How Accurate Are Big Five Personality Tests? Validity, Reliability, and Limits

7 days ago

18 min read

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

A well-developed Big Five personality test can provide a useful, reasonably dependable estimate of broad personality traits. Research supports the reliability and validity of several established questionnaires, particularly when their administration and interpretation resemble the conditions in which they were validated. Accuracy belongs to a specific instrument, score, population, and purpose. A website's claim that its quiz is “based on the Big Five” establishes none of those properties.

The key question is what kind of accuracy is being claimed. A questionnaire may give consistent scores yet have poor norms for the intended reader. It may predict an outcome statistically while offering little certainty about a particular individual. It may work well for research on broad trends but lack the precision required for consequential individual decisions. The original BFI-2 validation study and a 2025 meta-analysis of BFI/BFI-2 reliability supply strong foundations for specific Big Five assessments while demonstrating why scientific claims must be instrument-specific.

How Accurate Are Big Five Tests? The Evidence in Brief

Established multi-item Big Five inventories generally show meaningful reliability and validity for describing relative personality tendencies. Their performance varies by instrument, domain, version, language, sample, and intended use. Their strongest routine contribution is an estimate of continuous dimensions of personality, rather than a categorical label or a prediction of exactly how one person will act.

A 2025 meta-analysis combined 57 reliability estimates from 34 articles involving 43,715 participants across languages and settings. It reported generally satisfactory internal consistency for the 44-item Big Five Inventory (BFI) and 60-item BFI-2. Separately, a meta-analysis of short-term retesting synthesized 682 correlations from 74 samples and reported a median aggregated dependability coefficient of approximately .82. These statistics describe particular aspects of measurement; neither means that everyone's result is “82% accurate,” and neither certifies every internet quiz.

The parent guide, Big Five Personality Test: What It Measures and How to Interpret Your Profile, explains the assessment process and score interpretation. This article examines the scientific evidence behind the measurements, sources of error, and their practical limits.

What Does Accuracy Mean in Personality Assessment?

Personality questionnaires sample answers about characteristic behavior, emotional responses, preferences, and thought patterns. They estimate underlying dispositions by combining items according to established scoring rules. Unlike a question with one objectively correct answer, trait items involve interpretations of ordinary behavior, personal context, and response tendencies. Accordingly, accuracy is evaluated through several complementary kinds of evidence.

Reliability: Are the scores sufficiently dependable?

Reliability addresses consistency. Researchers examine whether items intended to measure a domain share relevant variance and whether comparable administrations tend to give similar results over short periods. Internal consistency and test–retest dependability are different concepts. A strong estimate of one does not automatically establish the other.

Construct validity: Does the test measure the intended traits?

Construct validity concerns whether interpreting a set of responses as Extraversion, Agreeableness, or another Big Five domain is supported. Relevant evidence includes questionnaire structure, agreement with independent instruments, relationships with narrower trait facets, and convergence with knowledgeable observers. Validity is an accumulated argument about the meaning of the scores, not a permanent certification given to an entire model once.

Criterion-related validity: Does the score relate to external outcomes?

A measure may show statistical relationships with relevant outcomes, such as a particular dimension of job performance. The relationship's size, uncertainty, and generalizability depend on how both the trait and outcome were measured and on the people and situations studied. Statistical prediction is different from causal explanation and from certainty about any one individual.

Interpretive validity: Is the proposed use justified?

A result adequate for reflection or research may be inadequate for employee selection, clinical judgment, or a decision about another person. Evaluating a test therefore includes its intended application. Even well-supported scale scores can be overinterpreted when the consequences exceed what their validation evidence permits.

The Big Five Framework and a Particular Test Are Different Things

The Big Five framework summarizes individual differences on five broad dimensions, often represented by OCEAN: Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. For a broader introduction, see Big Five Personality Traits: What They Are, What OCEAN Means, and How the Model Works.

An instrument translates those broad constructs into items, instructions, response choices, subscales, scoring algorithms, and norms. Two valid tests can emphasize somewhat different content within the same domain. They can also differ in score precision. The BFI-2 uses 60 items to assess five domains and 15 instrument-specific facets, whereas the NEO instruments and Big Five Aspect Scales use different subdivisions. The BFI-2 calls two of its domains Negative Emotionality and Open-Mindedness; those labels relate to conventional Neuroticism and Openness constructs, but instrument terminology must be respected.

Broad domains, intermediate aspects, and narrower facets serve different interpretive purposes. There is no one universally mandated set of facets shared by every Big Five measure. See Big Five Domains, Aspects, and Facets for the full hierarchy.

Internal Consistency: How Consistent Are Big Five Scale Items?

A multi-item Conscientiousness scale, for example, samples multiple descriptions of organization, dependability, and goal pursuit. Its internal-consistency estimate summarizes how much the items behave as related indicators in a particular sample. A higher coefficient usually indicates greater shared information, given the measure's assumptions, but it does not guarantee comprehensive construct coverage or useful individual decisions.

The 2025 reliability generalization meta-analysis by Husain and colleagues reported original-BFI pooled internal-consistency estimates of approximately .77 for Openness, .80 for Conscientiousness, .80 for Extraversion, .73 for Agreeableness, and .80 for Neuroticism. The corresponding BFI-2 estimates were approximately .83, .86, .85, .80, and .89. These were pooled estimates across studies, not fixed specifications of every administration.

The authors reported variation and heterogeneity across settings. Reliability depends partly on the sample and context in which it is estimated. Translating an instrument, shortening it, changing its response scale, or administering it to a population unlike the original validation sample may change its properties. A webpage that reproduces a published coefficient without identifying its own version and intended population has presented incomplete evidence.

Internal consistency is only one aspect of reliability. A scale can contain highly similar items that generate an impressive coefficient while overlooking meaningful parts of a domain. That is why a responsible evaluation considers content representation, structural evidence, and external convergence together.

Test–Retest Reliability: Will You Get the Same Result Tomorrow?

Trait measures are intended to capture characteristic tendencies, yet answers may be influenced by stress, fatigue, recent experiences, attention, context, and imperfect measurement. Repeating the same assessment can therefore produce somewhat different numbers even over short periods.

In a meta-analysis of 682 test–retest correlations from 74 samples involving 14,923 participants, Gnambs examined Big Five measures with retest intervals up to two months. The median aggregated dependability estimate was approximately .816. Extraversion scales were generally more dependable, whereas Agreeableness measures showed somewhat greater measurement error. Different retest intervals also influenced dependability estimates for some domains.

A correlation of .816 is a description of score consistency across participants, not a claim that a person will reproduce exactly 81.6% of their earlier answers. Test–retest correlation primarily concerns relative standing in a group. A group's average score can shift even if people largely retain their relative positions. On longer time scales, genuine personality development becomes an additional consideration.

A changed result needs interpretation through the particular scale, time interval, administration conditions, and expected error. For someone comparing results across years, the scientific question is different from that of two questionnaires completed over a single weekend.

Construct Validity: How Do We Know a Big Five Test Measures Personality?

Construct validity draws on the structure of responses and their relationships with other credible indicators. The purpose is to test whether the instrument behaves as expected if it measures its named traits. It cannot be established by the brand name of the questionnaire or a favorable testimonial.

Structural evidence and factor analysis

Factor-analytic methods assess whether item responses show a structure sufficiently compatible with the intended domains and narrower scales. In the BFI-2 development and validation program, Soto and John developed and tested a five-domain, 15-facet hierarchy, evaluated items with conceptual and empirical criteria, and compared scale scores with additional self- and peer-reported variables. Their work supported the BFI-2's reliability and validity and identified improvements over the original BFI.

Structural evidence is instrument-specific. A five-factor solution in one sample does not demonstrate that five factors will always be recovered identically from every language, item pool, or method. Factor structure alone does not tell us whether score differences across groups are comparable.

Convergence across independently developed questionnaires

When two well-developed scales aim to assess similar constructs, their results should generally show meaningful correspondence. Their agreement will be imperfect when their items emphasize different subcomponents or when their response and scoring systems differ. Convergence is one form of evidence, especially when examined alongside appropriate distinctions from other constructs.

A score should not be treated as invalid simply because another legitimate questionnaire yields a different raw number or percentile. Such comparisons require compatible scoring rules and reference populations. Agreement between broad domain measures is also a different question from agreement between narrow subscales.

Agreement with observer ratings

Self-report is not the only source of trait information. Friends, family members, and colleagues may observe characteristic patterns that provide partially independent evidence.

A 2007 meta-analytic review by Connolly, Kavanagh, and Viswesvaran found corrected mean correlations between self- and observer-rated Big Five traits ranging from approximately .46 for Agreeableness to .62 for Extraversion. These are meta-analytic statistical correlations, not percentages of truthful reporting. Other domain correlations fell between these values.

A larger 2010 meta-analytic integration, covering 263 independent samples and 44,178 rated people, showed that observer relationships and trait visibility mattered for the information contained in ratings. An observer may see outward sociability clearly yet have less access to a person's private anxieties, motivations, or self-evaluations. Disagreement therefore need not mean that one person is lying. Both self-ratings and informant ratings have method-specific strengths and limitations.

Criterion Validity: Can Big Five Scores Predict Real Behavior?

Some Big Five scores show reproducible relationships with outcomes, and this has made personality assessment useful to scientific research. Outcomes vary greatly, however, and the presence of a statistically significant relationship does not mean a test can forecast a particular event for an individual.

In a foundational 1991 meta-analysis, Barrick and Mount examined Five-Factor Model traits and several job-performance criteria across occupational groups. Conscientiousness showed consistent associations across those studied occupations and criteria; associations for other traits depended more strongly on the outcome and occupational context. A later meta-analysis of observer ratings found that observer-assessed traits contributed useful predictive information about job performance beyond self-reports in the included studies.

These findings illustrate criterion-related validity for particular questions. They do not establish that a personality score identifies an ideal occupation, determines someone's moral character, or reveals what they will do in every situation. A correlation can be practically meaningful in a large population while leaving extensive uncertainty about a single person.

Personality, opportunity, knowledge, ability, incentives, social relationships, and situations jointly shape behavior. Reliability and predictive utility must be evaluated separately. A consistent questionnaire can have modest predictive relevance to a specific outcome; a statistically predictive model can still make mistakes about many individuals.

How Much Precision Does One Score Really Have?

Every reported trait score is an estimate. An individual may lie near the middle of a distribution but receive slightly different ranks across repeated administrations or test versions. More items, when appropriately designed, may reduce uncertainty; they cannot remove it altogether.

Measurement precision may differ by trait and by the portion of the score distribution being assessed. Reliability coefficients from a research sample cannot be converted into universal personal confidence intervals without instrument-appropriate assumptions and data. In a high-stakes context, responsible interpretation may require a documented standard error, retesting strategy, multiple methods, or other relevant evidence.

The difference between two people with nearly identical percentiles may be too small for a confident substantive distinction. Labels such as “exceptional,” “extremely high,” or “very low” should not be applied using made-up thresholds. A well-validated instrument may provide norm-specific interpretive conventions; those conventions do not automatically transfer to other Big Five questionnaires.

Are Longer Big Five Tests More Accurate Than Short Tests?

A longer questionnaire can cover more of a domain and yield more dependable scores. Item quality and construct coverage matter as much as item number, and some short forms were developed specifically to retain useful broad-domain information under time constraints.

Soto and John's 2017 evaluation of abbreviated BFI-2 forms studied the 30-item BFI-2-S and 15-item BFI-2-XS relative to the 60-item original. Both shortened versions retained much of the full form's domain-level reliability and validity. The researchers also cautioned that the 15-item extra-short form should not be used to assess facets; facet-level use of the 30-item form was discussed primarily for sufficiently large samples.

The Ten-Item Personality Inventory (TIPI)00046-1) is an even briefer measure created when comprehensive assessment is impractical. A 2023 scoping review of 29 studies found acceptable retest reliability, mixed convergent and structural validity evidence, and poor internal consistency under conventional criteria. This illustrates the compromise of estimating five domains using only ten items.

A short test may be entirely sensible for a large survey measuring broad associations. Someone seeking a detailed individual profile should expect more evidence of precision and coverage. Brevity is a design choice, not a moral verdict on a test.

Are Big Five Scores and Percentiles Comparable Across Websites?

They are not automatically comparable. Different inventories use different item wording, scoring rules, norms, and sometimes different labels for closely related constructs. A raw score of four on one response scale is not necessarily equivalent to four on another instrument.

A percentile represents position relative to a specified reference distribution. The 75th percentile means a score falls above approximately 75% of that norm group under the given scoring procedure. It is not “75% of the personality trait” or “75% accurate.” Nor does a percentile indicate how often a person will behave a particular way.

Age, language, recruitment method, and cultural context affect the meaning of normative comparisons. A website that presents personal percentile rankings should identify the reference sample and what those rankings actually compare. For an in-depth explanation, see Big Five Test Scores and Percentiles: What High, Low, and Average Results Mean.

What Makes a Big Five Test Less Accurate?

Wording and translation

Statements about being “reserved,” “imaginative,” or “organized” may carry different meanings across linguistic and social contexts. A translation can be fluent while failing to preserve how questions function psychometrically. Valid adaptation involves evidence, rather than word replacement alone.

Acquiescence and other response styles

Some respondents tend to agree with items, prefer middle responses, or avoid extremes regardless of the underlying trait. Item balancing and scoring choices can help. The BFI-2 validation research explicitly examined and addressed acquiescent responding.

Social desirability and incentives

People may respond differently when they believe a desirable profile could improve their chances of employment or approval. Their answers may reflect self-presentation as well as characteristic tendencies. A 2017 meta-analysis of online and paper administration did not support the assumption that web-based questionnaires inherently reduce socially desirable responding.

Fatigue, temporary states, and setting

Poor sleep, acute stress, time pressure, distractions, recent disagreements, and shifting role expectations can affect how respondents recall their behavior. A trait score is intended to summarize a characteristic pattern; an emotional state can influence the answers used to estimate that pattern.

Careless or inconsistent responding

Rushing, random responding, inattention, skipped items, and misunderstanding reverse-worded questions can reduce score quality. An online platform must preserve validated instructions and scoring, and it should not mistake an elaborate generated report for evidence that the underlying response data were informative.

Questionable norms

A reference distribution recruited from a single class or website cannot necessarily describe all adults. Unrepresentative norms can turn an otherwise useful raw score into a misleading comparative claim.

Do Big Five Tests Work Equally Well Across Languages and Cultures?

Many personality studies provide substantial evidence for broad Big Five dimensions across languages and populations, but instrument validity and measurement equivalence cannot be assumed universal. There is a difference between studying similar constructs in different societies and demonstrating that a particular set of questions yields directly comparable scores across them.

A major 2019 analysis in Science Advances examined face-to-face personality surveys from 94,751 respondents in 23 low- and middle-income countries. The authors reported serious problems with commonly used Big Five questions' intended measurement properties under those survey conditions. Internet samples from the same countries showed a different pattern, pointing to survey mode, response styles, education, and interviewing circumstances as relevant factors.

The lesson is methodological precision. Researchers need evidence that items capture the intended constructs in the actual population and that a sufficient degree of measurement invariance holds before comparing means or drawing strong conclusions across groups. Findings from one English-speaking convenience sample do not automatically validate use in another language or under a different form of administration.

Which Big Five Instruments Are Well Supported?

The right choice depends on the needed scope, precision, population, and purpose, not on a single global ranking of “most accurate tests.”

Big Five Inventory and BFI-2

The BFI and BFI-2 have substantial scientific documentation, and the BFI-2 development paper provides instrument-specific evidence about five domains and 15 facets. The 2025 reliability synthesis adds broader evidence about internal consistency across studied languages. See Big Five Inventory: What the BFI and BFI-2 Measure.

NEO instruments

NEO measures belong to the Costa–McCrae Five-Factor Model tradition and include versions varying in depth. The extended NEO family includes a 30-facet structure specific to that tradition, while shorter forms offer less facet detail. Interpretation must follow the manual and norms for the actual version. See NEO Personality Inventory: NEO-PI-3 and NEO-FFI.

IPIP-based measures

The International Personality Item Pool is an ecosystem of publicly available personality items used in multiple scales. “IPIP Big Five” by itself does not identify a single validated questionnaire with one scoring key. A specific IPIP implementation requires evaluation of its items, psychometrics, and interpretation.

Big Five Aspect Scales

These scales operationalize ten narrower aspects under the five broad domains. Aspect scores should be evaluated using the measure's own evidence and should not be equated with BFI-2's 15 facets or the NEO facet set.

Ultra-short inventories

The TIPI and other brief measures may be useful when respondent burden must be minimized. They preserve some information about broad dimensions at the cost of narrower coverage and, frequently, less dependable individual interpretation.

Can Two Valid Big Five Tests Produce Different Profiles?

Yes. Instruments can vary in the subcomponents sampled, item weightings, norms, number of questions, and terminology. Temporary circumstances and measurement error can also shift results. Disagreement does not establish that either questionnaire is fraudulent.

Compare domain with domain, rather than a narrow facet with a whole trait. Check the scoring direction and whether labels refer to the same construct. Examine the norms and whether one report uses raw averages while another uses percentiles. Finally, consider the time interval and setting of the administrations.

If the discrepancy matters, examine the instruments' documentation and obtain additional evidence rather than averaging incompatible numbers or choosing the more flattering interpretation.

Does Retaking the Same Test Improve Accuracy?

Repeated administration may reveal stability when conditions are comparable, but repeating a questionnaire within minutes can introduce memory and confirmation effects. A retest is informative only if there is a reason for choosing its interval and a way to interpret the difference using instrument-specific evidence.

To judge genuine development over months or years, research must distinguish expected measurement variation from substantive trait change. A short-term retest coefficient cannot alone answer this longer-term question.

Are Free and Online Big Five Tests Trustworthy?

Some free online assessments derive from credible research instruments and can produce useful information. Others borrow the names of traits while changing items or scoring without validation. The price, popularity, graphic design, and length of the automated interpretation are separate from scientific quality.

A credible service identifies its exact questionnaire, cites studies involving that measure, explains its norms, describes scoring and limitations, and provides a clear statement about data handling. Respondents should know whether answers are stored, shared, sold, or used to train automated systems and should understand the service's retention and deletion practices.

Can AI Make Big Five Testing More Accurate?

Artificial intelligence can assist with questionnaire delivery, data quality review, statistical analysis, and inference of reported traits from text or digital traces. Each requires its own validation. Correlation between AI predictions and questionnaire scores is evidence about the model's predictive performance relative to those scores; it does not establish direct knowledge of a person's underlying personality.

A 2024 study of personality inference across two text domains reported that models performing well within one text domain generalized poorly or insignificantly to another. Results can depend on platform, language, training sample, genre, and the quality of the target questionnaire.

Use of digital traces also raises consent, privacy, fairness, and bias questions. A sophisticated algorithm cannot compensate for weak construct definitions or unrepresentative training and validation samples. Broad claims that AI “knows your personality” should be evaluated against out-of-sample evidence, measurement quality, and the relevant privacy context.

Can Big Five Scores Diagnose a Mental Health Condition?

Big Five inventories measure personality dimensions according to a particular questionnaire. Their results do not diagnose anxiety disorders, depressive disorders, personality disorders, or other clinical conditions. High or low scores cannot substitute for symptom history, impairment, differential assessment, or an appropriately qualified professional evaluation.

For example, Neuroticism-related scores and BFI-2 Negative Emotionality scores describe certain broad tendencies involving negative emotions. They are not diagnostic thresholds. Personality-trait results can sometimes provide context in a larger assessment, but their meaning remains bounded by the instrument and the purpose for which it was validated.

Should Big Five Tests Be Used for Hiring or Other High-Stakes Decisions?

Some trait scales have predictive relationships with particular work-related outcomes. Such research does not automatically validate a specific selection system for all applicants or occupations. The organization using a questionnaire needs evidence tailored to the actual job criteria, population, language, format, and consequences.

Response incentives, norm quality, accessibility, privacy, measurement error, and fairness can become especially important when test results influence real opportunities. High-stakes decisions should use relevant additional evidence and an explicitly evaluated procedure. A single profile should never be treated as a verdict on someone's capability, character, or future.

How to Evaluate a Big Five Test Before Taking It

First identify its full name and version. Next look for peer-reviewed validation involving the same instrument, rather than just research about the Big Five generally. Check the reliability and validity of the specific domains or facets you care about, the language and population studied, and how the questionnaire's norms were assembled.

Then examine whether the interpretation stays close to what scores actually mean. Continuous dimensions should remain continuous; no unexplained universal “high” or “low” threshold should be imposed. A scientifically literate report acknowledges measurement uncertainty, avoids diagnosing conditions, and uses probabilistic language for outcomes. Finally, inspect the provider's privacy information and consider whether the test is suitable for the decision you intend to make.

Frequently Asked Questions

Is the Big Five scientifically accurate?

Several established Big Five instruments have strong scientific support for particular interpretations. Their reliability and validity depend on the version, population, scale, and purpose. Endorsement of the general framework does not validate an unidentified online quiz.

What percentage accurate is a Big Five personality test?

There is no universal percentage. Reliability coefficients, agreement correlations, and predictive associations describe different statistical properties. A coefficient of .82 does not mean that a person's profile is 82% true.

Are Big Five test results stable over time?

Many well-developed inventories demonstrate substantial short-term retest dependability, but individual results can vary because of error, states, context, and genuine change over longer periods.

Which Big Five test is most reliable?

No one test is best for all people and uses. The BFI-2 has substantial published evidence, but an appropriate choice depends on desired detail, available norms, relevant validation, and the population being assessed.

Can I get a different result tomorrow?

Yes. A change may reflect measurement variability, the situation, interpretation of questions, or an actual difference in responses. It should not be interpreted as evidence of sudden personality transformation without additional information.

Do Big Five tests measure personality or mood?

They are designed to estimate characteristic dispositions. Current mood and other temporary states may influence some answers.

Does a high or low score indicate a problem?

No. The five domains are continuous descriptions of traits; score direction alone is neither a diagnosis nor a moral judgment. Interpretation depends on the instrument, context, and norms.

Are short Big Five tests useless?

No. Some short measures are efficient and useful for broad research purposes. They generally permit less detailed or less precise individual-level conclusions than appropriately validated longer instruments.

Can a Big Five result predict my career or relationship success?

It may be statistically related to certain outcomes in specific research samples, but a questionnaire cannot determine a person's future or establish compatibility or job suitability by itself.

Is an AI-generated Big Five profile more accurate?

Only if the measurement and resulting claims have been validated for that use. Automated writing can produce a persuasive narrative without improving the precision or validity of the underlying scores.

Conclusion: Trust the Evidence Behind the Instrument

A credible Big Five personality test is an estimate supported by evidence about how its scores behave. Strong assessment combines internal consistency, test–retest dependability, construct validity, cross-method convergence, relevant predictive evidence, and suitable norms. The findings collectively justify useful interpretations under defined conditions.

An accurate reading keeps the personality model, the questionnaire, and the proposed decision distinct. Trait scores describe tendencies along continuous dimensions, accompanied by uncertainty and context. Their scientific value comes from careful measurement and proportionate interpretation.

References

Barrick, M. R., & Mount, M. K. (1991). The Big Five personality dimensions and job performance: A meta-analysis. Personnel Psychology, 44(1), 1–26. https://doi.org/10.1111/j.1744-6570.1991.tb00688.x

Berggren, M., Kaati, L., Pelzer, B., Stiff, H., Lundmark, L., & Akrami, N. (2024). The generalizability of machine learning models of personality across two text domains. Personality and Individual Differences, 217, 112465. https://doi.org/10.1016/j.paid.2023.112465

Connelly, B. S., & Ones, D. S. (2010). An other perspective on personality: Meta-analytic integration of observers' accuracy and predictive validity. Psychological Bulletin, 136(6), 1092–1122. https://doi.org/10.1037/a0021212

Connolly, J. J., Kavanagh, E. J., & Viswesvaran, C. (2007). The convergent validity between self and observer ratings of personality: A meta-analytic review. International Journal of Selection and Assessment, 15(1), 110–117. https://doi.org/10.1111/j.1468-2389.2007.00371.x

Gnambs, T. (2014). A meta-analysis of dependability coefficients (test–retest reliabilities) for measures of the Big Five. Journal of Research in Personality, 52, 20–28. https://doi.org/10.1016/j.jrp.2014.06.003

Gnambs, T., & Kaspar, K. (2017). Socially desirable responding in web-based questionnaires: A meta-analytic review of the candor hypothesis. Assessment, 24(6), 746–762. https://doi.org/10.1177/1073191115624547

Gosling, S. D., Rentfrow, P. J., & Swann, W. B., Jr. (2003). A very brief measure of the Big-Five personality domains. Journal of Research in Personality, 37(6), 504–528. https://doi.org/10.1016/S0092-6566(03)00046-100046-1)

Husain, W., Haddad, A. J., Husain, M. A., et al. (2025). Reliability generalization meta-analysis of the internal consistency of the Big Five Inventory (BFI) by comparing BFI (44 items) and BFI-2 (60 items) versions controlling for age, sex, language factors. BMC Psychology, 13, 20. https://doi.org/10.1186/s40359-024-02271-x

Laajaj, R., Macours, K., Pinzon Hernandez, D. A., et al. (2019). Challenges to capture the big five personality traits in non-WEIRD populations. Science Advances, 5(7), eaaw5226. https://doi.org/10.1126/sciadv.aaw5226

Oh, I.-S., Wang, G., & Mount, M. K. (2011). Validity of observer ratings of the five-factor model of personality traits: A meta-analysis. Journal of Applied Psychology, 96(4), 762–773. https://doi.org/10.1037/a0021832

Soto, C. J., & John, O. P. (2017a). 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

Soto, C. J., & John, O. P. (2017b). Short and extra-short forms of the Big Five Inventory–2: The BFI-2-S and BFI-2-XS. Journal of Research in Personality, 68, 69–81. https://doi.org/10.1016/j.jrp.2017.02.004

Thørrisen, M. M., & Sadeghi, T. (2023). The Ten-Item Personality Inventory (TIPI): A scoping review of versions, translations and psychometric properties. Frontiers in Psychology, 14, 1202953. https://doi.org/10.3389/fpsyg.2023.1202953

Psychological Encyclopedia