Character Personality Statistics
Cited prevalence and distribution data for personality systems, updated June 2026
About these statistics
This page collects published prevalence data for the main personality frameworks. You'll find numbers for 16 personality types, the Enneagram, the Big Five, attachment styles, defense mechanisms, cognitive biases, and stress responses. Data quality varies — opt-in surveys and weak samples are noted. Use the numbers to make fictional characters feel realistic.
Key takeaways
- ISFJ is the most common 16-type type (~13.8%), and INFJ is the rarest (~1.5%), per the CPP 16-type Manual convenience sample.
- Secure attachment is the modal adult style (~52–56%), with anxious and avoidant each around 20–23% (meta-analytic, peer-reviewed).
- Big Five traits shift with age: Neuroticism, Extraversion, and Openness decline through the 20s; Agreeableness and Conscientiousness rise (Srivastava et al., 2003).
- Enneagram frequencies are community-sourced, not probability-sampled. Type 9 is most commonly reported (~13–15%); Type 4 among the least. Treat these as tendencies, not population statistics.
- Neuroticism is higher in women by about d ≈ 0.25–0.40, a well-replicated cross-cultural finding (Schmitt et al., 2008, 55 nations).
- Confirmation bias is near-universal, with over 90% of lab participants showing biased assimilation in classic studies (Lord, Ross & Lepper, 1979).
- ~70% of adults report physical or psychological stress symptoms in the past month (APA Stress in America, 2023).
- Big Five heritability averages ~40–60% across twin studies (Jang et al., 1996; Vukasović & Bratko, 2015 meta-analysis ≈ 0.40).
16-type personality
Estimated share of each of the 16 types in the general population. Figures come from the CPP 16-type Manual (3rd ed., 2003), a large self-selected convenience sample rather than a nationally representative probability sample. They are the most widely reproduced distribution in popular references but should be read as descriptive of that sample.
| 16-type Type | Share of population | Common name | Source & notes |
|---|---|---|---|
| ISFJ | ~13.8% | Defender | CPP 16-type Manual (3rd ed., 2003). Most common type. Self-selected sample. |
| ESFJ | ~12.3% | Consul | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ISTJ | ~11.6% | Inspector | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ISFP | ~8.8% | Adventurer | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ESTJ | ~8.7% | Executive | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ENFP | ~8.1% | Campaigner | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ISTP | ~5.4% | Virtuoso | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ESFP | ~8.5% | Entertainer | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ENFJ | ~2.5% | Protagonist | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| INTP | ~3.3% | Logician | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ESTP | ~4.3% | Entrepreneur | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| INTJ | ~2.1% | Architect | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ENTP | ~3.2% | Debater | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| ENTJ | ~1.8% | Commander | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| INFP | ~4.4% | Mediator | CPP 16-type Manual (3rd ed., 2003). Self-selected sample. |
| INFJ | ~1.5% | Advocate | CPP 16-type Manual (3rd ed., 2003). Rarest type. Self-selected sample. |
16-type dichotomy breakdowns: Introverts ~50.7%, Extraverts ~49.3%; Sensing ~73.3%, Intuition ~26.7%; Thinking ~40.2%, Feeling ~59.8%; Judging ~54.0%, Perceiving ~46.0%. Source: CPP 16-type Manual (3rd ed., 2003). The strong S/N and T/F skews are partly why 16-type's test-retest reliability is weaker than the Big Five's.
Test-retest reliability: Studies report ~36–50% of people get a different four-letter type on retest within five weeks (Pittenger, 2005; Stein & Swan, 2019). Caveat any single "type" as state-dependent.
Enneagram type frequency
No nationally representative probability sample exists for Enneagram type prevalence. The figures below are drawn from community and online self-selection surveys (e.g., the Enneagram Institute's large web samples and Truity's 2020 community report of ~54,000 respondents). They describe people who opt into Enneagram content, not the general population.
| Enneagram Type | Reported share | Name | Source & notes |
|---|---|---|---|
| Type 9 | ~13–15% | Peacemaker | Truity 2020 community survey (~54k respondents); Enneagram Institute web samples. Most commonly reported. |
| Type 6 | ~11–13% | Loyalist | Truity 2020. Some teachers argue Type 6 is under-reported because it is anxious about labels. |
| Type 7 | ~10–12% | Enthusiast | Truity 2020. |
| Type 2 | ~11–13% | Helper | Truity 2020. Higher in female respondents. |
| Type 3 | ~10–11% | Achiever | Truity 2020. |
| Type 1 | ~9–10% | Reformer | Truity 2020. |
| Type 8 | ~8–9% | Challenger | Truity 2020. Higher in male respondents. |
| Type 5 | ~8–10% | Investigator | Truity 2020. |
| Type 4 | ~10–11% | Individualist | Truity 2020. Among the most commonly reported in online Enneagram communities, possibly over-represented due to self-selection. |
Wing preferences: In Truity's 2020 sample, about 70–80% of respondents identified a clear wing, with the rest reporting balanced or no-wing. Self-selection bias is substantial: Type 4 appears over-represented online relative to clinical estimates.
Instinctual variants: The Enneagram Institute's web samples suggest self-preservation is reported by roughly 45–50% of respondents, social by 30–35%, and sexual/one-to-one by 20–25%. These are descriptive of the community, not validated population norms.
Big Five trait distributions
The Big Five (OCEAN) is the most empirically validated personality model. Norms below are illustrative population means from large-sample studies, with sex differences and heritability estimates. Big Five scores are continuously distributed (not categorical), so "prevalence" is expressed as means, standard deviations, and percentile norms.
| Trait & dimension | Population mean (norm) | Sex difference | Source & notes |
|---|---|---|---|
| Openness to Experience | Mean ≈ 3.5–3.7 / 5 (normative samples) | Small: d ≈ 0.05–0.10 (women slightly higher) | Soto & John (2017) Big Five normative data; Schmitt et al. (2008) cross-cultural. |
| Conscientiousness | Mean ≈ 3.4–3.6 / 5 | Small: d ≈ 0.05–0.15 (women slightly higher) | Soto & John (2017); increases with age through the 20s–30s. |
| Extraversion | Mean ≈ 3.0–3.3 / 5 | Small: d ≈ 0.10–0.15 (men slightly higher in assertiveness facet) | Soto & John (2017). Declines modestly with age. |
| Agreeableness | Mean ≈ 3.4–3.6 / 5 | Small–moderate: d ≈ 0.20–0.30 (women higher) | Soto & John (2017); Schmitt et al. (2008). |
| Neuroticism | Mean ≈ 2.7–3.0 / 5 | Moderate: d ≈ 0.25–0.40 (women higher) | Schmitt et al. (2008), 55 nations. Most replicated sex difference in personality. |
Age trajectories (Srivastava et al., 2003, n ≈ 132k web sample): Neuroticism declines most in the 20s and is lowest in the 50s+; Conscientiousness and Agreeableness rise through the 20s–40s; Openness peaks in the late teens/early 20s then declines slightly; Extraversion declines gradually across adulthood.
Heritability: Twin studies put Big Five heritability at roughly 40–60% (Jang et al., 1996, Canadian twins). A meta-analysis of 2,902 twin pairs across 24 studies estimated average heritability ≈ 0.40 (Vukasović & Bratko, 2015).
Cross-cultural stability: The Big Five structure replicates in ~50+ cultures (McCrae et al., 2005, 51-culture IPPD study), though mean levels differ (e.g., higher Neuroticism means in East Asian samples, higher Extraversion in some Western samples).
Attachment style prevalence in adults
Adult attachment is typically measured by self-report (ECR-R or the original Hazan & Shaver three-category measure). The meta-analytic estimates below are among the most robust personality prevalence figures on this page.
| Attachment style | Prevalence (adults) | Source & notes |
|---|---|---|
| Secure | ~52–56% | Hazan & Shaver (1987) community samples; meta-analytic update by Schindler et al. (2010, 4-category). Most prevalent. |
| Anxious / Preoccupied | ~19–21% | Schindler et al. (2010); ECR-R community norms (Sibley & Liu, 2004). |
| Dismissive-Avoidant | ~14–16% | Schindler et al. (2010, 4-category). Some 3-category studies fold this into a single "avoidant" bucket. |
| Fearful-Avoidant / Disorganized | ~5–7% | Schindler et al. (2010). Rates are higher (up to ~80%) in clinical and trauma-exposed samples (Bakermans-Kranenburg & van IJzendoorn, 2009). |
Childhood roots: A meta-analysis of 88 studies (van IJzendoorn & Bakermans-Kranenburg, 2003) found adult-attachment-classification concordance with the Strange Situation at about r ≈ 0.47, supporting intergenerational transmission. Disorganized attachment in infancy is ~15% in non-clinical samples but rises to ~80% in maltreated children (Carlson et al., 1989; Cyr et al., 2010 meta-analysis).
Stability: Adult attachment shows moderate rank-order stability (test-retest r ≈ 0.45–0.65 over months–years; Sibley & Liu, 2004) but is more state-responsive than Big Five traits.
Defense mechanism frequency
Defense mechanisms are unconscious strategies for managing internal conflict. The most-cited prevalence data come from the Vaillant (1992) longitudinal Grant Study cohort and the Defense Mechanism Rating Scale literature. Categories below follow Vaillant's hierarchy (mature → neurotic → immature).
| Defense mechanism (level) | Frequency / finding | Source & notes |
|---|---|---|
| Mature defenses (e.g., sublimation, humor, anticipation, suppression) | Used by ~30–40% of healthy adults as their modal defense | Vaillant (1992), 50-year Grant Study. Predicts better midlife adjustment. |
| Neurotic defenses (e.g., repression, displacement, reaction formation) | ~30–45% modal in non-clinical adults | Vaillant (1992); Perry & Henry (2004) review. |
| Immature defenses (e.g., projection, passive-aggression, acting out, splitting) | ~15–25% modal in community samples; >50% in personality-disordered samples | Vaillant (1992); Perry & Henry (2004). |
| Denial | Common in acute grief/medical crisis (situationally normal); rarely the modal defense in healthy adults (<5%) | DSM-IV-TR defense-functioning review (American Psychiatric Association, 2000, Appendix B). |
| Repression | Estimated 15–20% of adults show a repressive coping style (low anxiety + high defensiveness) | Weinberger et al. (1979); Derakshan & Eysenck (2001) review. Estimated (varies by measure). |
| Projection | Among the most common immature defenses; elevated in paranoia and certain personality disorders | Vaillant (1992); Semrad et al. reanalysis. |
| Humor (mature) | Correlates with longer life and lower midlife morbidity in the Grant Study cohort | Vaillant (2000). Descriptive, not a prevalence figure. |
Clinical note: Defense-style maturity predicts long-term adjustment better than many symptoms (Vaillant, 1992). In personality-disorder samples, immature defenses dominate modal use (Perry & Henry, 2004).
Cognitive bias prevalence
Cognitive biases are systematic deviations from rational judgment. Prevalence here means the share of typical lab participants showing the effect in classic studies; effect sizes (Cohen's d or r) are reported where available. These are the most-studied biases for character work (see the Cognitive Bias Generator).
| Bias | Prevalence / effect | Source & notes |
|---|---|---|
| Confirmation bias | Biased assimilation in >90% of participants (studies show the effect is near-universal) | Lord, Ross & Lepper (1979); Nickerson (1998) review. |
| Anchoring | Effect size large (η² ≈ 0.20–0.40); present in ~80%+ of participants when anchor is salient | Tversky & Kahneman (1974); Furnham & Boo (2011) review. |
| Availability heuristic | Drives ~60–75% of frequency estimates in classic word-list tasks (e.g., "r" first-letter vs third-letter) | Tversky & Kahneman (1973). |
| Dunning–Kruger effect | Bottom quartile overestimate performance by ~30–50 percentile points | Kruger & Dunning (1999). |
| Framing effect | ~70–80% flip preference between gain- and loss-framed versions of equivalent gambles | Tversky & Kahneman (1981). |
| Halo effect | Correlation between unrelated rated traits often r ≈ 0.30–0.50 when one salient trait is positive | Thorndike (1920); Nisbett & Wilson (1977). |
| Sunk-cost fallacy | ~50–70% continue an unprofitable project after investment, vs ~30% for new entrants | Arkes & Blumer (1985). |
| Base-rate neglect | ~60–85% ignore base rates in classic "Tom W." / engineer–lawyer problems | Kahneman & Tversky (1973). |
| Better-than-average effect | ~80%+ of drivers rate themselves "above average" (Illusory superiority) | Svenson (1981); Dunning et al. (1989). |
| Hindsight bias | Present in ~75–85% of participants across review of 120+ studies (mean effect d ≈ 0.39) | Blank et al. (2007) meta-analysis; Fischhoff (1975). |
Replication caveat: Many classic bias effects shrink under pre-registration and larger samples; the magnitudes above are from the original or canonical studies. Treat as ballpark, not fixed constants.
Stress response distribution
How people respond to stress varies by physiology and personality. Figures below mix population survey data (APA Stress in America) and laboratory findings (e.g., Taylor et al., 2000, on fight-or-flight vs tend-and-befriend).
| Stress response pattern | Prevalence / finding | Source & notes |
|---|---|---|
| Fight-or-flight (sympathetic activation) | Default autonomic pattern in acute threat; prevalence near-universal as a capacity, expressed situationally | Cannon (1932). Descriptive. |
| Tend-and-befriend | More frequently observed in female samples under stress (~60–70% in Taylor et al. samples) | Taylor et al. (2000). Oxytocin-modulated; sex difference is a tendency, not absolute. |
| Freeze / tonic immobility | Reported in ~10–20% of trauma-exposed individuals during the event | Marx et al. (2008); Heidt et al. (2005). Higher in assault survivors. |
| Adults reporting physical stress symptoms in past month | ~70% (2023 US adult survey) | APA Stress in America (2023). Self-report, non-probability online panel. |
| Adults reporting psychological stress symptoms in past month | ~60–65% (2023 US adult survey) | APA Stress in America (2023). |
| Adults reporting "a great deal of stress" the prior day | ~44% (Gallup, 2023 global) | Gallup Global Emotions Report (2023). 142-country probability survey (more robust than convenience samples). |
| Problem-focused coping (active) | Modal strategy in ~50–60% of community adults | Carver et al. (1989) COPE validation samples. Estimated. |
| Emotion-focused / avoidant coping | Modal in ~30–40%; rises under uncontrollable stressors | Carver et al. (1989). |
| Neuroticism amplifies perceived stress | Neuroticism correlates r ≈ 0.40–0.55 with perceived stress scales | Bolger & Eckenrode (1986); meta-analyses confirm the link. |
Sex difference: Taylor et al. (2000) argue "tend-and-befriend" is more prevalent in females, partly oxytocin-mediated; the effect is a tendency rather than a strict dichotomy. Use it cautiously for character work.
Values & moral foundations distribution
Personal values are commonly measured with Schwartz's circumplex or Haidt's Moral Foundations theory. Below are illustrative population-level findings, not fixed norms.
| Value / foundation | Finding | Source & notes |
|---|---|---|
| Self-transcendence (Benevolence + Universalism) | Endorsed more by women (d ≈ 0.20–0.30) cross-culturally | Schwartz & Rubel (2005), 70-country IPPD samples. |
| Self-enhancement (Power + Achievement) | Endorsed more by men (d ≈ 0.20–0.40) | Schwartz & Rubel (2005). |
| Openness to change (Stimulation + Self-direction) | Slightly higher in men and younger adults | Schwartz & Rubel (2005). |
| Conservation (Tradition + Conformity + Security) | Endorsed more by older adults and less-WEIRD samples | Schwartz (1992); cross-cultural replication. |
| Individualizing foundations (Harm / Fairness) | Higher in self-identified liberals | Haidt (2012); Graham et al. (2011). |
| Binding foundations (Loyalty / Authority / Purity) | Higher in self-identified conservatives | Haidt (2012); Graham et al. (2011). |
| Achievement value priority declines with age | Documented across ~60 cultures | Schwartz (2005) IPPD age-cohort data. |
WEIRD caveat: Most Schwartz value means are calibrated on WEIRD (Western, Educated, Industrialized, Rich, Democratic) student samples; means shift meaningfully in non-WEIRD cultures. Treat as tendencies, not universals.
Personality disorder & trait prevalence (DSM-5)
For clinical context, prevalence of personality-disorder trait patterns in the general US adult population. Useful for depicting characters at the extremes of trait distributions.
| Pattern | US adult prevalence | Source & notes |
|---|---|---|
| Any personality disorder (DSM-5) | ~9–15% | NESARC (2008); Lenzenweger (2008), with different studies converging around 9–15%. |
| Borderline personality disorder | ~1.4–1.6% | NESARC (2008); Tomko et al. (2014). |
| Narcissistic personality disorder | ~0.5–1.0% (higher in men) | NESARC (2008); Stinson et al. (2008). |
| Antisocial personality disorder | ~0.6–3.6% (varies; higher in men and substance-using samples) | NESARC (2008); Compton et al. (2005). |
| Avoidant personality disorder | ~1.2–2.4% | NESARC (2008). |
| Schizotypal personality disorder | ~0.6–1.2% | NESARC (2008). |
| Obsessive-compulsive personality disorder | ~2.1–2.4% | NESARC (2008). Most common Cluster C in NESARC. |
NESARC = National Epidemiologic Survey on Alcohol and Related Conditions, a large US probability sample. These are the most robust clinical prevalence figures on this page.
Methodology and limits
How reliable are these numbers? The Big Five and adult attachment stats are the most robust since they come from peer-reviewed studies and replicated samples. The 16-type distribution comes from the publisher's convenience sample—decent for general rank order (ISFJ is common, INFJ is rare) but not a perfect population census. Enneagram numbers rely entirely on self-selected online surveys, so treat them as community trends rather than hard demographics. Lastly, classic cognitive bias percentages come from original lab studies and often shrink in modern replications.
Why there is no official "Enneagram population" data: Nobody has run a representative, random-sample study on the Enneagram. Every stat we have comes from web surveys. Because introspective and individualistic people (like Type 4s) are much more likely to take personality tests online, they are heavily overrepresented. Treat these numbers as relative popularity in online communities, not census data.
Why 16-type frequencies look different depending on where you look: The publisher's official sample shows about 73% Sensing types and 60% Feeling types. But if you look at online test results (like 16Personalities), the share of Intuitive types sky-rockets. That’s not because the world changed, but because Intuitive types are far more likely to spend time taking personality tests online. Always consider who actually took the test.