INFORMATION TO USERS
This manuscript has been reproduced from the microfihn master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter free, while others may be from any type o f 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 o f 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 Zed) Road, Ann Aibor MI 48106-1346 USA 313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PERSONALITY CORRELATES OF MALE SEXUAL ORIENTATION
AND RECALLED CHILDHOOD SEX-TYPED BEHAVIORS
by
Chavis Alan Patterson
submitted to the
Faculty of the College of Arts and Sciences
of The American University
in Partial Fulfillment of
the Requirement for the Degree
of
Doctor of Philosophy
in
Psychology
Signatures of Committee:
Chair: Fantie, PhD.
Dean Hamer, Ph.D.
^ennis Murphy, M.Çt
lean of ^ C o lle g e Alan Silberberg, Ph D.
1996 American University llil Washington, D C. 20016 àXSBîCm UNIVERSITY LIBRARY
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 9704808
Copyright 1996 by Patterson, Chavis Alan All rights reserved.
mvn Microform 9704808 Copyright 1996, by UMI Company. All 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
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. © COPYRIGHT
by
CHAVIS ALAN PATTERSON
1996
ALL RIGHTS RESERVED
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PERSONALITY CORRELATES OF MALE SEXUAL ORIENTATION
AND RECALLED CHILDHOOD SEX-TYPED BEHAVIORS
BY
CHAVIS A. PATTERSON
ABSTRACT
I explored the relationships between sexual orientation, recalled childhood sex-typed
behavior (RCSTB), and personality traits in a group of 145 heterosexual and 114
homosexual men. The Kinsey scales assessed the participants’ sexual orientation a
questionnaire assessed recalled childhood sex-typed behaviors, and the Neuroticism
Extraversion and Openness Personality Inventory-Revised (NEO-PI-R) and 16 Personality
Factor Inventory (16PF) defined personality traits. In agreement with previous research,
there was a strong correlation between sexual orientation and recalled childhood sex-
typed behaviors. Heterosexual men recalled stereotypically masculine childhoods whereas
homosexual men described more diverse and generally more feminine activities. As
expected from the literature, significant correlations were obtained between sexual
orientation and three domains of adult personality. Compared to heterosexual men, gay
men were more tender-minded and ascetic; more open to different values, feelings, and
actions, and more anxious. In heterosexual men, childhood femininity was associated
with multiple adult characteristics including tender-mindedness, openness, introversion,
and lack of conscientiousness. In contrast, with gay men the main adult correlate of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. childhood femininity was anxiety. This provides the first indication that the origins or
development of sex-typed behaviors in gay men may be qualitatively different than in
heterosexual men. 1 discuss the implications of this finding for the further understanding
of the complex interactions among personality, sexuality, and childhood behavior.
Ill
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS
In such an undertaking, there are always a number of people who should be
acknowledged for their contributions. I would like to thank my committee members;
Dean Hamer, Bryan Fantie, Alan Silberberg, and Dennis Murphy. 1 would like to thank
Dean (and the lab of gene structure) for allowing me to work in the laboratory on the
NIH campus and assisting with the research. His consistent help, support, and
encouragement made this process extremely enjoyable. 1 would like thank Bryan Fantie
for his continued support and confidence in me during my development as a researcher.
1 would also like to thank Dennis Murphy and Aan Silberberg for taking time out of
their schedule to provide constructive and insightful feedback.
1 wish to thank a number of family, friends, and colleagues who made the process
a little more bearable in stressful times; Sylvia and Theodore Patterson, Mark Nelson,
J. Frederick Long, Nancy Shapiro, Phil Fisher, Tom Hays, Amy Sample, Michelle
Sedgewick, Judi Kinney, Scott Parker, Sara Holland, Terri Hazapis, Louis Murdock and
Mary Lanman.
And finally, 1 wish to thank my wife, Janet, for her enduring support through this
bizarre and foreign process, we call graduate school.
IV
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS
ABSTRACT...... ü
ACKNOWLEDGEMENTS...... iv
LIST OF TABLES...... vü
LIST OF FIGURES ...... vüi
CHAPTER
1. INTRODUCTION...... I
Early Psychodynamic and Social Learning Theories
Adult Sex-Typed Characteristics and Personality Structure
Childhood Sex-Typed Behaviors
Sexual Differentiation of the Brain
Genetic Factors
2. METHODS ...... 16 Subjects Recruitment Demographic Information
3. PROCEDURE...... 20
Instruments
General Questionnaire
Sexual Orientation Assessment
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Recalled Childhood Sex-Typed Behavior Scale
Personality Assessment
Factor Structure of the Personality Inventories
DNA Assessment
4. RESULTS...... 25 Sexual Orientation vs. Recalled Childhood Sex-Typed Behaviors
Sexual Orientation vs. 16 PF Personality Factor Traits
Sexual Orientation vs. NEO-PI-R
Personality Inventory Traits
Recalled Childhood Sex-Typed Behaviors vs. Personality Traits
5. DISCUSSION ...... 29 Sexual Orientation and Recalled Childhood Sex-Typed Behaviors
Sexual Orientation and Personality
Recalled Childhood Sex-Typed Behaviors and Personality
Limitations
Prospects
APPENDICES ...... 42
REFERENCES...... 55
VI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES
Demogiaphics...... 19
Cattell's 16 Personality Factor Inventory ...... 43
Factor Loadings for the NEO-PI-R and the 16 P F ...... 45
Correlations Between Sexual Orientation and Recalled Childhood Sex-Typed Behaviors ...... 47
Correlations Between Sexual Orientation and 16 PF Personality Factors ...... 48
CoiTelations Between Sexual Orientation and NEO-PI-R Personality Factors ...... 49
Conelations Between Recalled Childhood Sex-Typed Behaviors and 16 PF Personality Factors ...... 50
Con elations Between Recalled Childhood Sex-Typed Behaviors and NEO-Pl-R Personality Factors ...... 52
vu
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES
Recalled Childhood Sex-Typed Behaviors and Sexual Orientation ...... 54
vm
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1. INTRODUCTION
Personality has been defined as "the sum total of physical, mental, emotional, and
social characteristics of an individual," (Flexner & Stein, 1982), or as "the dynamic
organization within the individual of those psychophysical systems that determine his/her
characteristics, behavior, and thought" (Allport, 1937). In principle, these definitions include
virtually every aspect of human natme. In reality, the study of personality psychology is
focused largely on certain defined aspects of temperament and character such as sociability,
emotional stability, curiosity, conscientiousness, and cooperativeness. Personality
psychologists have defined such central traits using a combination of natural language and
factor analysis of those similarities and differences that are most often used to describe and
compare people (Cattell, Eber, & Tatsuoka, 1970; Costa & McCrae, 1991).
Although such traits tell a good deal about a person, they are by no means
comprehensive. For example, gender and sexual orientation are two undeniably important
aspects of a person's make up, yet most personality inventories do not ask any questions
directly related to these characteristics. The existence of male-female differences in
personality style is evident from the normative data available for standard inventories such as
the 16 Personality Factor Inventory (16 PF) and Neuroticism Extraversion & Openness
Personality Inventory-Revised (NEO-PI-R), which exhibit multiple significant differences
between men and women (Cattell, Eber, & Tatsuoka, 1970; Costa & McCrae, 1991).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2
Similar information in regard to heterosexuals compared to homosexuals is not generally
available.
A limitation of the dispositional approach to personality is that it does not explain how
individual differences arise. Some critics of personality psychology have pointed out its
reliance on circular arguments in which a person's behavior is used to infer the existence of
a trait, which in turn is used to explain the person's behavior (Carver & Scheier, 1988).
Thus, the motivation for the current work was two-fold. The first goal was to
understand the relationships between standard personality traits and gender-related behaviors
during development, versus sexual orientation in adulthood. To accomplish this, I collected
and analyzed information on each of these characteristics firom a substantial cohort of
heterosexual and homosexual men. The second aim was to set the stage for an integrated
analysis of the role of genes and other biological and environmental factors on possible
interrelated aspects of human nature. That molecular work is currently in progress m the NIH
laboratoiy where my research was conducted.
Early Psychoanalytic and Social Learning Theories
Many of the early studies of sexual orientation focused on demonstrating that
homosexuality is, or is not, a form of psychopathology resulting fi’om poor parenting or
abnomial childhood experiences. The most influential psychoanalytic theorist was Freud, who
proposed that homosexuality was the product of arrested development in the phallic stage of
psychosexual development. During this Oedipal period, the heterosexual male child's love and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3
desire for his mother are replaced with vicarious love for the mother through identification
with the father. This vicarious love eliminates the competition between the father and son.
The child's anxiety is reduced specifically around the fear of castration by the father. In
homosexual development, Freud hypothesized that a dominant mother and a detached father
lead to resolution of castration anxiety through identification with the mother. As a result, the
male child desires to be a mother-like figure, thus becoming attracted to men (Baron, Byrne
& Kantowitz, 1980; Ellis & Ames, 1987; Freud, 1953).
Freud also believed that homosexuality could be facilitated by early sex play between
a child and a same-sex playmate. This model of sexual orientation represents an early version
of social learning theory, which is based on the idea that any behavior that is rewarded will
be repeated. In other words, as a child encounters pleasiurable same-sex sexual activity, the
behavior is reinforced and is thereby maintained. Social learning theorists further hypothesized
that such pleasurable same-sex activities would have an especially strong effect if they
occun ed during puberty, reported to occur at an earlier than average age in pre-homosexual
boys (Wasserman & Storms, 1984). By contrast, negative experiences with an opposite sex
playmate lead to a rejection of heterosexuality as an adult (Ellis & Ames, 1987; Stoller &
Herdt. 1984).
These concepts of homosexuality as the product of arrested psychological
development and abnormal early sexual experiences dominated the field for over half a
century (Bancroft, 1994; Ellis & Ames, 1987). Then Hooker (1957) conducted the first
empirical study on the psychological characteristics of homosexual men. She administered
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4
projective tests, such as the Rorschach and the Thematic Apperception Test, to a series of gay
and heterosexual men and gave the responses to raters who were blind to the sexual
orientation of the subjects. Much to the surprise of the psychiatric community, the
homosexual subjects showed no increase in social or psychological maladjustment and no
distinguishing psychopathology.
Subsequently, many studies using both projective and objective tests have failed to
demonstrate any greater psychopathology in homosexual than in heterosexual individuals
(Isay. 1989; Riess, 1980). For example, Clark (1975) studied 20 age- and education-matched
males at each of the seven levels of the Kinsey Scale of sexual orientation. He found no
significant correlation between sexual orientation and psychopathology as measured by the
Tennessee Self-Concept Scale, which includes measures of self-criticism, defensiveness, self-
concept. general emotional maladjustment, neurosis, psychosis, personality-character
disorder, and overall personality integration. Similarly, Langevin ( 1978) found no differences
in Minnesota Multiphasic Personality Inventory (MMPI) measures of psychopathology in
non-ps)'chiatric heterosexual and homosexual male populations. Based on many studies of this
soit, the American Psychiatric Association removed homosexuality firom its list of disorders
in 1973 (Friedman & Downey. 1994; Isay, 1989).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5
Adult Sex-Typed Characteristics and Personality Structure
Studies of normal personality variation in sexual orientation have been relatively
sparse, and have largely focused on those traits that show a difference between males and
females (LeVay, 1993). The underlying assumption of these studies is that because gay men
are similar to heterosexual women in one particular characteristic, namely their sexual
attraction to males, there may be a general feminization of their personality structure. A
second line of research has examined potential sexual orientation-related differences in social
confoiTuity.This research is driven by the idea that homosexuals will display increased social
nonconformity as either a precursor or result of their "nonconforming" sexual orientation
(LeVay. 1993).
Several studies have produced partial support for both of these ideas. For example,
Duckitt and DuToit (1988) compared gay and heterosexual men using Cattell's 16 PF
inventory, which shows significant sex differences for 8 out of the 16 measured personality
traits. In partial agreement with the psychological feminization hypothesis, gay men were
significantly more sociable, tender-minded, apprehensive, and emotionally imstable than their
heterosexual coimterparts; however, there was no significant difference in the personality
factor of dominance. The gay men also showed increased levels of social nonconformity as
measured by the 16PF scales for bohemianism, shrewdness, and self-sufficiency; contrary to
tlie my expectation; however, there were no differences in radicalism or suspiciousness. Evans
(1970) obtained similar results, finding significant heterosexual-homosexual differences for
six out of the eight 16PF factors described by Duckitt and DuToit (1988).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6
Bernard ( 1982) compared heterosexual and homosexual men using scales derived
from the Bem Sex Role Inventory, The 16 PF Inventory, The Adjective Check List, the Social
Insight Test, and the Experience Inventory. Varimax analysis revealed nine major factors
accounting for most of the variance in the study population, of which only two were
significantly different between heterosexual and homosexual men; namely, feminine opeimess
and unconventional independence. The first factor supports the sex-difference hypothesis and
the second factor relates to social nonconformity. Manosevitz (1971) observed differences
between heterosexual and homosexual men using the Male-Female scale of the MMPI.
Childhood Sex-Typed Behaviors
The strongest evidence for psychological differences between heterosexual and
homosexual men comes firom studies of childhood rather than of adulthood, especially in
regard to sex-typed or gender-typical behaviors (Bailey & Zucker, 1994). These are
behaviors that are exhibited, on average, more frequently by one sex than the other. Although
there is considerable debate over whether such sex-typed behaviors result from biological
(LeVay. 1991; McCormick & Witelson, 1991), social (Kardiner, 1963), or cultural factors
(Freud. 1953), there is no doubt that they exist; boys and girls show statistical differences in
die fi equencies of many types of activities such as ball sports and rough-housing compared
to playing with dolls and cooking (Bailey & Zucker, 1994).
Pre-gay boys recall a more feminine pattern and more diversity (variance) than pre
heterosexual boys in many of these sex-typed behaviors. This has been observed in over 48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7
retrospective studies, including some 27,724 subjects, conducted in a variety of societies and
cultures (Bailey and Zucker, 1994). ha their meta-analysis of this literatme, Bailey and Zucker
( 1994) examined the effect size (D) of sexual orientation on recalled cross-gender behavior
as the difference between the mean scores for homosexual and heterosexual men divided by
their pooled standard deviation; thus positive values of D indicate greater cross-gender
behavior in the homosexual subjects. They foimd that D was significantly greater than zero,
t(47) = 19.2, g < 0.001, mean = 1.19, S.D. = 0.43, ranging from 0.5 to 2.09. Quite unusual
in the literature on sexual orientation, there was not a single study in which an opposite effect
was observed.
Although the correlation between adult sexual orientation and recalled childhood sex-
typed behaviors (RCSTB) might be an artifact of selective recall, there is no real evidence to
support tliis idea (Bailey & Zucker, 1994). Moreover, the correlation has also been observed
prospectively: boys identified as having very feminine development (gender dysphoria) have
a greatly elevated chance of identifying as gay or bisexual as adults (Green, 1987). These
studies suggest that sexual orientation has its developmental origins well before sexual
maturation occurs at puberty.
Sexual Differentiation of the Brain
The studies cited above show that homosexual and heterosexual males differ in both
their adult personality profiles and their recalled childhood behaviors, and that at least some
of these differences are in the same direction as those observed when comparing heterosexual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8
females to males. These observations have led to the proposal that variations in sexual
orientation may have a biological substrate in the sexual differentiation of the brain, which in
humans, as in other mammals, is largely controlled by sex hormones.
Early research in this area sought differences between adult heterosexual and
homosexual men in the levels of circulating sex hormones. Although there were some positive
reports in the early literature, these were traced subsequently to methodological artifacts.
Unfortunately, this was not realized until after certain physicians attempted to "treat"
homosexuality by injections of testosterone; the outcome was not a reverse to "normal' sexual
orientation, but instead, an increase in homosexual activity and desire (Hamer & Copeland,
1994; Tripp, 1987). The current consensus is that there is no significant correlation between
adult levels of circulating sex hormones (testosterone, estradiol) and sexual orientation
(Bancroft, 1994). The obseivation that adult males who have been castrated lose libido
(which can be restored by injected testosterone), but do not change sexual orientation,
suggests that the sex hormones play more of a role in the intensity than m the direction of
adult sexual interest (Tripp, 1987).
Tliese results do not, however, rule out the possibility of differences m sex hormone
synthesis during very early development, (i.e., prenatally or postnatally). This period in
dev elopment may represent a critical period for sexual differentiation of the brain (Ellis &
Ames, 1987). Because the hormonal enviromnent of human embryos and infants cannot be
manipulated experimentally, researchers have turned their attention to certain "accidents of
nature" in the form of genetically-determined metabolic disorders (LeVay, 1993). Genetic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9
(XY) males who completely lack androgen receptor activity due to an inherited mutation are
incapable of responding to testosterone both as fetuses and adults. As a result, they develop
the primary and secondary sexual characteristics of females. The observation that these
individuals usually develop a typical heterosexual female attraction to males has been taken
as evidence for the role of androgens in male psychosexual development; however, because
these indrviduals appear and are raised as females, it is difficult to separate purely biological
from social and cultmal influences.
Genetic males who lack the enzyme 5-alpha reductase, which converts testosterone
to the more potent analog 5-dehydro-testosterone, are bom with ambiguous genitalia and may
tlierefore be raised as girls. At puberty, the large surge of testosterone is sufficient to cause
descent of the testis and male genital development. Most of these individuals change their
gender role to male and develop sexual attractions to females. This has been cited as evidence
that the hormonal milieu of the brain is more important than the sex of rearing for adult sexual
orientation; however, this interpretation has been disputed since in some cases the parents
may have recognized that the child had ambiguous rather than tmly female genitalia (Wilson,
1995).
In females, the best studied condition is congenital adrenal hyperplasia, which leads
to increased prenatal androgen synthesis and varying extents of masculinization of the
external genitalia. Increased levels of childhood tomboyishness and of adult same-sex
attraction and behavior have been reported in these women, supporting the idea of a hormonal
influence. Taken together, these studies suggest that early hormonal events can have a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10
significant impact on adult psychosexual development; however, the interpretation of the
results is complicated by the rarity of the cases, the possibility of pleiotropic effects, and the
complex interaction between hormones, external genital appearance, and upbringing
(Bancroft, 1994; Ellis & Ames, 1987; McCormick & Witelson, 1991; Wilson, 1995).
A more direct approach has been to search for neuroanatomic differences related to
sex and sexual orientation. Swaab and Hofinan ( 1990) used vasopressin nemon staining to
show that the suprachiasmatic nucleus (SCN) was larger in gay men than in either
heterosexual men or women. Although, putatively, the SCN is primarily the biological clock
regulating hormone and physiological functions, it may also function in sexual behavior.
LeVay (1991) reported that the third interstitial nucleus of the anterior hypothalamus
(INAH3), a small group of cells in the hypothalamus, was larger in gay men and in women
than it was in a group of predominantly heterosexual men. This finding was of special interest
because the INAH3 lies in a region of the hypothalamus that has been implicated in sexual
behavior in rats and other non-human primates. Allen and Gorski (1991) foimd a difference
in the anterior commissme (AC), a band of coimecting fibers that was larger in gay men and
women than in heterosexual men. This finding is also of interest as Kolb and Whishaw ( 1985)
report that the AC has a large olfactory component which may serve some function in sexual
behavior, e.g. pheromones.
Cognitive and motor tasks also have revealed differences according to sexual
orientation. There are certain cognitive tasks on which men and women, as a group, differ on
average. In general, women excel at verbal tasks whereas men perform better on tasks that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11
deal with visuo-spatial abilities (LeVay, 1993; Maccoby & Jacklin, 1974; McCormick &
Witelson. 1991). For some (but not all) of these tasks there are also sexual orientation-related
differences, with gay men performing in the direction of heterosexual women (LeVay, 1993;
Sanders & Ross-Field, 1986; Willmott & Brierley. 1984). For example, McCormick &
Witelson (1991) claimed that they were able to distinguish between heterosexual men,
heterosexual women, and homosexual men using spatial ability and fluency tasks. On spatial
tasks, homosexual men scored lower than heterosexual men and higher than heterosexual
women. On word fluency tasks homosexual men scored higher than heterosexual men but
lower than heterosexual women. Other investigators, however, have found less consistent
results (LeVay, 1993). Recently, Hall and Kimura (1995) have found significant sexual
orientation-related differences in some sexually dimorphic motor tasks such as throwing-to-
target (performed more accurately by heterosexual men and lesbians than by homosexual men
and heterosexual females), but not in others such as the Purdue Pegboard (performed better
by women than by men but no sexual orientation difference). These results offer additional
support for the hypothesis of differences in the sexual differentiation of the brain between
heterosexual and homosexual individuals.
Genetic Factors
The role of heredity in individual differences in sexual orientation and its
psychological correlates have been investigated by both population and molecular genetic
approaches. Twin studies provide an especially powerful design for detecting the influence
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12
of genes on human behavioral and psychological traits (Chen, et al., 1990; Lykken et al.,
1993; Mulder, 1992). The basic premise behind twin research is that monozygotic (MZ) twins
share 100% of their genetic material whereas dizygotic (DZ) twins share about 50%. If
genetics play a role in the development of behavior or personality, then MZ twins should be
more similar to one another than are DZ twins. Fiuthermore, this should be true regardless
of whether the twins are raised together or apart. Over the past 40 years, there have been
several twin studies of sexual orientation in males. The largest and most systematic studies
are those of Bailey and Pillard (1993) and Whitam, Diamond, and Martin (1993). Both
studies foimd that the MZ twin of a gay man had a 40 to 65% chance of also identifying as
gay whereas the DZ twin of a gay man had a 25 to 30% chance of being homosexual. These
data produced a heritability index of approximately 50% for male sexual orientation. These
studies have been criticized because they did not include twins raised apart and because of the
possibility of volimteer bias. However, even when these possibilities were controlled for
statistically, heritability remained significantly greater than zero (Bailey and Pillard, 1993).
Family studies represent another approach for detecting the possible role of
inheritance in hmnan behavior. Hamer, Hu, Magnuson, Hu, and Pattatucci ( 1993) studied the
pedigrees of gay men and showed that they had more gay male relatives through matrilineal
lineages (maternal uncles and maternal cousins through an aimt), than through patrilineal
lineages (paternal uncles, paternal cousins, and maternal cousins through an uncle). This was
not due simply to differential knowledge of maternal compared to paternal relatives since the
gay men reported approximately equal numbers of lesbian relatives on both sides of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13
family. This pedigree pattem was interpreted to consistent with X-linked inheritance because
it showed the classical pattern of inheritance through the mother's side to males.
Molecular studies represent the most powerful and, potentially, the most informative
approach to studying the role of genes in complex behaviors and psychological processes.
Tlie aim of these studies is to identify individual genetic loci that are correlated to specific
behaviors and psychological traits. By isolating the genes and their products, and by
determining how they differ fi’om one person to another, it may be possible to deduce the
imderlying biochemical pathways that are partially responsible for individual differences in
human personality. An early example of such research is provided by the search for X-linked
genes that contribute to male sexual orientation. By DNA linkage analysis of gay brothers
from a selected population of families who showed the pattem consistent with a sex-linked
gene, Hamer et al. (1993) identified a small region of the X-chromosome (Xq28) that was
con elated with homosexual orientation. This region is thought to contain one or more genes
that increase the probability of predominant or exclusive same-sex orientation in at least some
men (Hamer & Copeland, 1994).
Twin and family studies have also been used to study the role of heredity in
personality traits. For example, Flodems-Myrhed, Pedersen, and Rasmuson (1980) used the
Eysenck Personality Inventory to show that correlated scores for MZ twins were twice those
of DZ twins for the personality traits of extroversion and neuroticism. Emde et al. ( 1992) and
Bouchard, Lykken, McGue, Segal, and Tellegen (1990) also found higher MZ than DZ
conelations for various measures of temperament, emotion, and cognition. Similarly, DeLeo,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14
Villa, and Capodieci ( 1992) found that MZ twins were more similar than DZ twins on a
variety of tests such as the Rorschach Test, the Lischer Test, the Cattell I6PF, The Bern Sex
Role Inventory, and the Symptom Distress Check List. The idea that heredity contributes to
personality traits has been further supported by studying twins raised apart. Bouchard et al.
(1990) compared 40 pairs of MZ twins raised apart to 48 pairs of MZ twins raised together.
Using the Multidimensional Personality Questionnaire and the California Personality
Inventory, they foimd substantial heritabUity for many dimensions of temperament, social
attitudes, and occupational and leisure-time activities. Plomin, Pedersen, McCleam,
Nesselroade, and Bergeman ( 1988) used a similar design to study the influence of genes and
environment on personality traits in the last half of the life span (mean age of 58.6 years).
They found significant genetic contributions to the temperaments of emotionality, activity
level, and sociability. By contrast, large differences in intratwin pair correlations dependent
on rearing stams were not foimd, indicating that shared environment has little effect on these
traits ( Plomin et al., 1988).
Genetic studies on tlie development of gender-related behaviors have been sparse.
Buhrich. Bailey, and Martin (1991) conducted a pilot twin study in Australia and fovmd
evidence for significant heritability of RCSTB. However, because the sample was small, the
conclusions were tentative. There have not been any molecular genetic studies of this aspect
ofbehavior.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15
In summary, there is increasing evidence that personality traits, sexual orientation, and
childhood sex-typed behaviors are aU influenced by heredity. There is little information,
however, on the extent and manner in which these characteristics are related to one another,
and whether they are influenced by distinct or related sets of genes. Thus the first goal of my
dissertation was to conduct a simultaneous analysis of personality traits, sexual orientation,
and sex-typed behaviors in a single group of gay and straight men. Also unknown is the
identity of the putative genes, how they differ flrom person to person, and how their products
interact with the development of the brain and its response to signals from the outside world.
Hence a second motivation for this work was to collect the subjects and data needed for an
integrated analysis of the role of genes in personality characteristics, sex-typed behaviors, and
sexual orientation. The data set that I collected for this dissertation has recently been used to
examine genetic links for both sexual orientation (Hu, Pattatucci, Patterson, Li, Fulker,
Chemey. Kmglyak, & Hamer, 1995) and the personality trait of Novelty Seeking (Benjamin,
Li. Patterson, Greenberg, Murphy, & Hamer, 1996).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2. METHOD
Subjects
Recruitment
Subjects were recruited through two protocols. Although both of these protocols
focus on pairs of brothers, which are required for DNA linkage analyses, this work treats
each subject as an independent datum.
Protocol I: Genetic Factors and Interrelationships for Sexual Orientation, Susceptibility to
HIV and Kaposi's Sarcoma, Alcoholism and Psychological Traits, and Histocompatibility
Antigens.
Pairs of homosexual brothers were sought for Protocol I. Other family members,
including heterosexual brothers, fathers, and mothers were also encouraged to participate.
Because this study was originally designed to evaluate the hypothesis of X-linked
transmission of predominant or exclusive male homosexuality (Hamer & Copeland, 1994;
Hamer et al., 1993), the following selection criteria were imposed;
1) Availability of DNA and interviews for two definitely homosexual brothers (or maternal half-brothers), 2) No indication of paternal transmission (i.e., no non-heterosexual fathers, sons, or paternal half brothers), 3) No more than one homosexual female, 4) No twins, 5) No bisexual or ambiguous sexual history.
Subjects were recruited through flyers to local and national homophile groups.
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17
advertisements in local and national gay and lesbian publications, and clinics at the
National Institutes of Health (NIH) and in Washington, D C. Interested parties were
instmcted to contact NIH. Upon telephone contact, the nature, procedure, and requirements
of the project were reviewed. Each participant received literatiue about the project and
mstmcdons to discuss it with participating members of their family. Appointments were made
to bring families to the NIH. On the day of the interview, the procedure and natiure of the
project were reviewed again. Participants signed NIH informed consent forms pre-approved
by The American University (A.U.) Institutional Review Board in lieu of a specific A.U.
consent form and were advised that they could withdraw from the study at any time without
negative consequences, loss of travel, or loss of monetary compensation. Participants were
interviewed separately, and given the 16 PF (Cattell et al., 1970), and NEO-PI-R, (Costa &
McCrae, 1991) to fill out. NIH phlebotomists performed venipunctiue.
Protocol II: Mapping Personality Traits to Chromosomes
Random pairs of brotlrers were sought for Protocol II through local university flyers,
newspapers, and on the NIH campus. The only criteria for participation in this part of the
study were availability of DNA samples and questionnaires for 2 brothers (or maternal half
brotiiers). No mention was made of sexual orientation. Participants either arranged to travel
to the NIH or investigators travelled to the university campus. After the procediue and nature
of the study were reviewed, the subjects were instmcted that they could withdraw at any time
without negative consequences or loss of compensation. Participants signed an informed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18
consent form. Blood was drawn on the premises and afterward, participants completed a
general questionnaire, the 16PF, and the NEO-PI-R. If time did not permit, they were given
a self-addressed, stamped envelope, and instructed to complete the forms and return them.
In cases where the brother was out of state, the consent form and the questionnaires were
provided with written instructions. Instructions were also given for the blood draw procedure.
Demographic Information
The 254 male participants were 91% Caucasian, 1% African-American, 2%
Latino/Hispanic, 0.4% Native American, and 5.5% Asian/Pacific Islander. Kinsey average
scores revealed that 140 participants described their sexual orientation as heterosexual
(Kinsey average score of 0.7, SD = .22) and 114 as homosexual (Kinsey average score of
5.7. SD = .44). Among the heterosexual participants, 14 subjects were from Protocol I and
126 subjects were from Protocol II. Among the homosexual participants, 112 were from
Protocol I and 2 were from Protocol I. As shown in table I, homosexual males were
significantly older, (t^,j,) = -9.10, p < .001) than the heterosexual males. The homosexual
males also had a significantly higher level of education level ( î^isd = -6.33, p < .001) than the
iieterosexual males. Therefore, all subsequent analyses were statistically adjusted for age,
age-squared, and education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19
TABLE I
Demographics
Homosexual Heterosexual n= 114 n= 140 Mean SD Mean SD t-vahie Age 36.8 8.3 26.2 10.2 -9.10** Education 17.7 2.3 15.8 2.3 -6.33** Income t 1.99 .4 1.87 .7 n.s. Kinsey Score 5.7 .44 0.7 .22 na
**p<001 t 1 = less than $ 15,000 per year 2 = $15,000 - $100,000 per year 3 = more than $100,00 per year
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3. PROCEDURE
Instruments
Questionnaire
A general questionnaire was administered to each participant (Hamer & Copeland,
1994; Hamer et al., 1993). The questionnaire covered demographics, recalled childhood sex-
typed behaviors, sexual development and orientation, brief medical and psychiatric history,
genetic screening, and tobacco, alcohol, and drug use.
Sexual Orientation Assessment
Sexual orientation was determined using questions developed by Kinsey, Pomeroy,
and Martin (1948). These questions asked about sexual self-identification, sexual attraction,
sexual behavior, and sexual fantasy. Responses were made using a 7-point Likert scale
ranging from 0, indicating exclusive heterosexuality or opposite sex attraction, to 6. indicating
exclusive homosexuality or same sex attraction. Kinsey average scores were computed using
the mean of the available scores. The internal consistency for the Kinsey questions for oiu
sample was Cronbach a = .99. For bivariate analyses, individuals with average Kinsey scores
of less than two were classified as heterosexual (n=l40) and those with scores greater than
four were classified as homosexuals (n=l 14).
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21
Recalled Childhood Sex-Typed Behavior Scale
The Recalled Childhood Sex-Typed Behavior (RCSTB) Scale was constructed on the
basis of previous research on childhood gender-related activities (Bailey & Zucker, 1994;
Evans, 1969; Freund, Nagler, Langevin, Zajac, & Steiner, 1974). Five questions were
employed and scored as follows:
1. As a young child (K through 7th grade, 6-12 yrs old), did you prefer to play with -1 boys +1 girls 0 both about equally 0 not to play with other children
2. As a child, did you like sports (i.e., basketball, baseball) -1 very much -.5 somewhat +.5 not very much +1 not at all
3. As a child, did you -1 prefer games like cops & robbers, soldiers, war, etc. +1 prefer games like dolls, cooking, sewing, etc. +.5 Both 0 neither 0 other games
4. As a child, did you think as yourself as "boyish" or "masculine" as the other boys your age? -1 more masculine than other boys -.5 about as masculine as other boys +1 less masculine than other boys +.5 don't remember
5. As a child, were you called/considered a "sissy" more than other boys your age? + l yes -1 no +.5 don't remember
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22
The scale score was the mean score for the 5 items. The five items were highly
intercorrelated, Cronbach a = .82.
Person alitv Assessment
Personality traits were assessed using two self-report, paper & pencil inventories: The
16PF, Version A and the NEO-PI-R. Both are supported by the literature to have high retest
reliability and longitudinal stability. The 16PF by Cattell consists of 187 items, each with 3
choices. The data were analyzed as described by Cattell to give 16 primary traits and 5
second-order factors (Table 2). The NEO-PI-R consists of 240 questions, each with 5
possible responses ranging fi’om strongly agree to strongly disagree. The data were scored
to give 5 personality factors (domains) each consisting of 6 traits (facets).
Factor Stnicture of Personality Inventories
The NEO-PI-R is based on the "Big 5" model of personality in which personality is
conceptualized as consisting of 5 major domains: Neuroticism, Extraversion, Openness,
Agreeableness, and Conscientiousness (Goldberg, 1981; McCrae & John. 1992). Each of
these domains is subdivided into 6 facets. The questions and scoring procedure are based on
extensive factor analytic research. To determine whether this factor model was appropriate
for the current population, a factor analysis with varimax rotation and 5 fixed factors was
performed (Table 3). This empirical analysis generated 5 factors that correspond almost
precisely to the 5 domains predicted by Costa and McCrae. The only exceptions were that the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23
facet of altruism (A3) was placed in the extraversion domain and the facet of excitement-
seeking (E5) was placed in the agreeableness domain. This reversal reflects the intrinsic
coiTelation of these two facets (Costa & McCrae, 1991). There is a certain amoimt of
extraversion necessary to be altruistic, and the domain of agreeableness contains certain
aspects of extraversion such as excitement-seeking.
The factor structure of the 16 PF is difficult to interpret and has been subject to
considerable debate (Howarth & Browne, 1971; Karson & O'Dell, 1974; SavHle & Blinkhom,
1981). While the second-order factors are derived fl-om the 16 primary factors, they are not
independent in that the primary factors contribute scores to more than one second-order
factor. While this may be somewhat confusing, Cattell defended his construction of the
second-order factors on tlie basis of factor analytic studies (Cattell & Krug, 1982). In an
attempt to replicate Cattell's structure, we factor analyzed the data for oim population and
anived at a similar distribution of the primary factors and second-order factors. Although otu"
analysis did not allow for multiple contributions to second-order factors by the same primary
factor, examination of the correlation matrix suggests that our data follow the overall scheme
advocated by Cattell (Table 3).
Analysis of the combined data for the 16 PF and the NEO-PI-R gave the results
shown in Table 3. These data demonstrate that there is a substantial overlap of the domains
measured by the 2 instruments. For example, the 16 PF second-order factor of'Anxiety' falls
in the same factor as the NEO-Pl-R domain of'Neuroticism', and the 16 PF second-order
factor of'Tough-Poise' is included in the NEO-PI-R domain of 'Openness'.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24
Data Analysis
The data was analyzed using SPSS-X statistical software. Because of the demographic
differences between the heterosexual and homosexual cohorts, all comparisons were adjusted
statistically for age, age-squared and education. Group comparisons of heterosexuals versus
homosexual were conducted by ANCOVA using age, age-squared and education as
covariates. Partial correlations were adjusted for the same demographic variables. Bonferroni
con certons for multiple comparisons were not used because of the nonindependence of the
dependent variables. Instead, a significance level of a < 0.005 was employed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4. RESULTS
We studied the relationship among sexual orientation, RCSTB, and personality traits
in a population o f254 adult males. A significance level of a = .005 was used for the primary
analyses due to the numerous correlations that were performed. Because the heterosexual and
homosexual group were not matched demographicaUy, aU comparisons were statistically
controlled for age, age-squared, and education; the age-squared correction compensated for
any non-linear effects of age.
Sexual Orientation vs. RCSTB
Gay males, on average, recalled lower levels of stereotypical male childhood
behaviors and more female-typical activities than did heterosexual men (Table 4). The
differences between the two groups were significant for each of the 5 individual items as well
as for the overall scale. Tire gay group also exhibited a greater diversity of behaviors as
indicated by the consistently larger standard deviations. The correlation between average
Kinsey score and the RCSTB scale score was r(247) = .58, p < .001 after correcting for age,
age-squared, and education.
Figure 1 shows a histogram of RCSTB scale scores for the heterosexual and
homosexual participants. The heterosexual men showed a relatively homogeneous distribution
of scores with a large peak close to the most male-typical score and a smaller tail that barely
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26
extended into the female-typical region of the graph. The homosexual men
showed a much broader distribution of scores with no clean peak and approximately equal
numbers of individuals in the male-typical and female-typical areas of the graph.
Sexual Orientation vs. 16 PF Personality Factor Traits
The homosexual and heterosexual subjects differed on several personality factors
measured by the 16 PF. Males who had high Kinsey scale scores reported being more Warm
(A) r(248) = .189, p < .005, Tender-Minded (I) r(248) = .338, p < .001, and Tense (Q4)
r(248) = .255, p < .001 than males with low Kinsey scores. The homosexual males also had
elevated scores on the second order factor of Tough Poise r(248) = -.337, p < .001 (Table
5).
Sexual Orientation vs. NEO-Pl-R Personality Traits
Differences between the homosexual and heterosexual subjects were also found using
the NEO-PI-R Questionnaire. High Kinsey scores correlated positively with the Openness
domain, r(248) = .247, p < .001 including 4 of its facets: Aesthetics (02) r(248) = .231, p <
.001. Feelings (03) r(248) = .243, p < .001, Actions (04) r(248) =. 177, p < .005, and Values
(06) r(248) = .289, p < .001. There was also a correlation between Kinsey scores and
Neuroticism, including Anxiety, but these were not significant at a = .005 (Table 6).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27
RCSTB vs. Personality Traits
The relationship between RCSTBs and personality traits was studied both in the entire
population and in the separate homosexual and heterosexual groups. Because of the reduction
in sample size, the secondary analysis in the separate groups was conducted at a significance
level of p < .02. Within the entire population RCSTBs correlated with a wide variety of 16
PF and NEO-PI-R factors: Affected by Feelings (C) r(247) = -.194, p < .005, Tender-
Minded (I) r(247) = .26, p < .001, Apprehensive (0) r(247) = .21, p < .001, Undisciplined
Self-Conflict (Q3) r(247) = -. 185, p < .005, Tense (Q4) r(247) = .283, p < .001, Anxiety (16
PF) r(247) = .285, p < .001, Tough Poise r(247) = -.23, p < .001, Neuroticism (N) r(247) =
.258, p < .001, Anxiety (NEO-PI-R) (Nl) r(247) = .22, p < .001, Depression (N3) [(247) =
.223, p < .001, Self Consciousness (N4) r(247) = .231, p<.001. Vulnerability (N6)
i(247)=.23, p< .001, Openness (O) r(247)=.252, p < .001, Fantasy (01) r(247) = .223, p <
.001, Aesthetics (02) r(247) = .234, p < .001, Feelings (03) r(247) = .206, p < .001, and
Values (06) r(247) = .264, p < .001 (Tables 7 & 8).
Among heterosexual men, RCSTBs were also correlated with many different
personality factors: Emotionally Stable (C) r( 135) = -.245, p < .005, Shy (H) r( 135) = -. 195,
p < .02, Tender-Minded (I) r( 135) = .262, p < .005, Self Assured (O) r(l35) = .212, p < .02,
Self-Sufficient (Q2) r(135) = .345, p < .001, Undisciplined Self-Conflict (Q3) r( 135) = -.343,
p < .001, Extroversion r(135) = -.244, p < .005, Anxiety (16 PF) r(135) = .263, p < .005,
Tougli Poise r( 135) = -.213, p < .02, Low Control r( 135) = -.282, p < .001; Depression (N3)
it 135) = .203, p < .02, Vulnerability (N6) r(135) = .206, p < .02, Extraversion (E) r(135) =
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28
-.238, p < .005, Gregariousness (E2) i(135) = -.212, p < .02, Assertiveness (E3) r(135) = -
.28, p < .001, Activity (E4) r(135) = -.22, p < .02, Openness (O) r(135) = .295, p < .001,
Fantasy (01) r(135) = .286, p < .001, Aesthetics (02) r(135) = .248, p < .005, Ideas (05)
r( I35) = .24, p < .005, Values (06) r(l35) = .207, p < .02, Conscientiousness (C) r(l35) =
-.367. p < .001, Competence (Cl) r(135) = -.353, p < .001, Dutifulness (C3) r(135) = -.30,
P < .001, Achievement Striving (C4) r(135) = -.38, p < .001, Self-Discipline (C5) r(135) =
-.325. p < .001, and Deliberation (C6) r(135) = -.215, p < .02.
By contrast, RCSTBs in homosexual men were significantly correlated to a more
restricted set of personality traits: Apprehensive (0) r(I07) = .221, p < .02, Tense (Q4) r( 107)
= .28. p < .005 and Anxiety ( 16 PF) r( 107) = .282, p < .005; Neuroticism r( 107) = .25, p <
.02. Anxiety (Nl, NEO-PI-R) r( 107) = .24, p < .02, Self-Consciousness (N4) r( 107) = .229,
p < .02. and Vulnerability (N6) r( 107) = .223, p < .02 (Table 7 & 8).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5. DISCUSSION
We assessed sexual orientation, RCSTBs, and personality traits in a population of 254
adult males. This appears to be the first time that these three dimensions of psychological
development have been measured in a single cohort.
Sexual Orientation and RCSTB
The strongest and most consistent correlate of adult sexual orientation was recalled
ciiildhood sex-typed behavior. As expected from previous research (Bailey & Zucker, 1994;
HaiTy, 1983; Whitman, 1977: Whitam & Mathy 1986), gay men recalled significantly fewer
male-typical and more female-typical behaviors than did their heterosexual counterparts. The
effect size (D = 1.4) was similar to that observed in other studies (D = .50 to 2.09, mean =
1.19. SD = .43; Bailey & Zucker, 1994) and the correlation of r(252) = .58, p < .001 between
tile Kinsey scale and the RCSTB scale indicates that more than one quarter of the variability
ill adult sexual orientation can be predicted from childhood activities and attitudes. Despite
tlie strength of the correlation between sexual orientation and RCSTBs, it is important to
recognize that there was significant overlap between the broad distribution of RCSTB scores
displayed by the gay men and the narrower distribution observed for the heterosexual men
(Figure 1). For example, wliile some gay men recalled playing female typical games with girls,
others engaged in solitary activities, fantasy, or rough and tumble games with boys. The
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30
heterosexual men recalled a less diverse repertoire of masculine and feminine childhood play,
but it is not clear whether this reflects a genuine difference between the groups or
compression at one end of the scale.
Our results are similar to those obtained in a substantial number of retrospective
studies comparing RCSTBs in adult homosexual and heterosexual males (Bailey & Zucker,
1994). Despite the consistency of the findings across studies, the validity of the correlation
between RCSTBs and adult sexual orientation has been questioned on several groimds. The
most common criticism concerns the possibility of selective recall. It has been hypothesized
that gay men may overreport, or heterosexual men may underreport, sex-atypical behaviors.
However, Bailey and Zucker (1994) found significant correlations between observer-reports
by mothers and self-reports by their homosexual sons concerning childhood sex-typed
behaviors. In addition, observational studies of unselected samples of boys, most of whom
w ill develop into heterosexual men, show that very few of them engage in female-typical
behaviors such as playing with girls. Therefore, the observed correlation of adult sexual
orientation to recalled childhood behaviors cannot be attributed solely to underreporting by
the heterosexual subjects since there was apparently little to report. Perhaps most important,
prospective studies of boys who display marked patterns of cross-gender behavior have
shown that they have a greatly increased probability of developing a homosexual orientation
in adolescence or adulthood (Green, 1985; Green, 1987). In sum, there is no empirical
ev idence to support the selective recall interpretation for the observed coiTelation between
adult sexual orientation and RCSTBs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31
A second potential criticism is that the correlation between sexual orientation and
childhood behavior may be culturally based rather than intrinsic. Specifically, it has been
iivpothesized that in cultures in which homosexuality is stigmatized, only the most feminine
men are likely to reveal their sexual orientation, thereby biasing the gay sample (Baüey &
Zucker, 1994; Whitman & Mathy, 1986). However, Whitam and Mathy ( 1986) found similar
levels of recalled childhood sex-atypical behaviors in gay men from four different societies
(United States, Brazil, Guatemala, and the Philippines) with varying attitudes toward
homosexuality. This suggests that the correlation between adult sexual orientation and
RCSTBs is not simply the product of cultural bias.
Lastly, we note that the correlation between adult sexual orientation and RCSTBs has
been obseived using several different approaches and instruments. For example, Freund et
al. ( 1974) demonstrated a difference between homosexual and heterosexual men using a
Feminine Gender Identity Scale that contained only female-specific items and was originally
designed to differentiate tianssexual and heterosexual males. By contrast, Hockenberry and
Billingliam (1987) were able to disciiminate significantly between gay and heterosexual men
using a scale that consisted predominantly of male-specific questions. Grelleit, Newcomb, and
Bentler ( 1982) asked homosexual men open-ended, non-directive questions about the types
of activities they participated in during specific time periods of their childhood. Homosexual
men recalled participating in fewer masculine sex-typed behaviors and more feminine sex-
typed behaviors as children compared to the heterosexual sample. The gay men were also
more likely to have played in mixed groups or with girls than were the heterosexual subjects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32
These similar results across studies and instruments indicate that the observed correlation
between adult sexual orientation and recalled childhood behaviors is not a measuring artifact.
The correlation between adult sexual orientation and RCSTBs raises two important
developmental questions. First, are pre-homosexual boys more feminine or less masculine
than pre-heterosexual boys? Our five item RCSTB scale was designed to measure the general
lev el of sex-typed behaviors rather than to distinguish between male-typical and female-
typical behaviors, and therefore we could not address this point definitively . It was
noteworthy, however, that the overall scale score, which combines male-typical and female-
tvpical responses, was more predictive of sexual orientation than any of the individual items.
Hockenberry and BiUingham ( 1987), who developed a Boyhood Gender Conformity Scale
with separate items for male-typical and female-typical behaviors, concluded that the absence
of masculine patterns was a more powerfiil predictor of adult homosexuality than feminine
or cross-sexed behaviors.
A second question concerns the temporal and etiological connection between
RCSTBs and adult sexual identification. Is childhood behavior the cause or result of a
person's sexual direction, or are both characteristics influenced by a common variable? Sex-
typed behaviors are evident by as young as two years of age whereas sexual orientation
becomes more salient in adolescence and young adulthood. Green (1974, 1987) sees gender
identity as a precursor to gender role, which in turn leads to adult sexual orientation. Isay
( 1989), however, believes that sexual orientation develops early enough to influence sex-
typed behaviors. For example, he has suggested that pre-homosexual boys adopt the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. behaviors of their mothers in order to sexually attract their fathers. Finally, Zucker (1990)
posits that both sets of behaviors originate from a common source but are expressed at
different stages because of a requirement for other developmental events such as puberty.
Unfortunately, these possibilities cannot be distinguished by retrospective and correlational
studies alone.
Sexual Orientation and Personality
Personality traits were assessed with two inventories, the NEO-Pl-R and the 16 PF,
that are based on distinct yet complementary perspectives of the underlying structure of
human temperament. We used both instruments in the hope of identifying the full range of
personality characteristics that differ between heterosexual and homosexual men. This
complementary analysis revealed sexual orientation-related differences in three broadly
defined domains of personality: anxiety, tender-minded, and openness.
Tlie main contributor to the anxiety domain is the 16 PF primaiy factor of Tenseness
(04). An example of a statement from the 16 PF Tenseness factor is. " I am always able to
keep the expression of my feelings under exact control." There was also a trend for gay men
to score higher than heterosexual men on the NEO-PI-R facet of Anxiety (Nl), which
includes statements such as, "1 am not a worrier." Factor analysis (Table 3) showed that Q4
and Nl are included in the same cluster of personality traits, supporting the validity of
considering these two different measures as related indices of anxiety. Homosexual men
scored higher than heterosexual men on both the Q4 and Nl scales, which reflects elevated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 34
levels of tension, apprehension, fear, worry, and frustration and decreased feelings of calm,
relaxation, and tranquility. Evans (1970) and Weinrich, Atkinson, McCutchan, Grant, and The
HIV Neurobehavioral Research Center ( 1995) have also found evidence for increased levels
of anxiety and tension in homosexual men.
We can consider three possible explanations for the increased anxiety levels of the
homosexual men. First, it could reflect the high prevalence of FHV infection and AIDS in the
gay community. However, in our data correlations between homosexuality and anxiety
remained highly significant (p < .001) even when the HIV seropositive subjects were removed
fr om the analysis. Also, Weinrich et al. (1995) has reviewed the evidence that increased levels
of anxiety and depression in gay men predate the AIDS epidemic. A second possibility is that
anxiety is a natural reaction to the social stigmatization of homosexuality in American culture.
Gay men face a variety of unique adaptive challenges in a society that is largely disapproving
of their sexuality and lifestyle, and thus increased anxiety could represent a purely reactive
phenomenon. A third model is based on the observation that anxiety, like depression, displays
a significant sex difference; in the 16 PF normative population, women score 0.47 standard
deviations liiglier than men on Tenseness, and in the NEO-PI-R normative sample, women
score 0.41 standard deviations higher than men on Anxiety. According to this model, the
increased anxiety of gay men could reflect, in part, an underlying feminization of their
personality style. It should be emphasized, however, that this model does not necessarily
imply a biological relationship since the underlying cause of increased anxiety in females is
still not understood.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35
Tlie tender-minded domain is defined by a combination of the 16 PF primary factor
of Tender-Minded (I), the 16 PF second-order factor of Tough Poise, and the NEO-PI-R
facet of Aesthetics (02). Gay men, who score high on these scales, appreciate art, beauty, and
culture and can be characterized as sensitive and refined. Heterosexual men. on average, are
more interested in practical, down to earth matters and tend to be rough and realistic. For
example, one of the questions for the 16 PF Tender-Minded factor is, "I admire the beauty
of a poem more than that of a weU-made gun."
The origins of tlie correlation between tender-minded and homosexual orientation,
which appears to be in accord with stereotypical views of gay men, are unknown. A socially-
based explanation is that gay men gravitate towards fields, such as fine aits and theatre, that
are accepting of diverse lifestyles. An alternative model is that tender-mindedness reflects a
partial feminization of personality style. In this regard, it is noteworthy that 16 PF-Tender-
minded shows the largest sex difference ( 1.28 standard deviations) of any of the personality
factors measured by the 16 PF and NEO-PI-R questionnaires. The underlying source of this
substantial sex difference is unknown.
Tlie openness domain compiises three facets of the NEO-PI-R Openness Factor (O)
- Feelings (03), Actions (04), and Values (06) — together with the 16 PF primary factor
of Warmth (A)'. An example of a question from the Openness Factor is, "I experience a wide
range of emotions or feelings." Homosexual subjects scored higher than
' Note that 16 PF Waimth is different than NEO-PI-R Warmth (a facet of NEO-PI-R Extiav ersion), and correlates 0.23 to 03, 0.23 to 04, and 0.12 to 06.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36
heterosexual subjects on each of these correlated personality traits. Openness implies an
awareness of one's environment, both internally and externally. Internally, individuals are
responsive to their thoughts and feelings. There is a flexible nature that lends itself to re
examination of social, political, and religious values. This aspect of openness is similar to
"social nonconformity," which Duckitt and DuToit ( 1988) found to be elevated in gay men.
Externally, individuals are receptive to novel and varied experiences. This openness extends
to relating to others.
Tlie increased opeimess of gay men could be either a cause or effect of their sexual
orientation. One possibility is that men who are in touch with their feelings are likely to
recognize and accept emotional or sexual attractions to other males. By contrast, emotionally
closed individuals may ignore or fail to value such feelings even if they arise. An alternative
possibility is that gay men develop an attitude of openness toward unconventional beliefs
because their sexual orientation runs contrary to accepted social, political, and religous
tenets. Heterosexual men miglit be less motivated to reexamine their beliefs in conventional
values that reafiflrm their own lifestyle. In contrast to anxiety and tender-mindedness, the
openness factors do not show large sex différences, and therefore the conelation between
openness and homosexuality is tmlikely to be the result of a feminized psychological profile.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37
RCSTBs and Personality
Tlie above analyses showed that homosexual and heterosexual men differ on average
in tiieir RCSTBs and adult personality traits, and that in both cases there is a tendency of the
gay men to show increased levels of female-typical characteristics. It was therefore
Inpothesized that personality scores would correlate with RCSTB scores in the same way as
with Kinsey scale measurements. That is, it was predicted that childhood femininity is a
precursor of adult female-like personality characteristics. Surprisingly, in heterosexual men
this hypothesis was at least partially supported whereas in homosexual individuals a different
pattern emerged.
Among heterosexual men, RCSTB was associated with a broad range of personality
factors. In certain respects, heterosexuals who recalled female-typical childhood activities
were similar to gay men. Tliey were tender-minded (I), aesthetic (02) and open to experience
(O) and values (06); they were not, however, tense (Q4) or anxious (N1 ). In other respects,
heterosexual men with higli RCSTB scores showed a distinct personality profile characterized
by low levels of extraversion (E), conscientiousness (C), and self-discipline (C6) and a high
degree of self-sufficiency (Q2).
The unexpected finding was that RCSTB scores in homosexual men were exclusively
associated with various facets of anxiety. Gay men who recalled paiticularly female-typical
childhood behaviors scored liigli on Neuroticism (N), anxiety (Nl, 16 PF-Anxiety), self-
consciousness (N4), vulnerability (N6), apprehension (O), and tenseness (Q4). They did not,
however, show increased levels of the various facets comprising tender-mindedness or
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38
openness.
These results show that childhood femininity is associated with different adult
personality profiles in heterosexual and homosexual men. Among heterosexuals, above
av erage childhood femininity is associated with certain adult female-typical traits such as
tender-mindedness and appreciation of aesthetics. However, these characteristics do not
cause anxiety, perhaps because these individuals have developed a higli degree of self-
sufficiency. Among homosexuals, who as a group have high baseline RCSTB scores, a
particularly feminine childliood is most strongly associated with anxiety, tenseness and
apprehension. This caimot simply be attributed to increased visible signs of adult femininity
since there were no significant correlations between RCSTB and female-typical personality
measures such as tender-mindedness and ascetics.
One interpretation of this data is that childhood sex-atypicality is a functionally
heterogeneous characteristic on which heterosexual and homosexual men differ qualitatively
as well as quantitatively. Specifically, I propose that there are at least two distinct pathways
for the development of female-typical behaviors in children, tlie first of which is found in both
pre-heterosexual and pre-homosexual boys and the second of which is specific for pre
homosexual boys. Althougli this model is admittedly speculative, it would nicely account for
two obscivations. First, it was shown that there was more variation in RCSTB scores in gay
men than in heterosexual men, which would be expected if there were two ratlier than one
ttnderlying factors. Second, it was shown that a fraction of the gay men had higher RCSTB
scores than any of the heterosexual men; perhaps these are the individuals in which both
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39
pathways are operative. Wliile further research is required to rigorously evaluate this
h\])othesis. it may eventually prove to be useful in experimentally testing the idea that there
are multiple pathways to sexual orientation.
Limitations
The main limitation of this study is the use of a convenience sample rather than a
population-based sample. This was necessary because the protocol required collecting
information from family members, who by definition are not independent of one another.
Fuithennore. because of the sensitive nature of the research, it was preferable to focus on
subjects who were willing to discuss their personal life and who felt that their relatives would
be similarly inclined. This generated a subject population typical for research projects without
direct benefits or remimeration to the participants; namely, a predominantly white, highly
educated, and financially secure group of males. Accordingly, the study results cannot
necessarily be applied to the population at large, or to all heterosexual or homosexual men.
A second limitation, common to all studies of sexual orientation, arises fi om the social
stigmatization of homosexuality. Because sexual orientation was assessed solely by self-
report, the homosexual population was necessarily selected for a willingness to disclose their
sexuality, which is neitlier a neutral nor an obvious trait. Being openly gay is more than just
an attraction to men; it also requires that a person self-identify as part of a sexual minority.
Accordingly, it is difficult to disciiminate between those psychological characteristics that are
associated with the direction of a person's sexual attraction as compared to those that are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40
associated with openness about sexuality. However, statistical analyses showed that the
associations between sexual oiientation, RCSTB, anxiety and tender-mindedness remained
significant even when corrected for opetmess personality scores (not shown). Tlierefore, it
seems unlikely that the differences tliat were identified between heterosexual and homosexual
men for these psychological characteristics were due solely to differences in opetmess.
A third limitation was that the groups were not matched for demographic variables
such as age and education. This was controlled for by statistical techniques and by using
consei\ ati\ e P values. In addition, tlie gay subjects were selected on the basis of criteria used
to study X-liiiked sexual orientation (Hamer & Copeland, 1994; Hamer et al., 1993; Hu et
al., in press), and therefore represent a selected subpopulation of homosexual males. This was
to allow the current results to be integrated with genetic mapping experiments, which are
cunently in progress.
Prospects
riiis study has shown that there are a complex series of interactions between sexual
orientation, RCSTBs, and adult personality traits. The challenge for the future is to determine
the de\ elopmental pathways that lead to these intercoirelations. One approach would be to
conduct longitudinal studies. It would be interesting, for example, to determine whether gay
men show increased levels of anxiety prior, simultaneously, or subsequent to acknowledging
their sexual orientation. Such data would be useful in determining whether anxiety is
associated with homosexuality tlirough a common underlying pathway, such as a biologically
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 41
mediated sex difference, or as the result of social stigmatization. An alternative approach
would be to search for innate factors that mediate the interactions between the measured
variables. For example, if it were discovered that a particular gene was involved in the
neurochemist ly of both anxiety and sexual attraction, the underlying pathway would become
readily amenable to study. Because the protocol allows genetic as well as psychological
analyses, such ideas should be testable in the near future.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX A
Tables
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 43
TABLE 2
CatteU’s 16 Personality Factor Inventory
Primary Factors Personality Component
Cool v5. Warm Reserved, Impersonal, Outgoing, Kindly, Easygoing, Detached, Formal, Aloof Participating, Likes People
B Concrete- Thinking V5. Abstract-Thinking Less Intelligent More Intelligent, Bright
Affected by Feelings vs. Emotionally Stable Emotionally Less Stable, Mature, Faces Reality, Calm Easily Annoyed
Submissive vs. Dominant Humble, Mild, Easily Led, Assertive, Aggressive, Accommodating Stubborn, Competitive, Bossy
Sober V5. Enthusiastic Restrained, Prudent, Spontaneous, Heedless, Taciturn, Serious Expressive, Cheerful
Expedient V5. Conscientious Disregards Rules, Conforming, Moralistic, Self-Indulgent Staid, Rule-bound
H Shy vs. Bold Threat-Sensitive, Timid, Venturesome, Uninhibited, Hesitant, Intimidated Can Take Stress
Tough-Minded vs. Tender-Minded Self-Reliant, No-nonsense, Sensitive, Overprotected, Rough, Realistic Intuitive, Refined
Trusting V5. Suspicious Accepting Conditions, Hard to fool. Distrustful, Easy to get on with Skeptical
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 44
Primary Factors Personality Component (con’t)
M Practical vs. Imaginative Steady, Concerned with Absentminded, Impractical, ’Down to Earth Issues’ Absorbed in thought
N Forthright V5. Shrewd Unpretentious, Open, Polished, Socially Aware, Genuine, Artless Diplomatic, Calculating
O Self-Assured V 5 . Apprehensive Secure, Feels Free of Guilt, Self-blaming, Guilt-prone, Untroubled, Self-satisfied Insecure, Worrying
Qi Conservative vs. Experimenting Respecting Traditional Liberal, Critical, Open to Ideas Change
Q2 Group-oriented vs. Self-sufficient A ’Joiner’ and Sound Follower, Resourceful, Prefers Own Listens to Others Decisions
Q3 Undisciplined Self-Conflict V5. Following Self-Image Lax, Careless of Social Rules Socially Precise, Compulsive
Q4 Relaxed V5. Tense Tranquil, Composed, Overwrought, Frustrated, Has Low Drive, Unfrustrated Has High Drive
Second-Order Factors Personality Component Extraversion Introversion vs. Extraversion Anxiety Low Anxiety vs. High Anxiety Tough Poise Emotional Sensitivity vs. Tough Poise Independence Subdueness vs. Independence Superego/Control Low Control vs. High Control
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45
TABLE 3
Factor Loadings for the NEO-PI-R and the 16PF '
Factor 1 Factor 2 Factor 3 Factor 4 Factor 5
NEUROTICISM / ANXIETY N1 -.07792 .82717 -.09150 .07288 -.00011 N2 -.06877 .61427 -.15306 -.39788 -.19187 N3 -.18579 .76081 -.26376 .08729 .10710 N4 -.23732 .76093 -.10693 .10344 -.04383 N5 .21220 .49188 -.41940 -.20346 .11401 N6 -.17536 .59650 -.46995 .22110 .01410
C .07238 -.65326 .07611 .00385 -.07257 0 -.05813 .69563 -.21199 .08222 -.01796 Q4 -.04334 .75789 -.03046 -.02837 .02669
EXTRAVERSION / EXTRA VERSION El .81980 -.08213 .03440 .16938 .19504 E2 .76687 -.11680 -.03733 -.17682 -.05454 E3 .46801 -.20532 .33066 -.45961 .15717 E4 .44329 -.14719 .40423 -.39400 .14142 E6 .65526 -.15147 .02617 -.02604 .38091 A3 .56029 -.04935 .05650 .45182 .08638
A .62473 .01033 .05133 -.03427 .07630 F .75747 -.11649 -.19990 -.26594 .02573 H .63935 -.33369 .04267 -.33361 .17217 Q2 -.65215 .02539 .08101 .09968 .21207
OPENNESS / TOUGH POISE 01 .17298 .16431 -.32141 -.13020 ,57609 02 .10156 .16064 .01420 -.01499 .79402 03 .37221 .28079 -.03098 -.08418 .64983 04 .29600 -.09805 -.14941 -.11758 .53957 05 -.08558 -.15679 -.02841 -.28421 .57480 06 .08365 .03417 -.05014 .10074 .70842
I .03134 .20374 -.05524 .24592 .60347 M -.10615 -.26611 -.01206 .12924 .50630 B -.17525 -.12115 -.05978 -.12050 .34300
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 46
AGREEABLENESS / INDEPENDENCE A1 .37431 -.25132 .19097 .51663 .23605 A2 -.03891 -.07808 .10385 .68233 .02206 A4 -.02341 -.13348 .02607 .76080 .20446 A5 -.16838 .18472 -.06659 .64735 -.09958 A6 .29466 .06976 .00281 .44680 .41161 E5 .40079 -.08592 -.16729 -.47510 -.15056
E .36310 -.10723 .07888 -.63490 .23788 L .22047 .29541 -.13964 -.52669 .03655 N -.17732 .24939 .20566 .44510 -.12116 Qi -.06840 -.13389 -.17481 -.54872 .23554
CONSCIENTIOUSNESS / SUPEREGO-CONTROL Cl .16259 -.32856 .71274 -.06073 .04590 C2 -.06582 .01513 .75043 .08312 .00500 C3 -.03080 -.14617 .73177 .16655 -.03471 C4 .20766 -.06264 .76751 -.21823 -.01572 C5 .04736 -.32960 .75616 -.00537 -.10247 C6 -.18272 -.11853 .65828 .25146 -.05745
G -.05663 .02000 .60489 .24172 -.26168 Q3 -.09001 -.22870 .59833 .12747 -.20300
NEO-PI-R Factors are listed Orst, followed by 16PF Factors.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 47
TABLE 4
Correlations Between Sexual Orientation and
Recalled Childhood Sex-Type Behaviors (RCSTB)
Heterosexual Homosexual R vs. Kinsey Mean Scores t Mean Scores t F Value^ Score- (n=140) (a =112)
Individual Items Play -.64 (.51) -.07 (.67) 34.6*=*= .36** Sports -.69 (.55) .38 (.76) 81.3*=* .49** Games -.97 (.22) -.02 (.78) 117.5** .58** Masculine -.47 (.57) .23 (.75) 48.9** .40** Sissy -.85 (.52) -.08 (.98) 33.6** .34**
RCSTB Scale Score -.72 (.26) .09 (.55) 127.6* .58*
‘ F value for One-way ANOVA of heterosexual group versus homosexual group. ■ Pearson Correlation coeffident for average Kinsey score versus individual and total RCSTB scores, t Mean Score (Standard Deviation) * p < .005 *♦ p<.001
Note. All scores have been corrected for age, age squared, and education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48
TABLE 5
Correlations Between Sexual Orientation and 16 PF Personality Factors
Heterosexual Homosexual Mean Score t Mean Score t F value' R vs. (n=140) (n=Il3) Kinsey Score* Primary Factors .\-Wannth 4.5 (2.0) 5.2 (2.2) 10.0** .189* B-Abstiact Thinking 7.6 (1.7) 7.8 (1.6) .061 -.027 C-Eniotional Stability 5.2 (1.9) 4.8 (1.9) .625 -.063 E-Dominant 6.7 (2.0) 7.0 (2.1) 1.09 .058 F-Enthusiatic 6.7 (2.3) 6.2 (2.3) .077 .002 G-Conscientious 4.4 (2.0) 4.7 (1.7) 2.3 -.107 n-Bold 5.1 (2 2) 5.6 (2.4) 1.5 .054 l-T ender-Minded 5.9 (2.1) 7.5 (1.9) 30.1** .34** L-Suspicious 6.6 (2.0) 6.3 (2.0) .28 .029 .M-Imaginative 5.2 (1-9) 6.1 (1.7) 5.2 .133 N-Shrewd 5.0 (1.8) 5.5 (2.0) 2.9 .118 O-Apprehensive 5.5 (1.9) 5.7 (1.8) 1.75 .094 QI-Experimenting 6.4 (1.8) 6.1 (2.0) .24 -.017 Q2-Self-Sufficient 6.7 (2.0) 7.3 (1.8) .66 .06 Q3-Following Self-Image 4.6 (1.8) 4.6 (1.6) .72 -.045 Q4-Tense 6.3 (1.8) 7.0 (1.7) 16.3** .256**
Second-Order Factors EXTRAVERSION 5.0 (2.4) 5.0 (2.3) 1.14 .05 ANXIETY 6.1 (1.7) 6.5 (1.7) 4.4 .145 TOUGH POISE 5.7 (1.9) 4.0 (1.7) 30.4** -.337** INDEPENDENCE 6.8 (2.3) 6.9 (2.3) .63 .039 S U P EREGO/CONTROL 4.3 (2.0) 4.5 (1.6) 2.1 -.095
' F value for one-way ANOVA of hetersexual group versus homosexual group. - Pearson correlation coefBcient for average Kinsey score versus personality test score, t Mean Score (Standard Deviation) * p- 005 ** p- 001 Note All scores have been corrected for age. age squared, and education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49
TABLE 6
Correlations Between Sexual Orientation and NEO-PI-R Personality Factors
Heterosexual Homosexual R vs. Kinsey FACTORS Mean Scores t Mean Scores t F Value^ Score* (N=140) (N=113) N - Neuroticism 83.0 (24) 84.3 (22.4) 5.5 .161 Nl: Anxiety 14.3 (5.5) 15.0 (5.5) 3.5 .13 N2: Angry Hostility 13.7 (5.3) 12.7 (4.6) .21 .032 N3: Depression 13.0 (6.2) 13.5 (5.8) 5.0 .157 N4: Selt-Consciousness 14.5 (5.0) 15.3 (4.8) 5.6 .164 N5: Impulsiveness 17.8 (5.0) 18.0 (4.3) 3.6 .123 N6: Vulnerability 9.7 (4.4) 9.8 (4.0) 3.2 .127 E - Extraversion 114.3 (20.7) 116.2 (20.5) .72 .043 El: Warmth 21.6(4.2) 23.0(4.2) 1.4 .067 E2: Gregariousness 17.5 (5.7) 17.6 (5.7) .34 .02 E3: Assertiveness 16.1 (4.8) 17.5 (4.8) 2.6 .099 E4: Activity 18.1 (4.5) 19.0 (4.4) .14 .021 E5: Excitement Seeking 20.7 (4.4) 17.6 (4.4) 3.2 -.117 E6: Positive Emotions 20.3 (5.3) 21.5 (5.3) 1.7 .071 O - Openness to Experience 119.1 (24.1) 129.4 (17.2) 16.8** .247' ** o f: Fantasy 20.3 (5.4) 21.1 (4.8) 6.5 .154 02: Aesthetics 18.0(6.6) 21.1(5.4) 13.7**13.7** .231**.231' 03: Feelings 21.6(4.7) 2 3(4.2 . 6 i6<**16.5** .243' i^^** 04: Actions 16.0 (4.0) 17.6 (3.9) 8.2* .177' 05: Ideas 21.2 (6.0) 20.2 (5.6) .09 .012 06: Values 22.1 (5.0) 25.7 (2.7) 23.0** .289 A - Agreeableness 109.4 (19.1) 118.6 (15.7) .82 .063 A l: Trust 18.5 (5.3) 21.6 (4.3) 3.9 .115 A2: Straightforwardness 17.4 (4.8) 18.3 (4.9) .69 -.048 A3: Altruism 22.7 (3.5) 23.4 (3.7) .18 -.03 A4: Compliance 16.3 (4.6) 18.4 (4.1) 1.5 .085 A5: Modesty 15.8(5.1) 16.4(4.6) .34 .05 A6: Tender-Minded 18.8 (4.6) 20.6 (3.6) .70 .06 C - Conscientiousness 110.8 (22.7) 118.9 (19.4) .014 .009 Cl: Competence 22.0 (4.0) 23.4 (3.5) .003 -.009 C2: Order 16.0 (5.2) 18.5 (4.8) 2.6 .111 C3: Dutifulness 20.7 (4.6) 22.1 (3.7) .024 -.014 C4: AchievementStriving 17.8 (5.1) 19.3 (5.2) .5 .045 C5: Self-Discipline 18.5 (5.5) 19.4 (5.0) .49 -.037 C6: Deliberation ______15.9 (4.4)______16.3 (4.0) 1.2 -.073 ' F value for One-way ANOVA of hetersexual group versus homosexual group. - Pearson Correlation coefGcient for average Kinsey score versus personality test score. t Mean Score (Standard Deviation) * p < .005 ** p<.001 Note. All scores have been corrected for age, age squared, and education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50
TABLE 7
Correlations Between Recalled Childhood Sex-Type Behaviors
and 16 PF Personalitv Factors
Overall’ Heterosexual* Homosexual Popouiation Group Group (N=252) (n=140) (n=112)
Primary Factors .A-Warmth .04 -.10 -.102 B-Abstract Thinking -.007 .051 -.049 C-Emotional Stability -.194** -.245** -.197 E-Dominant -.009 -.097 -.038 F-Enthusiatic -.08 -.116 -.123 G-Conscientious -.10 -.187 .047 H-Bold -.103 -.195* -.192 1-Tender-Minded .26** .262** -.028 L-Suspicious .08 .087 .07 .M-Imaginative .06 .056 -.10 N-Siirewd .16 .085 .155 O-Apprehensive .21** .212* .221* Q 1-Experimenting -.05 -.029 -.045 Q2-Self-Sufficient .12 .345** -.029 Q3-Following Self-Image -.19** -.343** -.111 Q4-T ense .28** .055 .28**
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51
Second-Order Factors
EXTRAVERSION -.09 -.244** -.132 ANXIETY .29** .263** .282** TOUGH POISE -.23** -.213* .067 INDEPENDENCE -.06 -.084 -.14
SUPEREGO/CONTROL -.15 -.282** -.021
' Pearson Correlation coefficient for Recalled Childhood Sex-typed Behaviors versus personality test score for the entire sample population. - Pearson Correlation coefficient for Recalled Childhood Sex-typed Behaviors versus personality test score within heterosexual group. ' Pearson Correlation coefficient for Recalled Childhood Sex-typed Behaviors versus personality test score within homosexual group. * p- 02 ** p- 005 Note .All scores have been corrected for age, age squared, and education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52
TABLE 8
Correlations Between Recalled Childhood Sex-Typed Behaviors
and NEO-PI-R Personality Factors
FACTORS Overall* Heterosexual* HomosexuaP (N=252) (n=140) (a =112) N - Neuroticism .258** .185 .254* N l: Anxiety .22** .098 .24* N2: Angry Hostility .111 .075 .16 N3: Depression .223** .203* .18 N4: Self-Consciousness .231** .131 .229* N5: Impulsiveness .173 .127 .146 N6: Vulnerability .23** .206* .223* E - Extraversion -.096 -.24** -.129 E l: Warmth -.037 -.189 -.068 E2: Gregariousness -.076 -.212* -.083 E3: Assertiveness -.07 -.280** -.110 E4: Activity -.109 -.218* -.119 E5: Excitement Seeking -.124 -.15 -.035 E6: Positive Emotions -.01 .014 -.13 O - Openness to Experience .252** .295** .024 01: Fantasy .223** .286** .096 02: Aesthetics .234** .248** .055 03: Feelings .206** .149 .028 04: Actions .17 .173 .031 05: Ideas .020 .242** -.140 06: Values .265** .207* .077 A - Agreeableness .04 -.001 .005 A l: Trust .014 -.144 -.038 A2: Straightforwardness .04 .034 .119 A3: Altruism -.06 -.128 -.034 A4: Compliance .081 .087 .018 A5: Modesty .033 .037 .002 A6: Tender-Minded .032 .095 -.085 C - Conscientiousness -.141 -.367** -.07 Cl: Competence -.123 -.353** -.031 C2: Order -.061 -.147 -.158 C3: Dutifulness -.144 -.301** -.106 C4: Achievement Striving -.07 -.380** .035 C5: Self-Discipline -.152 -.325** -.054 C6: Deliberation -.103 -.215* .002
' Pearson Correlation coefficient for Recalled Childhood Sex-typed Behaviors versus personality test score for the entire sample population. ‘ Pearson Correlation coefficient for Recalled Childhood Sex-typed Behaviors versus personality test score within heterosexual group. ' Pearson Correlation coefficient for Recalled Childhood Sex-typed Behaviors versus personality test score within homosexual group. • p < .0 2 •• p<.OOS Note. All scores have been corrected for age. age squared, and education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX B
Figure
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54
f i O • -W«-4 4 - 1 -S f i CQ « —4 *s U a. ■o f fi 00 T) O S O cZ oo J 3
c/3 Z2 > U PQ T34/ k
1
W3 O
SïiredppjBJ JO )U93J@j
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES
Allen, L. S. & Gorski, R. A. (1991). Sexual dimorphism of the anterior commissure and massa intermedia of the human brain. Journal of Comparative Neurology. 312. 97-104.
Allport, G. W. (1937). Personalitv: A psychological interpretation. New York: Holt.
Bailey, J. M. & Zucker, K. J. (1994). Childhood sex-typed behavior and sexual orientation: A conceptual analysis and quantitative review. Developmental Psychology.
Bailey, J. M. & Pillard, R. C. (1993). A genetic study of male sexual orientation (letter). Archives of General Psychiatry. 50. 240-241.
Bancroft, J. (1994). Homosexual orientation: The search for the biological basis. British Journal of Psychiatry. 164. 437-440.
Baron, R. A., Byrne, D. E., & Kantowitz, B. H. (1980). Psychology: Understanding behavior (2nd Ed.). New York: Holt, Rinehart, and Winston Publishing.
Benjamin, J., Li, L., Patterson, C., Greenberg, B., Murphy, D., & Hamer, D. (1996). Population and familial association between the D4 dopamine receptor gene and measures of Novelty Seeking. Nature Genetics. 12. i , 81-84.
Bernard, L. C. (1982). Sex-role factor identification and sexual preference of men. Journal of Personality Assessment. 46. 3, 292-299.
Bernard, L. C. & Epstein, D. J. (1978). Sex Role Conformity in homosexual and heterosexual males. Journal of Personalitv Assessment. 42. 5, 505-511.
Bouchard, T. J., Lykken, D. T., McGue, M., Segal, N. L., & Tellegen, A. (1990). Sources of human psychological differences: The minnesota study of twins reared apart. Science. 250. 223-228.
Buhrich, N., Bailey, J. M. & Martin, N. G. (1991). Sexual orientation, sexual identity, and sex-dimorphic behaviors in male twins. Behavior Genetics. 21. 75-96.
Carver, C. S. & Scheier, M. F. (1988). Perspectives on Personalitv. Boston: Allyn and Bacon Inc.
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56
Cattell, R. B., Eber, H. W., & Tatsuoka, M. M. (1970). Handbook for the 16PF. Champaign, II: Institute for Personality and Ability Testing.
Cattell, R. B. & Krug, S. (1982). The number of factors in the 16PF: A review of the evidence with special emphasis on methodological problems. Educational and Psychological Measurement. 46. 509-522.
Chen, C. J., Yu, M. W.. Wang, C. J., Tong, S. L., Tien, M., Lee, T. Y., Lue, H. C., Huang, F. Y., Lan, C. C., Yang, K. H., Wang, H. C., Shih, H. Y., Lui, C. Y., & Chen, J. S. (1990). Genetic variance and heritability of temperament among Chinese twin infants, Acta Genet Med Gemellol. 39: 485-490.
Clark, T. R. (1975). Homosexuality and psychopathology in nonpatient males. American Journal of Psychoanalysis. 35. 2 163-168.
Costa, P. T., Jr. & McCrae, R. R. (1991). Revised NEO Personality Inventory (NEO- PI-R) and NEO Five-Factor Inventory (NEO-FFI) professional manual. Odessa, FL: Psychological Assessment Resources, Inc.
De Leo, D., Villa, A. & Capodieci, S. (1992). Personality factors in monozygotic and dizygotic twins: A comparative study. Psychological Reports. 71. 1115-1122.
Duckitt, J. H. & DuToit, L. (1988). Personality profiles of homosexual men and women. The Journal of Psychology. 123: 497-505.
Ellis, L. & Ames, M. A. (1987). Neurohormonal functioning and sexual orientation: A theory of homosexuality-heterosexuality. Psychological Bulletin. 101. 2, 233-258.
Emde, R. E., Plomin, R., Robinson. J., Corley, R., Defries, J., Fulker, D., Reznick, J. S., Campos, J., Kagan, J., & Zahn-Waxier, C. (1992). Temperament, emotion, and cognition at fourteen months: The MacArthur longitudinal twin study. Child Development. 63. 1437-1455.
Evans, R. B. (1969). Childhood parental relationships of homosexual men. Journal of Consulting and Clinical Psychology. 33. (2), 129-135.
Evans, R. B. (1970). Sixteen personality factor questionnaire scores of homosexual men. Journal of Consulting and Clinical Psychology. 34: 212-215.
Flexner, S. B. & Stein, J. (Eds.). (1982). The random house college dictionary (Rev. ed.). New York: Random House, Inc.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57
Floderus-Myrhed, B., Pedersen, N. L., & Rasmuson, I. (1980). Assessment of heritability for personality, based on a short form of the Esyenck Personality Inventory: A study of 12,898 twin pairs. Behavior Genetics. 10. 153-162.
Freud, S. (1953). Three essays on the theory of sexuality. In J. Strachey (Ed. and Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 7, pp. 125-243). London: Hogarth Press.
Freund, K., Nagler, E., Langevin, R., Zajac, A., & Steiner, B. (1974). Measuring feminine gender identity in homosexual males. Archives of Sexual Behavior. 3, 3, 249- 260.
Friedman, R. C. & Downey, J. 1. (1994). Homosexuality. The New England Journal of Medicine. 331. 14, 923-929.
Goldberg, L. R. (1981). Language and individual differences: The search for universals in personality lexicons. In L. Wheeler (Ed.), Review of personalitv and social psychology (pp. 141-165). Beverly Hills: Sage.
Green, R. (1974). Sexual identity conflict in children and adults. Basic Books, New York, and Gerald Duckworth: London.
Green, R. (1985). Gender identity in childhood and later sexual orientation: Follow-up of 78 males. American Journal of Psychiatry. 142. 339-341.
Green, R. (1987). The sissy bov syndrome and the development of homosexuality. New Haven, CT: Yale University Press.
Grellert, E. A., Newcomb, M. D., & Bentler, P. M. (1982). Childhood play activities of male and female homosexuals and heterosexuals. Archives of Sexual Behavior. 11. 451-478.
Hall, J. A. & Kimura, D. (1995). Sexual orientation and performance on sexually dimorphic motor tasks. Archives of Sexual Behavior. 24. 4, 395-407.
Hamer, D. H. & Copeland, P. (1994). The science of desire. New York: Simon and Schuster.
Hamer, D. H., Hu, S., Magnuson,V., Hu, N., & Pattatucci, A. (1993). A linkage between DNA markers on the X chromosome and male sexual orientation. Science. 261. 321-327.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58
Harry, J. (1983). Defeminization and adult psychological well-being among male homosexual. Archives of Sexual Behavior. 12. 1-19.
Hockenberry, S. L. & Billingham, R. E. (1987). Sexual orientation and boyhood gender conformity: Development of the boyhood gender conformity scale (BGCS). Archives of Sexual Behavior. 16. 475-492.
Hooker, E. (1957). The adjustment of the male overt homosexual. Journal of Projective Techniques. 21. 18-31
Howarth, E. & Browne, J. A. (1971). An item factor analysis of the 16PF. Personalitv. 2, 117-139.
Hu, S., Pattatucci, A., Patterson, C., Li, L., Fulker, W., Cherny, S., Kruglyak, L., & Hamer, D. (1995). Linkage between Xq28 and sexual orientation in males but not in females. Nature Genetics. 11. 3, 248-256.
Isay, R. A. (1989). Being homosexual: Gav men and their development. New York: Farrar, Straus, & Giroux.
Kardiner, A. (1963) The Flight from masculinity. In H.M. Ruitenbeek (Ed.), The problem of homosexuality in modem society, (pp. 17-39). New York: Dutton.
Karson, S. & O’Dell, J. W. (1974). Is the 16PF factorially valid? Journal of Personalitv Assessment. 38. 104-114.
Kinsey, A. C, Pomeroy, W. B., & Martin, C. E. (1948). Sexual Behavior in the Human Male. Philadelphia: W. B. Saunders.
Kolb, B. & Whishaw, I. (1985). Fundamentals of Human Neuropsychology (2nd Ed.) New York: W. H. Freeman & Co.
Langevin, R. (1978). Personality characteristics and sexual anomalies in males. Canadian Journal of Behavioural Science. 10. 3, 222-238.
LeVay, S. (1991). A difference in hypothalamic structure between heterosexual and homosexual men. Science. 253. 1034-1037.
LeVay, S. (1993). The sexual brain. Cambridge, MA: MIT Press.
Lykken, D. T., Bouchard, T. J., Jr., McGue, M., & Tellegen, A. (1993). Heritability of interests: A twin study. Journal of Applied Psychology. 78. 4, 649-661.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59
Maccoby, E. E. & Jacklin, C. N. (1974). The Psychology of Sex Differences. Stanford: Stanford University Press.
Manosevitz, M. (1971). Education and MMPI Mf scores in homosexual and heterosexual males. Journal of Consulting and Clinical Psychology. 36. 3, 395-399.
McCormick, C. M. & Witelson, S. P. (1991). A cognitive profile of homosexual men compared to heterosexual men and women. Psychoneuroendocrinologv. 16. (6), 459-473.
McCrae, R. R. & John, O. P. (1992). An introduction to the five-factor model and its applications. Journal of Personalitv. 60. 175-215.
Mulder, R. (1992). The biology of personality. Australian and New Zealand Journal of Psychiatry. 26. 3, 364-376.
Plomin, R., Pedersen, N. L., McCleam, G. E., Nesselroade, J. R., & Bergeman, C. S. (1988). EAS temperament during the last half of the life span: Twins reared apart and twins reared together. Psychology and Aeine. 3, 1, 43-50.
Riess, B. F. (1974). Psychological test data on female homosexuality: A review of the literature. Journal of Homosexuality. 1(1), 71-85.
Sanders, G. & Ross-Field, L. (1986). Sexual orientation and visuo-spatial ability. Brain & Cognition. 5, 280-290.
Saville, P. & Blinkhom, S. (1981). Reliability, homogeneity, and the construct validity of Cattell’s 16PF. Personalitv and Individual Differences. 2, 325-333.
Stiller, R. J., & Herdt, G. H. (1985). Theories of origins of male homosexuality: A cross cultural look. Archives of General Psychiatry. 42. 399-404.
Swaab, D. F. & Hofman, M. A. (1990). An enlarged suprachiasmatic nucleus in homosexual men. Brain Research. 537: 141-148.
Tripp, C. A. (1987). The homosexual matrix (2nd Ed.). New York: Penguin Books USA, Inc.
Wasserman, E. B. & Storm, M. D. (1984). Factors influencing erotic orientation development in females and males. Women and Therapy. 3, 51-60.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60
Weinrich, J. D., Atkinson, J. H., Jr., McCutchan, J. A., Grant, I., & The HIV Neurobehavioral Research Center. (1995). Is gender dysphoria dysphoric? Elevated depression and anxiety in gender dysphoric and nondysphoric homosexual and bisexual men in an HIV sample. Archives of Sexual Behavior. 24. 1, 55-72.
Whitam, F. L., Diamond, M., & Martin, J. (1993). Homosexual orientation in twins: A report on 61 pairs and three triplet sets. Archives of Sexual Behavior. 22. (3), 187- 206.
Whitam, F. L. & Mathy, R. M. (1986). Male homosexuality in four societies: Brazil. Guatemala. Philippines, and the United States. New York: Praeger.
Whitman, M. (1977). Childhood indicators of male homosexuality. Archives of Sexual Behavior. 6, 89-96.
Willmott, M., & Brierley, H. (1984). Cognitive characteristics and homosexuality. Archives of Sexual Behavior. 13. 4, 311-319.
Wilson, J. D. (1995). Sex hormones and sexual behavior. In Abramson, P., Pinkerton, S. (Ed.), Sexual nature/sexual culture, (pp. 121-133). Chicago: The University of Chicago Press.
Zucker, K. J. (1990). Gender identity disorders in children: Clinical descriptions and natural history. In Blanchard, R. & Steiner, B. W. (Eds.), Clinical management of gender identity disorders in children and adults (pp. 1-23). Washington, D C.: American Psychiatric Press.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.