Ambivalent Sexism Inventory

1. Abstract

The Ambivalent Sexism Inventory (ASI), formulated by social psychologists Peter Glick and Susan T. Fiske (1996), represents a seminal paradigm shift within social psychology and gender studies. Moving beyond the historical conceptualization of prejudice strictly as antipathy or overt hostility, the ASI operationalizes sexism as a multidimensional construct characterized by a profound motivational ambivalence. The instrument consists of 22 self-report items scored on a 6-point Likert scale ranging from 0 (disagree strongly) to 5 (agree strongly), bifurcated into two overarching, complementary subscales: Hostile Sexism (HS) and Benevolent Sexism (BS). Hostile Sexism captures adversarial, overtly derogatory attitudes aimed at women who challenge traditional male hegemony, encapsulating dominant paternalism, derogatory beliefs, and competitive gender differentiation. Conversely, Benevolent Sexism embodies a subjectively positive, paternalistic orientation that casts women in stereotyped, restricted roles, venerating them as pure, fragile beings deserving of chivalrous protection and adoration, underpinned by protective paternalism, complementary gender differentiation, and heterosexual intimacy.

Across extensive psychometric investigations, including large-scale cross-national validations spanning dozens of countries, the ASI has established exceptional psychometric properties. Internal consistency typically yields Cronbach’s alpha coefficients between .80 and .92 for Hostile Sexism and between .73 and .85 for Benevolent Sexism. Confirmatory factor analyses systematically support a hierarchical structural model wherein HS and BS represent distinct yet positively correlated factors, each subsuming theoretically meaningful subcomponents. The ASI demonstrates robust convergent, discriminant, and predictive validity, systematically forecasting intimate partner violence, tolerance for sexual harassment, victim blaming in sexual assault scenarios, justification of structural gender inequality, and systemic opposition to progressive socio-political policies. This article provides a comprehensive academic review of the ASI’s conceptual architecture, theoretical foundation, psychometric validation, factor analytical structures, scoring protocols, and administration guidelines.

2. Keywords

Ambivalent Sexism Inventory, Hostile Sexism, Benevolent Sexism, Gender Attitudes, Paternalism, Ambivalent Sexism Theory, System Justification, Stereotype Content Model, Psychometrics, Gender Role Ideology, Structural Inequality, Chivalry

3. Authors

The Ambivalent Sexism Inventory was conceived, validated, and formalized by:

• Peter Glick, Ph.D. — Henry Merritt Wriston Professor of the Social Sciences and Professor of Psychology, Department of Psychology, Lawrence University, Appleton, Wisconsin, United States. Dr. Glick is internationally recognized for his pioneering scholarship on prejudice, stereotyping, and intergroup power dynamics, having authored foundational treatises on cross-cultural gender relations.
• Susan T. Fiske, Ph.D. — Eugene Higgins Professor of Psychology and Professor of Public Affairs, Department of Psychology, Princeton University, Princeton, New Jersey, United States. Dr. Fiske is an eminent figure in social cognition, widely lauded for developing the Stereotype Content Model and her paradigm-defining contributions to social categorization and intergroup bias.

Inquiries regarding the theoretical architecture and cross-cultural deployments of the instrument are historically maintained through academic repositories at Lawrence University and Princeton University.

4. Purpose

For decades, empirical assessments of gender-based prejudice—such as the Attitudes Toward Women Scale (AWS) or Modern Sexism Scales—presumed that sexist attitudes mirrored classic ethnic or racial prejudices, manifesting uniformly as visceral antipathy, derogation, and explicit antagonism. While these legacy scales successfully indexed institutional and blatant bias, they proved increasingly insensitive to the evolving contours of contemporary gender relations, failing to capture the uniquely intimate, interdependent nature of male-female interactions. The explicit purpose of the Ambivalent Sexism Inventory (ASI) is to capture the complex, dual-valence reality of gender prejudice: the co-existence of antagonistic hostility alongside subjectively warm, protective, and paternalistic attitudes toward women.

From a theoretical perspective, the purpose of the ASI is to uncover how gender inequality is preserved and reinforced not merely through violence, subordination, and overt hostility, but also through rewarding adherence to conventional patriarchal gender roles. Hostile and benevolent sexism function collaboratively as a complementary reward-and-punishment regulatory mechanism: hostile sexism punishes women who deviate from traditional gender arrangements (e.g., career-driven women, feminists, sexually autonomous women), whereas benevolent sexism rewards women who conform to traditional domestic and submissive roles (e.g., homemakers, chaste partners) with male protection, romantic affection, and material provision.

In research contexts, the ASI serves as an indispensable tool across social, organizational, clinical, and political psychology. It is widely utilized to assess:

• Prejudicial hiring biases and glass ceiling phenomena within organizational settings, revealing how benevolent sexism subtly prevents women from accessing challenging leadership assignments.
• Legal and forensic dynamics, particularly juror decision-making, rape myth acceptance, and victim blaming in cases of sexual assault and domestic violence.
• Interpersonal relationship satisfaction, romantic attachment styles, and the normalization of coercive control in intimate partnerships.
• Cross-national disparities in human development, macro-level economic stratification, and political attitudes toward reproductive autonomy.

5. Psychological Construct

The central construct measured by the ASI is Ambivalent Sexism, defined as an ideology characterized by deeply intertwined, evaluatively opposing orientations toward women. Rather than treating positive and negative feelings as mutually exclusive, the construct acknowledges that prejudice can take the form of subjectively positive patronization. The ASI bifurcates this ideology into two primary overarching dimensions, which are further divided into theoretical subcomponents:

Hostile Sexism (HS)

Hostile Sexism reflects explicit, adversarial antagonism toward women. It conceptualizes women as manipulative, deceitful, and overly sensitive individuals seeking to usurp male power, obtain illegitimate advantages, and exert control over men through sexuality or complaints of discrimination. Within the original theoretical formulation by Glick and Fiske (1996), HS encompasses three foundational thematic expressions:

• Dominant Paternalism: The conviction that men ought to command authority and retain ultimate structural power in societal and familial spheres, viewing women as inherently unfit for authority and naturally subordinate.
• Derogatory Beliefs and Hostile Heterosexuality: Viewing women as sexually manipulative manipulators who seek commitment to exploit men, using sexual allure to emasculate men or extract resources (e.g., Item 11: “Women seek to gain power by getting control over men”; Item 15: “Once a woman gets a man to commit to her, she usually tries to put him on a tight leash”).
• Competitive Gender Differentiation: The belief that men and women possess diametrically opposed traits, accompanied by the suspicion that women exploit gender grievances to secure unwarranted advantages over men in competitive settings (e.g., Item 2: “Many women are actually seeking special favors, such as hiring policies that favor them over men, under the guise of asking for ‘equality'”; Item 16: “When women lose to men in a fair competition, they typically complain about being discriminated against”).

Benevolent Sexism (BS)

Benevolent Sexism is defined as a set of interrelated attitudes toward women that are subjectively positive in affective tone yet inherently patronizing, patron-client oriented, and functionally disempowering. BS casts women in traditional pedestals of moral superiority and delicate beauty, contingent upon their adherence to traditional roles. Benevolent sexism is systematically delineated into three distinct sub-factors:

• Protective Paternalism: The conviction that men must serve as protectors and financial providers for women, who are perceived as fragile, vulnerable, and incapable of thriving independently in harsh societal environments (e.g., Item 9: “Women should be cherished and protected by men”; Item 20: “Men should be willing to sacrifice their own well-being in order to provide financially for the women in their lives”).
• Complementary Gender Differentiation: The idealization of women as possessing uniquely elevated moral purity, cultural refinement, and spiritual sensibility that men inherently lack (e.g., Item 8: “Many women have a quality of purity that few men possess”; Item 19: “Women, compared to men, tend to have a superior moral sensibility”; Item 22: “Women, as compared to men, tend to have a more refined sense of culture and good taste”). While ostensibly complimentary, this dimension confines women’s domain of excellence to interpersonal warmth and morality, implicitly delegitimizing their competence in domains involving agency, assertiveness, and analytical authority.
• Heterosexual Intimacy: The belief that psychological completeness and true fulfillment for a man are impossible without romantic union with a woman, viewing dyadic interdependence as a vital existential requirement (e.g., Item 1: “No matter how accomplished he is, a man is not truly complete as a person unless he has the love of a woman”; Item 12: “Every man ought to have a woman whom he adores”).

6. Theoretical Framework

The Ambivalent Sexism Inventory is anchored within Ambivalent Sexism Theory (AST) (Glick & Fiske, 1996; Glick et al., 2000), which derives from foundational sociological and social-psychological concepts of structural power, intergroup relations, and evolutionary interdependence. AST posits that male-female relations are universally marked by a profound structural paradox unique among social groups: male structural dominance (patriarchy) coexists with extreme dyadic dependence on women.

In standard intergroup conflict models, such as realistic group conflict theory or Social Identity Theory, dominant groups typically express antipathy toward subordinate groups, maintaining social distance and enforcing institutional exclusion. However, because men rely heavily on women for sexual reproduction, biological continuity, maternal caregiving, and intimate companionship, simple antipathy cannot effectively sustain male dominance. Absolute hostility would disrupt the dyadic intimacy and domestic stability that men require. Consequently, dominant groups resolve this tension by developing a bifurcated ideological system that employs both hostile and benevolent mechanisms.

This dynamic operates as a systemic carrot-and-stick paradigm:

• The “Stick” (Hostile Sexism): Functions as a reactive regulatory force. Hostile attitudes are activated when women reject subordinate positions, contest male privilege, or assert professional independence. Hostility punishes non-conforming women, characterizing them as untrustworthy, manipulative, and aggressive.
• The “Carrot” (Benevolent Sexism): Functions as a proactive ideological sedative. Benevolent attitudes incentivize women’s acquiescence to patriarchal structures by offering protection, emotional reverence, and material sustenance in exchange for compliance with traditional, submissive gender roles.

Furthermore, the inventory aligns with John Jost’s System Justification Theory, which elucidates how disadvantaged individuals often participate in the maintenance of unequal social systems that undermine their own group interests. Women frequently endorse benevolent sexism at rates comparable to, or occasionally exceeding, men’s endorsements, precisely because BS provides psychological reassurance and structural protections within patriarchal contexts. Simultaneously, the framework connects directly to Fiske’s Stereotype Content Model, wherein social targets are judged along the intersecting dimensions of warmth and competence: Hostile Sexism casts non-traditional women as competent but cold (inviting active hostility), whereas Benevolent Sexism casts traditional women as warm but incompetent (inviting patronizing paternalism).

7. Validity

The psychometric validity of the Ambivalent Sexism Inventory has been rigorously examined across hundreds of empirical studies spanning diverse cultural, socio-economic, and linguistic contexts:

Construct and Convergent Validity

In the foundational validation studies by Glick and Fiske (1996), Hostile Sexism exhibited strong, positive correlations with existing legacy measures of anti-female bias, including the Attitudes Toward Women Scale (AWS) (r values typically ranging from .50 to .65) and modern sexism scales. Conversely, Benevolent Sexism correlated only weakly or moderately with these traditional antipathy measures (r values typically between .10 and .30), confirming that BS captured a nuanced, distinct facet of gender attitudes unmeasured by conventional instruments. Importantly, across most male and female samples, HS and BS correlate positively with one another (average r between .35 and .45 in national samples; Glick et al., 2000), supporting the theoretical premise that hostile and benevolent ideologies are mutually reinforcing components of a single overarching system.

Predictive and Criterion Validity

The ASI demonstrates extensive predictive validity in diverse domains:

• Perceptions of Rape and Victim Blaming: Studies by Abrams et al. (2003) and Viki et al. (2004) demonstrate that Benevolent Sexism predicts blaming victims of acquaintance rape when the victim has violated traditional behavioral norms (e.g., accepting a drink, going to a man’s apartment), as BS sets rigid moral standards that strip “transgressive” women of male protection.
• Subtyping of Women: Research confirms that individuals with high HS scores express virulent negativity toward non-traditional female targets (career women, feminists, female politicians), while high BS scores predict positive evaluations of traditional targets (homemakers, chaste fiancées) alongside patronizing, condescending evaluations of professional women.
• Workplace Discrimination: Dardenne, Dumont, and Guignard (2007) established that exposure to benevolently sexist feedback undermined women’s cognitive performance and executive processing during complex cognitive tasks far more severely than hostile sexist commentary, operating via the induction of internal performance doubts and intrusive thoughts.

Cross-National and Ecological Validity

In a milestone cross-cultural examination involving over 15,000 participants across 19 nations, Glick et al. (2000) demonstrated that national averages of HS and BS systematically correlated with objective country-level indicators of gender inequality, such as the United Nations Gender Empowerment Measure (GEM) and Gender-related Development Index (GDI). Nations with higher average endorsements of BS and HS systematically exhibited lower female representation in political leadership, greater economic pay disparities, and poorer health outcomes for women, establishing profound ecological and macro-level validity.

8. Reliability

The Ambivalent Sexism Inventory consistently demonstrates robust internal consistency and temporal stability across diverse participant populations, including university undergraduates, working adults, and community samples worldwide:

Internal Consistency

Extensive psychometric evaluations report high internal consistency estimates:

• Hostile Sexism Subscale (11 items): Chronbach’s alpha coefficients reliably fall within the range of α = .80 to α = .92 across both male and female respondent cohorts. In the initial development samples (Glick & Fiske, 1996), alpha values ranged from .83 to .88. Cross-cultural adaptations (e.g., Spanish, German, Turkish, Japanese, and Chinese adaptations) have reported comparable alpha ranges from .81 to .89.
• Benevolent Sexism Subscale (11 items): Chronbach’s alpha coefficients typically fall between α = .73 and α = .85. The somewhat lower reliability of the full BS subscale relative to HS is attributable to its underlying multi-faceted nature, as it intentionally aggregates three conceptually distinct domains (Protective Paternalism, Complementary Gender Differentiation, and Heterosexual Intimacy).
• Benevolent Sexism Sub-factors: When assessed independently, the internal reliabilities for the individual sub-dimensions typically range from α = .65 to α = .78, which is deemed acceptable given that each sub-factor contains only 3 to 4 items.

Test-Retest Stability

Assessments of temporal stability indicate that the ASI captures enduring ideological belief systems rather than transient situational affective states. Test-retest reliability evaluations over intervals ranging from 2 weeks to 6 months demonstrate coefficients between r = .75 and r = .88 for Hostile Sexism, and between r = .70 and r = .82 for Benevolent Sexism. Long-term longitudinal studies spanning several years similarly illustrate notable stability in adult endorsement patterns, barring deliberate pedagogical or structural social interventions.

9. Factor Analysis

The internal latent structure of the Ambivalent Sexism Inventory has been subjected to exhaustive Exploratory Factor Analyses (EFA) and Confirmatory Factor Analyses (CFA), confirming the multidimensionality of the inventory:

Confirmatory Factor Analytical (CFA) Models

Glick and Fiske (1996) evaluated competing structural models via structural equation modeling:

• Unidimensional Model: A single global “sexism” factor yielded exceptionally poor model fit, demonstrating that hostile and benevolent sentiments cannot be modeled as endpoints of a single continuum.
• Orthogonal Two-Factor Model: Modeling HS and BS as completely uncorrelated constructs yielded inferior fit compared to models allowing correlated latent factors.
• Hierarchical / Multitrait-Multimethod Structural Model: The empirical data consistently favor a model with two primary, correlated higher-order factors (Hostile Sexism and Benevolent Sexism). Within this structure, Hostile Sexism operates largely as a single unified construct, whereas Benevolent Sexism is best modeled as a higher-order latent factor underpinned by three correlated first-order factors: Protective Paternalism (items 3, 9, 17, 20), Complementary Gender Differentiation (items 8, 19, 22), and Heterosexual Intimacy (items 1, 6, 12, 13).

Goodness-of-Fit and Item Loadings

In standard CFA assessments across international validation samples, this theoretical model routinely yields good to excellent fit indices:

• Comparative Fit Index (CFI): Values consistently range between .90 and .95.
• Root Mean Square Error of Approximation (RMSEA): Values typically range between .042 and .062, with 90% confidence intervals remaining below accepted cut-off thresholds of .08.
• Standardized Root Mean Square Residual (SRMR): Values routinely fall below .06.
• Standardized Factor Loadings: Nearly all items exhibit standardized loadings between .45 and .78 on their primary assigned latent factors. Cross-loadings between HS items and the primary BS factor, or vice versa, are negligible, confirming clean construct divergence.

10. Instrument / Measurement Tool

• Instrument Name: Ambivalent Sexism Inventory (ASI)
• Developer / Creators: Peter Glick, Ph.D. and Susan T. Fiske, Ph.D. (1996)
• Construct Assessed: Ambivalent Sexism toward women, bifurcated into antagonistic antipathy and benevolent paternalism
• Assessment Type: Self-report psychological questionnaire / attitude inventory
• Administration Modality: Paper-and-pencil or computerized self-administration (individual or group settings)
• Completion Duration: Approximately 5 to 10 minutes
• Target Population: Adults and adolescents aged 16 years and older; widely validated across clinical, educational, community, and cross-cultural cohorts
• Total Number of Items: 22 items
• Subscales / Factors:
  • Hostile Sexism (HS): 11 items (Items 2, 4, 5, 7, 10, 11, 14, 15, 16, 18, 21)
  • Benevolent Sexism (BS): 11 items (Items 1, 3, 6, 8, 9, 12, 13, 17, 19, 20, 22)
    • Protective Paternalism: Items 3, 9, 17, 20
    • Complementary Gender Differentiation: Items 8, 19, 22
    • Heterosexual Intimacy: Items 1, 6, 12, 13
• Response Scale: 6-point Likert rating scale:
  • 0 = Disagree strongly
  • 1 = Disagree somewhat
  • 2 = Disagree slightly
  • 3 = Agree slightly
  • 4 = Agree somewhat
  • 5 = Agree strongly
• Scoring and Transformation Procedures:
  • Reverse-Scoring: Six items are reverse-worded to control for acquiescence response set and must be reversed prior to computing subscale metrics: Items 3, 6, 7, 13, 18, and 21. The transformation rule is calculated as: New Score = 5 – Original Score (i.e., 0 becomes 5, 1 becomes 4, 2 becomes 3, 3 becomes 2, 4 becomes 1, and 5 becomes 0).
  • Hostile Sexism (HS) Score: Computed as the arithmetic mean of items 2, 4, 5, 7(reversed), 10, 11, 14, 15, 16, 18(reversed), and 21(reversed). Possible score range: 0.00 to 5.00.
  • Benevolent Sexism (BS) Score: Computed as the arithmetic mean of items 1, 3(reversed), 6(reversed), 8, 9, 12, 13(reversed), 17, 19, 20, and 22. Alternatively, subscale means for Protective Paternalism, Complementary Gender Differentiation, and Heterosexual Intimacy may be computed separately. Possible score range: 0.00 to 5.00.
  • Overall Ambivalent Sexism: While an overall mean score can be calculated across all 22 items, doing so is generally discouraged in empirical research because collapsing hostile and benevolent dimensions conceals meaningful theoretical divergence and interactions between the subscales.

11. Permissions & Fee and Test Year

The Ambivalent Sexism Inventory was officially published in 1996 by Peter Glick and Susan T. Fiske in the Journal of Personality and Social Psychology. The instrument is considered an open-access psychometric instrument for non-commercial, academic, scientific, and empirical educational research purposes. The scale authors have historically made the inventory freely accessible to academic scholars worldwide without licensing fees, provided that standard scientific attribution and proper formal citation are accorded to the original publication (Glick & Fiske, 1996).

For commercial use, proprietary organizational diagnostics, executive leadership consulting, or profit-generating digital platforms, formal written authorization must be obtained from the scale developers and/or the copyright holder (the American Psychological Association). Researchers adapting or translating the scale into additional languages are expected to adhere to standard cross-cultural translation back-translation guidelines and cite the original foundation studies.

12. References

• Abrams, D., Viki, G. T., Masser, B., & Bohner, G. (2003). Perceptions of stranger and acquaintance rape: The role of benevolent and hostile sexism in victim blame and rape proclivity. Journal of Personality and Social Psychology, 84(1), 111–125. https://doi.org/10.1037/0022-3514.84.1.111
• Dardenne, B., Dumont, M., & Guignard, S. (2007). Insidious dangers of benevolent sexism: Consequences for women’s performance. Journal of Personality and Social Psychology, 93(5), 764–779. https://doi.org/10.1037/0022-3514.93.5.764
• Flood, M. (2008). Measures for the assessment of dimensions of violence against women: A compendium. Sexual Violence Research Initiative (SVRI), 44–45. https://www.svri.org/
• Glick, P., & Fiske, S. T. (1996). The Ambivalent Sexism Inventory: Differentiating hostile and benevolent sexism. Journal of Personality and Social Psychology, 70(3), 491–512. https://doi.org/10.1037/0022-3514.70.3.491
• Glick, P., & Fiske, S. T. (1997). Hostile and benevolent sexism: Measuring ambivalent sexist attitudes toward women. Psychology of Women Quarterly, 21(1), 119–135. https://doi.org/10.1111/j.1471-6402.1997.tb00104.x
• Glick, P., & Fiske, S. T. (2001). An ambivalent alliance: Hostile and benevolent sexism as complementary justifications for gender inequality. American Psychologist, 56(2), 109–118. https://doi.org/10.1037/0003-066X.56.2.109
• Glick, P., Fiske, S. T., Mladinic, A., Saiz, J. L., Abrams, D., Masser, B., Adetoun, B., Osagie, J. E., Akande, A., Alao, A., Annetje, B., Willemsen, T. M., Chipeta, K., Dardenne, B., Dijksterhuis, A., Wigboldus, D., Eckes, T., Six-Materna, I., Exposito, F., … López, W. L. (2000). Beyond prejudice as simple antipathy: Hostile and benevolent sexism across cultures. Journal of Personality and Social Psychology, 79(5), 763–775. https://doi.org/10.1037/0022-3514.79.5.763
• Jost, J. T., & Banaji, M. R. (1994). The role of stereotyping in system-justification and the production of false consciousness. British Journal of Social Psychology, 33(1), 1–27. https://doi.org/10.1111/j.2044-8309.1994.tb01008.x
• Viki, G. T., Abrams, D., & Masser, B. (2004). Evaluating accusations of rape: The role of benevolent sexism in evaluating allegations against acquaintance and stranger rapists. Group Processes & Intergroup Relations, 7(3), 295–308. https://doi.org/10.1177/1368430204046146

13. Items of the Scale (Questionnaire)