to explain the acquisition of sex role behavior in
women. The social model attributes acquisition of sex role
behavior to family and childhood experiences. The biological
model attributes acquisition of sex role behavior to
reproductive hormone activity. The purpose of this project is
to use these two models to attem pt to create a biosocial
model of sex role behavior of women. A biosocial model is a
model that would simultaneously use, or incorporate, both
social and hormonal factors. The behavior that this paper
focused on is sex-atypicality of employment aspirations,
college majors and ISRO scores.
Sex role behavior is a subjective and biased term
because its description, and consequently the selection of data
11
in related research, is influenced by heterosexism (sexual
essentialism) or by the idea that there are sex-appropriate and
sex-inappropriate behaviors (Longino, 1990). Its definition
implies that people who take on roles that are not typical, or
"appropriate" for those of their sex, are somehow not normal.
Unfortunately, I have been unable to locate a term that is less
judgmental and that allows for descriptions which
differentiate sex role behaviors along a continuum rather than
being based on female or male stereotypes. Therefore, in this
paper I simply identified behaviors by their names (i.e.,
employment) wherever possible, rather than use the term sex
role behavior. I have found it useful to use the terms sex-
typical and sex-atypical to describe certain behaviors (Udry,
1988). The term sex-typical refers to those behaviors that are
typically exhibited by women. The term sex-atypical refers to
those behaviors that are typically exhibited by men. Appendix
B contains definitions of terms used in this paper.
This study focused on sex-atypicality of employment
aspirations of female college seniors. Sex-typical
employment by women, jobs that employ a high percentage of
females according to the 1980 U.S. Department of Labor, are
typically in low paying, nonprofessional jobs such as sales and
clerical or in the more nurturing professional jobs such as
nursing and elementary education (Lemkau, 1983; Williams &
12
McCullers, 1983). Sex-atypical employment by women are jobs
that employ a low percentage of females, according to the
1980 U.S. Department of Labor, such as law, engineering or
politics (Jones & Lamke, 1985). For this project, each
female's employment aspiration was given a numerical score
that corresponded to the percentage of males who held that
particular job according to the statistics compiled by the
1980 U.S. Department of Labor. This method enabled me to
examine employment aspirations of females on a continuum of
sex-typicality. This method was preferred over categorizing
employment aspirations into being either sex-typical or sex-
atypical as was the method used by the studies reviewed in
this paper. The studies reviewed in this paper categorized
sex-atypicality of employment using different percentages of
males in the job without validating this categorization (i.e.,
Lemkau (1983) used 75%; Jones & Lamke (1985) used 89%;
Williams & McCullers (1983) used 76%; and Houser & Garvey
(1985) used 80%).
Relevance of a Biosocial Perspective Towards Female
Employment Aspirations to Occupational Science and
Occupational Choice
Occupational science has been established at the
University of Southern California (USC). It is defined as the
13
"systematic study of the human as an occupational being"
(Clark, Parham, Carlson, Frank, Jackson, Pierce, Wolfe, &
Zemke, 1991), where occupation is defined as "chunks of
culturally and personally meaningful activity in which humans
engage" (USC, Department of Occupational Therapy, 1989, as
cited in Clark et al., 1991). Employment is one type of
occupation.
One purpose of occupational science is to investigate the
reasons why people choose to engage in certain occupations
(Clark et al., 1991). The faculty in the USC Department of
Occupational Therapy have developed a model consisting of six
subsystems that are expected to influence occupation:
physical; biological; information processing; sociocultural;
symbolic-evaluative; and transcendental. This model was not
intended to be equated with occupational science, but rather to
provide a coherent framework for the organization of
knowledge generated by occupational science research (Clark
et al., 1991). Two of the subsystems that this paper addressed
are the biological subsystem and the sociocultural subsystem.
Information pertinent to the biological subsystem includes
biological contributions to the engagement in occupation
(Clark et al., 1991), i.e., hormonal correlates of employment
aspirations. The sociocultural subsystem focuses on the
individual's perceptions of social and cultural expectations for
14
behavior (Clark e t al., 1991), and includes family
characteristics and childhood experiences. It was expected
that this project would contribute to the knowledge of
occupational science because it examined how biological (i.e.,
hormonal) and sociocultural (i.e., family characteristics and
childhood experiences) related to employment aspirations.
Examining the factors related to employment choice is
one area of interest to occupational science. Within the area
of occupational (i.e., employment) choice, there are four
different theoretical perspectives that have been applied to
women. Roe (1966) proposed a theory of male employment
choice that examined some biological and social factors after
a review of psychoanalytical studies, Maslow's "need" theory,
and genetic and evolutionary theories. Roe theorized that early
parent-child interactions mediate the “need” hierarchy, which
is thought to be related to later levels of motivation. She felt
that an individual had significant predispositions to certain
employment choices based on genetic influences on skills and
abilities. These skills and abilities, in conjunction with an
individual’s response to parenting style, were what Roe
believed would determine employment choice.
Holland (1963, 1985), a vocational counselor who made
frequent observations on men, proposed that there were six
specific personality types that corresponded to an employment
15
environment: realistic; investigative; artistic; social;
enterprising; and conventional. According to Holland, choosing
employment is a three-stage process where employment
selection is based on a self-evaluation of one's ability to be
successful in the chosen employment environment. First,
experience leads an individual to develop preferences and
interests in some activities. Second, an individual searches
for an environment that allows him/her to exercise their
capabilities and express their values and attitudes. Third, an
employment choice is reached by an interaction between the
individual's personality and the employment environment.
Ginzberg (1966, 1972) originally proposed that a
relatively irreversible process occurs from adolescence
through early adulthood, as an individual compromises between
societal and individual factors in choosing employment. His
original theory (1966) was based on a study of 60 upper class,
young men (aged 11-24). He later theorized (1972) that an
employment choice could be changed in adulthood. He felt that
an individual continually attem pted to find a balance between
abilities, interests and available employment opportunities
when making employment choices. His revised theory included
a study on educated women in which he found that women did
not fit the male model of transitioning from school to full
time employment. This was because women interrupted their
16
education for marriage, modified their career objectives after
marriage, and tended to shift their careers between work and
home.
Super (1966), who expanded upon Ginzberg's work through
a longitudinal study on 142 ninth-grade boys, proposed that in
choosing employment, males integrated their abilities,
interests and environmental demands and reached a
compromise between these three factors. This compromise in
choice was thought to be a continuous process that occurred
throughout life as men changed with time and experience.
Some of the above-mentioned theories (Ginzberg, 1966,
1972; Holland, 1963, 1985; Roe, 1966; Super, 1966) have
been generated with limited empirical observations. Except
for Ginzberg (1972), data collection has been solely on males.
Finally, no hormonal measures have been included in the
theories, although hormonal correlates of female employment
choice have been reported (Bancroft et al., 1983; Dongyun &
Yumin, 1990; Purifoy & Koopmans, 1980; Schindler, 1979).
This study attem pted to reduce the limitations of previous
studies while contributing to occupational science in the
following ways: (1) empirical observations on women were
used to create a biosocial model; (2) variables included in this
model were those that were already demonstrated to be
significant predictors of employment choice in women; and (3)
17
hormonal measures were included in this model.
Social Model
For the purposes of this study, the social model was
primarily concerned with social influences on the females
when they were between the ages 5-15. One of the social
learning theories maintains that learning occurs through
observation and imitation (Schiamberg, 1985). This age period
was chosen because: (1) it is a time when imitative behaviors
are most likely to be developing; and (2) it was felt that
subjects could recall parental and other social influences
during this stage of development.
Significant relationships have been reported between
social influences and sex-atypicality of female employment
aspirations. They include: ordinal position of birth; mother's
employment status; father's occupational prestige; parental
educational level; and academic and employment
encouragement from parents. The relationships between these
factors and sex-atypicality in employment aspirations are
outlined below.
As previously noted, a limitation of the following
studies was that employment choice was categorized as either
sex-typical or sex-atypical without empirical validation of
this method of categorization. Each researcher used different
18
criteria for categorizing employment choice; therefore, it is
possible that the results of these studies are indicative of a
trend, but are not conclusive.
Ordinal position (oldest, youngest, middle) or being an only
child
Lemkau (1983) found that women who were first-born or
only children were more likely to make sex-atypical
employment choices than women of other birth orders.
Williams & McCullers (1983) also found that women who
became physicians or lawyers, sex-atypical employment
choices, were more likely to be first born or only children as
compared to the women who became licensed practical nurses
or legal secretaries, which are sex-typical employment
choices.
Mother's employment status
Having a mother who worked outside the home has been
positively associated with whether or not her daughter will
enter the work force (Houser & Garvey, 1985). There are
discrepancies as to whether this correlates with the decision
to enter sex-typical or sex-atypical employment; however, in
several studies, the mothers of women who entered sex-
atypical employment were employed for more years than the
mothers of those women who entered sex-typical employment
19
(Houser & Garvey, 1985; Lemkau, 1983; O'Connell, Betz, &
Kurth, 1989).
Father's occupational prestige
Williams & McCullers (1983) found that fathers who
were employed in professional or managerial careers were
more likely to have daughters in sex-atypical employment than
fathers who were employed in other job categories.
Furthermore, prestige of father's job has been positively
associated with the occupational prestige of the job held by
his daughter (Johnson, Nagoshi, Ahern, Wilson, DeFries,
McClearn, & Vandenberg, 1983).
Parental educational level
Parents of women in sex-atypical employment have been
found to be better educated than parents of women in sex-
typical employment (O'Connell et al., 1989; Williams &
McCullers, 1983). In a study of 173 female students pursuing
sex-atypical employment (engineering and veterinary
medicine) and sex-typical employment (nursing), O'Connell et
al. (1989) found that the number of years the father attended
school was more positively related to sex-atypical
employment choices of their daughters than the number of
years the mothers attended school.
Williams & McCullers (1983) studied eighty women
employed in various positions from the fields of either
20
medicine or law. These positions ranged from sex-atypical
choices such as physicians or lawyers to sex-typical choices
such as certified medical assistants or legal secretaries.
They found that the number of years that both mothers and
fathers attended school was significantly, positively,
associated with sex-atypical employment choices of their
daughters.
Childhood encouragement
Parental values and encouragement appear to be related
to both educational and employment plans of daughters (for a
review of the literature, see Mendelson, 1985). Broverman
(1980) and Williams and McCullers (1983) found that a
mother's encouragement of her daughter's education positively
influenced her daughter's educational attainment. Houser and
Garvey (1985) found that women who aspired to sex-atypical
employment had consistently received more support from their
parents, peers, educators and counselors than women who were
planning to enter sex-typical employment.
Lemkau (1983) reported that women who received
positive support from men that they generally had regular
close contact with, such as their fathers and male teachers,
were likely to choose sex-atypical employment. In
comparison, women who received positive support from other
women outside their immediate family, such as nonparental
21
female relatives and female teachers, were likely to choose
sex typical employment. Recently, the Los Angeles Times
reported on a study that surveyed the top 25 women politicians
(Roan, 1992). The study found that almost all 25 women had
received positive encouragement, from one or both parents,
that they "could be anything they wanted" while they were
growing up.
Summary
It appears that there are consistent relationships
between early social factors and sex-atypicality of
employment aspirations of women. Women were more likely to
enter into sex-atypical employment if they: (1) were first
born or only children; (2) had fathers in professional or
managerial positions; (3) had parents with college-level
education; and (4) received positive encouragement in
academic and employment aspirations from their parents. The
relationship between having a mother who worked and the sex-
atypicality of her daughter's employment was less clear.
There was no attem pt made to investigate how oral
contraceptive use covaried with relationships between social
factors and sex-atypicality of employment choice.
22
Biological Model
For the purposes of this study the biological model was
concerned with the relationships between reproductive
hormones and sex-atypicality of employment aspirations.
"Sexually dimorphic" is a term most commonly used by
biological researchers to imply the existence of two different
forms of behavior exhibited by males and females (Goy &
McEwen, 1980). As this term is judgmental and does not allow
for behaviors to be examined on a continuum, I will continue to
use the terms sex-typical and sex-atypical.
A reproductive steroid hormone is a substance that is
synthesized in an endocrine gland (i.e., the female ovary or
adrenal gland) and is circulated through the body to a hormone
receptor cell. The steroid hormone passes through the
membrane and binds to a specific receptor protein in the
cytoplasm (Rosenzweig & Leiman, 1982). The hormone binding
then begins a receptor transformation process. The receptors
activate genes which regulate biochemical neurotransmitter
production.
There are believed to be two types of steroid hormonal
influences, organizational and activational, on sex specific
tissues and on behaviors (Feder, 1984). Prenatal hormones
organize central nervous system functions that are later
activated by circulating hormone levels in adulthood (MacLusky
23
& Naftolin, 1981). For instance, prenatal hormones organize
the brain and the internal genitalia of the female fetus in such
a way that these tissues, in adulthood, respond to circulating
hormones to initiate and maintain the female menstrual cycle.
Examples of organizational effects are the differentiation of
the internal and external genitalia, and sexual differentiation
of the brain. Organizational effects of hormones are thought
to be long-term, relatively permanent, and limited to a
specific time period in development (Feder, 1981).
Examples of activational effects of hormones on females
are the morphological and physiological changes that
accompany puberty: mammary gland enlargement; growth of
body hair; and onset of menses. Activational effects of steroid
hormones on females continue to be present long after puberty:
maintenance of menstrual cycles (Yen, 1991); adult sexual
behavior (Cutler, Garcia, Huggins, & Preti, 1986; Hedricks,
Piccinino, Udry, & Chimbira, 1987; Sherwin, Gelfand, &
Brender, 1985; Udry & Morris, 1968, 1970), and physical
activity levels (Morris & Udry, 1970). Activational effects of
hormones are considered reversible and not limited to a
specific time period in development (Feder, 1981).
Many animal studies have shown a direct influence of
perinatal hormones on later behaviors through experimental
manipulation of hormones, either by injection or castration.
24
There have been no experimental studies conducted on the
effects of early hormone exposure on adult human behavior.
However, researchers have shown relationships between
prenatal hormone exposure and later behaviors by studying
people with abnormal prenatal hormone exposure. Abnormal
prenatal hormone exposure can be due to either a genetic
hormonal abnormality or by maternal ingestion of a steroid
hormone during pregnancy (for a review of the literature, see
Hines, 1982). Relationships have also been found between
hormones circulating in the adult woman's bloodstream and
adult behaviors such as employment choice, childhood play,
aggressivity, parenting and sexual behaviors.
This project examined relationships between circulating
levels of adult female hormones and sex-atypicality of female
employment aspirations because: (1) prenatal hormone levels
were not available; and (2) adulthood is when actual
employment choices are being initiated. The hormones found
to have the most influence on sex-atypicality of behavior such
as those related to employment fall into three categories:
androgens, estrogens and progestogens.
Androgens
Androgenic hormones include testosterone (T),
androstenedione (A), dehydroepiandrosterone (DHEA), and
dehydroepiandrosterone sulfate (DHEAS). Androgens are
25
secreted by the adrenal cortices of both sexes. The female
ovaries also secrete T, A and DHEA. Total testosterone (TT) is
a global measure of the amount of the hormone in the blood, or
plasma. This includes: (1) the portion of T that is bound to sex
hormone binding globulin (SHGB); (2) the portion of T that is
bound to albumin; and (3) unbound or free testosterone (FT)
(Manni, Pardridge, Cefalu, Nisula, Bardin, Santner, and Santen,
1985). Only FT and a considerable fraction of T bound to
albumin are biologically active and can thus influence behavior
(Manni et al., 1985). The ratio of active to inactive hormone
varies according to the concentration of SHGB in plasma
(Vermeulen, 1983). Levels of SHBG vary between individuals
so that people with similar plasma levels of TT may actually
have different amounts of biologically available hormone.
Measurement of SHBG as well as TT allow one to calculate the
free testosterone index (FTI), which provides more complete
information on bioavailability of T (FTI = (TT/SHBG) (Clark,
Marcellus, DeLory, & Bird, 1975) than does measurement of
total testosterone only.
Behavioral correlates of prenatal exposure to androgen.
Congenital adrenal hyperplasia (CAH) is a cortisol deficiency
which results in high levels of adrenocorticotropic hormone,
which stimulates the adrenal cortices to secrete high amounts
of androgens and progestogens. Female infants with CAH are
26
usually born with masculinized genitalia (enlarged clitoris and
labial fusion), which is typically corrected by surgery shortly
after birth. Both males and females are usually treated with
cortisone from infancy.
Females with CAH have been reported to exhibit sex-
atypical behaviors. It should be noted, however, that the
following studies have been criticized due to small sample
sizes of less than 20 and the possibility that the interviewers'
biases were reflected in the behavioral observations (Bleier,
1991). For example, researchers relied on mothers'
descriptions of their daughters' play patterns, and mothers
were aware of the congenital condition and its potential
effects on later behavior. As young girls and adolescents, the
females with CAH were reported to display less wedding
fantasy, a decreased interest in infant care and doll playing,
and an increased interest in careers over marriage (Ehrhardt &
Baker, 1974). Play characteristics of these girls were also
reported to resemble those of boys, including aversion to doll
play, wearing of sex-typical clothing, preference for rough and
tumble play, self identification as a "tomboy", and preference
for boy playmates (Ehrhardt & Meyer-Bahlburg, 1981).
Behavioral correlates of circulating androgens in
adulthood. Research on biological sex differences has been
criticized for attempting to explain observed differences in
27
social roles or acheivements through untested assumptions,
inconclusive and contradictory findings, and misleading
interpretations (Bleier, 1991). Although these three studies
have not been directly critiqued, they have shown correlations
between adult androgen levels and employment choices of
women. Bancroft, Davidson, Warner, and Tyrer (1983) studied
55 women. They found that testosterone levels tended to be
higher in women who worked full-time outside the house than
in those women who did not work full-time outside the house.
Purifoy and Koopmans (1980) studied 55 women whose ages
ranged from 20 to 87. They found that the women with high
androgen levels had made sex-atypical employment choices.
Women with high T and A levels were more likely to be engaged
in professional, managerial, or technical positions, or to be
students. Women with low T and A levels were more likely to
be clerical workers or housewives. Schindler (1979) also
found that women who had chosen sex-atypical employment
had higher androgen levels than did women who had chosen
sex-typical employment. For example, women who were
lawyers had higher testosterone levels than women who were
nurses or teachers. It is of interest to note here that in China,
Dongyun and Yumin (1990) found that women who were pilots
(sex-atypical employment for women in China) had higher
levels of testosterone than women who were part of the
28
ground crew (sex-typical employment for women in China).
Estrogens
Estrogenic hormones include estradiol, estrone, and
estriol; however, only estradiol has been related to sex-
atypicality of behavior. Estrogens are secreted by the ovaries,
the placenta, and the adrenal cortices. A large percentage of
estradiol is synthesized in adipose tissue by conversion from
androgens that are secreted by the adrenal cortices
(Braunstein, 1986). In fact, it is believed that conversion of
androgen into estrogen is the primary mechanism responsible
for the masculinizing effects of prenatal androgen exposure.
The fetus is protected from maternal estrogens by an estrogen
binding factor; however, aromatized or converted estrogen
from testosterone within the brain, estrogen administered in
amounts sufficient to saturate the binding system, or potent
synthetic estrogens, such as diethylstybestrol (DES), can
bypass the binding system and have effects on the development
of sex-atypical behavior (Hines, 1982).
Behavioral correlates of prenatal exposure to estrogen.
Prenatal exposure to DES was studied by Erhardt, Meyer-
Bahlburg, Rosen, Feldman, Veridiano, Elkin, and McEwen (1989)
for relationships to sex-typical behaviors of children,
adolescents, and adults. Women who were prenatally exposed
to DES were closely matched with a control group of women
29
who were not exposed to DES prenatally, but who had an
abnormal pap smear. Their study revealed that DES-exposed
women showed less orientation toward parenting than did the
women in the control group. The investigators were unable to
show any consistent differences between the DES-exposed
females and the control females regarding other sex-role
behaviors such as sex-typical play patterns. The investigators
did not study sex-atypicality of employment choice of their
subjects.
Behavioral correlates of circulating estrogens in
adulthood. I have not been able to find any literature
investigating the relationship between adult estradiol levels
and sex-atypicality of employment aspirations of women.
However, estradiol and androgens have similar relationships
with sex-atypicality of behaviors (Feder, 1984; Goy &
McEwen, 1980; Hines, 1982; Sherwin, 1988), particularly play
behaviors (Goy & McEwen, 1980), suggesting that estradiol be
included in this study.
Progestogens
An example of a progestogenic hormone is progesterone.
Progestogens are secreted by the ovaries and adrenal cortices.
Progestogens are steroid hormones that assist in preparing the
uterus and placenta for pregnancy, and in the development of
mammary glands. It has been demonstrated in both sexes of
30
several animal species (rats, guinea pigs and monkeys) that
progestogens can counteract both the organizational and
activational effects of androgens and estrogens (Ehrhardt &
Meyer-Bahlburg, 1981; Ehrhardt, Meyer-Bahlburg, Feldman, &
Ince, 1984).
Behavioral correlates of prenatal exposure to
progestogens. Medroxyprogesterone acetate (MPA), which is
closely related to progesterone, is the active compound in a
drug that was previously administered to prevent premature
labor (Erhardt, Meyer-Bahlburg, Feldman, & Ince, 1984).
Erhardt e t al. (1984) studied 15 preteen girls who had been
prenatally exposed to MPA and 15 matched controls who had
not been prenatally exposed to MPA. Compared to the controls,
the MPA-exposed girls showed more sex-typical play
interests; specifically, more interest in doll play and sex-
typical clothing; less interest in sports and physically active
play.
Behavioral correlates of circulating progestogens in
adulthood. At this time I have been unable to locate any
research studies which correlated present levels of
progestogens to sex-atypicality of employment aspirations of
women. However, it is thought that prenatal activity of
progestogens may counteract the prenatal influences of
estrogens and androgens (Ehrhardt & Meyer-Bahlburg, 1981;
31
Ehrhardt et al., 1984; Feder, 1984). This study, therefore, also
included progesterone in its analysis, in the event that
circulating progesterone in adulthood was related to adult
occupational behavior.
Oral contraceptive use
Previous studies that have examined female employment
choice have not distinguished between women with respect to
oral contraceptive use. However, Alexander, Sherwin,
Bancroft, and Davidson (1990) and Bancroft, Sherwin,
Alexander, Davidson, and Walker (1991a, 1991b) reported
significant differences between the two groups in
testosterone levels, sexual behavior and correlations between
hormones and sexual behavior. That is, OC users were found to
have lower levels of testosterone and engage in a higher
frequency of sexual behavior than non OC users. Additionally,
there were correlations between testosterone and some sexual
behaviors with the OC users but not the non OC users. Hedricks
and Udry (1990) reported significant differences between the
two groups in in the relationships beween testosterone levels
and maternal aspirations. There was a significant correlation
between testosterone levels and maternal aspirations for the
OC users but not for the non OC users.
32
Summary
Based on previous studies of sex-atypicality of female
employment and adult circulating hormone levels, it was
expected that the more sex-atypical the woman's employment
aspiration, the higher her androgen levels would be. Also,
based on the studies that investigated behavioral correlates of
prenatal hormone exposure, it was expected that estrogen
levels would be positively correlated with sex-atypicality in
employment aspirations and progestogen levels would be
negatively correlated with sex-atypicality in employment
aspirations.
Based on previous studies of hormonal and behavioral
differences between non oral contraceptive and oral
contraceptive users, it was expected that these two groups of
women would have hormonal and behavioral differences.
Furthermore, it was expected that the nature of relationships
between hormones and behavior within OC users might be
different than that of the non OC users.
Previous studies that have examined sex-atypicality of
employment choice have found this behavior to have
significant social and hormonal correlates. However, as
previously discussed and as evidenced by this literature
33
review, social and biological scientists have not integrated
their findings. It was the intent of this project to examine
both social and hormonal correlates of female employment
aspirations.
Chapter III
Methodology
The study was concerned with social and hormonal
predictors of sex-atypicality of employment aspirations of
female college seniors. I attem pted to create a biosocial
model of sex-atypicality of employment aspirations from data
that have been previously collected. This section contains: (1)
subject recruitment and eligibility requirements; (2)
description of the instruments that were used; (3)
experimental protocol that was followed during the data
collection process in 1987; and (4) a data analysis section.
Subject Recruitment
Senior college women were studied because it was
expected that: (1) there would be more variability in their
employment aspirations then there would be in females with a
high school degree or less education and; (2) the hormonal
system of these women had not been influenced by the
experience of engaging in employment. The study (n=57) was
conducted at the Carolina Population Center at the University
of North Carolina at Chapel H ill (UNC-CH). The subjects were
selected from the 1986-1987 Student Listings of the UNC-CH
35
directory. The name of every 20th senior female was selected,
up to the 100th senior female. These women were mailed a
short letter explaining a study that was investigating the
relationship between women's sex role behaviors and their
physiology. It further explained that if they chose to
participate, their involvement would entail filling out several
questionnaires and giving two small blood samples. They
would receive a check for fifty dollars two weeks after the
second blood draw. Subjects were told that they would
receive a telephone call a few days later to request
participation or to explain more about the study. One week
after the mailing of this letter the females were contacted by
phone to determine their eligibility and willingness to
participate in the study.
Inclusion criteria consisted of the following
characteristics: (1) being a female college senior; (2) being
between 21-25 years of age; (3) not having any children; and
(4) having had a menstrual period within the last 40 days.
Fifty-seven (75%) of all the eligible females contacted agreed
to participate in the study.
Data were collected from 57 subjects. Of these
subjects, 84% (n=48) were Caucasian, 10% (n=6) were African
American and the remaining 6% were Hispanic, American Indian
and Spanish American (where n=1 for each ethnic group). Ages
36
ranged from 21-25 years; 91% of the subjects were between
21-22 years old. Forty-two percent (n=24) of the subjects
were taking oral contraceptives.
Instrumentation
The subjects, with the Principle Investigator (C. H.),
completed part A of the Female Senior Questionnaire (FSQ)
(Hedricks, 1987; a copy of the entire FSQ is available upon
request). Part A contained questions related to the household
environment when they were growing up, and marital status,
educational level, and occupation of parents. The subjects
then completed the following self-administered
questionnaires:
1. Female Senior Questionnaire (FSQ). part B. The FSQ
part B, contained a series of short questions about career or
educational plans upon graduation from college, academic
record (GPA and college major), parent's encouragement while
they were growing up, sex-role orientation, childhood play
behaviors, medical history, reproductive history and behavior,
handedness, and maternal aspirations and attitudes (on child
bearing and child care). With the exception of the Index of
Sex-Role Orientation (Dreyer, Woods, & James, 1981) and the
Oldfield Handedness Inventory (Oldfield, 1971) the items
within this questionnaire were not standardized nor tested for
37
reliability or validity. Because of its relevance to sex-
atypicality of behavior, the Index of Sex-Role Orientation will
be included in this project.
a. Index of Sex-Role Orientation (ISRO). The ISRO is a
scale which measures women's sex-role orientation. The scale
lists 16 attitude statem ents to be rated on a 5 point scale
(ranging from "strongly agree" to "strongly disagree"). The
higher the score, the more sex-atypical a person's sex-role
orientation. The scale was found to have a reliability
coefficient of .92 and a positive test-retest correlation of .62,
indicating internal consistency and temporal stability over 30
days. The ISRO correctly categorized 92.6% out of 26 women
as "feminists" (or as having a sex-atypical sex role
orientation) and 95.5% out of 22 women as "traditional" (or has
having a sex-typical sex role orientation), demonstrating high
sensitivity and high specificity construct validity (Dreyer,
Woods, & James, 1981). Sociodemographic variables have been
examined for impact on sex-role attitudes. Education was the
single most important correlate of sex-role orientation; that
is, a higher educational level was associated with sex-
atypical sex-role orientation in women (Dreyer, Woods, &
James, 1981).
3. Other inventories administered, but not analyzed in
the present study. In addition, Adjective Checklist (ACL, Gough
38
& Heilbrun, 1983), the Bern Sex Role Inventory (BSRI, Bern,
1974), the 16 Personality Factor Questionnaire (16PF, Cattell,
Eber, & Tatsuoka, 1980), and the Strong-Campbell Interest
Inventory (SCII, Hansen & Campbell, 1985) were administered.
These inventories were not addressed in this study.
Experimental Protocol
The subjects were individually interviewed at the
Carolina Population Center (CPC) by the Principle Investigator
(Dr. C. Hedricks). The interview lasted approximately one and
one-half hours. First the FSQ and the standardized
psychological inventories were explained to the subjects. Any
questions that they had regarding the forms were answered.
The Informed Consent Form (ICF) was then signed by the
subjects. A copy of this form was given to the subjects. The
subjects were then given unique, confidential identification
numbers. Only these numbers were used in identifying a
particular subject's data. The interviewer-administered
portion of the FSQ (part A) was then completed by both the
Principle Investigator and the subject.
Instructions on filling out the standardized psychological
inventories were reviewed. These instructions included being
able to ask the Principle Investigator any questions as they
arose during the process of completing the forms. The
39
subjects privately completed the FSQ (part B) and the three
standardized psychological assessments (ACL, BSRI, and SCII).
When they had completed the FSQ and the three
standardized psychological inventories, the subjects returned
to the Principle Investigator's office and dropped off the
assessments in a sealed envelope. They were given the 16
Personality Factor Questionnaire to complete at home along
with a self-addressed (to the Principle Investigator), stamped
envelope and instructions to mail back the completed inventory
within the next twenty-four hours. They were also given a
letter that verified their participation in the study that they
were to show to the lab technicians when they went to give
their blood specimens.
All the subjects were instructed to call the Principle
Investigator on the first day of their next menstrual bleeding
so that an appointment could be made for a blood sample to be
taken at the Student Health Services (SHS). All of the
appointments were made Monday through Friday, between
1500-1630 hours to control for circadian rhythmicity in
hormone levels. Many reproductive hormone levels vary across
the 24-hour day (Yen, 1986), requiring that all blood specimens
be taken within the same time period for the purpose of
comparing levels across subjects.
Additionally, all reproductive hormone levels vary across
40
the female's menstrual cycle (Yen, 1986). During menstrual
bleeding, hormone levels may be too low to reach the threshold
required for accurate measurement. Most hormone levels begin
to rise towards the end of menstrual bleeding (approximately
day 5 of the cycle, where the first day of menstrual bleeding is
day 1), peak close to the time of ovulation (which usually
occurs some time after day 9; the exact day is different
between individuals) and then decline (Yen, 1986). Several
studies have found that estrogen levels peak a day or two
before ovulation whereas progesterone levels peak shortly
after ovulation (Dupon, Hosseinian, & Kim, 1973; Frolich,
Brand, & Hall, 1976; Johansson, Wide, & Gemzell, 1971). Most
studies that have examined cyclic changes of testosterone and
androstenedione across the menstrual cycle have found trends
where T and A tend to be slightly elevated around the time of
ovulation (Dupon, Hosseinian, & Kim, 1973; Frolich, Brand, &
Hall, 1976; Judd & Yen, 1973; Tyler, Newton, & Collins, 1975).
Blood draws were taken between days 5 through 9 of the
women's menstrual cycles to assure that: (1) hormone levels
would be high enough to reach the threshold for measurement;
and (2) any differences found between hormonal levels
reflected individual differences and not differences related to
the phase of the menstrual cycle.
Appointments for blood draws at the Student Health
41
Service (SHS) were made by the Principle Investigator. The
last name and identification number of the subject was given
to the laboratory receptionist as each appointment was made.
The receptionist recorded this information onto a logsheet and
then recorded if the appointment was kept. The logsheet was
mailed back to the Principle Investigator on a weekly basis in
a self addressed, stamped envelope that was provided.
At the time of their scheduled appointment, the subjects
went directly to the SHS. They showed the lab technicians the
letter verifying their participation in the study. A total of
18m l of blood was drawn and evenly distributed into two glass
serum separator tubes. Each tube held approximately 10 ml
and contained a clotting gel that did not affect the steroid
hormone assays (Biomed, Burlington, North Carolina). The
specimen was centrifuged after the blood was given 25
minutes to clot. The blood specimens were stored at -18°C in
the freezer at the SHS. They were later transported to an
endocrine laboratory for radioimmunoassay (RIA) of the
hormones previously mentioned.
The subjects were to give a second blood specimen
during their next menstrual cycle by following the same
protocol that was used for their first blood draw. Most of the
subjects (n=46 or 81%) gave two blood specimens. These
women received a check for fifty dollars approximately two
42
weeks after the second blood donation. The remaining subjects
(n=11) were unable to schedule a second blood draw and
received a check for thirty-five dollars. The second blood
draw was used to te st reliability of the hormone levels
obtained from the first blood draw.
After the collection of the last blood sample, the
subjects mailed the Principle Investigator the date of onset of
their next menstrual bleeding on a stamped, self-addressed
postcard that was provided. The next menstrual onset date
was needed to clarify the length of the menstrual cycle during
which the second blood specimen was taken. Once this
postcard was received by the Principle Investigator, a copy of
the subject's Strong-Campbell Interest Inventory was mailed
to Student Development and Counseling if the subject
requested an interpretation of this psychological inventory.
Radioimmunoassay of Blood Samples
Subjects contributed at least one 18m l blood sample
that was subjected to radioimmunoassay (RIA) procedures.
RIA is a procedure that determines the concentration of
hormones in biological specimens (Thomas, 1985). The
following hormones were analyzed: total testosterone (TT, in
ng/m f)’ , sex hormone binding globulin (SHBG, in nmol/ \ );
androstenedione (A, ng/ml)', dehydroepiandrosterone (DHEA,
43
ng/m l); dehydroepiandrosterone sulfate (DHEA-SO4, in umol/l);
progesterone (P, in ng/ml); and estradial (E2, in umol/l)- A
free testosterone index (FTI) was calculated for each specimen
using the following formula: FTI = TT/SHBG (Clark et al.,
1975). Only FT and a considerable fraction of T bound to
albumin are biologically active and can thus influence behavior
(Manni et al., 1985).
Data Analysis
Predictor and criterion variables
Statistical analysis was performed on the data using
Statistical Analysis System (SAS, Cary, NC). The hormonal and
social variables of interest in this study are referred to as the
independent, or predictor variables. The sex-atypical
behaviors in this study are referred to as the dependent, or
criteria variables. (See Appendix C for a list of the variable
labels.) Statistical language uses the term, "predictor
variable," for analytical purposes. Prediction is the process of
estimating the values of one variable from knowledge of
values on another variable (Hinkle, Wiersma & Jurs, 1979). The
use of the term, "predictor," is not meant to imply cause and
effect.
44
Cycle to cycle differences
Group comparisons with t-test showed no significant
differences in hormone values between the two consecutive
menstrual cycles. Therefore, the mean for each hormone
across the two cycles was then calculated for those subjects
who had two blood draws (n=46), and this value was used for
statistical analyses. The sole hormone value obtained from
those subjects who had one blood draw (n=11) was used for
statistical analyses.
Non oral contraceptive versus oral contraceptive users
Independent t-tests and chi square analyses were
performed to compare non oral contraceptive versus oral
contraceptive users on all predictor and criterion variables.
Significant differences were found between the two groups
(see Results); thus, separate analyses were run on each
individual group.
Transform ations
A SAS univariate procedure was run to test for normality
of the predictor variables. This procedure showed the hormone
values to be skewed. To correct for skewness, the square root
was taken of each hormone value with the exception of
androstenedione (A), of which the digamma value was taken.
45
These transformed values were then used for data analysis
purposes. However, non-transformed values are used when
reporting descriptive statistics.
Relationships between variables: Single correlations
Single correlations use the score of the independent (or
predictor) variable to predict or estimate the score of the
dependent (or criterion) variable. Pearson Product Moment
Correlation Coefficients (r) were computed between each
social predictor and the behavioral criteria variables, and
between each hormonal predictor and the behavioral criteria
variables. There were a total of 13 social predictors
investigated. Only those social predictors that correlated
with the behavioral criteria variables (at p < .10) were
considered statistically significant and available for entry
into multiple regression models that investigated social and
biosocial models of sex-atypical behavior.
There were a total of 8 hormonal predictors investigated.
Only those hormonal predictors that correlated with the
behavioral criteria variables (at p < .10) were considered
statistically significant and available for entry into multiple
regression models that investigated biological (hormonal) and
biosocial models of sex-atypical behavior.
46
Biological, social, and biosocial models of sex-atypical
behavior: Multiple regression analysis
Multiple regression analysis is a method that analyzes
the variability in a criteria or dependent variable based on
numerical information from two or more predictor or
independent variables (Pedhazer, 1982). Multiple regression
analyses were computed between the set of statistically
significant social predictors and each of the six behavioral
criterion variables. Multiple regression analyses were also
computed between the set of statistically significant
hormonal predictors and each of the six behavioral criterion
variables. These procedures indicated which social and
hormonal predictor variables had significant relationships to
each of the behavioral criterion variables.
Biosocial multiple regression models were also
constructed. Both the social and the hormonal predictor
variables that were significant at the p < . 10 level in the
separate multiple regressions were entered together into a
regression model to create biosocial models. These biosocial
models determined whether hormonal or social variables were
the most significant predictors of the behavioral criterion
variables.
47
Chapter 4
RESULTS
Comparisons between non oral contraceptive users (Non OC)
versus oral contraceptive users (OC)
There were significant differences in hormone levels
between the non oral contraceptive users versus the oral
contraceptive users. Women who were taking oral
contraceptives had significantly lower levels of FTI, T, DHEA,
DHEAS, and E2, significantly higher levels of SHBG, and tended
to have lower levels of A and P (see Table 1).
There were also significant differences on two of the
behavioral criteria variables (RMALE80, UNCFIELD) between
non oral contraceptive users versus oral contraceptive users
(see Table 2). (See Appendix C for definitions of the variable
labels.) There were no significant differences between the
two groups with respect to the social predictor variables (see
Tables 3-4). Because of these many hormonal and behavioral
differences between the two groups, the sample was divided
into non oral contraceptive users (n=33) and oral contraceptive
users (n=24) for the purposes of analysis.
Non oral contraceptive users (Non OC)
Correlation analyses. Of the 13 social predictor
48
Table 1
Mean (+ SD) hormonal differences between non oral
contraceptive (Non OC) users (n=33) and oral contraceptive
(OC) users (n=24)
Non OC Users OC Users
var a> b mean SD mean SD t
P
TT 0.419 0.184 0.308 0.147 2.61 0.01
FTI 0.016 0.017 0.005 0.005 4.54 0.001
SHBG 38.96 20.23 77.87 31.09 5.74 0.001
DHEA 9.799 5.856 6.515 2.777 2.40 0.01
DHEAS 8.353 3.812 6.223 3.344 1.73 0.02
A 2.993 2.732 2.786 3.344 1.73 0.09
E2 0.080 0.042 0.063 0.100 2.66 0.01
P 0.324 0.127 0.303 0.236 1.105 0.27
aOperational definitions can be found in Appendix B
bDefinitions of the variable labels can be found in Appendix C
49
Table 2
Mean (+ SD) differences between non oral contraceptive (Non
OC) users (n=33) and oral contraceptive (OC) users (n=24) on
the behavioral criterion variables
Non OC Users OC Users
Vara> b n m ean SD n m ean SD t
P
RMALE
80
29 47.96 28.35 23 63.68 18.96 2.28 0.02
RPRES
T
29 57.37 12.90 22 59.13 10.96 0.51 0.85
UNCFIELD 33 0.33 0.15 24 0.40 0.13 1.90 0.06
ISRO
33 62.03 6.34 24 62.58 6.47 0.32 0.74
Operational definitions can be found in Appendix B
bDefinitions of the variable labels can be found in Appendix C
50
Table 3
Mean (+ SD) differences between non oral contraceptive (Non
the continuous social Dredictor variables
Non OC OC Users
Vara> b n mean SD n mean SD t
P
FATHERS
FEDUC
32 4.72 2.02 22 4.68 2.03 0.06 .94
FPRES
T
28 53.41 14.23 21 52.21 12.09 0.31 .75
FMALE
80
30 77.27 16.36 24 76.11 17.05 0.26 .79
FENC
33 1.76 0.61 24 1.88 0.45 0.7 .43
MOTHERS
MEDUC
33 3.48 1.77 24 4.04 1.83 1.15 .25
MPREST 30 29.48 28.14 20 35.77 25.39 0.80 .42
MMALE
80
33 27.20 30.31 24 23.66 25.32 0.46 .64
MENC
33 1.82 0.52 24 1.87 0.34 0.49 .62
operational definitions can be found in Appendix B
definitions of the variable labels can be found in Appendix C
51
Table 4
Differences between non oral contraceptive (Non OC) users
(n=33) and oral contraceptive (OC) users (n=24) on the
dichotomous social predictor variables
Non OC Users OC Users
Variable3-
b
n n
*2 P
ONLY 2 0 1.50 0.22
FIRST 12 10 0.16 0.68
MEMP 20 18 1.29 0.25
MGRAD 3 3 0.17 0.67
FGRAD 10 7 0.01 0.92
Operational definitions can be found in Appendix B
^Definitions of the variable labels can be found in Appendix C
52
variables, six were significantly correlated with the
behavioral criterion variables. Of the 8 hormonal predictor
variables, only one was significantly correlated with the
behavioral criterion variables (see Appendix D for the results
of all correlation analyses).
Employment aspirations. Sex-atypicality of the
woman's employment aspiration (RMALE80) was
positively correlated with DHEA (DHEA; p < 0.03) and
with her father's occupational prestige (FPREST; p<
0.04). The higher the woman's DHEA level and the higher
her father's occupational prestige, the more sex-atypical
her employment aspiration.
The occupational prestige of the woman's
employment aspiration (RPREST) was positively
correlated with whether her mother worked outside of
the home (MEMP; p < 0.01), with her mother's
occupational prestige (MPREST; p < 0.05), and with her
DHEA level (DHEA; p < 0.10). Contrary to expectations,
this behavior criterion variable tended to be negatively
correlated with whether her father had attended
graduate school (FGRAD; p < 0.06), and with her mother's
encouragement (MENC; p < 0.09).
In summary, whether the mother worked outside of
the home, the more prestigous the mother's occupation,
53
and a high DHEA level were related to greater sex-
atypicality in occupational prestige of the daughter's
employment aspiration. Greater mother's encouragement
and whether her father attended graduate school were
related to less sex-atypicality in occupational prestige
of the woman's future employment.
College maior. Sex-typicality of the woman's
college major (UNCFIELD) was positively correlated with
her father's occupational prestige (FPREST; p < 0.002)
and tended to be positively correlated with her father's
educational level (FEDUC; p < 0.09). Thus, the higher her
father's occupational prestige and the higher his
educational level, the more sex-atypical was his
daughter's college major.
ISRO. None of the social or hormonal predictor
variables correlated with this behavioral criterion
variable.
Multiple regression analyses: Social models. Correlation
analyses indicated that there were three behavioral criterion
variables (RMALE80, RPREST, UNCFIELD) that had significant
correlations with one or more social predictor variables.
Thus, only three multiple regression analyses were run using
social variables to predict the behavioral criterion variables
(see Table 5).
54
Table 5
Summary of significant correlations between social and
hormonal predictor variables3*hand behavioral criterion
variables: Non oral contraceptive users (n=33)
Behavioral Criterion Variables
RMALE80 RPREST UNCFIELD
Predictor DHEA M EM P FPREST
Variables r=0.40 r=0.45 r=0.55
p<0.03 p<0.01 p<0.002
FPREST MPREST
r=0.39 r=0.37
p<0.04 p<0.05
FGRAD
r=-0.34
p<0.06
M ENC
r=-0.31
p<0.09
DHEA
r=0.30
p<0.10
aOperational definitions can be found in Appendix B
bDefinitions of the variable labels can be found in Appendix C
55
Employment aspirations. The social model for sex-
atypicality of the employment aspiration (RMALE80) is
presented in Table 6. This model was significant (p<
0.04), and contained one significant social predictor
variable (FPREST). This model explained 12 percent of
the variance in the behavioral criterion variable.
The social model for occupational prestige of the
employment aspiration (RPREST) is presented in Table 7.
This model was significant (p<0.01). It contained one
significant social predictor variable (FGRAD), and three
nonsignificant social predictor variables (MPREST, MEMP,
MENC). It explained 33 percent of the variance in the
behavioral criterion variable. However, contrary to
expectations, the significant social predictor variable
was negative. That is, if the father attended graduate
school the daughter's occupational prestige was more
likely to be sex-typical.
College major. The social model for sex-typicality
of college major (UNCFIELD) is presented in Table 8.
This model was significant (p <0.002), and contained one
significant social predictor variable (FPREST) and one
nonsignificant social predictor variable (FEDUC). The
model explained 29 percent of the variance in this
behavioral criterion variable.
56
Table 6
Social modeia for sex-atvpicalitv of employment aspirations
(RMALE80) for non oral contraceptive (Non OC) users
Variablesb> c Param eter Standard T
P
Estimate Error
FPREST 02.22 00.56 3.90 0.04
aAdj. R2=.12, F=4.38, p<.04, n=25
bOperational definitions can be found in Appendix B
definitions of the variable labels can be found in Appendix C
57
Table 7
Social modeia for sex-atvpicalitv of occupational prestige of
employment aspirations (RPREST1 of non oral contraceptive
(Non OC) users
Variables1 3 -0 Param eter Standard T p
Estimate Error
FGRAD -10.58 04.92 -2.14 0.04
M EM P 20.53 12.82 1.60 0.12
M ENC -02.99 04.21 -0.71 0.48
MPREST -00.13 00.22 -0.61 0.54
aAdj. R2=.33, F=4.19, p<0.01, n=27
Operational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
58
Table 8
Social models for sex-atvpicalitv of college maior (UNCFIELD)
for non oral contraceptive (Non OC') users
Variables1 3 -0 Param eter Standard T
P
Estimate Error
FPREST 0.008 0.002 3.25 0.003
FEDUC 0.023 0.188 1.22 0.23
aAdj. R2=.28, F=6.68, p <0.004, n=28
bOperational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
59
Multiple regression analyses: Biological models. Based
on the correlational analyses, there were two behavioral
criterion variables (RMALE80, RPREST) that had a significant
correlation with one hormonal predictor variable (DHEA).
Employment aspirations. The biological model for
sex-atypicality of the woman's employment aspiration
(RMALE80) is presented in Table 9. This model was
significant (p < 0.03), and contained one hormonal
predictor variable (DHEA). This model explained 13
percent of the variance in the behavioral criterion
variable.
The biological model for occupational prestige of
the woman's employment aspiration (RPREST) is
presented in Table 10. This model showed a trend
towards significance (p < 0 .10), and contained one
hormonal predictor variable (DHEA). This model
explained 6 percent of the variance in the behavioral
criterion variable.
Multiple regression analyses: Biosocial models. There
were two behavior criterion variables (RMALE80; RPREST) that
were significantly predicted by both social and hormonal
predictor variables. Therefore, two multiple regression
analyses were run using both social and hormonal variables to
predict these behavioral criterion variables.
60
Table 9
Biological models for sex-atvpicalitv of employment
aspirations (RMALE80) for non oral contraceptive (Non OC)
users
Variablesb« c Param eter Standard T p
Estimate Error
DHEA 19.46 8.56 2.27 0.03
aAdj. R2=.13, F=5.16, p < 0.03, n=29
bOperational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
61
Table 10
Biological models for sex-atvpicalitv of occupational prestige
of employment aspirations (RPREST1 for non oral contraceptive
(Non QC1 users
Variablesb* c Param eter Standard T p
Estimate Error
DHEA 6.78 4.04 1.67 0.10
aAdj. R2=.06, F=2.81, p<0.10, n=29
bOperational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
62
Employment aspirations. The biosocial model for
sex-atypicality of the woman's employment aspiration
(RMALE80) is presented in Table 11. This model was
significant (p <0 .0 1 ), and contained one significant social
and one significant hormonal predictor variable (FPREST
and DHEA, respectively).
This biosocial model explained 27 percent of the
variance for this behavioral criterion variable.
The biosocial model for occupational prestige of
the woman's employment aspiration is presented in Table
12. This model was significant (p <0.01). It contained
one significant social variable (FGRAD). This model
contained three nonsignificant social variables (MEMP,
MPREST, MENC) and one nonsignificant hormonal variable
(DHEA). This model explained 35 percent of the variance
for this variable.
Summary. With respect to the non oral contraceptive
users, three regression models that contained social predictor
variables, two regression models that contained a biological
predictor variable, and two regression models that contained
both a social and a biological predictor variable emerged from
the data analyses.
With respect to social predictors studied, sex-
atypicality of the women's employment aspirations (RMALE80)
63
Table 11
Biosocial modeia for sex-atvpicalitv of employment
aspirations (RMALE80) for non oral contraceptive (Non PC)
users
Variables* 3 -0 Param eter Standard T
P
Estimate Error
DHEA 19.67 8.17 2.40 0.02
FPREST 00.71 0.33 2.10 0.04
aAdj. R2=.27, F= 5.54, p <0.01, n=25
Operational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
Table 12
Biosocial models for sex-atvpicalitv of occupational prestige
of employment aspirations (RPREST1 for non oral contraceptive
(Non PCI users
Variables1 3 -0 Param eter
Estimate
Standard
Error
T
P
FGRAD -9.47 04.92 -1.92 0.06
M EM P 20.88 12.61 1.65 0.11
DHEA 4.67 03.56 1.31 0.20
MPREST -0.15 00.21 -0.21 0.48
M ENC -2.29 04.18 -0.54 0.58
aAdj. R2=.35, F= 3.81, p <0.01, n=27
^Operational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
65
and the occupational prestige of the employment aspiration
(RPREST) were significantly predicted by the occupational
prestige of their fathers (FPREST). The occupational prestige
of their employment aspirations (RPREST) was inversely
predicted by whether their fathers had attended graduate
school (FGRAD).
With respect to biological predictors studied, sex-
atypicality of the employment aspiration (RMALE80) was
significantly predicted by the hormone DHEA. Occupational
prestige of their employment aspirations (RMALE80) also
tended to be predicted by DHEA.
A regression model that included both social and
biological predictors (i.e., a biosocial model) revealed that
sex-atypicality of
employment aspiration was significantly predicted by both
DHEA and FPREST. Individually, DHEA and FPREST explained 13
and 12 percent of the variance, respectively. However,
together they explain 27 percent of the variance for this
behavioral criterion variable ( p < 0.01).
The biosocial model used to explain hormonal and social
predictors occupational prestige of the employment aspiration
(RMALE80) was also significant (p < 0.01). Within this model,
whether the father had attended graduate school (FGRAD) and
whether the mother had worked outside the home (MEMP)
66
tended to be significant predictors, (p< 0.06 and p < 0.11,
respectively). The hormone DHEA, the mother's occupational
prestige (MPREST), and the mother's encouragement (MENC)
were not significant predictors; however, the relationship
between DHEA and this behavior criterion variable was in the
expected direction.
Individually, the social model explained 33 percent, and
the biological model explained 6 percent, of the variance in
occupational prestige of the employment aspiration. When
both types of predictors were included in a biosocial model,
35 percent of the variance in this behavioral criterion variable
was accounted for.
See Table 13 for a comparison of social, biological and
biosocial models to predict sex-atypicality in both
employment aspiration (RMALE80) and occupational prestige
(RPREST).
Oral contraceptive (OC) users
Correlation analyses. Of the 13 social predictor
variables, only five were significantly correlated with the
behavior criterion variables (p <.10). Contrary to
expectations, none of the 8 hormonal variables correlated with
any of the behavioral criterion variables (see Appendix E for
the results of the correlation analyses).
67
Table 13
Comparison of social, biological, and biosocial models to
predict sex-atvpicalitv in both employment aspiration
(RMALE801 and occupational prestige (RPREST1: Non oral
contraceptive users
Social
Model
n=28
Biological
Model
n=24
Biosocial
Model
n=24
RMALE80
F Value 4.38 5.16 5.54
P Value <0.04 <0.03 <0.01
% of variance
explained
12% 13% 27%
RPREST
F Value 4.20 2.81 3.81
P Value <0.01 <0.10 <0.01
% of variance
explained
33% 6% 35%
68
Employment aspirations. Contrary to expectations,
none of the social or hormonal variables correlated with
sex-atypicality of the woman's employment aspiration
(RMALE80).
The occupational prestige of the woman's
employment aspiration (RPREST) was positively
correlated with her mother's educational level (MEDUC)
and with whether her mother had attended graduate
school (MGRAD; p < 0.03 for both). Thus the higher the
mother's educational level, the more sex-atypical was
the occupational prestige of her daughter's future
employment choice.
College maior. The only significant correlation
with sex-atypicality of the woman's college major
(UNCFIELD) was opposite of what was expected. FMALE80
was negatively correlated with the sex-atypicality of
the woman's college major (p < 0.03); that is, the more
sex typical the father's employment, the less sex-
atypical was his daughter's college major.
ISRO. Sex-atypicality of the woman's score on the
Index of Sex Role Orientation (ISRO) was positively
correlated with the educational level of her father
(FEDUC; p < 0.05), with whether her father had attended
graduate school (FGRAD; p < 0.07), and with whether her
69
mother had attended graduate school (MGRAD; p < 0.03).
Thus, the more sex-atypical a woman's ISRO score, the
more likely it was that her father was highly educated,
and that both of her parents had attended graduate
school.
Multiple regression analyses: Social models. Based on
the results of the correlation analyses, there were three
behavioral criterion variables (RPREST, UNCFIELD, ISRO) that
were significantly correlated with one or more social
predictor variables (see Table 14). Thus, three multiple
regression analyses were run using social variables to predict
these three behavioral criterion variables.
Employment aspirations. The social model for sex-
atypicality of occupational prestige of the employment
aspiration (RPREST) is presented in Table 15. This model
showed a trend towards significance (p < 0.06), even though
both social predictor
variables (MEDUC, MGRAD) lost significance when they
were entered together into the model. This model
explained 17 percent of the variance for the behavioral
criterion variable.
College maior. The social model for sex-
atypicality of college major (UNCFIELD) is presented in
Table 16. This model was significant (p< 0.03), and
70
Table 14
Summary of significant correlations between social and
horm onal predictor variablesb^and behavioral criterion
variables: Oral contraceptive users (n=24)
Behavioral Criterion Variables
RPREST UNCFIELD ISRO
Predictor M EDUC FMALE80 MGRAD
Variables r=0.46 r=-0.42 r=0.44
p<0.03 p<0.03 p<0.03
MGRAD FEDUC
r=0.44 r=0.42
p<0.03 p<0.05
FGRAD
r=0.37
p<0.0 7
aNone of the hormonal predictor variables significantly
correlated with the behavioral criterion variables
bOperational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
71
Table 15
Social models for sex-atvpicalitv of occupational prestige of
employment asoirations (RPREST) of oral contraceptive (OC)
users
Variablesb< c Param eter
Estimate
Standard
Error
T
P
M EDUC 1.80 1.53 1.17 0.25
MGRAD 7.7 6 8.06 0.96 0.34
aAdj. R2=.17, F=3.18, p<.06, n=22
bOperational definitions can be found in Appendix B
definitions of the variable labels can be found in Appendix C
72
Table 16
Social modeia for sex-atvpicalitv of college maior (UNCFIELD)
of oral contraceptive (OC ) users
Variables1 3 .0 Param eter Standard T
P
Estimate Error
FMALE80 -0.003 0.001 -2.23 0.03
aAdj. R2=.15, F=4.98, p < 0.03, n=24
bOperational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
73
contained one significant social predictor variable
(FMALE80). It explained 15 percent of the variance for
the behavioral criterion variable. However, contrary to
expectations, the significant social predictor was
negative. That is, the more sex-atypical the father's
employment, the less sex-atypical was his daughter's
college major.
ISRO. The social model for sex-typicality of the
ISRO score is presented in Table 17. This model showed
a trend towards significance (p <0.09). However, all
three social predictor variables (FGRAD, MGRAD, FEDUC)
lost significance when they were entered together into
the social regression model. This model explained 17
percent of the variance for the behavioral criterion
variable.
74
Table 17
Social models for sex-atvpicalitv of the ISRO scores of oral
contraceptive ( 0 0 users
Variablesb> c Param eter
Estimate
Standard
Error
T
P
MGRAD 6.50 4.54 1.43 0.16
FEDUC 0.56 1.05 0.53 0.60
FGRAD 1.45 4.91 0.29 0.77
aAdj. R2=.17, F=2.48, p < .09, n=22
bOperational definitions can be found in Appendix B
cDefinitions of the variable labels can be found in Appendix C
75
Multiple regression analyses: Biological models. There
were no significant correlations between the hormonal
predictor variables and the behavioral criterion variables.
Therefore, no multiple regression analyses were run using the
hormonal variables to predict the
behavior criterion variables.
Multiple regression analyses: Biosocial models. Because
there were no biological models for the behavioral criterion
variables, there are no biosocial models for the behavioral
criterion variables.
Summary. Three regression models that contained only
social predictor variables emerged from the data analyses on
oral contraceptive users. The social model for occupational
prestige of the employment aspiration (RPREST) showed a
trend towards being predicted by whether the mother had
attended graduate school (MGRAD). Contrary to expectations,
the social model for sex-atypicality of college major
(UNCFIELD) was inversely predicted by the father's employment
choice (FMALE80). The social model for the ISRO score showed
a trend towards being predicted by whether the fathers and
mothers had attended graduate school (FGRAD, MGRAD) and by
the father's educational level (FEDUC).
76
Chapter 5
DISCUSSION
Introduction
The purpose of this study was to simultaneously evaluate
the relationships of social and hormonal correlates on sex-
atypicality of employment aspirations, college major and
score on the Index of Sex Role Orientation (ISRO, Dreyer et el.,
1981) of senior college women. The following variables were
expected to positively correlate with sex-atypicality of
employment aspirations, college major and ISRO scores: (1)
being the oldest or only child; (2) having a mother who worked;
(3) having a father in a high prestige job; (4) having one or both
parents with high levels of education; (5) having received
encouragement from parents to do well in school and to have a
career; (6) having a high free testosterone index (FTI); (7)
having high blood levels of androstenedione (A), estradiol (E2),
dehydroepiandrosterone (DHEA), and dehydroepiandrosterone
sulfate (DHEAS); and (8) having a low blood level of
progesterone (P).
Social variables that significantly correlated with the
behavioral criterion variables were used to construct social
models to predict the behavioral variables. Biological
(hormonal) variables that significantly correlated with the
77
behavioral criterion variables were used to construct
biological models to predict the behavioral variables. Finally,
significant social and biological predictor variables were used
to construct biosocial models to predict the behaviors of
interest. These biosocial models were used to evaluate
whether social variables, hormonal variables, or both social
and hormonal variables were the most significant predictors
of the behavioral criterion variables.
Additionally, this study was expected to contribute to
the knowledge of occupational science. One purpose of
occupational science is to examine the multiple reasons behind
why people choose to engage in certain occupations (Clark et
al., 1991). Employment is one form of an occupational activity.
The faculty in the USC Department of Occupational Therapy
have developed a model consisting of six subsystems that are
expected to influence occupation: physical; biological;
information processing; sociocultural; symbolic-evaluative;
and transcendental. This study examined how two subsystems
of occupational science: the biological subsystem (hormonal)
and sociocultural subsystem (family characteristics and
childhood experiences) related to employment aspirations.
It was also expected that this project would contribute
to current theories within the area of occupational (i.e.
employment) choice because this study: (1) was on women, who
78
are under-represented in research in this area; and (2)
examined hormonal predictors of occupational choice, which
none of the current theories of occupational choice include
(Ginzberg, 1966, 1972; Holland, 1963, 1985; Roe, 1966;
Super, 1966).
Previous studies that have examined female employment
choice have not distinguished between women with respect to
oral contraceptive use. However, studies that have focused on
sexual or maternal behavior have found hormonal and
behavioral differences between these two groups of women
(Alexander, Sherwin, Bancroft, & Davidson, 1990; Bancroft,
Sherwin, Alexander, Davidson, & Walker, 1991a, 1991b;
Harding, Vail, & Brown, 1985; Hedricks & Udry, 1990).
Additionally, one of these studies (Bancroft, Sherwin,
Alexander, Davidson, & Walker, 1991b) reports correlations
between testosterone levels and sexual behaviors in Oc users
but not for non OC users. In the present study, 42% of the
women were using oral contraceptives; these women were
found to have significant differences in hormonal levels and
sex-atypicality of employment aspirations and college majors
compared to those women not using oral contraceptives.
Therefore, the data from the two groups of women were
analyzed separately.
79
Comparisons between non oral contraceptive users (Non OC)
versus oral contraceptive users (OC)
There were significant differences between the non OC
and the OC users with respect to hormone levels and two of the
behavioral outcome variables. Specifically, the OC users were
more sex-atypical in their employment aspirations and college
majors than were the non OC users. The two groups of women
were also similar with respect to two of the behavioral
outcome variables. Specifically, they had similar occupational
prestige of their employment aspirations and they had similar
ISRO scores.
Also, hormone values for FTI, TT, DHEA, DHEAS, and E2
were significantly greater for non OC users compared to the OC
users. There was also a trend for A and P values to be greater
for the non OC users compared to the OC users. SHBG values
were significantly lower for the non OC users compared to OC
users.
There were no significant differences between the non
OC and the OC users with respect to the social predictor
variables studied.
Non oral contraceptive users
Employment aspirations. Sex-atypicality of the female's
employment aspiration was positively associated with her
80
father's occupational prestige and her adrenal androgen (DHEA)
level. The association between sex-atypicality of her
employment aspiration and the occupational prestige of her
father's employment was consistent with results presented by
Williams & McCullers (1983). They found that fathers who
were employed in professional or managerial careers were
more likely to have daughters in sex-atypical employment than
the fathers who were employed in other job categories.
The association between sex-atypicality of the female's
employment aspiration and the androgen, DHEA, was also
consistent with the findings of Purifoy and Koopmans (1980)
and Schindler (1979). They found that sex-atypicality of
female employment choice was positively related to androgen
levels. Historically it has been more sex-atypical for women
to choose to engage in full-time employment and sex-typical
for women not to work full-time outside the home. Bancroft
et al. (1983) found a trend between a woman's testosterone
level and her engagement in full-time employment outside the
home. It is relevant to note here that in China, Dongyun and
Yumin (1990) found that women who were pilots (sex-atypical
employment for women in China) had higher levels of
testosterone than women who were part of the ground crew
(sex-typical employment for women in China).
Although Dongyun and Yumin (1990), Purifoy and
81
Koopmans (1980), and Schindler (1979) found that sex-
atypicality of employment choice was related to high
testosterone levels, in this study, DHEA, and not testosterone,
was the androgen that significantly correlated with the
behavioral criterion variables. These previous studies did not
measure DHEA levels. The present study did not find
correlations between testosterone and the sex-atypicality of
the female's employment aspiration. Although there were
significant differences between the non OC and OC users in
their testosterone levels, there was limited variability within
each group of women as compared to the variability in DHEA
levels. This might have reduced the likelihood that significant
correlations would be found for either group of women.
Additionally, these previous studies were on women who were
already engaged in employment, while the present study
examined data that were collected before engagement in
employment. It may also be possible that testosterone is
greatly influenced by actual engagement in employment.
The biological model for sex-atypicality of the female's
employment aspiration was slightly stronger than the social
model for this same variable. When a biological predictor
(DHEA) was included with a social predictor (father's
occupational prestige) into one biosocial regression model, the
explanation of variance for sex-atypicality of the female's
82
employment aspiration more than doubled (from 12% to 27%).
Thus, the biosocial model for sex-atypicality of the female's
employment aspiration explained considerably more of the
variance for this variable than either the social or biological
model alone.
Occupational prestige of the female's employment
aspiration was positively associated with whether her mother
was employed outside the home, her mother's occupational
prestige, and her own DHEA level. This variable was also
negatively associated with whether her father had attended
graduate school, and her mother's encouragement. However,
when these four variables were included together into one
regression model of social predictors, the only variable that
remained a significant negative predictor was whether her
father had attended graduate school.
The finding of a negative association between the father
having attended graduate school and the sex-atypicality of the
occupational prestige of his daughter's employment aspiration,
is opposite to that reported by O'Connell et al. (1989) and
Williams & McCullers (1983). They found that fathers of
women in sex-atypical employment were better educated than
the fathers of women in sex-typical employment. Perhaps one
reason for the discrepancy between the results of this study
and the two previous studies is that these two studies had
83
larger samples than the present study. Additionally, the
samples from the two previous studies were chosen from
specific fields (i.e., engineering and nursing) whereas the
present study examined women with a variety of different
employment aspirations.
Johnson et al. (1983) found that prestige of the father's
employment was positively associated with the occupational
prestige of his daughter's employment choice. This was not
found in the present study. The mean age of the women in
Johnson et al.'s study was 43.6 years. In this study, the mean
age of the women was considerably younger, at 21.5 years. It
is probable that the women in Johnson e t al.'s study were
established in their current employment, whereas there was a
greater possibility of substantial change in employment choice
later on in life for the women in the present study. It could be
expected that a change in employment would be in an upward
direction towards higher occupational prestige (i.e., through
employment promotions based on experience).
When the hormone, DHEA, was included with the four
social variables into one biosocial model of occupational
prestige, it made only a slight addition (from 33% to 35%) to
the explanation of variance of sex-atypicality of the
occupational prestige of the female's employment aspiration.
College major. Sex-atypicality of the female's college
84
major was positively associated with the occupational
prestige of her father's employment. Her father's educational
level had a tendency to be negatively associated with the sex-
atypicality of her college major; however, when included with
his occupational prestige into a social model, his educational
level was no longer significant.
No previous studies have examined the relationship
between sex-typicality of a woman's college major and the
occupational prestige of her father's employment. These
variables were included in this study for the following
reasons: (1) there was a positive relationship between sex-
atypicality of college major and sex-atypicality of the
employment aspiration for this group of women (p<0.001); and
(2) previous studies had found a positive association between
the father's occupational prestige and his daughter's
occupational prestige (Johnson et al., 1983). As expected, this
study found that the higher the father's occupational prestige,
the more sex-atypical was his daughter's college major.
Summary. The social predictor variables that had the
most significant impact on sex-atypicality of the behavioral
criterion variables were a high occupational prestige of the
father, and the father not attending graduate school. The
social predictor variables that tended towards significance
were the mother working outside the home, and a high
85
occupational prestige of the mother's employment. Mother's
encouragement tended towards a sex-typical choice in her
daughter's college major.
When DHEA and the father's occupational prestige were
put into a biosocial regression model, the explanation of
variance in sex-atypicality of the female's employment
aspiration more than doubled. The inclusion of DHEA into a
biosocial regression model for occupational prestige of the
female's employment aspiration also made a slight influence
to the explanation of variance for this variable. Therefore, the
biosocial models were superior to either the social or
biological models by themselves.
Oral contraceptive users
Employment aspirations. The occupational prestige of
the female's employment aspiration was positively associated
with her mother's educational level and with whether her
mother had attended graduate school.
O'Connell et al. (1989) and Williams & McCullers (1983)
also found positive associations between a mother's
educational level and the sex-atypicality of her daughter's
employment. These two previous studies analyzed the
relationships between mother's education and sex-atypicality
of their daughter's employment utilizing single correlations.
86
The present study included single correlations and social
regression models. The two social variables were more
strongly related to employment when separately correlated
with this variable than when entered together into a
regression model that explored effects of social predictors on
sex-atypicality of the female's employment aspiration.
College maior. The percentage of males in the field of
the father's employment was negatively associated with sex-
atypicality of his daughter's college major. That is, the higher
the percentage of males in the father's employment, the lower
the percentage of males in his daughter's college major.
No previous studies have examined sex-atypicality of
female college major, or sex-atypicality of the father's
employment. It may be the case that factors that correlate to
sex-atypicality of college major are different than those
factors that correlate to sex-atypicality of employment
aspirations.
ISRO. Father's educational level, whether he attended
graduate school, and whether the mother attended graduate
school, were positively associated with sex-atypicality of
their daughter's ISRO scores. However, when entered together
into the same regression model, they lost significance and only
showed a non-significant tendency to predict ISRO scores.
Based on the findings of Williams and McCullers (1983), it had
87
been expected that a high educational level of the parents
would be associated with sex-atypicality in their daughter's
ISRO scores. A reason for the discrepancy between the results
of the present study and the previous study may be the mean
age difference of the women between the two studies. The
mean age of the women in the study by Williams & McCullers
(1983) was 42.2 years, while in this study the mean age of the
women was 21.5 years.
Summary. The social variable that had the most
significant predictive relationship to a behavioral criterion
variable; specifically, sex-atypicality of the female's college
major, was the percentage of males in the field of the father's
employment. Both parents' educational level and whether both
parents had attended graduate school tended to be associated
with the occupational prestige of their daughter's employment
aspiration and the sex-atypicality of their daughter's ISRO
score. For this group of women, none of the hormonal variables
was significantly related to sex-atypicality of the female's
employment aspiration, her occupational prestige, her college
major, or her ISRO score.
Summary
Comparisons between the non OC users versus the OC users
Hormone values for FTI, TT, DHEA, DHEAS, and E2 were
88
significantly greater for non OC users compared to the OC
users. There was also a trend for A and P values to be higher
for the non OC users. SHBG values were significantly lower for
the non OC users.
The non OC users had less sex-atypical employment
aspirations and less sex-atypical college majors than did the
OC users.
Social predictor variables
There were no differences between the non OC users and
the OC users with respect to the social variables studied.
However, there were significant differences between the two
groups of women in the ability of these social variables to
predict sex-atypicality of employment aspiration, college
major and ISRO scores.
The social variables that had the highest correlation for
both groups of women were two variables associated with
their fathers' employment: the percentage of males in the
field of his employment and his occupational prestige.
Additionally, for the non OC users, the mother's employment
and the father's education also demonstrated a correlation
with the sex-atypicality of their daughters' employment
aspiration and her college major. For the OC users, the
parents' educational level and whether both parents attended
graduate school also had a correlation with the occupational
89
prestige of their daughters' employment aspiration and the
sex-atypicality of her ISRO score.
Hormonal predictor variables
As expected, there were hormonal differences between
the non OC and OC users in this study. OC users were found to
have lower levels of FTI, TT, A, DHEA, DHEAS and higher levels
of SHBG than non OC users. Thus, these findings were in
agreement with other reports in the literature (Alexander,
Sherwin, Bancroft, and Davidson, 1990; Bancroft, Sherwin,
Alexander, Davidson, and Walker, 1991a, 1991b; Hedricks and
Udry, 1990). These authors reported that OC users were found
to have lower levels of testosterone than non OC users.
Behavioral criterion variables
Employment aspirations. Sex-atypicality of employment
aspirations was correlated with an adrenal androgen, DHEA,
but only for the group of non OC users. Part of the explanation
of this finding could lie in the fact that the variances in sex-
atypicality of the employment aspiration and DHEA levels were
much greater for non OC users compared to OC users, thus
increasing the likelihood that significant correlations would
be found between these two variables.
Additionally, in one study (Bancroft, et al., 1991b),
frequency of sexual behaviors were positively correlated with
free testosterone levels, but only within the group of OC users.
90
The authors suggested, at least with respect to sexual
behavior, that psychosocial influences may be more powerful
than hormones in contributing towards this behavior, in the
group of women who were not using oral contraceptives. In
fact, these may obscure a hormone-behavior relationship. As
supporting evidence for this notion, they reported that non OC
users had lower levels of partner-related sexual activity,
lower psychosexual motivation and enjoyment, less positive
evaluation of their partners, and more restrictive attitudes
with respect to sexual morality, compared to the OC users.
Furthermore, many of the differences in sexual behavior and
attitudes between the groups was apparent prior to the use of
oral contraceptives.
Dongyun and Yumin (1990), Purifoy and Koopmans (1980),
and Schindler (1979) found that sex-atypicality of employment
choice was related to high androgen levels, specifically
testosterone. Bancroft et al. (1983) found a tendency for full
time employment outside the home to be associated with
higher androgens, specifically, testosterone levels. In the
present study, and only for the women not using oral
contraceptives, DHEA was the only androgen to significantly
correlate with sex-atypicality of the female's employment
aspiration and the occupational prestige of her employment
aspiration.
91
There are several possible explanations for the
differences in the results of the present study and those of the
previous studies. The present study examined four different
androgens (FTI, A, DHEA, DHEAS). Bancroft et al. (1983),
Dongyun and Yumin (1990), and Schindler (1979) examined only
testosterone. Purifoy and Koopmans (1980) examined
testosterone and androstenedione. Unlike these previous
studies, the participants in the present study had not been
engaging in their occupation, which may affect testosterone
and androstenedione levels. Additionally, any psychosocial
influences that obscured hormone-behavior relationships with
the OC group, if indeed there were any, remain to be further
investigated.
The hormone DHEA, for non OC users, was slightly more
influential (13% of the variance) than the social variable
(father's occupational prestige, 12% of the variance) in
predicting sex-atypicality of the female's employment
aspiration. When both the hormonal and social variables were
examined together, this biosocial model explained 27% of the
variance in this behavioral criterion variable.
For non OC users, the social model was stronger than the
biological model in predicting occupational prestige of the
employment aspiration (33% of the variance versus 6% of the
variance). However, the biosocial model was slightly stronger
92
than the social model, accounting for 35% of the variance in
occupational prestige of the employment aspiration.
College maior. The non OC users tended to have less sex-
atypical college majors than did the OC users. None of the
hormonal variables were significantly related to sex-
atypicality of the female's college major. For the non OC
users, the father's occupational prestige and his educational
level were positively associated with his daughter choosing a
sex-atypical college major, explaining almost a third of the
variance for this behavioral variable.
For the OC users, sex-atypicality of the father's
employment was a significant indicator of his daughter
choosing a sex-atypical college major. No previous studies
have examined sex-atypicality of the female's college major.
However, it was included in this study because of the positive
relationship between sex-atypicality of the female's college
major and the sex-atypicality of her employment aspiration
for the non OC users. This same relationship was not found for
the OC users.
ISRO scores. There were no significant differences
between the two groups of women with respect to their ISRO
scores. Overall, the ISRO scores were high, indicating a sex-
atypical sex role orientation. There was not a significant
social model of sex-atypicality of the female's ISRO scores for
93
the non OC users. For the OC users, the educational level and
both parents attending graduate school, entered together into a
social model, had a notable influence on the sex-atypicality of
the ISRO scores.
Previous research found that the single-most important
factor influencing sex-role orientation in women was their
level of education (Dreyer, et al., 1981). In evaluating the
validity of the ISRO, these investigators (Dreyer et al., 1981)
noted a strong association between formal education and sex-
atypical ISRO scores. In the present study, the lack of
significance in predicting sex-typicality of ISRO scores may
have been due to the fact that all the subjects were seniors in
college; thus, there was no variability in their educational
level.
Conclusions
Information pertinent to the sociocultural subsystem of
the occupational science model includes the individual's
perceptions of social and cultural expectations for behavior
(Clark e t al., 1991), and includes family characteristics and
childhood experiences. The biological subsystem in the
occupational science model focuses on the biological
contributions to occupation (Clark et al., 1991), i.e., hormonal
correlates of sex-atypicality of the female's employment
94
aspiration. The results of the present study suggest that the
interactional effects of both the sociocultural and the
biological subsystems may play a more important role in sex-
atypicality of female employment aspiration than either of
these subsystem alone.
Previous models of occupational choice (Ginzberg, 1966,
1972; Holland, 1963; Roe, 1966; Super, 1966) proposed that
employment choice is a function of inter-individual
differences such as early parent-child interactions,
personality, abilities and interests, and environmental
demands. The results from the present study suggest that
hormonal factors may also be related to employment choice.
Furthermore, both social and hormonal factors should be
considered simultaneously, in the same model, when
attempting to explain female employment choice.
These results indicate that there may be interaction
effects of both social and hormonal influences on sex-
atypicality and occupational prestige of female employment
aspirations. It is interesting to note the behavioral
differences in both groups of women. It may have been useful
to have examined the reasons why the women were or were not
taking oral contraceptives. Additionally, it is recommended
that relationships between social factors, hormones and
occupational choice be assessed in the two groups prior to OC
95
use. Suggestions for further research are to conduct a
longitudinal study that examines biological and social factors
related to females' employment choices before and after OC
use.
96
References
Alexander, G. M ., Sherwin, B. B., Bancroft, J., & Davidson, D. W.
(1990). Testosterone and sexual behavior in oral
contraceptive users and nonusers: A prospective study.
Hormones and Behavior. 24. 388-402.
Bancroft, J., Davidson, D. W., Warner, P., & Tyrer, G. (1983).
Androgens and sexual behaviour in women using oral
contraceptives. Clinical Endocrinology. 12. 327-340.
Bancroft, J., Sherwin, B. B., Davidson, D. W., & Walker (1991a).
Oral contraceptives, androgens, and the sexuality of
young women: I. A comparison of sexual experience,
sexual attitudes,and gender role in oral contraceptive
users and nonusers. Archives of Sexual Behavior. 20.
105-120.
Bancroft, J., Sherwin, B. B., Davidson, D. W., & Walker (1991b).
Oral contraceptives, androgens, and the sexuality of
young women: II. The role of androgens. Archives of
Sexual Behavior. 20. 121-135.
Bern, S. L. (1974). The measurement of psychological
androgyny. Journal of Consulting and Clinical
Psychology. 42(2). 155-162.
Bleier, R . (1991). Sex differences research: Science or belief?
In R . Bleier (Ed.), Feminist approaches to sciencefpp.
147-164). New York, New York: Teachers College Press.
Braunstein, D. (1986). The testes. In F. S. Greenspan & P. H .
Forsham (Eds.). Basic and clinical endocrinology (dp.
351-385). Connecticut: Appleton-Century Crofts.
Broverman, I. (1980). Sex-role stereotypes: A current
appraisal. Journal of Social Issues. 28. 59-78.
97
Cattell, R . B., Eber, H . W., & Tatsuoka, M . M . (1980). Handbook
for the sixteen personality factor questionnaire. (1980
ed.). Champaign, Illinois: Institute for Personality and
Ability Testing, Inc..
Clark, A. F., Marcellus, S., DeLory, B., & Bird, C. E. (1975).
Plasma testosterone free index: A better indicator of
plasma androgen activity. Fertility and Sterility. 26(10).
1001-1005.
Clark, F. A., Parham, D., Carlson, M . E., Frank, G., Jackson, J.,
Pierce, D., Wolfe, R. J., & Zemke, R. (1991). Occupational
science: Academic innovation in the service of
occupational therapy's future. American Journal of
Occupational Therapy. 45(4). 300-310.
Cutler, W. B., Garcia, C. R , Huggins, G. R., & Preti, G (1986).
Sexual behavior and steroid levels among gynecologically
mature premenopausal women. Fertility and Sterility.
45, 496-502.
Dongyun, S., & Yumin, W. (1990). Flight influence on the plasma
level of sex hormones of women pilots. Military
Medicine. 155(6). 262-264.
Dreyer, N. A., Woods, N. F., & James, S. A. (1981). ISRO: A scale
to measure sex-role orientation. Sex Roles. 7(2). 173-
182.
Dupon, C., Hosseinian, A., & Kim, M . H . (1973). Simultaneous
determination of plasma estrogens, androgens, and
progesterone during the human menstrual cycle.
Steriods. 22. 47-61.
98
Ehrhardt, A. A., & Baker, S. W. (1974). Fetal androgens, human
central nervous system differentiation and behavior sex
differences. In R . C. Friedman, R . M . Richart, & R . L. Vande
Wiele (Eds.). Sex differences in behavior. New York:
Wiley.
Ehrhardt, A. A., & Meyer-Bahlburg, H. F. L. (1981). Effects of
prenatal sex hormones on gender-related behavior.
Science. 211(201. 1312-1318.
Ehrhardt, A. A., Meyer-Bahlburg, H . F. L., Feldman, J. F., & Ince,
S. E. (1984). Sex-dimorphic behavior in childhood
subsequent to prenatal exposure to exogenous
progestogens and estrogens. Archives of Sexual
Behavior. 13(51. 457-477.
Ehrhardt, A. A., Meyer-Bahlburg, H . F. L., Rosen, L. R., Feldman, J.
F., Veridiano, N. P., Elkin, E. J., & McEwen, B. S. (1989).
The development of gender-related behavior in females
following prenatal exposure to diethylstylbestrol (DES).
Hormones and Behavior. 23. 526-541.
Feder, H. H. (1981). Perinatal hormones and their role in the
development of sexually dimorphic behaviors. In N. T.
Adler (Ed.), Neuroendocrinoloav of reproduction(pp. 127-
157). New York: Plenum Press.
Feder, H. H . (1984). Hormones and sexual behavior. Annual
Review of Psychology. 35. 165-200.
Frolich, M ., Egenius, C. B., & van Hall, E. V. (1976). Serum levels
of unconjugated aetiocholanolone, androstenedione,
testosterone, dehydroepiandrosterone, aldosterone,
progesterone and oestrogens during the normal menstrual
cycle. ACTA Endocrinology. 81. 548-562.
99
Ginzberg, E. (1966). Toward a theory of occupational choice.
In H. J. Peters & J. C. Hansen (Eds.). Vocational guidance
and career development (2nd ed.. pp. 127-134). New York:
Macmillan.
Ginzberg, E. (1972). Toward a theory of occupational choice: A
restatement. Vocational Guidance Quarterly. 20. 169-
176.
Gough, H. G., & Heilbrun, A. L. (1983). The Adjective Checklist
Manual (1983 ed.). Palo Alto, California: Consulting
Psychologists Press.
Goy, R . W., & McEwen, B. S. (1980). Sexual differentiation of
the brainfpp. 1-80). Cambridge, Massachusetts: M IT
Press.
Hansen, J., & Campbell, D. (1985). Manual for the Strong
Vocational Interest Blank-Strong Campbell Interest
lnventorv(4th ed.). California: Consulting Psychologists
Press.
Harding, C. M ., Vail, C., & Brown, R . (1985). Effect of oral
contraceptives and some psychological factors on the
menstrual cycle. Journal of Biosocial Science. 17. 291 -
304.
Hedricks, C. (1987). Female Senior Questionnaire. Unpublished
manuscript, University of North Carolina-Chapel H ill.
Available upon request.
Hedricks, C., Piccinino, L. J., Udry, J. R., & Chimbira, T. H. K .
(1987). Peak coital rate coincides with onset of
lutenizing hormone (LH) surge. Fertility and Sterility.
45j 234-238.
100
Hedricks, C., & Udry, J. R . (1990). Hormonal correlates of
female sexual behavior, adornment, and maternal
motivation. Presentation to the Conference on
Reproductive Behavior, Atlanta, GA, June 6-8.
Hines, M . (1982). Prenatal gonadal hormones and sex
differences in human behavior. Psychological Bulletin.
92(1), 56-80.
Hinkle, D.E., Wiersma, W., & Jurs, S. G. (1979). Applied
statistics for the behavioral sciences. Chicago, Illinois:
Rand McNally College Publishing Company, 368-389.
Holland, J. L. (1963). Explorations of a theory of vocational
choice and achievement. Psychological Reports. 12. 547-
594.
Houser, B. B., & Garvey, C. (1985). Factors that affect
nontraditional vocational enrollment among women.
Psychology of Women Quarterly. 9. 105-117.
Johansson, E. D. B., & Gemzell, C. (1971). Luteinizing hormone
(LH) and progesterone in plasma and LH and oestrogens in
urine during 42 normal menstrual cycles. ACTA
Endocrinology. 68. 502-512.
Johnson, R . C., Nagoshi, C. T., Ahern, F. M ., Wilson, J. R., DeFries,
J. C., McClearn, G. E., & Vandenberg, S. G. (1983). Family
background, cognitive ability, and personality as
predictors of educational and occupational attainment.
Social Biology. 30(1). 86-100.
Jones, S. L., & Lamke, L. K . (1985). The relationship between
sex role orientation, self esteem, and sex-typed
occupational choice of college women. Psychology of
Women Quarterly. 9. 145-152.
101
Judd, H. L., & Yen, S. S. C. (1973). Serum androstenedione and
testosterone levels during the menstrual cycle. Journal
of Clinical Endocrinology and Metabolism. 36(31. 4 75-
481.
Kaufman, D. (1989). Professional women: How real are the
recent gains? In Freeman, J. (Ed.), Women: A feminist
perspective. (4th ed.). California: Mayfield.
Lemkau, J. P. (1983). Women in male-dominated professions:
Distinguishing personality and background
characteristics. Psychology of Women Quarterly. 8(2).
144-165.
Longino, H . E. (1990). Science as knowledge (pp. 3-15, 133-
161). New Jersey: University Press.
MacLusky, N. J., & Naftolin, F. (1981). Sexual differentiation of
the central nervous system. Science. 211(20). 1294-
1301.
Manni, A., Pardridge, W., Cefalu, W., Nisula, B., Bardin, C.,
Santner, S., & Santen, R. (1985). Bioavailability of
albumin-bound testosterone. Journal of Clinical
Endocrinological Metabolism. 61. 705-710.
Mendelson, R. (1985). Female career development.
Occupational Health Nursing.(April). 194-196.
Morris, N. M ., & Udry, J. R. (1970). Variations in pedometer
activity during the menstrual cycle. Obstetrics and
Gvnecoloav. 35. 199-201.
O'Connell, L., Betz, M . & Kurth, S. (1989). Plans for balancing
work and family life: Do women pursuing nontraditional
and traditional occupations differ? Sex Roles. 20(1/21.
35-45.
102
Oldfield, R . C. (1971). The assessment and analysis of
handedness: The Edinburgh Inventory. Neuropsvcholoaia.
9, 97-113.
Pedhazur, E. J., (1982). Multiple regression in behavioral
research(2nd ed., pp. 5-6, 722). New York, New York: CBS
College Publishing.
Perun, P. J., & Bielby, D. D. (1981). Towards a model of female
occupational behavior: A human development approach.
Psychology of Women Quarterly. 6(2). 234-252.
Purifoy, F. E., & Koopmans, L. H . (1980). Androstenedione,
testosterone, and free testosterone concentration in
women of various occupations. Social Bioloav. 26(31.
179-188.
Reskin, B. F. & Roos, P. (1987). Status hierarchies and sex
segregation. In C. Bose, & G. Spitze (Eds.), Ingredients for
women's employment policy. New York: State University
of New York Press.
Roan, S. (1992, August 17). Women's powers of persuasion.
The Los Angeles Times, pp. E l, E2.
Roe, A. (1966). Early determinants of vocational choice. In H .
J. Peters, & J. C. Hansen (Eds.). Vocational guidance and
career development (2nd ed., pp. 101-108). New York:
Macmillan.
Rosenzweig, M . R . & Leiman, A. L. (1982). Hormones: A
chemical communication system. In M . R. Rosenzweig &
A. L. Leiman (Eds.), Physiological psvcholoovfpp. 179-
207). Lexington, Massachusetts: D . C. Heath and
Company.
Schiamberg, L. B., (1985). Human development (2nd ed., pp. 23-
25). New York: Macmillan.
103
Schindler, G. L., (1979). Testosterone concentration.
personality patterns, and occupational choice in women.
Unpublished doctoral dissertation, University of Houston,
Texas.
Sherwin, B. B. (1988). A comparative analysis of the role of
androgen in human male and female sexual behavior:
Behavioral specificity, critical thresholds, and
sensitivity. Psvchobioloav. 16(41. 416-425.
Sherwin, B. B., Gelfand, M . M ., & Brender, W. (1985). Androgen
enhances sexual motivation in females: A prospective,
crossover study of sex steroid administration in the
surgical menopause. Psychosomatic Medicine. 47. 339-
351.
Stevens, G., & Hoisington, E. (1987). Occupational prestige.
Social Science Research. 16. 74-105.
Super, D. E. (1966). A theory of vocational development. In H .
J. Peters, & J. C. Hansen (Eds.). Vocational guidance and
career development (2nd ed., pp. 144-153). New York:
Macmillan.
Tabachnick, B. G., & Fidell, F. S. (1983). Using multivariate
sta tistic s . New York, NY: Harper&Row, Publishers, 146-
148, 248-249.
Thomas, C. L. (Ed.). (1985). Taber's cyclopedic medical
dictionary (15th ed.l. Philadelpia: F. A. Davis Company.
Tyler, J. P. P., Newton, J. R., & Collins, W. P. (1975). Variations
in the concentration of testosterone in peripheral venous
plasme from healthy women. ACTA Endocrinology. 80.
542-550.
Udry, J. R. (1988). Untitled and unpublished manuscript.
104
Udry, J. R., & Morris, N. M . (1968). The distribution of coitus in
the menstrual cycle. Nature. 220. 593-596.
Udry, J. R., & Morris, N. M . (1970). Effect of contraceptive pills
on the distribution of sexual activity in the menstrual
cycle. Nature. 227. 502-503.
Vermeulen, A. (1983). Androgen secretion by adrenals and
gonads. In V. B. Mahesh and R . B. Greenblatt (Eds.).
Hirsutism and virilism(pp. 17-34). Boston: John Wright
PSG Inc.
Williams, S. W., & McCullers, J. C. (1983). Personal factors
related to typicalness of career and success in active
professional women. Psychology of Women Quarterly.
7(4), 343-357.
Yen S. S. C. (1986). Steroid hormones. In Yen, S. S. C., & Jaffe,
R. B. (Eds.), Reproductive endocrinoloav(2nd ed., pp. 140-
153). Philadelphia: W. B. Saunders Co..
105
APPENDIX A
Definition of Terms
1. Congenital adrenal hyperplasia (CAH). CAH is a genetic
disorder which prevents the synthesization of cortisone by the
adrenal cortices. This results in high levels of
adrenocorticotropic hormone which stimulate the secretion of
high amounts of androgens and progestogens pre- and
postnatally. The syndrome occurs in males and females.
Female infants with CAH are usually born with masculinized
genitalia (enlarged clitoris and labial fusion), which is
typically surgically corrected shortly after birth. Males and
females are usually treated with cortisone from infancy.
2. Diethvlstilbestrol (PEST DES is a synthetic estrogen that
was administered to pregnant women for medical reasons (i.e.,
to prevent miscarriages) in the early 1940's to 1971. Exposure
is limited to the prenatal period and individuals were usually
born with normal external genitalia and normal health.
However, exposure was found to be associated with the
occurrence of vaginal and uterine abnormalities in the
prenatally exposed daughters. DES is currently used to treat
disorders resulting from estrogen deficiency.
106
3. Medroxyprogesterone acetate (MPA). MPA is the active
compound in Provera (a drug administered for medical reasons
during pregnancy) and is chemically related to progesterone. It
acts as a suppresser of androgen activity.
4. Sex-atvpical behavior for women. This is a term I have
incorporated into my paper to describe behaviors that are
typically exhibited by men.
5. Sex-role behavior. This is a term that has been used by
researchers to describe behaviors that are thought to be
appropriate according to one's sex (or gender). I have chosen
not to use this term in my paper because it is subjective and
biased. Its description, and consequently the selection of data
in related research, is influenced by the idea that there are
sex-appropriate and sex-inappropriate behaviors.
6. Sex-tvpical behavior for women. This is a term I have
incorporated into my paper to describe behaviors that are
typically exhibited by women.
7. Sex-atvpicalitv. This is a term that I have incorporated
into my paper that examines behaviors along a continuum
ranging from sex-typical to sex-atypical.
107
APPENDIX B
Operational Definitions of Variables
I. Predictor Variables
1. Only child status (ONLY). Only child status was coded as a
dummy variable, where a "0 " was assigned if the subject was
not an only child and a " 1" was assigned if the subject was an
only child.
2. Ordinal position of birth (FIRST). Ordinal position of birth
of the subject was an ordinal level variable measured by
comparing the subject's reported birth year and the birth years
of her siblings, with "1"=first born; "2"=second born; "3"=third
born, etc. If she was an only child, she was considered to be
first born.
3. Father's occupational prestige (FPREST). Occupational
prestige was established for the subjects' fathers using
Stevens' & Hoisington's (1987) scale, which ranged from a low
of "0 " to a high of "100."
4. Mother's employment status (MEMP). The mother's
employment status was coded as a dummy variable, where a
108
"0" was assigned if the subject's mother did not work outside
the home and a "1" was assigned if the subject's mother did
work outside the home.
5. Father's educational level (FEDUC). Father's educational
level was measured using an ordinal scale of measurement,
with "0 "=did not complete high school; " 1"=high school
graduate; "2"=technical school graduate; "3"=some college;
"4"=college graduate or some graduate school; "5"=at least a
master's degree.
6. Father attended graduate school (FGRAD). Father attending
graduate school was coded as a dummy variable, where a "0 "
was assigned if the father did not attend graduate school and a
"1" was assigned if the father did attend graduate school.
7. Mother's educational level (MEDUC). Mother's educational
level was measured using an ordinal scale of measurement,
with "0 "=did not complete high school; "1"=high school
graduate; "2"=technical school graduate; "3"=some college;
"4"=college graduate or some graduate school; "5"=at least a
master's degree.
8. Mother attended graduate school (MGRAD). Mother attending
109
graduate school was coded as a dummy variable, where a "0 "
was assigned if the mother did not attend graduate school and
a "1" was assigned if the mother did attend graduate school.
9. Encouragement from fathers (FENC). Encouragement from
fathers was measured using a behavioral checklist. Subjects
were given 2 statem ents that asked if their fathers
"encouraged educational achievement" and "encouraged you to
have a career." Subjects indicated with a check whether their
fathers had exhibited the particular behavior, coded as "0 "=no
and "1"=yes. The total number of "yes" responses was tallied.
10. Encouragement from mothers (MENC). Encouragement from
mothers was measured using the same behavioral checklist
that was used to measure encouragement from fathers.
Subjects indicated with a check whether their mothers had
exhibited the particular behavior, coded as "0 "=no and "1"=yes.
The total number of "yes" responses was tallied.
11. Hormones. The following hormones were measured by
radioimmunoassay: total testosterone (TT); sex hormone-
binding globulin (SHBG); androstenedione (A);
dehydroepiandrosterone (DHEA); dehydroepiandrosterone
sulfate (DHEA-SO4 ); estradial (E2); and progesterone (P). A
110
free testosterone index (FTI) was calculated with the formula:
FTI=TT/SHGB (Clark et al., 1975). The hormone values for TT,
A, DHEA, and P are reported in ng/ml. The hormone values for
SHBG are reported in nm/l. The hormone values for E2 and
DHEA-SO4 are reported in^umol/l.
II. Behavior Criterion Variables
1. College maior (UNCFIELD). The subject's reported college
major was given a numerical value that reflected the
percentage of males registered with that major in the 1985
graduating class at UNC-Chapel Hill.
2. Employment aspiration (RMALE80). The type of employment
the respondents would like to have in the future was assigned
a numerical value that reflected the percentage of males in
that job according to the U.S. Department of Laborl 980.
3. The Index of Sex Role Orientation (ISRO). The ISRO is a
summated rating scale that purports to measure a women's
sex-role orientation. Subjects were given 16 statem ents
related to sex role orientation, such as "women should take
care of running their homes and leave running the country up to
men," or "having a job means a life of your own." Level of
111
agreement with these statem ents was indicated on a five point
ordinal scale: (1) agree very much; (2)agree somewhat; (3)
neither agree nor disagree; (4) disagree somewhat; or (5)
disagree very much. A high score indicated a sex-atypical, sex
role orientation (Dreyer et al., 1981).
112
APPENDIX C
Definitions of the variable labels
Variable Label Definition
I. Social Variables
A. Fathers
1 . FEDUC
2. FGRAD
3. FPREST
4. FMALE80
5. FENC
B. Mothers
1. M EDUC
2. MGRAD
3. MPREST
4. MMALE80
5. M EM P
6. MENC
educational level
attended graduate school
occupational prestige, according to
Stevens' & Hoisington's (1987) scale
% of males in occupation, according
to 1980 U.S. Department of Labor
encouragement
educational level
attended graduate school
occupational prestige, according to
Stevens' & Hoisington's (1987) scale
% of males in occupation, according
to 1980 U.S. Department of Labor
worked outside the home
encouragement
113
II. Hormonal Variables
1. FTI
2. TT
3. A
4. DHEA
5. DHEAS
6. E2
7. P
8. SHBG
III. Outcome Variables
1. RMALE80
2. RPREST
3. UNCFIELD
4. ISRO
free testosterone index (TT/SHBG)
total testosterone, in ng/ml
androstenedione, in ng/ml
dehydroepiandrosterone, in ng/ml
dehydroepiandrosterone sulfate, in
ng/ml
estradiol, injimol/L
progesterone, in ng/ml
sex hormone binding globulin, in nmol/l
% of males in employment aspiration,
according to 1980 U.S. Department of
Labor
occupational prestige of aspiration,
according to Stevens' & Hoisington's
(1987) scale
% of males in college major at UNC-CH
Index of Sex Role Orientation Score
114
Appendix D
Correlation analyses between predictor and criterion
variables3’* 3 for the non oral contraceptive users
RMALE80 RPREST UNCFIELD ISRO
PREDICTOR
VARIABLES
FPREST
r 0.39 - 0.24 0.55 0.05
P<
0.04 0.24 0.002 0.78
n 25 25 28 28
MPREST
r 0.12 0.37 0.01 0.03
P<
0.54 0.05 0.94 0.86
n 25 25 30 30
M ENC
r 0.03 - 0.31 0.03 - 0.14
P<
0.87 0.09 0.85 0.41
n 29 29 33 33
M EM P
r 0.14 0.45 - 0.09 0.12
P<
0.45 0.01 0.59 0.49
n 29 29 33 33
FGRAD
r 0.03 - 0.34 0.08 - 0.09
P<
0.86 0.06 0.63 0.58
n
FIRST
29 29 33 33
r 0.07 - 0.07 - 0.23 - 0.12
P<
0.70 0.71 0.18 0.48
n 29 29 33 33
115
Appendix D - continued
RM ALE80 RPREST U N C FIELD ISRO
PREDICTOR
VARIABLES
ONLY
r 0.12 0.25 - 0.14 0.16
P<
0.52 0.17 0.42 0.36
n
29 29 33 33
MGRAD
r 0.02 - 0.05 0.11 - 0.03
P<
0.93 0.79 0.50 0.84
n 29 29 33 33
M EDUC
r 0.25 - 0.07 0.20 0.11
P<
0.17 0.70 0.25 0.55
n 29 29 33 33
FEDUC
r 0.17 - 0.26 0.30 - 0.07
P<
0.37 0.17 0.09 0.67
n 28 28 32 32
FMALE80
r 0.02 0.12 - 0.26 - 0.07
P<
0.91 0.56 0.16 0.70
n 26 26 30 30
MMALE80
r 0.22 0.04 - 0.20 - 0.08
P<
0.24 0.82 0.25 0.63
n 29 26 33 33
FENC
r 0.12 - 0.19 0.15 - 0.21
P<
0.52 0.31 0.40 0.23
n 29 29 33 33
116
Appendix D - continued
RM ALE80 RPREST U N C FIELD ISRO
PREDICTOR
VARIABLES
DHEA
r 0.40 0.30 0.16 0.03
P<
0.03 0.10 0.37 0.82
n 29 29 33 33
TT
r 0.13 0.01 0.03 - 0.09
P<
0.48 0.95 0.85 0.59
n 29 29 33 33
FTI
r 0.16 - 0.02 0.09 - 0.17
P<
0.40 0.88 0.59 0.31
n 29 29 33 33
SHBG
r 0.15 0.004 - 0.07 0.15
P<
0.41 0.98 0.66 0.38
n 29 29 33 33
DHEAS
r 0.17 - 0.02 0.16 ■ 0.09
P<
0.37 0.88 0.37 0.60
n 29 29 33 33
A
r 0.01 - 0.11 0.05 - 0.05
P<
0.93 0.57 0.78 0.77
n 29 29 33 33
117
Appendix D - continued
RMALE80 RPREST UNCFIELD ISRO
PREDICTOR
VARIABLES
E2
r 0.05 0.02 0.05 - 0.03
P<
0.79 0.89 0.77 0.83
n 29 29 33 33
P
r 0.02 0.03 - 0.12 - 0.02
P<
0.91 0.85 0.47 0.89
n 29 29 33 33
aOperational definitions can be found in Appendix B
bDefinitions of the variable labels can be found in
Appendix C
Appendix E
Correlation analyses between predictor and criterion
variables3.1 3 for the oral contraceptive users
RMALE80 RPREST UNCFIELD ISRO
PREDICTOR
VARIABLES
FEDUC
r 0.18 0.35 0.22 0.42
P<
0.41 0.13 0.31 0.05
n 21 20 22 22
FMALE80
r 0.09 0.07 - 0.42 - 0.28
P<
0.66 0.76 0.03 0.17
n 23 22 24 24
M EDUC
r 0.32 0.46 0.003 0.25
P<
0.13 0.03 0.98 0.24
n 23 22 24 24
MGRAD
r 0.11 0.44 - 0.15 0.44
P<
0.59 0.03 0.47 0.03
n 23 22 24 24
FGRAD
r 0.16 0.29 - 0.16 0.37
P<
0.44 0.18 0.43 0.07
n 23 22 24 24
FIRST
r 0.20 0.03 0.19 0.18
P<
0.34 0.87 0.35 0.37
n 23 22 24 24
119
Appendix E - continued
RM ALE80 RPREST U N C FIELD ISRO
PREDICTOR
VARIABLES
ONLY
r
P<
n 23 22 24 24
FPREST
r 0.07 0.29 - 0.07 0.03
P<
0.74 0.21 0.76 0.86
n 20 19 21 21
MPREST
r 0.09 0.04 - 0.10 0.11
P<
0.70 0.85 0.66 0.62
n 20 19 20 20
MMALE80
r 0.07 0.18 - 0.25 0.20
P<
0.74 0.42 0.22 0.33
n 23 22 24 24
FENC
r 0.04 - 0.008 - 0.06 0.02
P<
0.85 0.97 0.75 0.90
n 23 22 24 24
M ENC
r 0.28 0.10 - 0.32 0.11
P<
0.18 0.63 0.11 0.59
n 23 22 24 24
120
Appendix E - continued
RM ALE80 RPREST U N C FIELD ISRO
PREDICTOR
VARIABLES
M EMP
r 0.14 - 0.15 - 0.11 - 0.14
P<
0.51 0.51 0.59 0.50
n 23 22 24 24
TT
r 0.31 0.18 - 0.23 - 0.08
P<
0.14 0.41 0.27 0.70
n 23 22 24 24
FTI
r 0.16 0.10 - 0.09 0.04
P<
0.46 0.64 0.66 0.81
n 23 22 24 24
SHBG
r - 0.05 - 0.08 - 0.10 - 0.11
P<
0.79 0.71 0.63 0.58
n 23 22 24 24
DHEA
r 0.18 0.004 - 0.19 - 0.18
P<
0.39 0.98 0.35 0.38
n 23 22 24 24
DHEAS
r 0.16 - 0.01 - 0.17 - 0.09
P<
0.45 0.95 0.42 0.66
n 23 22 24 24
A
r 0.16 0.24 0.04 0.14
P<
0.44 0.27 0.82 0.50
n 23 22 24 24
121
Appendix E - continued
RMALE80 RPREST UNCFIELD ISRO
PREDICTOR
VARIABLES
E2
r 0.12 0.32 0.10 0.08
P<
0.58 0.14 0.63 0.68
n 23 22 24 24
P
r 0.07 0.19 - 0.28 - 0.01
P<
0.73 0.37 0.17 0.94
n 23 22 24 24
aOperational definitions can be found in Appendix B
bDefinitions of the variable labels can be found in
Appendix C

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A Comparison Of Prenatally Drug Exposed Preschoolers To Non-Drug Exposed Preschoolers Using The Miller Assessment For Preschoolers

PDF

Coping Strategies Of Three Adolescents With Disabilities

PDF

Fine motor skills of two- to three-year-old drug exposed children

PDF

An investigation of the relationship between measures of social, academic and global self-concept

PDF

A follow up study of adapted computer technology training at the High Tech Traning Center of Santa Monica College

PDF

A study of a pilot sensory history questionnaire using contrasting groups

PDF

Performances of orthodox and liberal Jewish children in third grade on praxis subtests of the SIPT: a cross cultural study

PDF

The application of Sarbin's theory of emotions as narrative emplotments to stories of two men diagnosed with cancer

PDF

The use of occupational therapists or interdisciplinary teams in the evaluation of assistive technology needs of children with severe physical disabilities in Orange County schools

PDF

The relationship of demographic status, educationl background, and type and degree of disability to transition outcomes in young adults with disabilities: a quantitative research synthesis

PDF

Hand function in older adults: the relationship between performance on the Jebsen Test and ADL status

PDF

A comparison of the play performance of boys with autism and that of boys without disabilities in Taiwan

PDF

A model for integrating occupational therapy procedures

PDF

An examination of the health-related quality of life and functional skills as reported by the parents of young children with developmental delays

PDF

Interrater reliability of the Evaluation of Sensory Processing (ESP)

PDF

Lesbian identities, daily occupations, and health care experiences

PDF

A qualitative study on the relationship of future orientation and daily occupations of adolescents in a psychiatric setting

PDF

A Model To Determine Effectiveness Of School Community Advisory Councils Of The Los Angeles Unified School District

PDF

Constructing identities in social worlds: Stories of four adults with autism

PDF

Variables Related To Life Satisfaction In Persons With Spinal Cord Injuries

Asset Metadata

Creator Mckenna, Dawn Marie (author)

Core Title A Biosocial Model Of Sex-Atypicality In Female Employment Aspirations

Degree Master of Arts

Degree Program Occupational Therapy

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,psychology, industrial,women's studies

Format masters theses (aat)

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Hedricks, Cynthia A. (committee chair), Neville-Jan, Ann (committee member), Parham, Linda Diane (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-9117

Unique identifier UC11357675

Identifier 1378426.pdf (filename),usctheses-c18-9117 (legacy record id)

Legacy Identifier 1378426-0.pdf

Dmrecord 9117

Document Type Thesis

Format masters theses (aat)

Rights Mckenna, Dawn Marie

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA