Sex and Gender Differences in the Big Five Personality Traits

6 days ago

19 min read

Author: Ukrainian Psychological Hub · Published: October 9, 2026 · Editorial Policy

What Research Actually Finds About Gender and the Big Five

Research on the Big Five finds recurring average differences between groups labeled women and men, especially in Agreeableness and Neuroticism or closely related negative-emotionality measures. The size of a reported difference depends on the questionnaire, the population, the age range, the language, and the level of personality being studied. A group average offers information about a sample; it does not describe every member of either group or determine an individual's personality.

A major 2026 study of 143,313 adults across 27 countries reported the most pronounced average differences in Neuroticism (d = 0.40) and Agreeableness (d = 0.35), with women scoring higher on both in the study's comparisons. It found differences on other domains except Extraversion, and the size of the associations varied by country. These numbers belong to that study, not to a universal set of female and male personality scores.

Earlier evidence sometimes finds larger or smaller estimates. For example, Costa, Terracciano, and McCrae's 26-culture analysis described sex differences as small relative to variation within each sex, while an online U.S. sample of 320,128 respondents found domain-level differences of d = 0.58 for Agreeableness and d = 0.40 for Neuroticism. The contrast reveals why sampling and instruments matter.

The Big Five personality traits form a descriptive framework for variation in openness, conscientiousness, extraversion, agreeableness, and neuroticism. The present article owns the narrower question of sex- and gender-associated differences across those traits. It examines how researchers measure those differences, how large they are, where they vary, and what they can legitimately tell us about people.

Sex and Gender: What Is Actually Being Compared?

Sex and gender identify related but distinct constructs. Sex-related variables can refer to biological characteristics, reproductive anatomy, chromosomal patterns, or a classification recorded at birth. Gender-related variables can refer to identity, social positioning, roles, or a self-description. In many classic personality datasets, the survey recorded only a binary male/female response. Such a variable may have been labeled sex or gender by the authors, without measuring the different concepts separately.

Therefore, a paper reporting male–female mean differences usually establishes a difference between the two categories that it actually sampled and recorded. It does not automatically establish that biological sex caused the difference, that gender identity explains it, or that the results generalize to transgender, nonbinary, and intersex people. Researchers must document how a demographic variable was assessed, and what participants could select.

Methodological guidance on gender-inclusive psychological measurement recommends distinguishing gender constructs explicitly and avoiding the routine assumption that all participants fit into two categories. When discussing older findings here, the terms women and men follow the categories used in the source; this reporting convention should not be mistaken for a complete scientific map of sex and gender diversity.

A second distinction is essential: a Big Five trait is a relatively enduring dimension of characteristic patterns, while a state is a more temporary expression of emotion, thought, or behavior. A study showing an average group difference in a trait questionnaire cannot tell us how a particular person will feel or behave in a particular encounter. Personality dimensions are continuous and are not male and female personality types.

The Five Domains: Which Average Differences Recur?

Researchers usually compare standardized average scores within each trait rather than assigning people to fixed categories. Results across studies are not numerically identical because a Big Five domain combines narrower features and different measures sample those features in different ways. The following evidence summary describes typical directions, alongside their principal qualifications.

Agreeableness

Agreeableness captures tendencies involving compassion, cooperation, interpersonal consideration, and related patterns of social interaction. Women have tended to report higher average Agreeableness scores in many of the best-known male–female comparative studies. This result appears in the 2001 NEO analysis across 26 cultures, the 2008 Big Five Inventory study across 55 cultures, and the 2026 27-country investigation.

Agreeableness also illustrates the difference between a domain and its subcomponents. Compassion, politeness, trust, altruism, compliance, and tenderness are related but not interchangeable labels across personality instruments. A study of ten Big Five aspects found notable differences in the Agreeableness aspects of Compassion and Politeness, supporting the value of examining narrower features instead of treating a broad score as one homogeneous behavior. See Weisberg, DeYoung, and Hirsh (2011).

These results do not establish that any woman is kinder, more dependable, or more ethically developed than any man. Agreeableness is a probabilistic description of socially relevant tendencies. Both groups contain people throughout the distribution, and the social meaning of an observed act depends on circumstances and cultural norms.

Neuroticism and Negative Emotionality

Neuroticism broadly describes differences in the tendency to experience negative emotions or heightened emotional reactivity. Several independent cross-national and U.S. investigations report higher mean Neuroticism scores among respondents classified as women. In the 2026 cross-national analysis the standardized difference was d = 0.40; in the large U.S. online study it was also d = 0.40. Both are sample-specific statistical estimates, not thresholds or individual-level assessments. Rammstedt and colleagues (2026); Kajonius and Johnson (2018).

Terminology should follow the instrument. The Big Five Inventory-2 uses the domain name Negative Emotionality, which is closely related to the traditional Neuroticism domain. It is inaccurate to substitute the names of particular scales as though their items, facets, and score distributions were identical. The established BFI-2 measurement framework and the dedicated Neuroticism coverage address these measurement distinctions in greater depth.

Higher measured Neuroticism is a trait score. It is not a diagnosis of anxiety, depression, a personality disorder, or any other clinical condition. Neither sex nor gender identity justifies making such an inference. Questions about the definition, facets, and clinical associations of Neuroticism belong to the existing Neuroticism Master rather than a duplicate explanation in this population-comparison article.

Extraversion

Extraversion contains partly distinct tendencies, including sociability, social energy, positive affect, and assertive behavior. A simple question such as 'which gender is more extraverted?' can mislead because some of these elements may move in opposite directions. Feingold's meta-analysis identified higher average female scores on some extraversion-related indicators and higher male scores on assertiveness. Costa and colleagues (2001) likewise identified a higher male mean on Assertiveness and a higher female mean on Warmth.

Some earlier cross-national questionnaire studies found higher overall Extraversion among women. By contrast, the 2026 nationally oriented 27-country study did not find a comparable aggregate gender difference in Extraversion. This is substantive variation across research designs, not a contradiction that can be resolved by selecting one headline. The questionnaire's emphasis on assertiveness versus affiliation affects the broad domain.

An outgoing, warm person may score differently from a dominant, socially forceful person even if both appear highly extraverted in everyday speech. Population comparisons become more informative when authors report which narrower components drove the mean difference.

Conscientiousness

Conscientiousness covers organized, responsible, persistent, and self-regulated tendencies. Findings about average sex or gender differences at the broad domain level are generally less uniform than those for Agreeableness and Neuroticism. Some surveys report higher average scores among women, including the 55-culture Big Five Inventory comparison and a U.S. twin investigation, while effect sizes vary by sample and instrument.

At a narrower level, orderliness and industriousness may not show the same pattern. Weisberg and colleagues' aspect-level work demonstrates why domain averages can obscure component differences. A 2026 Brazilian investigation found higher mean Conscientiousness and Orderliness in women in its sample; its age and education adjustments were relevant to interpretation. Schirmer and colleagues (2026).

Conscientiousness does not measure a person's worth, moral reliability, or aptitude for any particular occupation. It captures a family of typical behaviors and preferences whose expression can depend on the opportunities and obligations built into a person's environment.

Openness to Experience

Openness includes curiosity about ideas, aesthetic sensitivity, imagination, intellectual exploration, and emotional or experiential receptivity. Men and women may differ on selected components while having relatively similar broad Openness averages. Earlier NEO findings reported higher male means on Openness to Ideas and higher female means on Openness to Feelings. Costa and colleagues (2001). These facets belong to the NEO framework; they should not be treated as the universal facet vocabulary of every Big Five test.

The Big Five Aspects Scale distinguishes the Openness and Intellect aspects. A 2011 aspect study and its 2026 Brazilian replication and extension both illustrate how group differences can be clearer in specific aspects than in a combined domain. In the Brazilian study, women scored higher on its Openness aspect while men scored higher on Intellect.

A population difference in an openness component does not establish a difference in measured general intelligence, creativity, artistic merit, or willingness to learn. Each outcome demands its own operational definition and empirical test.

A Research Snapshot: Sample Sizes, Findings, and Boundaries

Costa, Terracciano, and McCrae (2001) analyzed NEO-PI-R data from 23,031 participants in 26 cultures. They identified recurring differences in selected traits and facets, emphasized extensive within-sex variation, and reported cross-cultural differences in effect magnitude. Their NEO facet findings should be interpreted within that inventory.

Schmitt, Realo, Voracek, and Allik (2008) examined Big Five Inventory responses from 17,637 respondents in 55 nations. Women reported higher means on several domains across most nations. The authors also reported larger group differences in nations with higher human development. The study's country-level pattern does not itself determine the causes of individual personality differences.

Weisberg, DeYoung, and Hirsh (2011) studied the Big Five and ten narrower aspects and demonstrated that broad domain totals can conceal opposing aspect differences. The paper's illustrated overlap in Agreeableness is a helpful visual corrective to the assumption that different means imply separate personality categories.

Kajonius and Johnson (2018) analyzed 320,128 U.S. online self-reports on the IPIP-NEO-120, reporting women higher on mean Agreeableness (d = 0.58) and Neuroticism (d = 0.40). The considerable sample size increases precision, while online self-selection still limits claims about national representativeness.

Rammstedt, Roth, Roemer, and Lechner (2026) assessed 143,313 people across 27 countries using adult-population datasets and systematically adapted questionnaires. The reported standardized differences included d = 0.40 for Neuroticism and d = 0.35 for Agreeableness, with variation across countries and facets. The authors made their analysis code publicly available, supporting scrutiny of the estimates.

Schirmer and colleagues (2026) studied 1,200 Brazilian adults using the Big Five Aspects Scale. Their results partly reproduced earlier aspect findings while demonstrating moderation by age and education. Its sample composition and single-country setting should be considered when comparing estimates with national surveys.

How Big Are the Differences? Effect Sizes and Distribution Overlap

Statistical significance answers whether a study provides evidence of an association or difference under its model. Practical magnitude asks a different question. Cohen's d expresses the distance between two group means in standard-deviation units. A difference of d = 0.40 means the means differ by about four-tenths of a within-group standard deviation under the study's standardization. It does not mean that 40% of women or men have a trait, that 40% of behavior is explained, or that a score predicts a person's gender correctly 40% of the time.

When two roughly bell-shaped groups have similar variances, the shared area under the two estimated distributions is approximately 84% for d = 0.40 and 86% for d = 0.35. These are illustrative normal-distribution calculations, not observed percentages from the 2026 paper; real distributions may depart from those assumptions. Most people occupy ranges where the groups overlap extensively, even when the group means are reliably different.

The 2015 metasynthesis by Zell, Krizan, and Teeter aggregated 106 meta-analyses spanning many psychological domains and found an average absolute gender difference of d = 0.21. That figure is a broad synthesis of psychology, not a pooled five-domain estimate for Big Five traits. It provides context for thinking about psychological similarity while leaving room for dimensions whose differences are larger or smaller.

A multivariate personality profile creates another statistical question: how far apart are two groups when many correlated characteristics are considered simultaneously? Del Giudice, Booth, and Irwing (2012) reported a large multivariate distance using the 16PF, a distinct assessment framework. That conclusion attracted methodological debate about how to interpret combined dimensions. A multivariate index cannot be substituted for a Big Five domain effect or turned into evidence that people form two separate personality species.

No effect size permits a confident individual classification from an ordinary trait score alone. Base rates, the questionnaire, population characteristics, response errors, and calibration all matter. Reporting group distributions and uncertainty is more informative than sorting participants into stereotypical male and female personalities.

Why Domains, Aspects, and Facets Change the Answer

A Big Five domain is a broad cluster of correlated tendencies. An aspect is an intermediate level used in some models, while a facet is a narrower measured tendency within an instrument's hierarchy. The Big Five domain–aspect–facet guide explains the hierarchy in detail. The number and names of narrower traits depend on the instrument and the theoretical model.

The NEO-PI-R organizes five domains into six facets per domain, making 30 NEO facets. The BFI-2 contains three facets per domain, for 15 facets, and labels one domain Negative Emotionality. The Big Five Aspects Scale measures two aspects within each domain, for ten aspects. These are different measurement designs. A difference in a NEO Openness to Feelings scale cannot automatically be called a difference in the identically named BFI-2 or BFAS score. See Soto and John (2017).

Domain aggregation can average out distinct effects. Imagine two narrower components with differences in opposite directions. A domain total may appear similar across groups even when the components have consistent differences. Researchers should present domain and narrower-scale estimates where the instrument is designed and validated to support them. Post hoc subdivisions of a short questionnaire should not be presented as validated facets.

Scores themselves are not universal units. A 'high' score is meaningful only relative to a test's scoring framework and, where used, an appropriate normative sample. There is no single high-versus-low cutoff shared by all Big Five instruments, genders, or cultures. Readers seeking guidance on scores should use the English Hub guide to Big Five scores and percentiles.

Are the Findings the Same Across Cultures?

A substantial body of Big Five research spans multiple countries and languages, and recurring average gender-associated findings have been documented in more than one instrument. That gives cross-national results greater relevance than evidence from one college sample. It does not mean every lexical personality study reproduces an identical five-factor structure or every translated questionnaire functions identically across groups.

The 2008 investigation covering 55 nations reported larger sex differences in personality in countries with higher human-development indicators. The 2026 27-country analysis found country-to-country variation and reported that combined differences across domains tended to be larger in more individualistic countries, while Openness followed a different moderation pattern.

Cross-national correlations invite multiple explanations: social roles, schooling, language use, reference-group effects, response styles, sample composition, opportunities for personal expression, and biological hypotheses may all be relevant research candidates. Nation-level associations do not show which mechanism operates within any one person. Nor do they imply that greater gender equality produces personality differences directly.

For multilingual measurement, translation quality and invariance testing are central. If items carry different meanings or thresholds in different populations, differences in the observed totals may partly reflect the instrument. Country comparisons should report construct comparability, sampling procedures, and uncertainty before drawing broad conclusions.

What Changes Across Age and Development?

Age can shape both mean personality scores and their relationship to sex or gender categories. Some studies focus on college students, others on working-age adults, and still others on adolescents or older adults. Pooling them without attention to age composition risks producing a finding that describes the mixture of ages rather than a stable group contrast.

A longitudinal investigation of adolescents and young adults illustrates the importance of tracking development rather than assuming the same trait differences remain constant from youth through adulthood. Study results in this area depend on assessment format, developmental stage, and time interval. An adolescent sex difference in a brief personality inventory cannot be generalized to every later stage.

The distinction between mean-level developmental change and rank-order stability matters. Mean levels may shift across age groups while people's relative positions retain some consistency, and both phenomena can coexist with sex-associated differences. The English Hub lifespan stability review and personality change review develop those broader topics.

Age effects and cohort effects also need separation. A survey comparing today's 20-year-olds with today's 70-year-olds combines life-stage differences with differences in historical experience. Repeated studies and longitudinal designs improve inference, although selective attrition and changing survey measures remain concerns.

Can Tests Compare Men, Women, and Other Gender Groups Fairly?

A fair group comparison requires adequate construct validity and evidence that a measure is operating comparably across the groups involved. Measurement invariance evaluates whether the latent structure, item relations, and sometimes item intercepts support comparisons. Differential item functioning asks whether an item may behave differently for respondents with similar levels of the underlying trait.

The empirical answer varies by test and population. Shchebetenko and colleagues (2019) found favorable sex and age invariance evidence for the Russian BFI-2 in two investigated samples. Ock and colleagues (2020) studied gender invariance in the Mini-IPIP and BFI facet measures, demonstrating why the test and analytic framework must be specified. These results do not grant every Big Five translation universal invariance.

Questionnaire answers can also reflect impression management, the social desirability of certain descriptors, self-knowledge, stereotype expectations, and the respondent's comparison group. Self-ratings and observer ratings need not agree perfectly. For a specific methodological comparison see Self-Report vs Observer Ratings in Big Five Personality Assessment.

Researchers should report the measure, version, language, recruitment method, response options, group definitions, age distribution, missing data, measurement-invariance evidence, and uncertainty intervals. A sample drawn from a voluntary online personality-test platform should not be described as a probability sample of an entire country.

What About Transgender, Nonbinary, and Intersex People?

Classic Big Five sex-difference datasets often used two response categories and did not separately measure gender identity, sex assigned at birth, or intersex variation. Their binary comparisons provide little direct evidence about personality distributions within transgender, nonbinary, or intersex populations. The absence of sufficient data is a limitation of those particular designs; it is not evidence that these populations have any particular Big Five profile.

Research that aims to study gender diversity must decide whether its scientific question concerns identity, lived experience, social treatment, biology, or some combination. It should permit suitable response choices and preserve participants' distinctions where sample size and confidentiality allow. The methodological recommendations by Cameron and colleagues address such decisions in psychological research.

No individual's gender identity can be validated, inferred, or discredited from their Big Five scores. For the same reason, a questionnaire profile is not a clinical gender assessment. The Big Five measures selected personality constructs; their scores should be interpreted only for that purpose.

Why Do Differences Appear? Competing Explanations and Causal Limits

Possible explanations include socialization, social role expectations, unequal experiences, occupational selection, developmental pathways, differences in opportunities, genetic and hormonal hypotheses, and interactions among biological and social factors. The observed average score is the outcome of a measurement process within a particular society, not a mechanism by itself. Cross-sectional surveys establish group associations and cannot partition those influences causally.

The well-known 'gender-equality paradox' narrative often starts with cross-national correlations in which some measured differences were larger in economically developed or more gender-egalitarian countries. Those correlations allow several rival interpretations. Changes in the meanings of self-description, differences in national samples, cultural reference frames, and confounding country characteristics can contribute to the observed pattern. Schmitt and colleagues (2008) reported a development association; Rammstedt and colleagues (2026) provided a newer cross-national comparison with important domain-specific variation.

Behavioral genetic evidence likewise requires precise reading. A U.S. twin study of 973 pairs found some sex differences in observed mean trait levels but did not identify qualitative or quantitative sex differences in genetic and environmental variance for the Big Five domains in its analyses. This finding neither proves that biology is irrelevant nor establishes a biological cause for the observed mean differences.

Heritability estimates refer to how variation is statistically partitioned in particular populations under particular conditions. They do not specify the cause of a group mean difference. A high heritability estimate for a trait cannot establish that a difference between sex-labeled groups is inherited, biologically fixed, or resistant to change.

Causal claims would require research designs that can identify the proposed pathways with stronger assumptions and tests than simple male–female score comparisons. At present, the evidence supports robust description of selected average differences and substantially more limited conclusions about their origin.

Scientific Disagreements: Similarity, Difference, and Multivariate Profiles

Hyde's gender similarities review and the later Zell, Krizan, and Teeter metasynthesis emphasized that numerous psychological sex/gender effects are small on average, while allowing larger differences for particular variables. Personality research adds important nuance because aggregating all psychological phenomena can conceal the distinct behavior of a few Big Five domains.

Other investigators argue that considering multiple narrower personality characteristics together can reveal larger group separations than examining each dimension on its own. Del Giudice, Booth, and Irwing (2012) advanced this approach using the 16PF, and Del Giudice (2013) defended multivariate measures against methodological criticism. The competing approaches answer different statistical questions, and the results depend strongly on which attributes and assumptions enter the calculation.

The best empirical description can therefore contain both propositions: several traits show reproducible mean group differences, and within-group variation and overlap are substantial. The first describes detectable population patterns; the second describes the information those patterns leave unresolved about individuals. Neither requires a claim that men and women have separate personality architectures.

Practical Meaning for Relationships, Work, Education, and Assessment

The practical starting point is to evaluate people as individuals. A manager cannot reliably infer a candidate's conscientiousness from gender, a teacher cannot assign curiosity from sex, and a partner cannot know another person's agreeableness from group averages. When an assessment has an appropriate purpose, use a suitable validated measure, informed consent, relevant local norms, and attention to the decision's consequences.

For relationships, a person's repeated behavior, communication, circumstances, and personal history provide information that a broad demographic comparison does not. A group tendency cannot tell us whether one partner will cooperate, express feelings, or seek social contact more readily. Gender stereotypes can distort communication by making disconfirming evidence less visible.

In hiring or education, sex- or gender-based shortcuts can create unfairness and poor prediction. A statistically reliable population association does not justify substituting a demographic label for direct, job-relevant or learning-relevant evidence. Measurement quality, institutional rules, accessibility, and the validity of the specific decision procedure require independent evaluation.

Clinical interpretation requires even greater discipline. Neuroticism or negative emotionality can be studied in relation to mental-health outcomes, but a trait result is not a screening diagnosis, a diagnostic determination, or a treatment plan. Group-level population differences also do not establish an individual disorder risk without a validated predictive model and appropriate context.

The same reasoning applies to data-driven profiling. A digital platform's prediction of a Big Five score from online behavior is a prediction generated by its model and training data; it is not proof that the platform 'knows' the personality of the person. Accuracy, construct validity, demographic bias, privacy, and out-of-sample performance must be evaluated separately.

How to Read a Headline Claim About Gender and Personality

First identify the exact outcome: is the reported number an overall Big Five domain, an aspect, a NEO facet, or a separate construct such as empathy, assertiveness, or self-esteem? Next identify the groups, the actual demographic question, the sample, and the country. Then locate the effect size with its uncertainty and examine whether the distributions overlap.

Ask whether the study used a validated instrument in the relevant language and tested group measurement comparability. Distinguish an analysis of women and men in one cohort from evidence intended to apply across populations, ages, or genders. Finally, read causal explanations as hypotheses unless the research design specifically identifies them.

The strongest summary of the literature combines converging studies, transparent uncertainty, instrument-specific detail, and the recognition that traits are continuous. Sex and gender differences can be legitimate subjects of population research without becoming fixed prescriptions for how anyone should think, work, relate, or live.

Frequently Asked Questions

Are women higher in Neuroticism than men?

Across many binary-category Big Five studies, respondents classified as women have shown higher average scores on Neuroticism or closely related negative-emotionality measures. The 2026 27-country analysis estimated d = 0.40. The inference applies to sample-level means, with substantial overlap between individuals and no diagnostic implication.

Which Big Five trait shows the clearest sex or gender difference?

Agreeableness and Neuroticism are among the most consistently reported broad-domain differences in this literature. Magnitudes and even domain rankings vary across instruments and populations, so no trait has one universal effect size. Narrower facets may differ more strongly than their parent domains.

Are men more extraverted or more open?

Broad Extraversion and Openness scores do not yield a single stable universal answer. Findings at the narrower level may diverge: assertiveness and warmth differ from one another, as do the NEO's Openness to Ideas and Openness to Feelings or the BFAS's Intellect and Openness aspects. Recent cross-country and aspect studies demonstrate why the measure needs to be named.

Do gender differences get larger in more equal countries?

Some cross-national studies report larger aggregate differences in countries with higher development or individualism indicators. The strength and direction of moderation varies by domain, as shown in the 2026 international study. Country-level correlations do not establish that equality causes the differences.

Are personality differences innate or learned?

Current descriptive comparisons alone cannot divide a group difference into innate and learned components. Developmental, biological, cultural, and social processes may interact. Twin studies partition sources of variance under assumptions but do not automatically identify causes of mean differences between demographic groups.

Do men and women have separate Big Five personality types?

No categorical male or female type is defined by the Big Five. Each domain is a continuous distribution with extensive within-group variation and overlap. A person's combination of traits is their measured profile, not evidence of membership in a sex-specific personality category.

Can a Big Five test identify someone's gender?

A Big Five score is a personality measurement, not an identity test. Statistical associations within binary research categories cannot establish an individual's gender identity. Any classifier trained on such data must be assessed for validity, privacy, bias, generalizability, and harmful applications.

Should Big Five scores be compared with sex-specific norms?

Use the validated scoring and interpretation guidance for the specific instrument and purpose. Sex-specific reference groups can change what a percentile means, while common reference groups allow different comparisons. Neither is universally correct; the choice depends on the intended interpretation, fairness considerations, and validated norms.

Does a gender difference in Neuroticism mean a difference in mental illness?

No. Neuroticism is a continuous personality domain. Diagnoses are determined through condition-specific clinical criteria and assessment, and a mean difference in personality scores does not establish a mean difference in a particular disorder, much less a diagnosis for an individual.

What is the most useful conclusion for an individual?

Use validated personality information when relevant, notice real behavior over time, and interpret any score in the context of its instrument, norms, and uncertainty. Population research can enrich understanding of social patterns while an individual's personality remains a matter for individual evidence.

References

Cameron, J. J., & Stinson, D. A. (2019). Gender (mis)measurement: Guidelines for respecting gender diversity in psychological research. Social and Personality Psychology Compass, 13(11), e12506. https://doi.org/10.1111/spc3.12506

Costa, P. T., Jr., Terracciano, A., & McCrae, R. R. (2001). Gender differences in personality traits across cultures: Robust and surprising findings. Journal of Personality and Social Psychology, 81(2), 322–331. https://doi.org/10.1037/0022-3514.81.2.322

Del Giudice, M. (2013). Multivariate misgivings: Is D a valid measure of group and sex differences? Evolutionary Psychology, 11(5), 1067–1076. https://doi.org/10.1177/147470491301100511

Del Giudice, M., Booth, T., & Irwing, P. (2012). The distance between Mars and Venus: Measuring global sex differences in personality. PLOS ONE, 7(1), e29265. https://doi.org/10.1371/journal.pone.0029265

Feingold, A. (1994). Gender differences in personality: A meta-analysis. Psychological Bulletin, 116(3), 429–456. https://doi.org/10.1037/0033-2909.116.3.429

Hyde, J. S. (2005). The gender similarities hypothesis. American Psychologist, 60(6), 581–592. https://doi.org/10.1037/0003-066X.60.6.581

Kajonius, P. J., & Johnson, J. A. (2018). Sex differences in 30 facets of the five factor model of personality in the large public (N = 320,128). Personality and Individual Differences, 129, 126–130. https://doi.org/10.1016/j.paid.2018.03.026

Ock, J., McAbee, S. T., Mulfinger, E., & Oswald, F. L. (2020). The practical effects of measurement invariance: Gender invariance in two Big Five personality measures. Assessment, 27(4), 657–674. https://doi.org/10.1177/1073191119885018

Rammstedt, B., Roth, M., Roemer, L., & Lechner, C. (2026). Revisiting gender differences in personality: New evidence on Big Five domains and facets with large-scale samples from 27 nations. Personality and Individual Differences, 257, 113806. https://doi.org/10.1016/j.paid.2026.113806

Schirmer, L. R., Garzoni, D. L. C., da Silva, J. W. G., dos Santos, A. L., & Natividade, J. C. (2026). Revisiting gender differences in 10 aspects of personality. Personality and Individual Differences, 258, 113852. https://doi.org/10.1016/j.paid.2026.113852

Schmitt, D. P., Realo, A., Voracek, M., & Allik, J. (2008). Why can't a man be more like a woman? Sex differences in Big Five personality traits across 55 cultures. Journal of Personality and Social Psychology, 94(1), 168–182. https://doi.org/10.1037/0022-3514.94.1.168

Shchebetenko, S., Kalugin, A. Y., Mishkevich, A. M., Soto, C. J., & John, O. P. (2019). Measurement invariance and sex and age differences of the Big Five Inventory-2: Evidence from the Russian version. Assessment. https://doi.org/10.1177/1073191119860901

Soto, C. J., & John, O. P. (2017). The next Big Five Inventory (BFI-2): Developing and assessing a hierarchical model with 15 facets to enhance bandwidth, fidelity, and predictive power. Journal of Personality and Social Psychology, 113(1), 117–143. https://doi.org/10.1037/pspp0000096

South, S. C., Jarnecke, A. M., & Vize, C. E. (2018). Sex differences in the Big Five model personality traits: A behavior genetics exploration. Journal of Research in Personality, 74, 158–165. https://doi.org/10.1016/j.jrp.2018.03.002

Weisberg, Y. J., DeYoung, C. G., & Hirsh, J. B. (2011). Gender differences in personality across the ten aspects of the Big Five. Frontiers in Psychology, 2, 178. https://doi.org/10.3389/fpsyg.2011.00178

Zell, E., Krizan, Z., & Teeter, S. R. (2015). Evaluating gender similarities and differences using metasynthesis. American Psychologist, 70(1), 10–20. https://doi.org/10.1037/a0038208

Psychological Encyclopedia