The Multiple Dimensions of Gender Stereotypes: A Current Look at Men's and Women's Characterizations of Others and Themselves

The study investigates the current state of gender stereotypes using a multidimensional view of the core constructs of agency and communality. Motivated by persistent gender inequality in pay and leadership positions and mixed evidence about changes in stereotype content over time, the authors argue that inconsistencies in prior findings may stem from measuring different facets of agency and communality across studies. The research aims to: (1) develop and validate a multidimensional framework distinguishing specific dimensions of agency and communality; (2) examine how male and female raters currently characterize men and women; (3) assess how men and women characterize themselves; and (4) compare self-characterizations to characterizations of one's gender group. The overarching purpose is to provide a nuanced picture of contemporary stereotypes and their internalization, clarifying where stereotypes persist or have shifted.

The paper reviews theory and evidence on gender stereotypes as generalizations rooted in social role distributions (social role theory), wherein men are viewed as agentic (taking charge, controlling) and women as communal (relational, nurturing). Agency and communality have been operationalized variously (e.g., competence vs dominance, warmth vs morality), leading to inconsistent results regarding stereotype change. Prior measures span explicit trait ratings, implicit associations, and distinctions among traits, roles, occupations, and emotions. Research shows self-stereotyping contributes to gender identity formation, with evidence both for persistence and some shifts over time in self-perceived agency/communality. The authors highlight mounting evidence that agency and communality are multifaceted (e.g., competence separate from other agentic aspects; communion subdivided into warmth and morality) and propose a more granular approach. They also consider rater gender effects, arguing women may be more receptive to evolving roles while men may resist changes. Finally, they note theoretical reasons (attribution and construal level theories) for possible divergence between self- and group-based characterizations, with mixed prior empirical support.

Design: Experimental between-subjects design with two independent variables: rater gender (male, female) and target group (men in general, women in general, or self). Target was randomly assigned. Participants: 629 U.S. MTurk participants (61% female), ages 19–83 (M = 34.5, SD = 13.1); education: 17% no college, 33% some college, 37% college grads, 13% graduate degrees; race/ethnicity: 77.6% White, 8.4% Asian, 7.0% African American, 4.8% Hispanic, 2.2% other. One respondent who reported responding dishonestly was excluded (final N = 628). Inclusion criteria included >95% prior MTurk acceptance rate. Procedure: Participants rated either men in general (N = 215), women in general (N = 208), or themselves (N = 205) on a 7-point scale (1 = not at all to 7 = very much) using an attribute inventory assessing multiple dimensions of agency and communality. Item order was randomized. Scale construction: Inductive, four-step process. (1) Item pool of 74 attributes drawn from classic and contemporary stereotyping measures to broadly capture agentic and communal content. (2) Three expert judges sorted items into conceptually distinct categories; consensus yielded seven categories: four agency dimensions (instrumental competence, leadership competence, assertiveness, independence) and three communality dimensions (concern for others, sociability, emotional sensitivity). Items lacking consensus were dropped and categories labeled. (3) Three new independent judges re-sorted items into labeled categories; misclassified items were eliminated. (4) Confirmatory factor analysis (CFA) refinement removed poorly fitting items. Final scales had 3–4 items each, with coefficient alphas > 0.75 and corrected item–total correlations > 0.40. Reported alphas: instrumental competence α = 0.88; leadership competence α = 0.80; assertiveness α = 0.80; independence α = 0.82; concern for others α = 0.91; sociability α = 0.77; emotional sensitivity α = 0.75. Overall scales were also computed: overall agency (15 items; α = 0.93) and overall communality (11 items; α = 0.93). CFA: Agency four-factor model fit: χ²(84) = 370.224, p < 0.001; χ²/df = 4.41; CFI = 0.947; RMSEA = 0.076; SRMR = 0.045; superior to one-factor model χ²(90) = 813.318; CFI = 0.866; RMSEA = 0.116; SRMR = 0.068; Δχ²(6) = 443.09, p < 0.001. Communality three-factor model fit: χ²(41) = 326.000, p < 0.001; CFI = 0.931; RMSEA = 0.108; SRMR = 0.048; superior to one-factor model χ²(44) = 359.803; CFI = 0.924; RMSEA = 0.110; SRMR = 0.048; Δχ²(3) = 33.80, p < 0.001. Despite high intercorrelations, dimensions were empirically distinguishable. Preliminary analyses: Age (≤39 vs ≥40) and education (college grad/advanced vs not) showed no main or interactive effects on outcomes; data were pooled across these factors. Analytic strategy: ANOVAs addressed three questions. For stereotypes about others: 2 (rater gender) × 2 (target group: men vs women in general) ANOVAs on overall scales and mixed-model ANOVAs with within-subjects factor of dimension (4 agency; 3 communality). For self-characterizations: one-way ANOVAs by self-rater gender on overall scales plus mixed-model ANOVAs with dimension as within-subjects factor. For self versus same-gender group: 2 (rater gender) × 2 (target: self vs same-gender in general) ANOVAs on overall scales plus mixed-model ANOVAs with dimension as within-subjects factor. Planned comparisons used Fisher’s LSD.

Validation of multidimensional framework: - CFA supported four distinct agency dimensions and three distinct communality dimensions, each outperforming single-factor alternatives (agency Δχ²(6) = 443.09, p < 0.001; communality Δχ²(3) = 33.80, p < 0.001). Do men and women differ in their gender stereotypes (others’ ratings): - Agency (2×2×4): Main effects of rater gender, F(1,418) = 15.55, p < 0.001, η² = 0.04; target group, F(1,418) = 5.51, p = 0.019, η² = 0.01; dimensions, F(3,1131) = 9.49, p < 0.001; Dimensions × Target Group interaction, F(3,1131) = 23.65, p < 0.001. • Male raters: Women in general rated lower than men in general on overall agency (p = 0.013) and on leadership competence (p = 0.024), assertiveness (p < 0.001), and independence (p = 0.002), but not instrumental competence (p = 0.805). • Female raters: No difference in overall agency (p = 0.512); women lower than men on assertiveness (p = 0.004) but equal on leadership competence (p = 0.620), independence (p = 0.776), and instrumental competence (p = 0.216). • Rater-gender differences: Women in general were rated lower in agency by male vs female raters across all four agency dimensions; ratings of men in general largely similar except female raters rated men higher in leadership competence than male raters (p = 0.040). - Communality (2×2×3): Strong target group effect, F(1,418) = 93.10, p < 0.001, η² = 0.18; interactions indicate some rater-gender and dimension nuances. • Both male and female raters rated women in general higher than men in general on overall communality and on concern for others, sociability, and emotional sensitivity (all p < 0.001). • Female raters rated women in general higher than male raters did on overall communality (p = 0.036), concern for others (p = 0.048), and emotional sensitivity (p = 0.029); sociability showed no rater-gender difference for women in general. Do men and women differ in their self-characterizations: - Agency: Overall agency did not differ by self-rater gender (p = 0.215). However, men rated themselves higher in leadership competence (p = 0.012) and assertiveness (p = 0.017), with no gender differences in instrumental competence (p = 0.961) or independence (p = 0.910). - Communality: Women rated themselves higher overall (p = 0.013); specifically higher in concern for others (p = 0.009) and emotional sensitivity (p = 0.002), with no gender difference in sociability (p = 0.223). Do self-characterizations differ from characterizations of one’s gender group: - Men: Self rated more agentic overall than men in general (p = 0.046); higher in instrumental competence (p < 0.001) and independence (p = 0.039), no differences in leadership competence (p = 0.180) or assertiveness (p = 0.353). Self rated more communal overall (p < 0.001) and on all dimensions: concern for others (p < 0.001), sociability (p < 0.001), emotional sensitivity (p = 0.001). - Women: Overall agency self vs women in general did not differ (p = 0.883), but self rated higher in instrumental competence (p = 0.038) and lower in leadership competence (p = 0.004) and assertiveness (p = 0.016); independence marginal (p = 0.051). Overall communality did not differ (p = 0.367); self rated higher in concern for others (p = 0.008), with no differences in sociability (p = 0.943) or emotional sensitivity (p = 0.539). Illustrative means (Table 4/6): For others’ ratings, male raters: overall agency men 4.58 vs women 4.15; female raters: men 4.85 vs women 4.75. Overall communality male raters: men 4.01 vs women 4.86; female raters: men 4.04 vs women 5.17. For self-ratings, men vs women: leadership competence 4.92 vs 4.45; assertiveness 4.56 vs 4.14; concern for others 5.13 vs 5.60; emotional sensitivity 4.64 vs 5.20 (all on 1–7 scales).

The findings address the research questions by showing that while communal stereotypes remain robust for both male and female perceivers, agency-related stereotypes are more nuanced and depend on the specific dimension assessed and on rater gender. Male raters continued to view women as less agentic than men across most agency facets, whereas female raters perceived women as equally independent and leadership-competent as men but less assertive. Self-characterizations mirrored traditional stereotypes in selective ways: women self-rated lower in leadership competence and assertiveness but not in instrumental competence or independence; men self-rated lower in communality overall but not in sociability. Crucially, comparisons of self versus group revealed asymmetries: men portrayed themselves as more communal than men in general (and more agentic on some facets), suggesting reduced stereotype adherence or self-enhancement in self-views. In contrast, women portrayed themselves as less assertive and less leadership-competent than women in general, indicating stronger internalization of stereotype-consistent deficits in power-related agency facets when evaluating self. These patterns underscore the value of a multidimensional approach: overall agency and communality scores would have obscured key differences concentrated in assertiveness and leadership competence. The results highlight the persistence of gendered expectations with implications for evaluations, career advancement, and self-limitation, and they suggest that power-linked aspects of agency are a critical locus where stereotypes continue to constrain perceptions, especially among women.

The study develops and validates a multidimensional framework for agency (instrumental competence, leadership competence, assertiveness, independence) and communality (concern for others, sociability, emotional sensitivity) and demonstrates its utility for clarifying contemporary gender stereotypes in others’ and self-judgments. Key contributions include evidence that communal stereotypes remain strong, and that agency stereotypes vary by facet and rater gender: women are seen (and see themselves) as equally independent and instrumentally competent as men but lower in assertiveness and leadership competence. Self versus group comparisons further reveal that men depict themselves as more communal than men in general, whereas women self-describe more negatively than women in general on power-related agency. Future research should: (1) refine competence facets, including consideration of intellectual brilliance; (2) investigate differential predictive validity of specific agency/communality dimensions for outcomes such as selection, evaluation, rewards, and career aspirations; (3) broaden item pools to include morality within communality and potential cognitive agency facets; (4) test moderators (e.g., race, SES) and conduct cross-cultural and intersectional replications; and (5) examine mechanisms behind women’s lower self-assessments in assertiveness and leadership competence and men’s higher self-ascribed communality.

- Scale construction followed a top-down expert-driven process and relied on existing items; alternative bottom-up approaches and inclusion of additional facets (e.g., morality within communality; cognitive agency) may yield different or broader dimensions. - MTurk convenience sample of U.S. residents limits generalizability; replication with representative and cross-cultural samples is needed. - Potential moderators (e.g., race/ethnicity, socio-economic status) were not examined due to sample size constraints in subgroups. - Communality and agency scales were intercorrelated; while CFA supported distinct facets, overlap may still influence interpretations. - Targets were broad categories (men/women in general); intersectional subtypes were not assessed.