See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/225665711

Gender-role conflict scale: College men's fear of

femininity

Article in Sex Roles · March 1986

DOI: 10.1007/BF00287583

CITATIONS READS

773 16,684

5 authors, including:

Barbara J. Helms Lawrence Wrightsman

9 PUBLICATIONS 830 CITATIONS

University of Kansas
86 PUBLICATIONS 5,058 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Barbara J. Helms on 05 July 2017.

The user has requested enhancement of the downloaded file.

Sex Roles, Vol. 14, Nos. 5/6, 1986

Gender-Role Conflict Scale: College

Men's Fear of Femininity 1
James M. O'Neil, 2 Barbara J. Helms, Robert K. Gable
University o f Connecticut

Laurence David
Iowa State University

Lawrence S. Wrightsman
University o f Kansas

Gender-role conflict exists when gender roles have negative consequences for
people. This research reports initial validity and reliability data on measures
of gender-role conflict for men. Two measures, Gender Role Conflict Scale
I and H (GRCS-I and GRCS-II) were constructed to assess patterns o f
gender-role conflict described in the literature. GRCS-I assesses men's per-
sonal gender-role attitudes, behaviors, and conflicts. GRCS-H assesses men's
gender-role conflicts in specific gender-role conflict situations. Both GRCS
measures and the Personal Attributes Questionnaire (PA Q) were administered
to male college students (N = 52 7). Initial factor-analytic data for GRCS-I
and GRCS-H demonstrated eight meaningful factors. Acceptable test-retest
and internal consistency reliabilities were found for both measures.
M A N O VA, A N O VA, and Tukey procedures indicated differences for sub-
jects across the four PAQ categories. Significant gender-role conflict dif-
ferences across the factors were found for men who were instrumental,
expressive, or both instrumental and expressive. Results of these differences
are reported, as well as implications for future development of both scales.

IThe authors are indebted to Dr. Nancy Betz (Ohio State University) who generously gathered
data for us during the Spring Semester, 1982. This research was supported by a grant from
the General Research F u n d of the University of Kansas, Lawrence, Kansas. This paper was
presented at the American Psychological Association A n n u a l Convention, Washington, DC,
August 26, 1982.
2Address correspondence to James M. O'Neil, Department of Educational Psychology, Box
U-64, 249 Glenbrook Road, University of Connecticut, Storrs, Connecticut 06268.
335
0360-0025/86/0300-0335505.00/0© 1986PlenumPublishingCorporation
336 O'Neil, Helms, Gable, David, and Wrightsman

The negative effects of socialized gender role continues to be an area of scien-

tific inquiry. These negative effects have been described as gender- or sex
role conflict and sex role strain (Garnets & Pleck, 1979; O'Neil, 1981a,b,
1982; Pleck, 1981)? Gender-role conflict is a psychological state where gender
roles have negative consequences or impact on a person or others. The
ultimate outcome of this conflict is the restriction of the person's ability to
actualize their human potential or the restriction of some else's potential.
Sex role strain has also been described as an intrapsychic process that can
lead to a poor psychological adjustment, particularly low self-esteem (Garnets
& Pleck, 1979).
A recent sex role strain (SRS) paradigm (Pleck, 1981) provides addi-
tional understanding of the negative effects of socialized gender roles. This
paradigm enumerates ten propositions indicating that gender roles produce
conflict and strain for both sexes. Pleck's paradigm indicates that: (1) viola-
tion of gender roles can lead to negative psychological consequences, (2) cer-
tain gender-role characteristics are psychologically dysfunctional (3) both sexes
experience strain and conflict because of gender roles. Furthermore, these
assumptions are based on two theories of sex role strain: self-role discrepan-
cy theory and socialized dysfunctional characteristics theory. The former
theory suggests that individuals suffer negative consequences when they
fail to live up to sex roles. The latter theory suggests that because of
sex roles, individuals are socialized to have personality characteristics that
are dysfunctional. Pleck's sex role strain paradigm provides a theoretical base
to empirically study gender-role conflict in men's and women's lives.
Men's gender-role conflicts have been topics of discussion for some time
in the popular literature (Farrell, 1974; Fasteau, 1974; Nichols, 1975; Pleck
& Sawyer, 1974). Conceptual analysis of men's problems have also emerged
in the professional literature (David & Brannon, 1976; Doyle, 1983; O'Neil,
1981a, 1982; Pleck & Pleck, 1980; Scher, 1981; Skovholt, Gormally, Schau-
ble, & Davis, 1980; Solomon & Levy, 1982). Additionally, three case studies
of men's conflict have been reported in the literature (Komarovsky, 1976;
Levinson, Darrow, Klein, Levinson, & McKee, 1978; Vaillant, 1974). There
have been few empirical studies documenting men's conflict around socialized
gender roles. This lack of empirical research on gender-role conflict delays scien-
tific understanding of how gender roles produce conflict in men's lives.
One impediment to assessing men's gender-role conflict has been few
unifying constructs that capture its factors and patterns. One unifying theme
of gender-role conflict has been hypothesized as men's fear of femininity

3For purposes of this research, the terms gender-role conflict and strain and sex role conflict
and strain are used synonymously throughout the manuscript.
Gender-Role Conflict 337

(David & Brannon, 1976; Farrell, 1974; O'Neil, 1981b,c, 1982). The fear of
femininity is defined as a strong, negative emotion associated with stereotypic
feminine values, attitudes, and behaviors. These emotional reactions are learn-
ed primarily in early childhood when gender identity is being formed by
parents, peers, and societal values. Men's fear of their feminine sides and
women have been noted in the theoretical literature for many years (Boehm,
1930; Hays, 1964; Horney, 1967; Jung, 1953, 1954; Lederer, 1968; Menn-
inger, 1970). Most of these analyses of men's fears about femininity have
a psychodynamic foundation. Jung's archetype in men, the anima, is a well-
known concept about men's difficulty integrating their feminine sides. Reviews
of mythology (Lederer, 1968; Johnson, 1977) provide even more substantial
evidence that threats and fears of femininity have existed over the centuries.
More recently, Levinson et al. (1978), in their case study of men, found that
men (1) neglected or repressed the feminine sides of self, or (2) regarded
those parts of themselves as feminine as being dangerous. Men's fear about
their femininity has direct relevance to patterns of gender-role conflict and
strain.
Patterns of gender-role conflict and strain associated with the fear of
femininity have been described in the literature (David & Brannon, 1976;
O'Neil, 1981a,b, 1982; O'Neil, Helms, Gable, David, & Wrightsman, 1985).
Figure 1 shows a model of gender-role conflict and strain that includes the
following six patterns: (1) restrictive emotionality, (2) homophobia, (3)
socialized control, power, and competition, (4) restrictive sexual and affec-
tional behavior, (5) obsession with achievement and success and (6) health
care problems. The relationship between these gender-role patterns and the
fear of femininity has been discussed elsewhere (O'Neil, 198 lb, 1982; O'Neil
& Fishman, in press). Foremost to the model is the hypothesis that men's
fears of femininity contribute to the six patterns of gender-role conflict found
in Figure 1.
Another reason for limited research on men's gender-role conflict has
been the lack of psychometric instruments to assess it. Other measures of
gender-role attributes, orientations, and attitudes have been developed (Bern,
1974; Doyle & Moore, 1978; Spence & Helmreich, 1978; Spence, Helmreich,
& Stapp, 1974, 1975). The Personal Attitudes Questionnaire (PAQ; Spence
& Helmreich, 1978) and the Bern Sex Role Inventory (BSRI; Bern, 1974) have
been developed to assess androgyny or expressiveness-instrumentality. The
Attitudes Toward Women Scale (AWS; Spence & Helmreich, 1972) assesses
the rights and roles of women in society. The Attitudes Toward the Male's
Role Scale (AMR; Doyle & Moore, 1978) measures attitudes toward the male's
sex role in contemporary society.
Currently, there is no instrument to assess men's personal gender-role
attitudes, behaviors, and conflicts. Research instruments have not been
338 O'Neil, Helms, Gable, David, and Wrightsman

Restrictive
Emotionality
, ' ea'th
Homophobia FEAR OF Care
Problems

Socialized
Control, Power,~ 3bsession
~nd Competition q). FEMININITY with
Issues _ . . . . ~ ~Achievement
and Success
Restricted Sexua
and Affectionate
Behavior

Fig. 1. Six patterns of gender-roleconflictand strain emanating

from men's socialization and their fear of femininity. (From
Solomon & Levy, 1982, p. 7).

developed because gender roles, and their subsequent strain and conflict, are
multifaceted. Gender-role conflict can be conceptualized from four overlap-
ping and complex dimensions: cognitions, affective experience, behaviors,
and unconscious experience. Cognitive experience of gender-role conflict is
how we think about our gender roles and aspects of masculinity, femininity,
and androgyny. Affective experience of gender-role conflict represents our
emotional feelings around our gender-role issues. Behavioral aspects of
gender-role conflict includes how we act, respond, and interact with ourselves,
and others, around gender-role issues. Unconscious aspects of gender role
includes the intrapsychic and repressed aspect of our gender roles beyond
our conscious awareness. The multidimensionality of gender-role conflict and
men's individual differences provide challenges for assessing it through
psychometric instruments.
Three approaches to assessing men's gender-role conflict are apparent
from the dimensions described above. First, gender-role conflict can be assess-
ed by asking men about the specific ways they think and feel about their
gender-role behaviors. Second, men can report the degree of conflict and
Gender-Role Conflict 339

comfort in particular gender-role situations. These two approaches provide

an analysis of how men see themselves in terms of gender roles, as well as
how they might act in gender-role conflict situations. Thirdly, the assessment
of men's gender-role conflict will vary according to the man's personal at-
tributes. For example, men who describe themselves differentially in terms
of gender-role characteristics (i.e., expressiveness or instrumentality) may
show differential aspects of gender-role conflict.
The purpose of this study is to present initial construct validity data
on two scales to measure men's gender-role conflict patterns found in Figure
1. Through a series of methodological and statistical procedures, two scales
were developed: Gender Role Conflict Scale I and II (GRCS-I and GRCS-
II). One scale assessed men's thoughts and feelings about their gender-role
behaviors, and the other assessed men's degree of conflict and comfort in
particular situations.
Cronbach and Meehl (1955) discuss five methods of experimentation
to investigate construct validity including factor analysis, test-retest reliability,
internal consistency reliability, group differences, and studies of process. This
study assessed four of the five above methods by testing three hypotheses.
First, it was hypothesized that items from the instrument would cluster into
the six gender-role factors found in Figure 1. Second, it was hypothesized
that these factors and the instrument would demonstrate moderately high
reliability and validity. Third, it was hypothesized that subjects with different
gender-role orientations, measured by the PAQ, would differ with respect
to the degree to which they express patterns of gender-role conflict. Specifical-
ly, it was hypothesized that men who describe themselves as expressive, in-
strumental, Or both, would express differential degrees of gender-role conflict.
The overall purpose of the research was to gain empirical support for the
construct validity and overall reliability of gender-role conflict in men's lives
using GRCS-I and II.

METHOD

Subjects

Undergraduate men (N = 527) at two Midwestern universities enroll-

ed in introductory psychology classes were the subjects. The mean age of
the subjects was 19.8 years. Class levels of subjects included 57°70 freshmen,
27°70 sophomores, 11 °70 juniors, and 507o seniors. The marital status of sub-
jects indicated 95°70 single and 5°70 married.
340 O'Neil, Helms, Gable, David, and Wrighlsman

Procedure

Operational definitions of the six patterns of gender-role conflict shown

in Figure 1 were developed from previous theoretical literature (O'Neil,
1981a,b, 1982). The six patterns are defined below:
1. Restrictive emotionality-having difficulty expressing one's feelings
or denying others their rights to emotional expressiveness.
2. Homophobia-having fear of homosexuals or fear of being a
homosexual including beliefs, myths, and stereotypes about gay
people.
3. C o n tr o l - t o regulate, restrain, or to have others or situations under
one's command. P o w e r - t o obtain authority, influence, or ascen-
dancy over others. Competition-striving against others to win
or gain something.
4. Restricted sexual and affectionate behavior-having limited ways
of expressing one's sexuality and affection to others.
5. Obsession with achievement and success- having a disturbing and
persistent preoccupation with work, accomplishment and eminence
as a means of substantiating and demonstrating value.
6. Health care problems-having difficulties maintaining positive
health care in terms of diet, exercise, relaxation, stress, and a healthy
life style.
From these definitions and the available literature, GRCS-I and GRCS-II
were developed. For GRCS-I, 85 items were generated with the following
number of items per gender-role pattern in Figure 1: (1) restrictive emo-
tionality (N = 15), (2) health care problems (N = 14), (3) obsession with
achievement and success (N = 16), (4) restrictive sexual and affectionate
behavior (iV = 17), (5) control, power, and competition (N = 14), (6)
homophobia (N = 9). For GRCS-II, 51 items were generated with the follow-
ing number of items per gender-role pattern in Figure 1 : (1) restrictive emo-
tionality (N = 12), (2) health care problems (iV = 4), (3) obsession with
achievement and success (iV = 7), (4) restrictive sexual and affectionate
behavior (N = 6), (5) control power, and competition (N = 15), (6)
homophobia (N = 7).
Subjects were given these two scales, along with the PAQ (Spence &
Helmreich, 1978). These instruments are described below.

Instrumentation

Gender-Role Conflict Scale I (GRCS-1). The original scale was an

85-item self-report instrument designed to assess aspects of gender-role con-
Gender-Role Conflict 341

flict described in the literature (O'Neil, 1981a,b, 1982) and in Figure 1.

Respondents are asked to report the degree to which they agree or disagree
about their personal gender-role attitudes, behaviors, and conflicts. Each item
provides an assessment of men's conflict by the respondent's self-rating of
attitude or behavior previously categorized as gender-role conflict. Respondents
are asked to report the degree to which they agree or disagree with statements
using a 6-point Likert scale, ranging from strongly agree (6) to strongly
disagree (1). Example items from GRCS-I include "I have difficulty express-
ing my tender feelings," "I strive to be more successful than others," Find-
ing time to relax is difficult for me." Each statement expresses a gender
role conflict patterns found in Figure 1. Thus, a high score assumes and
reflects an expression of gender-role conflict and fear about femininity
(O'Neil, 1981b, 1982).
Gender-Role Conflict Scale II (GRCS-II). The original scale was a
51-item self-report instrument designed to assess situational dimensions of
gender-role conflict patterns found in Figure 1. This scale asks subjects to
report their degree of comfort or conflict in specific gender-role conflict situa-
tions. Each item provides an assessment of men's conflict by the respondent's
self-rating of conflict or comfort in a concrete gender-related situation.
Respondents are asked to rate the degree of comfort or conflict in specific
situations using a 4-point Likert scale, ranging from very much conflict-
very uncomfortable (4) to no conflict-very comfortable (1). Example items
from GRCS-II include (1) "Your best friend has just lost his job at the fac-
tory where you work. He is obviously upset, afraid, and angry but he has
these emotions hidden. How comfortable/uncomfortable are you to respond-
ing to your friend's intense emotions and fear about unemployment?" (2)
"There's a guy you've idolized since grade school. He's three years older than
you are. In high school he was the star quarterback, valedictorian, and very
active in the Young Methodist Fellowship. Last year he graduated from col-
lege. You have just learned he is a homosexual. How much conflict do you
feel between your admiration for this person and the fact that he is a homosex-
ual?" Each statement expresses a gender-role conflict pattern found in Figure
1. Thus, a high score reflects and assumes an expression of gender-role con-
flict and fear about femininity (O'Neil, 1981b, 1982).
Personal Attributes Questionnaire (PAQ). The PAQ (Spence &
Helmreich, 1978) measures the psychological dimension of masculinity and
femininity. The PAQ is a 24-item self-report instrument consisting of a
number of trait descriptions, each set upon a 5-point bipolar scale. These
trait descriptions of dispositional properties makes no reference to overt
behavior or to the situations in which these dispositions are manifested. Each
item describes a characteristic stereotypically believed to differentiate the
sexes. The PAQ is divided into three separate scales labeled Masculinity (M),
Femininity (F), and Masculinity-Femininity (MF). In content, the M scale
342 O'Neil, Helms, Gable, David, and Wrightsman

contains socially desirable traits stereotypically more characteristic of males

than females and refer to self-assertive, instrumental attributes. The F scale
contains socially desirable traits stereotypically more characteristic of females
and refer to interpersonally oriented expressive qualities. The M-F scale con-
sists of traits dimensions for which the social desirability ratings were related
to gender, the ideal woman falling toward the stereotypically feminine pole
and the ideal man toward the stereotypically masculine pole (Spence &
Helmreich, 1980). Median split of subjects' PAQ scores produced medians
for masculinity (M) and femininity (F) both at 23. These medians were used
to classify subjects to masculine (high masculinity, low femininity), feminine
(low-high), androgynous (high-high), or undifferentiated (low-low), gender-
role categories. As defined by Spence and Helmreich (1980), the PAQ assesses
respondents' self-reported attributes of expressiveness and instrumentality,
In this study, the man categorized as "masculine" has instrumental traits but
is low on expressive traits. The "feminine" man possesses expressive traits
but is low on instrumental traits. The "androgynous" man is high on both
instrumental and expressive traits. Finally, the "undifferentiated" man is low
on both instrumental and expressive traits. In short, the PAQ represents
men's self-reported attributes of expressiveness and instrumentality. Cron-
bach's alphas for the short form of the PAQ are .85, .82, and .78 for M,
F, and M-F, respectively (Spence & Helmreich, 1978).

Data Analys&

The analysis of the data and construct validity of the GRCS-I and II
was determined by the following approaches: item-reduction procedures, fac-
tor analysis, reliability data analysis, and multivariate analysis of variance
(MANOVA). Each will be described below.
Item-Reduction Procedures. A systematic procedure was utilized to
reduce the number of items on the GRCS. The goal of this systematic reduc-
tion of items was to obtain the best items and factor structure from the
original items. A three-step procedure was employed to decide which items
to retain for further analysis. First, all items were submitted to three dif-
ferent raters to assess the degree to which the items were gender related rather
than related to other kinds of human conflicts. For GRCS-I, raters assessed
the items by indicating on a 1-5 point scale that item definitely does not relate
to men's gender role (1) to item definitely does relate to man's gender role
(5). For GRCS-II, raters assessed the items by indicating on a 1-5 point scale,
(1) item definitely relates to human conflict, not male gender role to (5) item
definitely relates to conflict in male gender roles. Using an arbitrary 3.5
cut off, all items met the criteria for being assessed as a gender-role related
Gender-Role Conflict 343

item. Second, all items were excluded that did not meet the criterion of hav-
ing a standard deviation of at least 1.00; these items were essentially answered
the same way by most respondents and hence did not discriminate. Third,
each item had to correlate with at least one other item at the .30 level to
be retained.
Factor Analysis. Factor analyses, employing both principal components
and c o m m o n factor models, with both orthogonal and oblique rotations,
were completed on GRCS-I and II. The goal of these numerous analyses was
to determine the best simple structure of observed factors for the items (Rum-
reel, 1970). All items with factor loadings less than .35 were excluded from
the scale (Nunnally, 1978) as well as items that had loadings of greater than
.30 on two or more factors. Consequently, no item was allowed to cross-
load on any other factor in the construction of the scales.
Reliability Data. Internal consistency reliabilities for each scale score
were calculated using Cronbach's alpha procedure. Using the described sam-
ple, four-week test-retest reliabilities (N = 17) for scale scores were calculated
for GRCS-I using Pearson product m o m e n t correlations. Four-week
test-retest reliabilities (N = 14) for G R C S - I I were calculated from students
in an introductory educational psychology class using Pearson product mo-
ment correlations.
Multivariate Analysis o f Variance. Additional exploratory construct
validity evidence was obtained through multivariate analysis of variance
(MANOVA). Scale scores were developed by summing the responses to the
items defining each factor. M A N O V A was performed on derived factors using
the P A Q classification for all subjects. Univariate analysis of variance and
Tukey post hoc procedures were used to further analyze significant M A N O V A
findings.

RESULTS

Using the item-reduction procedure described above, 37 of the original

85 items met the criteria for inclusion in GRCS-I. C o m m o n factor analysis
with oblique rotation yielded the most significant factor composition for the
GRCS-I. Table I lists and names four emergent factors as follows: Factor
1 - s u c c e s s , power, competition (13 items); Factor 2 - restrictive emotionali-
ty (10 items); Factor 3--restrictive affectionate behavior between men (8
items); Factor 4-- conflicts between work and family relations (6 items). The
four factors explained 36°7o of the total variance.
Assessments of the scales' reliabilities found internal consistency scores
using Cronbach's alpha ranged f r o m .75 to .85. Four-week test-retest
reliabilities (N -- 17) ranged f r o m .72 to .86 for each factor.
344 O'Neil, Helms, Gable, David, and Wrightsman

Table I. Factor Structure for Gender-Role Conflict Scale I

Factor
Items loadings
Factor 1--success, power, competition (13 items)
1. Moving up the career ladder is important to me. .64
5. Making money is part of my idea of being a successful man. .52
8. I sometimes define my personal value by my career success. .54
12. I evaluate other people's value by their level of achievement and success. .54
14. I worry about failing and how it affects my doing well as a man. .45
18. Doing well all the time is important to me, .43
21. I often feel that I need to be in charge of those around me. .49
23. Competing with others is the best way to succeed. .58
24. Winning is a measure of my value and personal worth. .57
28. I strive to be more successful than others. .72
32. I am often concerned about how others evaluate my performance at
work or school .41
34. Being smarter or physically stronger than other men is important to me. .61
37. I like to feel superior to other people. .53

Factor 2--restrictive emotionality (10 items)

2. I have difficulty telling others I care about them. .70
6. Strong emotions are difficult for me to understand. .35
9. Expressing feelings makes me feel open to attack by other people. .37
13. Talking (about my feelings) during sexual relations is difficult for me. .52
15. I have difficulty expressing my emotional needs to my partner. .78
19. I have difficulty expressing my tender feelings. .76
22. Telling others of my strong feelings is not part of my sexual behavior. .44
25. I often have trouble finding words that describe how I am feeling. .41
29. I do not like to show my emotions to other people. .43
30. Telling my partner my feelings about him/her during sex is difficult
for me. .75

Factor 3-restrictive affectionate behavior between men (8 items)

3. Verbally expressing my love to another man is difficult for me. .50
7. Affection with other men makes me tense. .69
10. Expressing my emotions to other men is risky. .58
16. Men who touch other men make me uncomfortable. .67
20. Hugging other men is difficult for me. .71
26. I am sometimes hesitant to show my affection to men
because of how others might perceive me. .52
33. Being very personal with other men makes me feel uncomfortable. .66
35. Men who are overly friendly to me, make me wonder about their
sexual preference (men or women) .48
Factor 4--conflicts between work and family relations (6 items)
4. I feel torn between my hectic work schedule and caring for my health. .45
11. My career, job, or school affects the quality of my leisure or family life. .65
17. Finding time to relax is difficult for me. .57
27. My needs to work or study keep me from my family or leisure more
than I would like. .70
31. My work or school often disrupts other parts of my life (home, health,
leisure). .58
36. Overwork, and stress, caused by a need to achieve on the job or in
school, affects/hurts my life. .46
Gender-Role Conflict 345

For GRCS-II, 16 of the original 51 items met the criteria for inclusion.
Principal component analysis with oblique rotations yielded the most mean°
ingful factor composition for GRCS-II. Table II lists and names four
emergent factors for the instrument as follows: Factor 1-success, power,
and competition (6 items); Factor 2-homophobia (4 items), Factor 3 - l a c k
of emotional response (3 items); Factor 4-public embarrassment from

Table II. Factor Structure for Gender Role Conflict Scale II

Factor
Items ~ loadings
Factor 1-success, power, competition (6 items)
3. How much conflict do you feel about your brother's success compared
to your own job as a plumber's supply truck driver? .58
6. How much conflict do you feel between your poor performance and
your desired level of performance? .68
8. How conflicted do you feel that her salary is much higher than yours
during a conversation with this new couple you have just met? .43
13. How conflicted would you feel with your desire to have intercourse
and your inability to achieve an erection? .59
15. How comfortable/uncomfortable do you feel with your lower grades
compared to your friends' grades? .59
16. How conflicted do you feel about your low ranking and that they
will be known by your other colleagues? .73

Factor 2 - h o m o p h o b i a (4 items)
4. How comfortable/uncomfortable would you feel talking during
intermission to this person who is a known homosexual? .71
7. At the bar you notice that an unknown man is staring at you and
then he comes over to introduce himself. How comfortable/
uncomfortable would you feel talking to this man? .67
11. Under these conditions, how conflicted would you feel as a heterosexual
male going out with a man thought to be gay? .67
14. How much conflict do you feel between your admiration for this
person and the fact that he is a homosexual? .82

Factor 3--lack of emotional response (3 items)

5. How comfortable/uncomfortable do you feel responding to her
sadness, emotions, and tears? .71
1. How comfortable/uncomfortable are you responding to your friends'
intense emotions and fears about employment? .71
9. Disregarding your sadness of his dying state, how comfortable/
uncomfortable are you specifically with your father's expression of
love for you? .63

Factor 4 - p u b l i c embarrassment from gender-role deviance (3 items)

2. How comfortable/uncomfortable do you feel with this public
display of affection? .62
10. How comfortable/uncomfortable do you feel carrying a woman's
purse in front of people in the restaurant? .62
12. How conflicted do you feel about what your male co-worker might
think about your contact and relationship with your intimate friend? .60
"Each item is preceded by a descriptive situation that ends with the above questions for each factor.
346 O'NeU, Helms, Gable, David, and Wrightsman

Table !II. Factor Composition and Reliability Data for GRCS-I and II
Number Internal consistency Test-retest
Factors of items reliabilities (N = 527) reliabilities a
GRCS (personal self-report)
1. Success, power, and competition 13 .85 .84
2. Restrictive emotionality 10 .82 .76
3. Restrictive affectionate behavior
between men 8 .83 .86
4. Conflicts between work and
family relations 6 .75 .72

GRCS-II (situational self-report)

1. Success, power, and competition 6 .70 .79
2. Homophobia 4 .76 .78
3. Lack of emotional response 3 .51 .85
4. Public embarrassment from gender
role deviance 3 .59 .83
"Test-retest for GRCS-I was with 17 subjects, whereas test-retest for GRCS-II included 14 subjects.

gender-role deviance (3 items). The four factors explained 48°70 of the total
variance.
Assessments of the scale reliabilities found internal consistency scores
using Cronbach's alpha ranged from .51 to .76. Four week test-retest
reliabilities (N = 14) ranged from .79 to .85. Table III summarizes the factor
reliabilities and validity data for both scales.
Tables IV and V contain the means and standard deviations for the
four factors of GRCS-I and GRCS-II according to the P A Q categories. For
GRCS-I, multivariate analysis of variance (MANOVA) indicated an overall
significant main effect, F(12, 1535) = 9.87, p < .000. Univariate analysis
of variance on the four factors indicated significant differences on three of
the four factors. Differences between P A Q categories were found for Fac-
tor 1, F(3, 519) = 7.86, p < .000; Factor 2, F(3, 522) = 20.43, p < .001;
Factor 3, F(3, 522) = 9.69, p < .000. No significant differences were found
for Factor 4.

Table IV. Means and Standard Deviations for PAQ Groups and Factors of the GRCS-I °'b
GRCS GRCS GRCS GRCS
Factor 1 Factor 2 Factor 3 Factor 4
PAQ Groups N X SD X SD X SD X SD
Masculine 120 56.68 8.97 30.98 8.08 30.33 7.85 20.87 5.88
Feminine 118 50.28 11.77 29.69 9.24 27.63 7.68 21.25 5.88
Androgynous 129 53.12 10.76 26.33 8.52 27.04 9.22 21.70 6.60
Undifferentiated 160 52.66 9.59 34.09 8.10 31.39 6.87 21.95 5.32
~Factor 1, success, power, competition; Factor 2, restrictive emotionality; Factor 3, restrictive"
affectionate behavior between men; Factor 4, conflicts between work and family relations.
bN = 527. High score indicates greater self-concern regarding that attribute.
Gender-Role Conflict 347

Table V. Means and Standard Deviations for PAQ Groups and Factors of the GRCS-IIa'b
GRCS GRCS GRCS GRCS
Factor 1 Factor 2 Factor 3 Factor 4
PAQ Groups N X SD X SD X SD X SD
Masculine 120 15.64 2.95 1 1 . 9 9 2.82 5.93 1.78 6.73 1.80
Feminine 118 15.33 3.18 1 0 . 9 6 2.61 5.44 1.69 6.36 1.89
Androgynous 129 14.88 3.44 11.23 2.89 5.37 1.76 6.04 1.79
Undifferentiated 160 15.60 3.11 1 1 . 6 5 2.49 6.52 1.69 7.12 1.67
~Factor 1, success, power, competition; Factor 2, homophobia; Factor 3, lack of emotional
response; Factor 4, public embarrassment from gender-role deviance.
bN = 527. High score indicates greater self-concern regarding that attribute.

F o r G R C S - I I , m u l t i v a r i a t e analysis o f v a r i a n c e i n d i c a t e d an overall
significant m a i n effect, F(12, 1541) = 5.53, p < .000. U n i v a r i a t e analysis
o f v a r i a n c e on the f o u r f a c t o r s i n d i c a t e d significant differences on three o f
the four factors. Significant differences between P A Q categories were f o u n d
for F a c t o r 2, F(3, 522) = 3.51, p < .05; F a c t o r 3, F(3, 522) = 13.43, p <
.000; F a c t o r 4, F(3, 522) = 9.67, p < .000. N o significant differences were
f o u n d for F a c t o r 1.
Significant univariate results were followed up with the T u k e y procedure
at the .05 level. F o r G R C S - I , i n s t r u m e n t a l m e n (masculine) r e p o r t e d
significantly higher scores on F a c t o r 1 t h a n feminine, androgynous, or undif-
ferentiated men. F o r F a c t e r 2 instrumental and expressive m e n (androgynous)
h a d significantly lower scores t h a n masculine, feminine, or u n d i f f e r e n t i a t e d
men. A l s o , m e n w h o were neither expressive n o r i n s t r u m e n t a l ( u n d i f f e r e n -
tiated) h a d significantly higher scores o n restrictive e m o t i o n a l i t y t h a n the
three o t h e r types. F o r F a c t o r 3 i n s t r u m e n t a l m e n (masculine) a n d m e n w h o
were neither instrumental nor expressive (undifferentiated) had significantly
higher scores t h a n a n d r o g y n o u s a n d f e m i n i n e types.
F o r G R C S - I I i n s t r u m e n t a l m e n (masculine) h a d significantly higher
scores t h a n expressive m e n (feminine) on F a c t o r 2. F o r F a c t o r 3, m e n w h o
neither were i n s t r u m e n t a l o r expressive ( u n d i f f e r e n t i a t e d ) h a d significantly
higher scores t h a n masculine, feminine, a n d a n d r o g y n o u s men. F o r F a c t o r
4, men without instrumental or expressive traits (undifferentiated) were
significantly higher t h a n b o t h f e m i n i n e a n d a n d r o g y n o u s men. F i n a l l y , in-
strumental m e n (masculine) h a d significantly higher scores t h a n a n d r o g y n o u s
men on this factor.

DISCUSSION

This r e s e a r c h tested three h y p o t h e s e s f o c u s e d o n the c o n s t r u c t v a l i d i t y

o f the G R C S a n d men's g e n d e r - r o l e conflict. T h e first h y p o t h e s i s , stating
348 O'Neil, Helms, Gable, David, and Wrightsman

that GRCS-I and GRCS-II items would cluster into the six gender-role pat-
terns, was partially supported. Some of the patterns were clearly evident and
two new factors emerged from the factor analysis. Other factors were renam-
ed, combined, and one factor (health care problems) did not emerge from
the analysis. Overall, eight related patterns of gender-role conflict were iden-
tified.
The patterns of control, power, competition and obsession with
achievement and success in GRCS-I and GRCS-II (see Figure 1) were com-
bined into one factor named "success, power, and competition." The restric-
tive emotionality pattern was clearly evident in GRCS-I, and a similar pattern,
named "lack of emotional response," was evident from GRCS-II. The pat-
tern of restrictive sexual and affectionate behavior was renamed "restrictive
affectionate behavior between men" because the items retained reflected on-
ly affectional exchanges between men. A separate but related pattern of
H o m o p h o b i a was evident in GRCS-II. Two new patterns not hypothesized
in Figure 1 were named "conflict between work and family relations" and
"public embarrassment from gender-role deviance."
In summary, GRCS-I and GRCS-II identified gender-role conflict reflec-
ting men's thoughts, feelings, and behaviors, as well as conflicts in particular
gender-role situations. The patterns of success, power, and competition,
restrictive emotionality, lack of emotional response, were clearly evident in
men's self-report and also in a situational context. The subjects also self-
reported restrictive affectionate behavior with other men and h o m o p h o b i a
was identified as a situational gender-role conflict issue. Likewise, conflict
between work and family relations was self-reported and public embarrass-
ment from gender role deviance was identified as a situational problem area.
The second hypothesis stating that the derived scales would demonstrate
moderately high reliability and validity was supported. Internal consistency
reliabilities ranged between .51 to .85 for GRCS-I and GRCS-II, with the
reliabilities of the latter scale being somewhat lower than the former.
Test-retest reliabilities on each scale ranged between .72 to .86 for the deriv-
ed dimensions from GRCS-I and II measures. These initial reliability and
validity data are adequate for the earlier stages of instrument development.
Further research and item development is needed to raise the reliabilities of
each scale and add to the construct validity of GRCS-I and GRCS-II.
The third hypothesis, stating that men describing themselves as ex-
pressive, instrumental, or both, would express differential degrees of gender-
role conflict, received strong support. The post hoc analysis indicated
numerous differences among the eight observed factors across the four gender
types assessed by the PAQ. Men who described themselves as expressive, in-
strumental, both expressive and instrumental, or neither expressive nor in-
strumental expressed differential degrees of gender-role conflict. Two patterns
Gender-Role Conflict 349

of significant results are apparent across the two scales on related factors.
First, men reporting neither instrumental or expressive characteristics (un-
differentiated) reported significantly higher scores on both restrictive emo-
tionality and lack of emotional response. Second, men reporting themselves
as instrumental (masculine) reported significantly higher scores on restric-
tive affectionate behavior between men and homophobia than did expressive
(feminine) men.
These initial results support further research of the GRCS-I and GRCS-
II. The present results indicate that further empirical research is needed on
both scales. GRCS-I needs to be validated on older adult men across dif-
ferent racial and socioeconomic groups. GRCS-II needs the same validation,
but more importantly, more items need to be added to each factor.
For over ten years, the professional literature has made claims that men
experience strains and conflict from their socialized gender roles. Much of
the evidence documenting this conflict has emanated from theoretical analysis
in the professional and popular literature. Only recently has a sex role strain
(SRS) paradigm (Pleck, 1981) provided a coherent statement that critically
reviews the previous literature and provides direction for research. This
paradigm allows researchers to move from theory to empirical assessment
of gender-role strain and conflict. Pleck (1981) believed "[t]his is the research
agenda for the future, and it carries the promise of more humane and
egalitarian scientific study of the sexes" (p. 160).
The present research represents preliminary efforts to move beyond
theoretical speculation about gender-role conflict to empirical and construct
validation. Overall, the results do provide some empirical support for the
sex role strain paradigm in college-aged men. Additional empirical research
would allow more authoritative statements on the negative effects of gender-
role conflict for men, women, and children.

REFERENCES

Bem, S. L. The measurement of psychological androgyny. Journal o f Consulting and Clinical

Psychology, 1974, 42, 115-162.
Boehm, F. The femininity-complex in men. International Journal of Psycho-Analysis, 1930,
11, 444-469.
Cronbach, L. J., & Meehl, P. E. Construct validity in psychological tests. Psychological Bul-
letin, 1955, 52, 281-302.
David, D. S., & Brannon, R. Theforty-ninepercent majority." The male sex role. Reading, Mass.:
Addison-Wesley, 1976.
Doyle, J. A. The male experience. Dubuque, Ia.: William C. Brown, 1983.
Doyle, J. J., & Moore, R. J. Attitudes toward the male's role scale (AMR): A n objective in-
strument to measure attitudes toward the male's sex role in contemporary society. JSAS
Catalog o f Selected Documents in Psychology, 1978, 8, 35.
Farrell, W. The liberated man. New York: Bantam Books, 1974.
 350 O'Neil, Helms, Gable, David, and Wrightsman

Fasteau, M. F. The male machine. New York: McGraw-Hill, 1974.

Garnets, L., & Pleck, J. H. Sex role identity, androgyny, and sex role transcendence: A sex
role strain analysis. Psychology o f Women Quarterly, 1979, 3,270-283.
Hays, H. R. The dangerous sex: The myth o f feminine evil. New York: Pocket Books, 1964.
Homey, K. Feminine psychology. New York: W. W. Norton, 1967.
Johnson, R. A. He: Understanding masculine psychology. New York: Harper & Row, 1977.
Jung, C. G. Animus and anima. Collected works (Vol. 7). New York: Pantheon, 1953.
Jung, C. G. Concerning the archetypes, with special reference to the anima concept. Collected
works (Vol. 9, Part I). New York: Pantheon, 1954.
Komarovsky, M. Dilemmas o f masculinity: A study o f college youth. New York: W. W. Nor-
ton, 1976.
Lederer, W. The fear o f women. New York: Harcourt, Brace and Jovanovich, 1968.
Levinson, D. J., Darrow, C. H., Klein, E. B., Levinson, M. H. & McKee, B. The seasons o f a man's
life. New York: Ballantine Books, 1978.
Menninger, K. Love against hate. New York: Harcourt, Brace and Jovanovich, 1970.
Nichols, J. Men's liberation: A new definition o f masculinity. New York: Penguin Books, 1975.
Nunnally, J. C. Psychometric theory (2nd Ed.). New York: McGraw-Hill, 1978.
O'Neil, J. M. Male sex-role conflicts, sexism, and masculinity: Psychological implications for
men, women and the counseling psychologist. The Counseling Psychologist, 1981, 9,
61-80. (a)
O'Neil, J. M. Patterns of gender role conflict and strain: The fear of femininity in men's lives.
The Personal and Guidance Journal, 1981, 60, 203-210. (b)
O'Neil, J. M. A measure o f men's fear o f femininity." A rationale and initial statement. Paper
presented at the annual convention of the American College Personnel Association
(ACPA), Cincinnati, Ohio, March 1981. (c)
O'Neil, J. M. Gender and sex role conflict and strain in men's lives: Implications for psychiatrists,
psychologists, and other human service providers. In K. Solomon & N. Levy (Eds.), Men
in transition: Theory and therapy. New York: Plenum Press, 1982.
O'Neil, J. M., & Fishman, D. Adult men's career transitions and gender role themes. In Z.
Liebowitz & D. Lea (Eds.), Adult career development: Concepts, issues, and practices.
Alexandria, Va.: American Association for Counseling and Development Press, in press.
O'Neil, J. M., Helms, B. A., Gable, R., David, L., & Wrightsman, L. Data on college men's
gender role conflict and strain. Storrs, Conn.: University of Connecticut, Department
of Educational Psychology, 1985. (ERIC Document Reproduction Service No. 248488)
Pleck, E. H., & Pleck, J. The American man. Englewood Cliffs, N.J.: Prentice Hall, 1980.
Pleck, J. The myth o f masculinity. Cambridge, Mass.: The MIT Press, 1981.
Pleck, J. H., & Sawyer, J. Men and masculinity. Englewood Cliffs, N.J.: Prentice Hall, 1974.
Rummel, R. J. Applied factor analysis. Evanston, Ill.: Northwestern University Press, 1970.
Scher, M. (Ed.). Counseling men [Special Issue]. Personnel and Guidance Journal, 1981, 60(4).
Skovholt, T., Gormally, A., Schauble, P., & Davis, R. Counseling men. Monterey, Calif.:
Brooks/Cole, 1980.
Solomon, K., & Levy, N. Men in transition: Theory and therapy. New York: Plenum Press, 1982.
Spence, J. T., & Helmreich, R. L. The attitudes toward women scale: An objective instrument
to measure attitude toward rights and roles of women in contemporary society. JSAS
Catalog), o f Selected Documents in Psychology, 1972, 2, 66.
Spence, J. T., & Helmreich, R. L. Masculinity and femininity: Their psychological dimensions,
correlates, and antecedents. Austin, TX: University of Texas Press, 1978.
Spence, J. T., & Helmreich, R. L. Masculine instrumentality and feminine expressiveness: Their
relationship with sex role attitudes and behaviors. Psychology o f Women Quarterly, 1980,
5, 147-163.
Spence, J. T., Helmreich, R. L., & Stapp, J. The personal attributes questionnaire: A measure
of sex-role stereotypes and masculinity-femininity. JSAS Catalog o f Selected Documents
in Psychology, 1974, 4, 43.
Spence, J. T., Helmreich, R. L., & Stapp, J. Ratings of self and peers on sex role attributes
and their relation to self-esteem and conceptions of masculinity and femininity. Journal
o f Personality and Social Psychology, 1975, 32, 29-39.
Vaillant, G. E. Natural history of male psychological health. Archives o f General Psychiatry,
1974, 31, 15-22.

View publication stats