Escolar Documentos
Profissional Documentos
Cultura Documentos
4, 1983
'The authors would like to thank Dr. Sandra Bem for sending them the short BSRI which
was, at the time of this study, nol generally available. Order of aulhorship is alphabetical;
both authors contributed equally to the project.
2Correspondence should be sent to lhe first author, Deparlmenl of Psychology, The
College of Wooster, Wooster, Ohio 44691.
'Now a! the University of North Carolina School of Law.
441
0360 0025/83/0400 044I$03.00/() ]983 Plezlunl Put~lishirl~ ('orporatil~n
442 McPherson and Spetrino
(Bem, 1974, 1975, 1977, 1979; Spence & Helmreich, 1978, 1979). The Bern
Sex Role Inventory (BSRI; Bern, 1974, 1979) is widely used in the assess-
ment of androgyny despite controversy over its construction and scoring
(e.g., Pedhauzer & Tetenbaum, 1979; Spence, Helmreich, & Stapp, 1975).
As a result of discussion in the literature, Bem (1979) revised the BSRI to
eliminate items correlated with sex and feminine items found to be low in
social desirability and/or not loading on the feminine factor identified in
various factor-analytic studies of the instrument (e.g., Gaudreau, 1977).
The newer, short form of the BSRI contains the 10 items that represent the
most desirable personality characteristics usually associated with each sex
(Bern, 1979) and 10 filler items not found to be sex-linked (Pedhauzer &
Tetenbaum, 1979; Walkup & Abbott, 1978).
At the same time that Bern (1979) discussed her revision of the BSRI,
she also made explicit various assumptions that have been implicit in her
research. She asserts that sex-typed and androgynous individuals, in
addition to differing in the extent to which they have internalized socially
prescribed gender roles, also differ in the extent to which they believe
the sexes are basically different. She suggests that this difference in beliefs
and expectations regarding a basic "gender polarity" leads to both the dis-
crepancy in subjects' self-ratings on the BSRI and, consequently, the
hypothesized differences in their behavior. She also contends that as a result
of this variation in belief in gender polarity, androgynous and sex-typed
persons will differ in the extent to which they utilize sex as a factor in
processing information, making personality attributions, and inferring
causality of behavioral outcomes. These cognitive differences are believed
by Bern to represent the fundamental difference between androgynous and
sex-typed individuals, rather than variation in degrees of femininity and
masculinity.
Two recent investigations provide initial support for the hypothesis
that sex-typed individuals will more readily differentiate along a sex-
related dimension. Deaux and Major (1977) found that sex-typed subjects
differentiated between male and female actors more than did androgynous
subjects in a unitization task, and Lippa (1977) found that sex-typed
subjects utilized a masculinity-femininity dimension more than did an-
drogynous subjects in a handwriting analysis study. However, whether
androgynous and sex-typed individuals differ in the content of their
beliefs regarding what the two sexes are like has not been empirically
tested.
Bem's hypotheses suggest that androgynous and sex-typed individuals
should differ in the kinds of personality assumptions they make about
women and men. Specifically, sex-typed individuals should believe that
Androgynyand Sex-Typing 443
METHOD
Subjects
Design
Materials
The short form of the BSRI was used to initially classify subjects
as sex-typed or androgynous. At a later date the same form was admini-
stered with the directions altered so that subjects were asked to imagine the
ideal woman or the ideal man and to rate her or him on the character-
istics provided.
Procedure
RESULTS
Descriptive Information
Cell means for the dependent measure, ratings of the ideal woman
and ideal man, are presented in Table I. Although Bern assigns values
Androgyny and Sex-Typing 445
The major prediction was that sex-typed subjects would rate the
ideal man and woman differently, while androgynous subjects would
not. This pattern was expected to hold for subjects of both sexes. Or-
thogonal contrasts of means employing t-ratio tests of significance (Kirk,
1968) were performed to test this hypothesis. Results revealed that, as
predicted, the difference between ratings of ideal men (X = .90) and
ideal women (X = 1.49) made by androgynous subjects was not significant.
Contrary to prediction, the difference between ratings of ideal men (X =
1.1) and ideal w o m e n (X = 2.14) made by sex-typed subjects also failed
to reach significance, although it was larger than the difference for an-
drogynous subjects. This is also reflected in the nonsignificant sex role x
target sex interaction.
When sex of subject was considered, however, an interesting pattern
emerged. Androgynous women's ratings of ideal men and women did not
differ, while, contrary to prediction, the difference between androgynous
men's ratings of ideal men and women did differ; t(72) = 3.03, p < .01.
On the other hand, while masculine men's ratings of the ideals differed,
t(72) = 4.24, p < .001 as hypothesized, there was no significant dif-
ference between feminine women's ratings. In other words, the subject sex
x target sex interaction breaks down to reveal that rather than sex-typed
individuals rating a significant difference between ideal women and men
and androgynous persons not doing so, the difference is between men and
women. When considered separately, neither group of women rated a sig-
nificant difference between ideal men and women, while both groups of
men did. Cell means collapsed over sex role are presented in f a b l e Ill.
These collapsed means were further compared using Tukey's honest
significant difference (HSD) method. Men rated the ideal woman to be
more feminine than women did (p < .01), and also more feminine than
they rated the ideal man to be (p < .01). Women, on the other hand, rated
the ideal man as more feminine than men did (p < .01), and also more
feminine than the ideal woman, although this latter difference barely
reached significance at the .05 level and is based on twice the n as the major
analyses.
DISCUSSION
x=
<
0
L
0 {,-,4
- . - . ".--" II -~ II ~,=
450 MePherson and Spetrino
CONCLUSION
Regardless of the apparent feminine bias in the short BSRI and despite
the tendency for absolute scores to fall into the feminine category, all
analyses performed in this study involved relative differences, so that the
findings are still significant. The hypothesis derived from Bem's work was
not supported, but the finding that women are less likely than men to use
implicit personality schema based on traditional sex stereotypes is very
intriguing and should stimulate further research. This finding lends support
to the argument that being classified as either androgynous or sex-typed
does not necessarily mean the same thing for women and men (e.g., Jones,
Chernovitz, & Hansson, 1978). Sex differences are perhaps more im-
portant than androgyny researchers have realized, particularly during a
time of change.
It also seems that conceptions of the desirability of certain traits may
be changing rapidly and reassessment of trait desirability should be carried
out regularly if instruments such as the BSRI are to maintain validity over
time. The suggestion of a feminine bias in the short BSRI indicates that
large-scale testing of the instrument is necessary.
REFERENCES
Gilbert, L., Deutsch, C., & Strahan, R. Feminine and masculine dimensions of the typical,
desirable, and ideal woman and man. SexRoles, 1978, 4, 767-778.
Gonen, J., & Lansky, L. Masculinity, femininity, and masculinity-femininity: A phenom-
enological study of the MF scale of the MMPI. Psychological Reports, 1968, 23,
183-194.
Jenkin, N., & Vroegh, K. Contemporary concepts of masculinity and femininity. Psy-
chological Reports, 1969, 25, 679-697.
Jones, W., Chernovitz, M., & Hansson, R. The enigma of androgyny: Differential im-
plications for males and females? Journal of Consulting and Clinical Psychology,
1978, 46, 298-313.
Kirk, R. Experimental design: Procedures" for the behavioral sciences. Belmont, Calif.:
Brooks/Cole, 1968.
Lippa, R. Androgyny, sex-typing, and the perception of masculinity-femininity in hand-
writing. Journal of Research in Personality, 1977, 11, 21-37.
Locksley, A., & Colten, M. Psychological androgyny: A case of mistaken identity. Journal of
Personality andSocialPsychology, 1979, 37, 1017-1031.
Orlofsky, J., Aslin, A., & Ginsburg, S. Differential effectiveness of two classification
procedures on the Bern Sex Role Inventory. Journal of Personality Assessment,
1977, 41, 414-415.
Pedhauzer, E., & Tetenbaum, T. Bern Sex Role Inventory: A theoretical and methodological
critique. Journal of Personality and Social Psychology, 1979, 3 7, 996-1016.
Sherman, J. A. Social values, femininity, and the development of female competence.
Journal of Social Issues, 1976, 32, 181-195.
Spence, J., & Helmreich, R. Masculinity and femininity: Their psychological dimensions,
correlates, and antecedents. Austin: University of Texas Press, 1978.
Spence, J., & Helmreich, R. The many faces of androgyny: A reply to Locksley and Colten.
Journal of Personality and Social Psychology, 1979, 3 7, 1032-1046.
Spence, J., Helmreich, R., & Stapp, J. Ratings of self and peers on sex role attributes
and their relation to self-esteem and conceptions of masculinity and femininity.
Journal of Personality and Social Psychology, 1975, 32, 29-39.
Walkup, H., & Abbott, R. Cross-validation of item selection on the Bem Sex Role Inventory.
Applied PsychologicalMeasurement, 1978, 2, 63-71.