Effects of Three Conditions of Administration on Bem Sex Role Inventory Scores

In order to assess the validity of the technique employed in selecting items for the Bern Sex Role Inventory (BSRI) 58 male and 62 female undergraduates took the BSRI under three different sets of instructions. Both males and females were able to produce extremely masculine and extremely feminine BSRI profiles which differed (p < .01) from their own BSRI profiles. The results support Bem's (1974) assertion that the BSRI Masculinity and Femininity scales are comprised of items which are consistent with widely held sex-role sterotypes.     

Equating the social desirability of Bem Sex-Role Inventory masculinity and femininity subscales

Two experiments investigated the social desirability of Bem Sex-Role Inventory (BSRI) items. Mean social desirability of the masculinity subscale was significantly higher than that of the femininity subscale in both experiments. A revision of the BSRI femininity subscale was suggested which replaced three items of questionable social value with three more socially desirable items and virtually eliminated the social desirability difference between masculinity and femininity subscales.     

Construct validity of the bem sex role inventory (BSRI): Does the BSRI distinguish between gender-schematic and gender-aschematic individuals?

Factor analyses were performed on the masculine, feminine, and neutral items of the Bem Sex Role Inventory (BSRI) for gender-schematic and gender-aschematic individuals separately. The analyses revealed bipolar masculinity/femininity factors for the gender-schematic group and two unipolar “aschematicity” factors for the gender-aschematic group. The results support gender schema theory's claim that the two types of individuals respond differently to the items of the BSRI. Thus, the BSRI is a valid measure for distinguishing gender-schematic from gender-schematic individuals.     

Factor and discriminant analysis of the Bem Sex-Role Inventory

Several studies have recently recommended that items which fail to load substantially (< .40) on one of several primary factors characterizing the Bem Sex-Rote Inventory (BSRI) could be deleted without substantial loss of interpretability. This recommendation is questioned on the basis that the extent of interrelatedness of the items is an insufficient indicator of the contribution each item makes towards the classification (identifying sex-types by Bem's median split technique) efficiency of the inventory. Empirical support of this contention was provided by conducting several discriminant analyses (predicting sex-types) on a sample of 1117 BSRI protocols and excluding the various items suggested by the earlier research. These results were then compared with the results of a discriminant analysis using the 40 masculine and feminine items. The analyses using less. than all 40 items produced substantially poorer discrimination than when all 40 items were included. The results Suggest that decisions to exclude items on the basis of factor analysis may be inappropriate and risk serious loss of power in classifying individuals by sex-type.     

Reexamination of the Bem Sex-Role Inventory: The Interpersonal BSRI

The 40 masculine and feminine items of 746 Bern Sex-Role Inventory (BSRI; Bern, 1974) were subjected to a principal axis factor analysis with varimax rotation. Results revealed a four-factor solution, indicating that the original BSRI items do not provide two unidimensional ratings of femininity and masculinity. However, some support is given to Bern's initial two-factor thesis in that two principal factors emerged: one consisting of items that were originally identified as feminine, the other consisting of items that were originally identified as masculine. Content examination of the two principal factors resulted in an interpretation of the factors as Interpersonal Sensitivity and Interpersonal Potency. Internal consistency and validity of these two scales were examined and found to be adequate.     

INSTRUMENTAL AND EXPRESSIVE TRAITS, TRAIT STEREOTYPES, AND SEXIST ATTITUDES What Do They Signify?

College students rated die typical male and female student and themselves on 22 instrumental (I) and 16 expressive (E) items from the PAQ (Spence & Helmreich, 1978) and the BSRI (Bern, 1974), as well as on the BSRI items "masculine" and "feminine." They also completed measures of gender stereotypes and sexist attitudes. Significant gender stereotypes were found on all but two I and E items in both genders. Significant gender differences in self-report were found on all the E items but on only 41% of the I items, confirming our hypotheses that societal changes have led women to develop more agentic self-conceptions. The pattern of relationships found between the self-report, stereotype, and attitude measures supports the utility of a multidimensional approach to gender. Responses to the items "masculine" and "feminine" confirm the implications of our hypothesis that these items primarily assess men's and women's basic sense of gender identity.     

Effects of instructions and rating scales on item selection for the BSRI scales

Walkup and Abbott (1978) stated that Edwards and Ashworth's (1977) failure to replicate Bem's (1974) selection of items for the Masculinity and Femininity Scales of the Bern Sex Role Inventory (BSRI) may be attributed to differences in the instructions and anchored rating scales used in the two studies. The present study tested the hypothesis that presence of various interaction effects involving instructions and rating scales would influence the acceptability of items for the BSRI Masculinity and Femininity Scales. Results based on the evaluation of individual items by Bem's item selection criteria in each of the four experimental conditions obtained by systematically manipulating two instructions (Bem's and Edwards' instructions) and two rating scales (Bem's and Edwards' rating scales) and also those based on the analysis of variance of item mean desirability ratings from the four experimental conditions supported the hypothesis.     

Effects of Instructions and Rating Scales on Item Selection for the BSRI Scales

Walkup and Abbott (1978) stated that Edwards and Ashworth's (1977) failure to replicate Bem's (1974) selection of items for the Masculinity and Femininity Scales of the Bern Sex Role Inventory (BSRI) may be attributed to differences in the instructions and anchored rating scales used in the two studies. The present study tested the hypothesis that presence of various interaction effects involving instructions and rating scales would influence the acceptability of items for the BSRI Masculinity and Femininity Scales. Results based on the evaluation of individual items by Bem's item selection criteria in each of the four experimental conditions obtained by systematically manipulating two instructions (Bem's and Edwards' instructions) and two rating scales (Bem's and Edwards' rating scales) and also those based on the analysis of variance of item mean desirability ratings from the four experimental conditions supported the hypothesis.     

Evaluation of the BSRI masculine and feminine items using desirability and stereotype ratings

This study empirically tested Bem's (1974) assumption that the BSRI Masculinity and Femininity scales measure sex-typed standards of desirable behavior for men and women in American society. The adequacy of items in the two scales was evaluated by Bem's (1974) criteria using two types of desirability and stereotype ratings. Results obtained in all the experimental conditions except one involving Bem's desirability instructions and Bem's rating scale did not support the tested assumption. Implications of these results for revising the BSRI scales were discussed.     

Gender schema theory and sex role inventories: some conceptual and psychometric considerations

In this study we assessed whether the Bem Sex Role Inventory (BSRI) and the PRF ANDRO are appropriate for investigations of gender schema theory (Bem, 1981a). Because these instruments were developed for entirely different theoretical purposes, it is important to empirically examine the validity of these measures for investigating the construct of gender schema. On the basis of the propositions of gender schema theory, we made several predictions about the psychometric properties that should be exhibited by a valid measure of this construct. Responses to the PRF ANDRO and the BSRI were factor analyzed separately for sex-typed and non-sex-typed groups. Results show consistent and theoretically predictable differences in the factor solutions of these two groups. The sex-typed or gender-schematic group obtained bipolar factors, with masculine items loading with one sign, whereas feminine items loaded with the other sign on each factor. Also, sex of subject loaded highly on almost every factor for this group. The non-sex-typed group, however, obtained few such distinctly dichotomous factors, and sex of subject loaded only on the weaker factors. Results are interpreted as providing support for the construct validity of at least the BSRI and the PRF ANDRO for use in researching the implications of this approach.     

THE FACTOR STRUCTURE OF THE BEM SEX-ROLE INVENTORY (BSRI)

Research on the factor structure of the Bem Sex-Role Inventory generally indicates the existence of a number of smaller factors in addition to a major Masculine factor and a major Feminine factor. This has led to a suggestion that contrary to Bern's proposal, the scale is multidimensional. The present study attempted to examine further the factor structure of BSRI when applied to a non-Western sample. The results of a factor analysis appear to confirm the suggestion that BSRI is multidimensional. There are also indications that both the Masculine and the Feminine scales have high internal consistencies which can be improved further if certain items were removed.     

Desirability of the Bem Sex-Role Inventory Items for Women and Men: A Comparison Between African Americans and European Americans

The Bem Sex-Role Inventory (BSRI) is a widely used instrument for measuring gender role perceptions, but questions have been raised regarding whether its items reflect contemporary views on gender. A recent study (Holt & Ellis, 1998) revalidated all but 2 of the 40 adjectives included in the masculine and feminine BSRI indices for a predominantly European American undergraduate sample in a rural Southern town. We examined whether European Americans in a different geographical area and 2 samples of African Americans would show similar findings. Study participants were recruited in undergraduate management courses in 2 universities and included 62 European American women, 69 European American men, 40 African American women, and 31 African American men in a large Northeastern city and 56 African American women and 33 African American men in a small Southern city. Findings indicated that European American men in the urban Northeast and African American men in the South gave the most traditional ratings, whereas European American women in the urban Northeast expressed the most liberal views. European American women considered only 4 of the 40 BSRI items to be differentially desirable for women and men, a considerable departure from the findings of Holt and Ellis (1998) as well as Bem (1974).     

Using Reliability Generalization Methods to Explore Measurement Error: An Illustration Using the MMPI–2 PSY–5 Scales

This study empirically tested Bem's (1974) assumption that the BSRI Masculinity and Femininity scales measure sex-typed standards of desirable behavior for men and women in American society. The adequacy of items in the two scales was evaluated by Bem's (1974) criteria using two types of desirability and stereotype ratings. Results obtained in all the experimental conditions except one involving Bem's desirability instructions and Bem's rating scale did not support the tested assumption. Implications of these results for revising the BSRI scales were discussed.     

Gender and gender role differences in student–teachers’ commitment to teaching

Low commitment to teaching amongst teachers is a problem facing the teaching profession in many countries. Gender might be an important factor in explaining what kinds of prospective teachers are attracted to teaching. This empirical study examined the relationship between student–teachers’ gender, gender roles and commitment to teaching within the context of a large, university-based teacher education program in Tanzania. A self-report questionnaire was administered, comprising commitment to teaching items, gender as a demographic variable and items from Bem sex role inventory (BSRI). Cluster analysis on the masculine and feminine scale of the BSRI indicated three clusters: (1) highly androgynous students with high scores on both masculine and feminine scales, (2) medium androgynous students with relatively high scores on both masculine and feminine scales, and (3) low androgynous students with low scores in relation to the other two groups on both masculine and feminine scales of the BSRI. More female than male student–teachers reported to be highly androgynous. Covariance analysis showed that gender roles were significantly related to commitment to teaching and to intention to enter the teaching profession. Highly androgynous student–teachers reported significantly more commitment to teaching and higher intentions to enter the teaching profession compared to medium and low androgynous student–teachers. No significant relationships were found between gender, on the one hand, and commitment to teaching and the intention to enter the teaching profession, on the other hand. Findings are discussed in the context of teacher and teacher education in Tanzania.     

Self-rated adjustment and sex-typing on the Bem Sex-Role Inventory: Is masculinity the primary predictor of adjustment?

Questions concerning the relationship between self-rated adjustment and the Bem Sex-Role Inventory (BSRI) were examined in two studies. In Study I, whether sex-typing was defined by Bem's original t-test criterion or the newer median-split criterion, superior adjustment was associated with androgynous vs. traditional typing only among women, not men. Also in conflict with the androgyny position, adjustment differences among sex-types were accounted for by differences in masculinity, not in femininity or androgyny per se. The greater contribution of masculinity vs. femininity to self-rated adjustment may be exaggerated because BSRI femininity includes items which do not load on a unidimensional femininity factor. In Study II, inspection of these unrelated items suggested that they reduce the desirability of BSRI femininity, thereby reducing its relationship to adjustment. Comparisons between standard BSRI scoring and an alternative based on unidimensional subscales revealed that the alternative femininity subscale was judged to be more desirable, and it reduced the degree of difference between masculinity and femininity in their relationships to adjustment. Superior adjustment, however, was again associated with androgynous vs. traditional sex-typing only among women, and adjustment differences among sex-types were again accounted for by differences in masculinity.The authors acknowledge the assistance of the following persons who helped in the administration, scoring, or preparation of questionnaire materials: Jill Jamison, Eric Isberg, Kay Peter, Steve Zalinger, Walter Davis, James Ball, Mark Jensen, and Heidi Sanft. Correspondence should addressed to Dr. Louise Silvern, Psychology Department, University of Colorado, Boulder, Colorado 80309.     

Gender-schematic processing: An extended replication of the Larsen and Seidman study

This study is an extended replication of the Larsen and Seidman study (1986) on the Bem Sex-Role Inventory (BSRI) as a measurement instrument for gender-schematic processing. Using a Dutch sample and the GRAS (Groninger Androgyny Scale)—a Dutch sex-role inventory—responses to this inventory were factor-analysed separately for sex-typed and non-sex-typed groups. The extremity of response style on GRAS items and the degree of self-indicated cross-situational variability on GRAS items were used as additional indices for gender-schematic processing. Similar to Larsen and Seidman's (1986) study, bipolar factors were found for the sex-typed groups and unipolar factors for the non-sex-typed groups. Similar to a study conducted by Bem (1981) into differences among the sex-types in their response latencies to the attributes of the BSRI, sex-typed subjects showed more extremity of response style on attributes of the GRAS when making schema-consistent judgements about themselves, and showed less extremity of response style when making schema-inconsistent judgements about themselves. The data on cross-situational variability were less equivocal. It is concluded that the degree of gender-schematic processing in individuals can be well measured by means of a sex-role inventory.     

Androgyny and sex-typing: Differences in beliefs regarding gender polarity in ratings of ideal men and women

To test Bem's (1979) hypothesis that androgynous and sex-typed individuals are differentiated by the presence or absence of beliefs in “gender polarity,” 200 students rated themselves on the short Bem Sex Role Inventory (BSRI). Of these 200, 20 androgynous women, 20 androgynous men, 20 feminine sex-typed women, and 20 masculine sex-typed men were randomly selected and asked to rate either the ideal woman or the ideal man on the short BSRI. As predicted, androgynous women did not rate the ideal man and woman differently, but, contrary to prediction, neither did feminine-typed women. As predicted, masculine-typed men rated the ideal man and woman differently, but, contrary to prediction, so did androgynous men. Thus, sex rather than sex-role distinguished subjects' beliefs in gender polarity. This finding may reflect women's desires for, and men's fears of, abandoning traditional modes of sex differentiation. The large initial subject pool was necessary because an extremely high percentage of subjects were classified as feminine sex-typed. An apparent feminine bias in the short BSRI is discussed.     

Sex role patterns in an Irish student sample as measured by the bem sex role inventory (comparisons with an American sample)

Sex roles in 171 male and 180 female university students in Ireland were examined by the Bem Sex Role Inventory (BSRI). The distribution of scores for masculinity and femininity showed the same general pattern as in American samples. However, the Irish males rated themselves significantly lower in masculinity and significantly higher in femininity than the American males, indicating a more androgynous profile. Irish females rated themselves significantly lower than American females in both masculinity and femininity, and were less androgynous relative to Irish males. The distribution of subjects on sex role types showed a larger percentage of androgynous males and undifferentiated females in the Irish sample than in American samples. Both the male and the female Irish students scored lower than American students on the Social Desirability Scale. Separate scores were calculated for the positive and the negative items of the Social Desirability Scale, and it was found that positive items were assigned considerably higher ratings than the negative (reversed) items in both sexes. When controlled for social desirability, the masculinity and femininity scores were found to be uncorrelated. Factor analysis of the BSRI revealed six factors, which correspond to previous findings. In general, the present results support the view that masculinity and femininity, as measured by the BSRI, can be treated as separate dimensions, but that a more complex model than the two-factor structure is required.This is a joint report from the Psychosomatic Unit of the Irish Foundation for Human Development, Dublin; the Department of Psychiatry and Psychology, Karolinska Institutet, Stockholm; and the Department of Psychology, University of Stockholm. This collaboration was made possible by a twinning grant to Dr. John Cullen and Professor Marianne Frankenhaeuser from the European Training Programme in Brain and Behavior Research, European Science Foundation. Financial support was also obtained as grants from the Swedish Medical Research Council (Project No. 979) and the Swedish Council for Research in the Humanities and Social Sciences. The authors are indebted to Professor Desmond Swan, University College, Dublin, for his generous support, particularly during the initial stages of our work, and to Professor Liam Ryan, Maynooth College, who also gave valuable assistance.     

Construct Validation of the Bem Sex Role Inventory in Taiwan

It has been over 30 years since the creation of the Bem Sex Role Inventory (BSRI; Bem in Journal of Counseling and Clinical Psychology, 42, 155-162, 1974). Although the BSRI is the most frequently tested instrument for measuring gender orientation, no researchers to date have examined its psychometric properties in Taiwan. Using the short version of the instrument, I asked a total of 381 men and 450 women college students, nurses, police officers, or managers in Taiwan to evaluate themselves on the BSRI. Among the findings are (1) satisfactory internal consistency of the BSRI, (2) more complex structure of masculinity and femininity than originally reported, and (3) doubtful ability of the BSRI to differentiate gender role orientation. These results suggest use of the BSRI in Taiwan with extreme caution before further validation of its applicability is conducted.     

Dimensionality of the bem sex role inventory: A multidimensional analysis

Bem has criticized psychological scales which categorize individuals as masculine or feminine and has developed the Bem Sex Role Inventory (BSRI) to measure androgyny as well as femininity and masculinity. Bem's work on androgyny has stimulated not only a growing research tradition on androgyny but also methodological critiques of the BSRI instrument and scoring procedure. The purpose of this research is to replicate in part a study done by Pedhazur and Tetenbaum which raises serious questions about the unidimensionality of the feminine and masculine subscales in the BSRI. The alternative multidimensional techniques of factor analysis (the method used by Pedhazur and Tetenbaum) and smallest space analysis are applied to the BSRI data set. Both techniques indicate that the feminine and masculine subsets are not unidimensional, but they yield quite different results concerning the actual nature of the dimensions in the BSRI. The results of the factor analysis, but not the smallest space analysis, are consistent with the findings of Pedhazur and Tetenbaum, thus yielding mixed evidence about the validity of their conclusions based solely on factor-analyzed data.The author wishes to acknowledge the constructive comments by Professors Milton Bloombaum, John Gartrell, Louis Guttman, and Martha Mednick during various stages of this research, and the assistance of Dr. Jerry Brennan with the data analysis. Support for this project was provided by the Social Science Research Institute, University of Hawaii at Manoa, Honolulu, Hawaii.