The writer may make a research claim only if a paper here supports it. Each says what it supports and what it does not, and the second list does more work than the first: it names the boundary a confident writer would otherwise walk straight across.
Core — in front of the model for every report
These carry the effect-size discipline, the replication discount, the dimensional structure and the anti-Barnum rules. On any conflict with a retrieved paper, the core wins.
Johnson, J. A. (2014). Measuring thirty facets of the Five Factor Model with a 120-item public domain inventory. Journal of Research in Personality, 51, 78–89.
johnson-2014 · doi:10.1016/j.jrp.2014.05.003
Sample. Internet sample N = 619,150 with validation subsamples.
Findings. Domain alphas: O .81, C .90, E .89, A .86, N .90. IPIP-NEO scales correlate on average r ≈ .73 with corresponding NEO-PI-R scales (.94 corrected for attenuation).
Licenses. Facet-level scoring and interpretation. Use of Johnson's sex and age norms. The claim that the IPIP-NEO is a valid public-domain NEO-PI-R analogue.
Does not license. Item-level clinical inference. Claims of superiority over the commercial NEO. Treating the instrument as diagnostic.
Caveat. Documents the 120-item version. This engine administers the 300, whose facet scales are longer and generally more reliable — cite for method and validity logic, not for exact 300-item alphas.
DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between facets and domains: 10 aspects of the Big Five. JPSP, 93(5), 880–896.
deyoung-2007 · doi:10.1037/0022-3514.93.5.880
Sample. Eugene-Springfield community sample N = 481; validation sample N = 480.
Findings. Two correlated aspects within each domain: Volatility/Withdrawal, Enthusiasm/Assertiveness, Intellect/Openness, Compassion/Politeness, Industriousness/Orderliness.
Licenses. Interpreting within-domain facet splits as two coherent aspects. The aspect labels used by the salience layer.
Does not license. Treating aspects as fully independent. Presenting the biological-substrate claim as established — it rests on secondary genetic data.
Caveat. Community panel, not nationally representative.
Soto, C. J. (2019). How replicable are links between personality traits and consequential life outcomes? The Life Outcomes of Personality Replication Project. Psychological Science, 30(5), 711–727.
soto-2019 · doi:10.1177/0956797619831612
Sample. >6,100 adults across 4 online samples; 78 preregistered trait-outcome associations.
Findings. 87% of replication attempts were significant in the expected direction. Replication effects were typically 77% as strong as the original effects.
Licenses. Trusting main-effect trait-outcome associations. Discounting headline effect sizes to roughly 77% of their original magnitude.
Does not license. Trait-by-trait interaction claims — these replicate poorly. Causal language.
Gerlach, M., Farb, B., Revelle, W., & Amaral, L. A. N. (2018). A robust data-driven approach identifies four personality types across four large data sets. Nature Human Behaviour, 2, 735–742.
gerlach-2018 · doi:10.1038/s41562-018-0419-z
Sample. >1.5 million participants across four datasets.
Findings. Four replicable density clusters within a fundamentally continuous space; most clustering solutions are spurious. A published commentary (PMC7931870) shows a skewed distribution with no true cluster structure can produce these apparent types.
Licenses. Discussing type-like patterns cautiously. Serving flat or average profiles honestly — "average" is the largest region of the space.
Does not license. Assigning users to discrete types as if personality were categorical.
Forer, B. R. (1949). The fallacy of personal validation: a classroom demonstration of gullibility. Journal of Abnormal and Social Psychology, 44(1), 118–123.
forer-1949 · doi:10.1037/h0059240
Sample. N = 39 students, each rating an identical 13-statement generic sketch; mean accuracy 4.26/5.
Findings. Acceptance rises with perceived relevance, specificity and favourability.
Licenses. The core anti-Barnum safeguard: forbid vague, universally acceptable statements and require score-anchored, discriminating claims.
Does not license. Any claim about whether people rate AI-generated descriptions as accurate — that extension is untested.
Mõttus, R., Bates, T. C., Condon, D. M., Mroczek, D., & Revelle, W. (2017). Leveraging a more nuanced view of personality: narrow characteristics predict and explain variance in life outcomes.
mottus-2017 · doi:10.31234/osf.io/4q9gv
Findings. Item-based nuance models explained a median 9.7% of outcome variance vs 4.2% for domain-based and 5.9% for facet-based models. Nuances show rank-order stability (median .77) and heritability (median .52) comparable to higher-order traits.
Licenses. The item-level salience layer. The claim that narrow characteristics carry unique, valid predictive signal.
Does not license. Over-interpreting any single item as diagnostic — nuance reliability is lower than domain reliability, so item-level claims must be probabilistic and hedged.
Lahey, B. B. (2009). Public health significance of neuroticism. American Psychologist, 64(4), 241–256.
lahey-2009 · doi:10.1037/a0015309
Findings. Neuroticism is a robust correlate and predictor of many mental and physical disorders, comorbidity, service use, and quality and longevity of life.
Licenses. Taking high Neuroticism seriously as a health-relevant dimension. Justifying gentle safety routing.
Does not license. Diagnosing any disorder. Equating high Neuroticism with illness — it is a dimensional risk factor, not a diagnosis.
Roberts, B. W., Walton, K. E., & Viechtbauer, W. (2006). Patterns of mean-level change in personality traits across the life course. Psychological Bulletin, 132(1), 1–25.
roberts-2006-change · doi:10.1037/0033-2909.132.1.1
Sample. 92 longitudinal samples.
Findings. The maturity principle: increases in social dominance, Conscientiousness and emotional stability, especially ages 20–40.
Licenses. Telling users traits normatively shift with age; framing current scores as not fixed.
Does not license. Promising change to an individual — these are mean-level trends.
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.
weisberg-2011 · doi:10.3389/fpsyg.2011.00178
Findings. Women scored higher on E, A and N at the domain level; more extensive differences appeared at the aspect level, where aspects can diverge and cancel out at the domain level.
Licenses. Justifying sex- and age-specific norms. Interpreting aspect-level sex differences.
Does not license. Any claim about an individual from group means.
Kajonius, P. J., & Johnson, J. A. (2019). Assessing the structure of the Five Factor Model of personality (IPIP-NEO-120) in the public domain. Europe's Journal of Psychology, 15(2), 260-275.
kajonius-johnson-2019 · doi:10.5964/ejop.v15i2.1671
Sample. US N = 320,128.
Findings. Hierarchical and bifactor models showed tolerable fit; some facets are substantially more representative of their domain than others.
Licenses. Domain and facet structural claims for the IPIP-NEO specifically.
Does not license. Assuming all facets are equally strong markers of their domain.
Roberts, B. W., Luo, J., Briley, D. A., Chow, P. I., Su, R., & Hill, P. L. (2017). A systematic review of personality trait change through intervention. Psychological Bulletin, 143(2), 117-141.
roberts-2017-intervention · doi:10.1037/bul0000088
Sample. 207 studies tracking personality change during intervention; average duration 24 weeks (median 13). Predominantly clinical — only 19 studies used nonclinical samples.
Findings. Interventions were associated with change in personality trait measures of about d = .37 over an average of 24 weeks. Emotional stability changed most, followed by extraversion; type of therapy was not strongly associated with the amount of change. Changes replicated across experimental and nonexperimental designs and persisted in longitudinal follow-ups beyond the intervention. Corrected for small-study bias (PEESE), the experimental estimate falls to d = .13, 95% CI [-.10, .36].
Licenses. The claim that personality traits can change through deliberate intervention, rather than only drifting normatively with age. Naming Neuroticism as the domain with the best evidence for change, since emotional stability was the only trait robust to the bias corrections. Saying that change observed during intervention tends to persist afterwards rather than snapping back.
Does not license. Transferring the headline d = .37 to a reader who is not in treatment. 188 of the 207 studies treated psychopathology; only 19 used nonclinical samples. Presenting .37 as settled. The publication-bias-corrected experimental estimate is d = .13 with a confidence interval crossing zero, and the paper says so. Claims that Extraversion, Openness, Agreeableness or Conscientiousness are reliably movable — the paper states only emotional stability survived its small-study bias analyses and that changes in the rest should be read with caution. Any claim that a particular therapy or technique is better at changing traits; type of therapy did not predict the amount of change. Treating the change as demonstrated at the level of the trait rather than the measure. The data set is almost entirely self-report, which the paper names as its salient limitation because response sets cannot be ruled out.
Caveat. The one paper here that speaks to the premise of the "What to do with this" section, which every report contains. Before it, deliberate trait change rested on stieger-2021 — a single RCT.
Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2023). Revisiting the design of selection systems in light of new findings regarding the validity of widely used predictors. Industrial and Organizational Psychology, 16(3), 283-300.
sackett-2023 · doi:10.1017/iop.2023.24
Sample. Focal article restating the revised operational validity estimates and their applied implications.
Findings. Carries the revised validity matrix: general mental ability revised to about .31, and conscientiousness to about .19. Frames the revisions as a correction to range-restriction practice rather than a dismissal of the predictors.
Licenses. Presenting personality-outcome effects as real but modest. Explicitly correcting inflated older numbers.
Does not license. Dismissing personality — effects remain meaningful and incremental. Transferring workplace validities to non-work life outcomes.
Caveat. Replaced sackett-2022 in core, which carried the same revised matrix at three times the length. The 2022 paper is the primary source and remains the citation of record for the range-restriction argument itself; this is the companion that states the conclusions.
Vazire, S. (2010). Who knows what about a person? The self-other knowledge asymmetry (SOKA) model. Journal of Personality and Social Psychology, 98(2), 281-300.
vazire-2010 · doi:10.1037/a0017908
Sample. N = 165, round-robin design with up to four friends and four strangers, plus behavioural criteria.
Findings. The self is the better judge of traits low in observability, such as anxiety and mood. Others are the better judges of traits high in evaluativeness, such as intellect. For extraversion-related traits, self and other perspectives are roughly equally accurate.
Licenses. The "What this cannot tell you" section: naming which parts of a self-report are most and least trustworthy, and why. Saying that a reader own view is the better source for internal states and the weaker one for evaluative traits.
Does not license. Population estimates of self-other agreement; this is a single laboratory study and does not quantify them. Any claim that self-report is invalid. It describes where each perspective is stronger, not that either fails.
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.
gnambs-2014 · doi:10.1016/j.jrp.2014.06.003
Sample. 682 test-retest correlations from 74 samples, total N = 14,923, over intervals of up to two months.
Findings. The median aggregated dependability estimate across the five traits was .816. Transient error accounted for about 10% of the observed variance in Big Five scores.
Licenses. Stating honestly that a score would move somewhat on a retest, and by roughly how much. Distinguishing measurement noise from real change, which the report needs before it advises anyone to act on a number.
Does not license. Any claim about long-term stability. The intervals here are two months or less; roberts-delvecchio-2000 is the source for rank-order consistency over years. Treating .816 as this instrument reliability. It is a cross-measure median, and student-heavy samples inflate it relative to a general population.
Retrieved — matched to your scores
Each declares plain threshold conditions. Where your profile satisfies one, that paper is the first place the writer is told to look. Your report’s source list names every paper it actually drew on.
Laajaj, R., et al. (2019). Challenges to capture the Big Five personality traits in non-WEIRD populations. Science Advances, 5(7), eaaw5226.
laajaj-2019 · doi:10.1126/sciadv.aaw5226
Sample. Face-to-face N = 94,751 across 23 low/middle-income countries; internet N = 198,356.
Findings. In less-educated face-to-face samples the Big Five structure often failed to replicate; internet samples in the same countries replicated well. The failure is driven by respondent characteristics and administration, not nationality.
Licenses. Caution about applying US norms to non-native-English respondents. The caveat that profile shape is more trustworthy than absolute level for such readers.
Does not license. Dismissing a non-US reader's results as invalid.
Retrieved when: English is not your first language OR you are outside the US
Anglim, J., Horwood, S., Smillie, L. D., Marrero, R. J., & Wood, J. K. (2020). Predicting psychological and subjective well-being from personality: a meta-analysis. Psychological Bulletin, 146(4), 279–323.
anglim-2020 · doi:10.1037/bul0000226
Sample. k = 462 samples, N = 334,567 at domain level; facet analyses include an IPIP-NEO sample (N = 903).
Findings. N, E and C are the strongest personality predictors of well-being; facets add meaningful incremental prediction over domains. Extraversion maps to positive affect; Neuroticism to negative affect and life satisfaction.
Licenses. All wellbeing-component interpretation; facet-level subjective wellbeing claims.
Does not license. Causal or deterministic wellbeing claims.
Caveat. Supersedes Steel 2008 and DeNeve & Cooper 1998. Very long (~45k tokens) — prefer a condensed digest of the facet and wellbeing tables for routine retrieval.
Retrieved when: N >= 75 OR N <= 25 OR E >= 75 OR E <= 25
Judge, T. A., Rodell, J. B., Klinger, R. L., Simon, L. S., & Crawford, E. R. (2013). Hierarchical representations of the Five-Factor Model in predicting job performance. Journal of Applied Psychology, 98(6), 875–925.
judge-2013 · doi:10.1037/a0033901
Findings. Meta-analyzes all 30 NEO facets against overall job performance, integrating DeYoung aspects and Costa-McCrae facets.
Licenses. Facet-level occupational interpretation across all five domains. The bandwidth-fidelity rationale for reporting facets.
Does not license. Treating any single facet as a strong standalone predictor — most facet validities are modest.
Caveat. Very long (~45k tokens); prefer a condensed facet-table digest.
Retrieved when: C1 >= 75 OR C4 >= 75 OR C >= 75 OR C2 >= 75
Credé, M., Tynan, M. C., & Harms, P. D. (2017). Much ado about grit: a meta-analytic synthesis of the grit literature. JPSP, 113(3), 492–511.
crede-2017 · doi:10.1037/pspp0000102
Sample. 584 effect sizes, 88 independent samples, N = 66,807.
Findings. Grit's higher-order structure is not confirmed; grit correlates with Conscientiousness at ρ = .84 (k = 22, N = 18,826) — largely a jangle fallacy. Perseverance of effort has stronger criterion validity than consistency of interest.
Licenses. Interpreting high Achievement-striving and Self-discipline as the substance behind "grit". Debunking grit as a distinct super-trait.
Does not license. Recommending grit-building as an evidence-based intervention. Treating grit as incrementally valid over Conscientiousness.
Caveat. This is the corrective meta-analysis; it supersedes Duckworth 2007.
Retrieved when: C4 >= 75 OR C5 >= 75
Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: a meta-analytic review. PLoS Medicine, 7(7), e1000316.
holt-lunstad-2010 · doi:10.1371/journal.pmed.1000316
Sample. 148 studies, N = 308,849, mean follow-up 7.5 years.
Findings. OR = 1.50 (95% CI 1.42–1.59) for survival with stronger social relationships; strongest for complex measures of social integration (OR = 1.91).
Licenses. Framing social connection as a health-relevant asset.
Does not license. Equating introversion with social isolation. The construct is relationship quantity and quality, not trait Extraversion. This is the boundary a naive reader misses.
Retrieved when: E <= 25 OR E2 <= 25
Nguyen, T. T., Ryan, R. M., & Deci, E. L. (2018). Solitude as an approach to affective self-regulation. PSPB, 44(1), 92–106.
nguyen-2018 · doi:10.1177/0146167217733073
Findings. Solitude produces a deactivation effect, lowering both high-arousal positive and high-arousal negative affect. Autonomy is the moderator distinguishing restorative solitude from lonely isolation.
Licenses. Framing chosen solitude as adaptive for low-Extraversion profiles.
Does not license. Claiming solitude raises happiness — the effect is arousal reduction, not valence improvement.
Caveat. Per-study Ns should be confirmed from the full text before quoting.
Retrieved when: E <= 30 OR E2 <= 30 OR E5 <= 30
Asendorpf, J. B. (1990). Beyond social withdrawal: shyness, unsociability, and peer avoidance. Human Development, 33(4-5), 250–259.
asendorpf-1990 · doi:10.1159/000276522
Findings. Three withdrawal subtypes from approach × avoidance motivation: shyness, unsociability, avoidance. Unsociability is comparatively benign.
Licenses. Distinguishing chosen solitude (unsociability) from anxious withdrawal (shyness) — critical for not pathologizing low affiliation.
Does not license. One-to-one mapping onto adult Big Five facets — the original sample was children. Use as a conceptual lens.
Retrieved when: E <= 30 AND N4 <= 60 OR E1 <= 25
Turiano, N. A., Graham, E. K., Weston, S. J., et al. (2020). Is healthy neuroticism associated with longevity? A coordinated integrative data analysis. Collabra: Psychology, 6(1), 33.
weston-2020 · doi:10.1525/collabra.268
Sample. 12 cohort studies, total N = 44,702.
Findings. Conscientiousness consistently predicted lower mortality hazard; Neuroticism showed no consistent pattern. NO study provided statistical evidence of a Neuroticism × Conscientiousness interaction.
Licenses. Bluntly stating that the "healthy neuroticism" / "anxious achiever" interaction is not empirically supported.
Does not license. Telling a high-N/high-C reader their combination is protective. This paper licenses the negative claim only.
Retrieved when: N >= 75 AND C >= 75
Kaufman, S. B., et al. (2016). Openness to Experience and Intellect differentially predict creative achievement in the arts and sciences. Journal of Personality, 84(2), 248–258.
kaufman-2016 · doi:10.1111/jopy.12156
Sample. 4 demographically diverse samples, N = 1,035.
Findings. The Openness aspect predicts creative achievement in the arts; the Intellect aspect predicts achievement in the sciences.
Licenses. The Intellect vs Openness aspect-split interpretation. Differentiating artistic from scientific creative potential.
Does not license. Claiming high Openness causes creative achievement — correlations are modest and achievement is multiply determined.
Retrieved when: O2 >= 75 OR O5 >= 75 OR O1 >= 75
Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: a meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69–119.
gollwitzer-sheeran-2006 · doi:10.1016/S0065-2601(06)38002-1
Sample. 94 independent tests, N > 8,000.
Findings. If-then plans improve goal attainment with d = .65 (95% CI 0.6–0.7). A similar effect (d = .77) was found for preventing derailment of goal striving.
Licenses. Recommending implementation intentions as the best-supported behaviour-change tool.
Does not license. Assuming the effect is uniform — plan format and motivation moderate it.
Retrieved when: C5 <= 40 OR C <= 40
Galla, B. M., & Duckworth, A. L. (2015). More than resisting temptation: beneficial habits mediate the relationship between self-control and positive life outcomes. JPSP, 109(3), 508–525.
galla-duckworth-2015 · doi:10.1037/pspp0000026
Sample. 6 studies, total N = 2,274.
Findings. Habits and automaticity — not effortful inhibition — mediate the self-control to outcome link.
Licenses. Advising low-Self-discipline profiles toward habit and structure design rather than willpower exhortation.
Does not license. Any ego-depletion mechanism. Strong causal claims — the mediation is cross-sectional.
Retrieved when: C5 <= 40 OR C2 <= 40
Steel, P. (2007). The nature of procrastination: a meta-analytic and theoretical review of quintessential self-regulatory failure. Psychological Bulletin, 133(1), 65–94.
steel-2007 · doi:10.1037/0033-2909.133.1.65
Sample. Meta-analysis of 691 correlations.
Findings. Strongest predictors of procrastination: task aversiveness, task delay, self-efficacy, impulsiveness and Conscientiousness facets. Neuroticism showed only a weak connection.
Licenses. Interpreting procrastination as a Conscientiousness and impulsiveness phenomenon, not primarily a Neuroticism one.
Does not license. Framing procrastination as anxiety-driven for all readers.
Retrieved when: C5 <= 35 OR N5 >= 70
Hittner, J. B., & Swickert, R. (2006). Sensation seeking and alcohol use: a meta-analytic review. Addictive Behaviors, 31(8), 1383–1401.
hittner-swickert-2006 · doi:10.1016/j.addbeh.2005.11.004
Sample. 61 studies, approximately 93% cross-sectional.
Findings. Sensation seeking and alcohol use: mean weighted r = .263; Disinhibition subcomponent strongest at r = .368.
Licenses. Flagging high Excitement-seeking as a modest risk marker for substance use.
Does not license. Deterministic risk claims — cross-sectional, small-to-moderate effects. A gambling link — MacLaren 2011 found essentially no effect (d ≈ .04).
Retrieved when: E5 >= 80 OR E5 >= 75 AND C6 <= 30
Moshagen, M., Hilbig, B. E., & Zettler, I. (2018). The dark core of personality. Psychological Review, 125(5), 656–688.
moshagen-2018 · doi:10.1037/rev0000111
Findings. Defines the Dark Factor (D); D relates to aggression (.65–.67) and crime/delinquency (.32–.37).
Licenses. Interpreting very low Agreeableness (low Morality + low Altruism + low Sympathy) as overlapping the aversive-personality space.
Does not license. Labelling a low-Agreeableness reader "dark" or "psychopathic". Low A is common and non-pathological. This boundary is safety-relevant.
Retrieved when: A <= 15 OR A2 <= 15 AND A6 <= 15
Ashton, M. C., Lee, K., & de Vries, R. E. (2014). The HEXACO Honesty-Humility, Agreeableness, and Emotionality factors. PSPR, 18(2), 139–152.
ashton-2014 · doi:10.1177/1088868314523838
Findings. FFM Agreeableness under-represents the Honesty-Humility variance, especially the Modesty and Straightforwardness content.
Licenses. Explaining why IPIP-NEO Modesty behaves idiosyncratically; interpreting low Modesty and low Morality cautiously.
Does not license. Treating IPIP-NEO Modesty as a valid Honesty-Humility proxy.
Retrieved when: A5 <= 25 OR A5 >= 75 OR A2 <= 30
Ferrari, M., et al. (2019). Self-compassion interventions and psychosocial outcomes: a meta-analysis of RCTs. Mindfulness, 10, 1455–1473.
ferrari-2019 · doi:10.1007/s12671-019-01134-6
Sample. 27 RCTs.
Findings. Improvements in rumination (g = 1.37), stress (g = 0.67), depression (g = 0.66), self-criticism (g = 0.56), anxiety (g = 0.57).
Licenses. Recommending self-compassion practice for high Self-consciousness or elevated Depression-facet profiles.
Does not license. Clinical claims. Many trials are short-term and some are waitlist-controlled, which inflates effects relative to active controls.
Retrieved when: N4 >= 75 OR N3 >= 70 OR N >= 75
Stieger, M., et al. (2021). Changing personality traits with the help of a digital personality change intervention. PNAS, 118(8), e2017548118.
stieger-2021 · doi:10.1073/pnas.2017548118
Sample. N = 1,523 adults, RCT with waitlist control, 3-month intervention and 3-month follow-up.
Findings. Self-reported change in the goal direction: d = 0.52 (increase goals), d = −0.58 (decrease goals). Observer-reported: d = 0.35 for increase (significant); d = −0.22 for decrease (not significant).
Licenses. The strongest evidence that non-clinical trait change via structured digital intervention is achievable and persists about three months.
Does not license. Durability claims beyond three months. Claiming observer-confirmed decrease — that result was not significant.
Retrieved when: C <= 40 OR N >= 70
Chida, Y., & Steptoe, A. (2009). The association of anger and hostility with future coronary heart disease. JACC, 53(11), 936–946.
chida-steptoe-2009 · doi:10.1016/j.jacc.2008.11.044
Sample. 25 studies in initially healthy populations; 19 in existing-CHD populations.
Findings. Anger/hostility to incident CHD in healthy populations: combined HR = 1.19 (95% CI 1.05–1.35). Associations were NOT significant in the subset of studies controlling for behavioural covariates.
Licenses. Noting trait anger and hostility as a modest cardiovascular risk correlate.
Does not license. Telling a reader their anger will give them heart disease. HR 1.19 is modest and likely partly behaviourally mediated.
Retrieved when: N2 >= 75 OR N2 >= 75 AND A4 <= 30
Knowles, K. A., & Olatunji, B. O. (2020). Specificity of trait anxiety in anxiety and depression: Meta-analysis of the State-Trait Anxiety Inventory. Clinical Psychology Review, 82, 101928.
knowles-olatunji-2020 · doi:10.1016/j.cpr.2020.101928
Sample. Meta-analysis of 388 published studies (N = 31,021) comparing STAI-T scores across anxiety-disorder, depressive-disorder and nonclinical groups.
Findings. A total of 388 published studies (N = 31,021) were included. Anxiety and depressive symptom severity were similarly strongly correlated with the STAI-T (mean r = .59 to .61). People with a depressive disorder scored higher on the STAI-T than those with an anxiety disorder (Hedges g = 0.27).
Licenses. That a high self-report trait-anxiety score reflects broad negative affectivity rather than any specific condition. A dimensional, non-pathologising reading of a high Anxiety facet.
Does not license. Any clinical claim about an individual — this is a measurement critique, not a study of the trait itself. The claim that trait anxiety is adaptive; it shows non-specificity, not benefit. Generalising beyond the STAI. The evidence is cross-sectional and self-report throughout.
Caveat. The copy held here is the journal pre-proof. Its two group-comparison effect sizes against nonclinical samples sit in tables that were not verified against the final text; do not quote those.
Retrieved when: N1 >= 75
Barlow, D. H., Sauer-Zavala, S., Carl, J. R., Bullis, J. R., & Ellard, K. K. (2014). The nature, diagnosis, and treatment of neuroticism: Back to the future. Clinical Psychological Science, 2(3), 344-365.
barlow-2014 · doi:10.1177/2167702613505532
Sample. Theoretical and conceptual review; no empirical sample.
Findings. Defines neuroticism as the tendency to experience frequent and intense negative emotions in response to stress, together with a perception that the world is dangerous and a belief in an inability to cope. Places clinical presentations at the pathological extreme of a continuous trait rather than in a separate category. Argues neuroticism may be more malleable than previously thought and possibly amenable to direct intervention.
Licenses. The continuum framing: a high score is a position on a dimension, not a disorder. Saying the trait is not fixed, which the gentle template needs and no other paper here states conceptually.
Does not license. Any effect size, sample or causal claim — it is a conceptual review and reports none. Heritability figures, which belong to a companion paper not held here. Its clinical vocabulary. Only the continuum and malleability arguments may be surfaced; the empirical evidence for change is roberts-2017-intervention.
Retrieved when: N1 >= 75 OR N6 >= 75
Ames, D. R., & Flynn, F. J. (2007). What breaks a leader: The curvilinear relation between assertiveness and leadership. Journal of Personality and Social Psychology, 92(2), 307-324.
ames-flynn-2007 · doi:10.1037/0022-3514.92.2.307
Sample. Four studies using self-report, peer and coworker ratings, and dyadic face-to-face negotiation behaviour; working adults and MBA students.
Findings. Both markedly low and markedly high assertiveness are appraised as less effective; effectiveness peaks in a middle range. The pattern reflects a trade-off: low assertiveness limits instrumental outcomes, high assertiveness costs social ones.
Licenses. Reading Assertiveness as a trade-off with a functional middle rather than a more-is-better scale. A non-deficit framing of both ends of the facet.
Does not license. Treating the curvilinear effect as large; quadratic effects of this kind are small. Generalising past leadership perception to life outcomes at large. Equating interpersonal assertiveness exactly with the Assertiveness facet — it is adjacent, not identical.
Retrieved when: E3 <= 40 OR E3 >= 75
Dong, Y., Zhao, M., Li, Y., Lin, J., Fang, Y., & Yang, Y. (2024). The relationship between trust and well-being: A meta-analysis. Journal of Happiness Studies, 25, 56.
dong-2024 · doi:10.1007/s10902-024-00737-8
Sample. Meta-analysis of 132 primary studies, total N = 1,060,174, ages 6 to 84.
Findings. A moderate correlation between trust and well-being, rho = 0.255, 95% CI [.240, .269]. Well-being type moderated the association, strongest for social well-being and weakest for physical.
Licenses. That higher dispositional trust is associated with modestly higher well-being.
Does not license. Any causal reading; the evidence is correlational. Treating this as facet-specific. Trust here spans interpersonal, institutional and generalised forms, and the Trust facet is narrower. Any claim about being trusted, or about the risks of very high trust; neither is studied.
Retrieved when: A1 >= 75
Hui, B. P. H., Ng, J. C. K., Berzaghi, E., Cunningham-Amos, L. A., & Kogan, A. (2020). Rewards of kindness? A meta-analysis of the link between prosociality and well-being. Psychological Bulletin, 146(12), 1084-1116.
hui-2020 · doi:10.1037/bul0000298
Sample. Meta-analysis of 201 independent studies, N = 198,213.
Findings. A modest but significant overall association between prosocial behaviour and well-being across 201 studies. The association is heterogeneous: some prosociality-wellbeing correlations are weak, null or negative.
Licenses. That finding helping intrinsically rewarding is modestly associated with greater well-being.
Does not license. The claim that helping reliably raises happiness; the paper documents weak and null cases explicitly. Any causal reading, and any facet-specific claim — this is prosociality broadly, not the Altruism facet.
Retrieved when: A3 >= 75
Sibley, C. G., & Duckitt, J. (2008). Personality and prejudice: A meta-analysis and theoretical review. Personality and Social Psychology Review, 12(3), 248-279.
sibley-duckitt-2008 · doi:10.1177/1088868308319226
Sample. Meta-analysis of 71 studies, N = 22,068.
Findings. Right-wing authoritarianism was predicted by low Openness and by Conscientiousness; social dominance orientation by low Agreeableness and weakly by low Openness. The association between Openness and ideology is reliable across samples.
Licenses. Reading Liberalism, and low Openness generally, as an orientation toward order, security and convention rather than as a deficit.
Does not license. Any evaluative claim about the reader politics. Only the values-and-authority framing may be surfaced, never the prejudice outcomes. A causal reading; later work shows the Openness-ideology link is context-moderated.
Caveat. The only source here that speaks to the low end of Openness, which is otherwise the least covered region of the instrument.
Retrieved when: O6 <= 40 OR O6 >= 75
Judge, T. A., & Bono, J. E. (2001). Relationship of core self-evaluations traits with job satisfaction and job performance: A meta-analysis. Journal of Applied Psychology, 86(1), 80-92.
judge-bono-2001 · doi:10.1037/0021-9010.86.1.80
Sample. Meta-analysis based on 274 correlations.
Findings. True-score correlations with job satisfaction: self-esteem .26, generalized self-efficacy .45, internal locus of control .32, emotional stability .24. True-score correlations with job performance: self-esteem .26, generalized self-efficacy .23, internal locus of control .22, emotional stability .19.
Licenses. Reading a low Self-efficacy facet as associated with lower job satisfaction and performance, dimensionally.
Does not license. Equating generalized self-efficacy with the Self-efficacy facet; it is a core-self-evaluations construct and the bridge costs confidence. Any outcome beyond work, and any causal claim.
Retrieved when: C1 <= 40
de Ridder, D. T. D., Lensvelt-Mulders, G., Finkenauer, C., Stok, F. M., & Baumeister, R. F. (2012). Taking stock of self-control: A meta-analysis of how trait self-control relates to a wide range of behaviors. Personality and Social Psychology Review, 16(1), 76-99.
deridder-2012 · doi:10.1177/1088868311418749
Sample. Meta-analysis of 102 studies, total N = 32,648, using the Self-Control Scale, the Barratt Impulsiveness Scale and the Low Self-Control Scale.
Findings. A small to medium positive effect of trait self-control on behaviour across all three scales. Associations were significantly stronger for automatic than for controlled behaviour, indicating that trait self-control works by avoiding temptation more than by resisting it.
Licenses. Reading a low Immoderation facet as an asset rather than merely the absence of a problem. The constructive point that self-control operates through habit and avoidance rather than effortful resistance.
Does not license. Anything resting on ego depletion or willpower-as-a-resource, which failed replication. This paper is here because its finding contradicts that account, not because it supports it. Equating trait self-control with the Immoderation facet inverted; it is a distinct scale.
Caveat. Overlaps galla-duckworth-2015, which makes a similar habits-over-willpower argument for low Self-discipline. Kept for the different cell — low Immoderation as an asset — not as general self-control evidence.
Retrieved when: N5 <= 40
Liu, Q., & Nesbit, J. C. (2024). The relation between need for cognition and academic achievement: A meta-analysis. Review of Educational Research, 94(2), 155-192.
liu-nesbit-2024 · doi:10.3102/00346543231160474
Sample. Meta-analysis of 136 independent effect sizes, N = 53,258.
Findings. Need for cognition is positively associated with academic achievement across 136 effect sizes. The authors argue need for cognition is malleable and can be cultivated in intellectually stimulating environments.
Licenses. A non-pejorative reading of a low Intellect facet: enjoyment of effortful thinking is one dimension of engagement, associated with achievement rather than required for it, and trainable.
Does not license. Equating a low Intellect score with low ability. Need for cognition is a preference, not a capacity, and the outcome measured is achievement rather than intelligence. Extending beyond academic outcomes.
Retrieved when: O5 <= 40
Malouff, J. M., Thorsteinsson, E. B., Schutte, N. S., Bhullar, N., & Rooke, S. E. (2010). The Five-Factor Model of personality and relationship satisfaction of intimate partners: A meta-analysis. Journal of Research in Personality, 44(1), 124-127.
malouff-2010 · doi:10.1016/j.jrp.2009.09.004
Sample. Meta-analysis of 19 samples, total N = 3,848, heterosexual intimate partners.
Findings. Partner-effect correlations with a partner relationship satisfaction: neuroticism -.22, agreeableness .15, conscientiousness .12, extraversion .06. Openness showed no significant partner effect.
Licenses. That a reader own personality relates modestly to their partner satisfaction, in the direction of lower Neuroticism and higher Agreeableness and Conscientiousness.
Does not license. Any claim about the reader own satisfaction. These are partner effects, which is a different quantity. Treating the associations as more than small; none exceeds .22 in absolute value. Generalising past heterosexual samples, or reading the association as causal. Facet-level claims about Friendliness. The E1 trigger reaches this paper because warmth is the interpersonal core of Extraversion, but what was measured is the domain, and the extraversion partner effect is the smallest of the four at .06.
Caveat. The only paper here on romantic relationships, and short enough to load on almost every profile it matches.
Retrieved when: N <= 40 OR A >= 60 OR C >= 60 OR E >= 60 OR E1 >= 70
Bogg, T., & Roberts, B. W. (2004). Conscientiousness and health-related behaviors: A meta-analysis of the leading behavioral contributors to mortality. Psychological Bulletin, 130(6), 887-919.
bogg-roberts-2004 · doi:10.1037/0033-2909.130.6.887
Sample. Meta-analysis synthesising 194 studies.
Findings. Conscientiousness-related traits were negatively related to all risky health behaviours and positively related to all beneficial ones. The largest single effect was drug use at r = -.28 and the smallest physical activity at r = .05; tobacco, alcohol, unhealthy eating, risky driving, risky sex, suicide and violence fell between -.12 and -.25.
Licenses. Associating a low Conscientiousness or low Self-discipline score with a modestly raised rate of health-risk behaviour, dimensionally.
Does not license. Presenting the association as deterministic for an individual; the effects run from .05 to .28. A causal claim, or generalisation past the largely US evidence base.
Retrieved when: C <= 40 OR C5 <= 40
Wilmot, M. P., & Ones, D. S. (2022). Agreeableness and its consequences: A quantitative review of meta-analytic findings. Personality and Social Psychology Review, 26(3), 242-280.
wilmot-ones-2022 · doi:10.1177/10888683211073007
Sample. Second-order meta-analysis of 142 meta-analyses reporting effects for 275 variables, representing N > 1.9 million participants from k > 3,900 studies.
Findings. Agreeableness has effects in a desirable direction for 93% of the 275 variables reviewed, grand mean rho = .16. Lower-order evidence is summarised for 42 variables drawn from 20 meta-analyses.
Licenses. The high end of Agreeableness and its facets, which had no dedicated source before this.
Does not license. Treating .16 as a large effect. It is a small grand mean across 275 variables and hides wide variation between them. Any causal claim; the underlying evidence is predominantly correlational and self-report. Reading "desirable direction" as a value judgement about the reader. Desirability here is the direction of the correlation, not a verdict.
Retrieved when: A >= 70 OR A2 >= 70 OR A3 >= 70 OR A4 >= 70 OR A6 >= 70
Judge, T. A., Livingston, B. A., & Hurst, C. (2012). Do nice guys - and gals - really finish last? The joint effects of sex and agreeableness on income. Journal of Personality and Social Psychology, 102(2), 390-407.
judge-livingston-hurst-2012 · doi:10.1037/a0026021
Sample. Four studies of working adults, including longitudinal panel data spanning roughly two decades.
Findings. Less agreeable individuals earned more than more agreeable ones, and the difference was substantially larger for men than for women. Agreeableness was positively associated with life satisfaction, social networks and community involvement.
Licenses. Naming the honest cost of high Agreeableness in earnings terms, so the high end is not written as uniformly advantageous.
Does not license. Presenting the earnings penalty as universal. It is strongly sex-moderated and much weaker for women. A causal reading, or advice to become less agreeable — the paper notes the trade-off runs the other way on satisfaction and relationships.
Retrieved when: A >= 70
Chiaburu, D. S., Oh, I.-S., Berry, C. M., Li, N., & Gardner, R. G. (2011). The five-factor model of personality traits and organizational citizenship behaviors: A meta-analysis. Journal of Applied Psychology, 96(6), 1140-1166.
chiaburu-2011 · doi:10.1037/a0024004
Sample. 87 statistically independent samples.
Findings. Emotional stability, extraversion and openness showed incremental validity for citizenship behaviour over and above conscientiousness and agreeableness.
Licenses. Discussing Cooperation and the helping side of Agreeableness at work, with the correction that Agreeableness is not the dominant driver of citizenship behaviour.
Does not license. Generalising past workplace criteria to prosociality in ordinary life. Treating the coefficients as observed; they are corrected.
Retrieved when: A >= 70 OR A4 >= 70
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.
connelly-ones-2010 · doi:10.1037/a0021212
Sample. 44,178 targets across 263 independent samples.
Findings. Observer accuracy varies by trait: extraversion and conscientiousness are rated most accurately by others, neuroticism and agreeableness least, especially by unfamiliar observers. Well-acquainted informants add substantial predictive validity over self-report.
Licenses. Quantifying what vazire-2010 describes: which traits other people can see, and which they cannot.
Does not license. Any claim about the validity of self-report itself. This is about observer ratings, which this instrument does not collect.
Retrieved when: E >= 75 OR E <= 25 OR A >= 75 OR N >= 70
Ward, M. K., & Meade, A. W. (2023). Dealing with careless responding in survey data: Prevention, identification, and recommended best practices. Annual Review of Psychology, 74, 577-596.
ward-meade-2023 · doi:10.1146/annurev-psych-040422-045007
Sample. Review of methods for preventing, detecting and handling careless survey responding.
Findings. Careless responding is common enough to bias results and is detectable by several complementary methods, none of which is sufficient alone.
Licenses. Explaining to a reader whose validity checks fired what careless responding is and why it matters for their scores.
Does not license. Any accusation. A validity flag is a property of a response pattern, not a judgement of the person. Treating any single detection method as conclusive.
Retrieved when: N >= 95 OR N <= 5
Jones, A., Earnest, J., Adam, M., Clarke, R., Yates, J., & Pennington, C. R. (2022). Careless responding in crowdsourced alcohol research: A systematic review and meta-analysis of practices and prevalence. Experimental and Clinical Psychopharmacology, 30(4), 381-399.
jones-2022 · doi:10.1037/pha0000546
Sample. Preregistered systematic review and meta-analysis of 96 crowdsourced alcohol-related studies, about 126,130 participants; 51 of them screened for careless responding.
Findings. Only 53.2 per cent of crowdsourced studies used any measure of careless responding (95 per cent CI 42.7 to 63.3). Among studies that screened, the pooled prevalence of careless responding was about 11.7 per cent (95 per cent CI 7.6 to 16.5). MTurk studies identified more careless responders than other platforms, and prevalence rose with the number of careless-response items used. The most common screening method was an attention-check question.
Licenses. Telling a reader whose validity checks fired roughly how common careless responding is in online samples, so a flag reads as ordinary rather than as an accusation. The point that detection rates depend on how hard you look, which is why a flag is evidence about a response pattern and not a verdict.
Does not license. Applying the 11.7 per cent figure to this instrument. It comes from crowdsourced alcohol studies with paid workers; this engine is answered by invited readers with no incentive to rush, and the rate here is unknown and probably lower. Any accusation, for the same reason as ward-meade-2023. Treating attention checks as validated against this instrument, which uses different validity indices entirely.
Retrieved when: N >= 95 OR N <= 5
Terracciano, A., Schrack, J. A., Sutin, A. R., Chan, W., Simonsick, E. M., & Ferrucci, L. (2013). Personality, metabolic rate and aerobic capacity. PLOS ONE, 8(1), e54746.
terracciano-2013 · doi:10.1371/journal.pone.0054746
Sample. Community-dwelling adults with measured metabolic rate and aerobic capacity, assessed on the NEO facets.
Findings. Among all thirty facets, Activity was the strongest personality correlate of aerobic capacity and related measures, with correlations of about .20, .18, -.24 and .21 across the outcomes examined.
Licenses. Facet-level interpretation of Activity Level against objectively measured fitness, rather than self-reported exercise.
Does not license. A causal claim in either direction; fitness and trait activity are measured concurrently. Generalising to exercise behaviour as such — the outcomes here are metabolic and aerobic measures.
Caveat. The only source here with genuine facet-level numbers for Activity Level. wilson-dishman-2015 is domain-level and reports personality against self-reported physical activity instead.
Retrieved when: E4 >= 70 OR E4 <= 30
Wilson, K. E., & Dishman, R. K. (2015). Personality and physical activity: A systematic review and meta-analysis. Personality and Individual Differences, 72, 230-242.
wilson-dishman-2015 · doi:10.1016/j.paid.2014.08.023
Sample. Systematic review and meta-analysis of 64 studies relating the Big Five to physical activity.
Findings. Extraversion and conscientiousness were the traits most consistently associated with physical activity across the 64 studies. Conscientiousness associations from prospective studies ran nearly twice as large as those from cross-sectional ones.
Licenses. Domain-level statements about Extraversion, Conscientiousness and physical activity.
Does not license. Facet-level claims about Activity Level. This is domain-level and about physical-activity behaviour, not the pace-and-vigour facet — terracciano-2013 is the facet source. Quoting specific correlation values from this entry. The frequently-cited domain coefficients were not verified against the primary text and are deliberately absent here.
Retrieved when: E >= 70 OR E4 >= 70 OR E4 <= 30
Smillie, L. D., DeYoung, C. G., & Hall, P. J. (2015). Clarifying the relation between extraversion and positive affect. Journal of Personality, 83(5), 564-574.
smillie-deyoung-hall-2015 · doi:10.1111/jopy.12138
Sample. Study 1 N = 437 Australian students; Study 2 N = 262 US online participants.
Findings. Extraversion correlated .65 with activated pleasant affect and .31 with simple pleasant affect. The Enthusiasm aspect correlated .53 with pleasant valence.
Licenses. Reading Cheerfulness as the activated-positive-affect end of Extraversion rather than as happiness in general.
Does not license. Any causal claim that extraversion produces happiness. Reading the low end as sadness. Low positive activation is not negative affect, and the paper is explicit about that. Generalising from young, female-skewed samples totalling about seven hundred people.
Retrieved when: E6 >= 70 OR E6 <= 30
Salminen, J. K., Saarijarvi, S., Aarela, E., Toikka, T., & Kauhanen, J. (1999). Prevalence of alexithymia and its association with sociodemographic variables in the general population of Finland. Journal of Psychosomatic Research, 46(1), 75-82.
salminen-1999 · doi:10.1016/S0022-3999(98)00053-1
Sample. N = 1,285, general Finnish population, TAS-20.
Findings. Alexithymia was normally distributed in the population in both genders, confirming that it is a personality dimension. Prevalence was 13% overall, 17% in men and 10% in women.
Licenses. A dimensional, non-pathologising reading of low Emotionality: difficulty naming feelings is a continuum, not a condition.
Does not license. Any suggestion that a low Emotionality score indicates alexithymia or any disorder. The paper own finding is that the construct is dimensional, and the report must not turn it into a category. Generalising past a single-country sample, or treating the prevalence figure as anything but cut-off dependent. Saying anything about the high end; the paper does not characterise emotionally attuned respondents.
Retrieved when: O3 <= 30
Cross, C. P., Cyrenne, D.-L. M., & Brown, G. R. (2013). Sex differences in sensation-seeking: A meta-analysis. Scientific Reports, 3, 2486.
cross-cyrenne-brown-2013 · doi:10.1038/srep02486
Sample. Meta-analysis of studies using Zuckerman Sensation Seeking Scale.
Findings. Sex differences in total sensation-seeking scores were stable over time, while the difference on thrill and adventure seeking declined across decades.
Licenses. Framing Adventurousness as a normally varying preference for novelty and intensity rather than as recklessness.
Does not license. Criterion validity for Adventurousness. This is a study of sex differences over time, not of what the facet predicts. Equating sensation seeking with the Adventurousness facet; it is a proxy.
Retrieved when: O4 >= 70 OR O4 <= 30
Ng, V., Tay, L., & Kuykendall, L. (2021). The nomological network of openness facets. Frontiers in Psychology, 12, 707652.
ng-2021 · doi:10.3389/fpsyg.2021.707652
Sample. Review of the facet structure of Openness and its correlates.
Findings. The facets of Openness relate differently to their correlates, so domain-level Openness conceals meaningful facet-level variation.
Licenses. Reading the Openness facets separately rather than as one dimension, at either end.
Does not license. Specific effect sizes from this entry; none are quoted here because they were not verified against the primary text. Any claim that low Openness is advantageous, which this does not establish.
Retrieved when: O >= 70 OR O <= 30
Dudley, N. M., Orvis, K. A., Lebiecki, J. E., & Cortina, J. M. (2006). A meta-analytic investigation of conscientiousness in the prediction of job performance: Examining the intercorrelations and the incremental validity of narrow traits. Journal of Applied Psychology, 91(1), 40-57.
dudley-2006 · doi:10.1037/0021-9010.91.1.40
Sample. Meta-analysis separating global Conscientiousness into achievement, dependability, order and cautiousness.
Findings. Narrow conscientiousness traits incrementally predict performance beyond the global domain, to a degree that depends on the criterion and the occupation.
Licenses. Reading Dutifulness as the dependability component of Conscientiousness, which is the closest thing here to a facet-specific source.
Does not license. Quoting dependability-specific validity coefficients from this entry; they were not verified against the primary tables and are deliberately absent. Generalising past workplace criteria to promise-keeping in ordinary life.
Retrieved when: C3 >= 70 OR C3 <= 30
Ones, D. S., Viswesvaran, C., & Schmidt, F. L. (1993). Comprehensive meta-analysis of integrity test validities: Findings and implications for personnel selection and theories of job performance. Journal of Applied Psychology, 78(4), 679-703.
ones-1993 · doi:10.1037/0021-9010.78.4.679
Sample. 665 validity coefficients across 576,460 data points.
Findings. The estimated mean operational predictive validity of integrity tests for supervisory ratings of job performance is .41.
Licenses. Supporting a Dutifulness reading with a large, well-established integrity-test literature.
Does not license. Equating integrity tests with the Dutifulness facet. They are a composite of Conscientiousness, Agreeableness and Emotional Stability, which is a substantial bridging inference. Any use of .41 outside supervisory performance ratings; it is a corrected operational validity, not an observed correlation.
Retrieved when: C3 >= 70 OR C3 <= 30
Poropat, A. E. (2009). A meta-analysis of the five-factor model of personality and academic performance. Psychological Bulletin, 135(2), 322-338.
poropat-2009 · doi:10.1037/a0014996
Sample. Meta-analysis with a cumulative N above 70,000.
Findings. Academic performance correlated significantly with Agreeableness, Conscientiousness and Openness. The Conscientiousness association was largely independent of intelligence.
Licenses. Discussing Conscientiousness and academic attainment, including the point that it adds to rather than proxies for ability.
Does not license. Generalising to work performance, which sackett-2023 covers and revises downward. Ignoring the sample skew — a large share of the tertiary samples were psychology students.
Retrieved when: C >= 70 OR C4 >= 70
Alderotti, G., Rapallini, C., & Traverso, S. (2023). The Big Five personality traits and earnings: A meta-analysis. GLO Discussion Paper 902 (working paper version of the Journal of Economic Psychology article).
alderotti-2023 · doi:10.1016/j.joep.2022.102570
Sample. Working-paper version: 65 peer-reviewed articles published 2001-2020, from which 936 partial effect sizes were retrieved.
Findings. Positive associations between earnings and Openness, Conscientiousness and Extraversion; negative associations for Agreeableness and Neuroticism. No evidence of substantial publication bias.
Licenses. Discussing personality and earnings, and pairing with judge-livingston-hurst-2012 on the Agreeableness penalty.
Does not license. Any causal claim about a reader own earnings. Citing the published article counts. The copy held here is the working paper and reports 65 articles and 936 effect sizes, where the published version reports 62 and 896.
Retrieved when: C >= 70 OR A >= 70 OR N >= 70
Roberts, B. W., & DelVecchio, W. F. (2000). The rank-order consistency of personality traits from childhood to old age: A quantitative review of longitudinal studies. Psychological Bulletin, 126(1), 3-25.
roberts-delvecchio-2000 · doi:10.1037/0033-2909.126.1.3
Sample. 152 longitudinal studies yielding 3,217 correlation coefficients.
Findings. Rank-order consistency rose from .31 in childhood to .54 at college age, .64 by age 30, and plateaued around .74 between ages 50 and 70, at a fixed interval.
Licenses. Saying how stable a reader position relative to others is likely to be, and that it is higher later in life.
Does not license. Reading rank-order consistency as absence of change. It describes order, not level; roberts-2006-change covers mean-level shift. Treating .74 as a ceiling on personal change.
Retrieved when: N >= 0
Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection. Journal of Applied Psychology, 107(11), 2040-2068.
sackett-2022 · doi:10.1037/apl0000994
Findings. Conscientiousness operational validity revised down to r = .19 (from .31). GMA revised from .51 to .31; structured interviews from .51 to .42.
Licenses. The range-restriction argument itself, in technical detail — why the older estimates were inflated rather than merely that they were.
Does not license. Dismissing personality — effects remain meaningful and incremental. Transferring workplace validities to non-work life outcomes.
Caveat. Demoted from core when sackett-2023 took its place there: the companion carries the same revised matrix at a third of the length, and this is the primary source kept for the technical argument. It stays in the corpus because the stored demo report cites it, and a citation that no longer resolves is worse than the tokens.
Retrieved when: C >= 70 OR C <= 30
Le, H., Oh, I.-S., Robbins, S. B., Ilies, R., Holland, E., & Westrick, P. (2011). Too much of a good thing: Curvilinear relationships between personality traits and job performance. Journal of Applied Psychology, 96(1), 113-133.
le-2011 · doi:10.1037/a0021016
Sample. Two independent employee samples. Study 1 a concurrent validation in a large Midwestern public organisation spanning low- to high-complexity jobs; Study 2 the ACT Talent Assessment. Outcomes were task performance, organisational citizenship behaviour and counterproductive work behaviour.
Findings. The quadratic term for Conscientiousness predicting task performance was significant and negative (beta = -.12), an inverted U: the benefit flattens and then disappears rather than continuing to rise. Quadratic terms were also significant for Conscientiousness predicting citizenship (.10) and counterproductive behaviour (.14). Emotional Stability showed significant quadratic terms in all three models, negative for task performance (-.11) and citizenship (-.11). Job complexity moved the inflection point. For Emotional Stability and citizenship it stood at -0.10 SD in low-complexity jobs and 1.75 SD in high-complexity ones, so the level at which more stops helping depends on the work. Conscientiousness and Emotional Stability correlated .62 in Study 1.
Licenses. That deliberation and calm are beneficial up to a point rather than without limit, and that the point where the benefit stops depends on the demands of the situation. A non-deficit reading of a middling score on either: the curve is flat near the top, so the difference between high and very high is often nothing.
Does not license. Any facet-level claim. This measures the Conscientiousness and Emotional Stability domains. C6 and N1 are single facets inside them, and the paper never separates either one. That a high score is harmful. Two of the four inverted-U effects flatten rather than turn down; the paper reports the benefit disappearing, not reversing. Generalising past job performance. There is nothing here about health, relationships or wellbeing. Reading the inflection points as thresholds for an individual. They are standardised positions in a regression on samples of employees, not a level a reader can be said to have crossed.
Caveat. Employee samples with the selection effects that implies: people far below the hiring bar are not in them, which is exactly the range where the curve would be steepest.
Retrieved when: C6 >= 70 OR N1 <= 30
Clarke, S., & Robertson, I. T. (2005). A meta-analytic review of the Big Five personality factors and accident involvement in occupational and non-occupational settings. Journal of Occupational and Organizational Psychology, 78(3), 355-376.
clarke-robertson-2005 · doi:10.1348/096317905X26183
Sample. Meta-analysis of Big Five traits and accident involvement across occupational and non-occupational (mainly traffic) settings.
Findings. Low Conscientiousness and low Agreeableness were valid and generalizable predictors of accident involvement, corrected mean validities .27 and .26. Low Conscientiousness: k = 18, N = 4,550, corrected r = .273, 47 per cent of variance attributable to artefacts. Extraversion predicted traffic accidents but not occupational ones; overall corrected r = .164 across 30 samples, with a credibility interval spanning zero. Context moderated the pattern, so different traits mattered for work and for driving.
Licenses. That careful, deliberate conduct is associated with fewer accidents, and that the association is one of the more generalizable findings in the applied literature.
Does not license. A facet-level claim about Cautiousness. This is the Conscientiousness domain. An asset claim framed positively. The paper measured the low pole predicting a bad outcome; reading it as "high Cautiousness prevents accidents" is an inversion, and the corpus holds it only as corroboration for the deliberation reading, never as the primary source. Any personal risk statement. A corrected validity of .27 across occupational samples says nothing about whether a given reader will have an accident.
Retrieved when: C6 >= 70 OR C6 <= 30
Helgeson, V. S., & Fritz, H. L. (1998). A theory of unmitigated communion. Personality and Social Psychology Review, 2(3), 173-183.
helgeson-fritz-1998 · doi:10.1207/s15327957pspr0203_2
Sample. Theoretical synthesis drawing on multiple cross-sectional and longitudinal studies.
Findings. Unmitigated communion, a focus on others to the exclusion of the self, is distinguished from communion, an ordinary caring orientation toward others. Unmitigated communion is related to psychological distress including depressive symptoms; communion is not. The distinction resolves a long-standing puzzle: traditional measures of female gender-related traits do not correlate with depressive symptoms, so they could not explain the sex difference in depression.
Licenses. That other-directedness has a self-neglecting form that is distinct from ordinary warmth, and that only the self-neglecting form tracks distress. The reading that a lower score on other-focus is not a deficit in caring, because the documented cost sits at the extreme where the self drops out entirely.
Does not license. That low Altruism is an asset. The paper studies the harms of the extreme high end. It licenses "not being self-erasing is protective", never "self-focus is good". Facet-level claims. Unmitigated communion is a separate scale, not IPIP A3 or A6, and the two have never been mapped onto each other. Any claim about a reader sex. The construct is gender-linked and the effects are sex-moderated, but this engine reports on an individual. Causal language. The theory paper synthesises correlational work.
Caveat. A theory paper. The effect sizes live in the empirical companion, fritz-helgeson-1998, which is in the corpus alongside it.
Retrieved when: A3 <= 30 OR A3 >= 80
Fritz, H. L., & Helgeson, V. S. (1998). Distinctions of unmitigated communion from communion: Self-neglect and overinvolvement with others. Journal of Personality and Social Psychology, 75(1), 121-140.
fritz-helgeson-1998 · doi:10.1037/0022-3514.75.1.121
Sample. Four studies. Two questionnaire studies distinguishing the constructs, and two laboratory studies in which participants were exposed to a stranger in distress.
Findings. Unmitigated communion and communion are correlated but distinct: only unmitigated communion involves a negative view of the self, reliance on others for self-evaluative information, and psychological distress. Self-neglect and overinvolvement with others both contribute to the link between unmitigated communion and distress. The laboratory studies show the distress arises through overinvolvement in another person problems, not merely through caring about them.
Licenses. That the cost of extreme other-orientation runs through self-neglect and overinvolvement specifically, which is a mechanism rather than a correlation. A non-pejorative reading of a lower other-focus score, on the same logic as the theory paper.
Does not license. That low Altruism or low Sympathy is beneficial. The evidence is about the high extreme. Facet-level mapping onto A3 or A6. Generalising the laboratory findings. They used a confederate in a controlled setting and measured immediate distress, not life outcomes.
Retrieved when: A3 <= 30 OR A3 >= 80
Kashdan, T. B., Barrett, L. F., & McKnight, P. E. (2015). Unpacking emotion differentiation: Transforming unpleasant experience by perceiving distinctions in negativity. Current Directions in Psychological Science, 24(1), 10-16.
kashdan-2015 · doi:10.1177/0963721414550708
Sample. Narrative review of clinical, social and health-psychology studies. Not a primary data set.
Findings. People who describe their feelings with more granularity are less likely to fall back on maladaptive regulation such as binge drinking, aggression and self-injury, show less neural reactivity to rejection, and have less severe anxiety and depressive disorders. Reviewing Pond et al. 2012, people better at differentiating negative feelings were 20 to 50 per cent less likely to retaliate aggressively against someone who had hurt them. Reviewing Barrett et al. 2001, high differentiators reported using nearly 30 per cent more regulation strategies across two weeks of daily diaries. Knowing that someone feels frequent, intense negative affect is not enough to predict whether they cope well; the granularity of that affect carries separate information.
Licenses. That being able to tell your feelings apart is a regulatory asset, and that it is distinct from simply feeling less. The framing that awareness of fine emotional distinctions does work, rather than merely accompanying distress.
Does not license. Any pooled effect size. This is a review; the numbers above belong to the primary studies it cites, and it reports no meta-analytic estimate of its own. Equating emotion differentiation with the O3 facet. Differentiation is measured by experience sampling across many reports; O3 is a self-report of how aware of feelings a person believes they are. Someone can score high on the second without the first. Clinical framing of a low score. The paper describes a skill that varies, not a deficit.
Caveat. Short, so it loads on almost every profile that matches. Paired with seah-coifman-2022, which supplies the meta-analytic effect size this review lacks.
Retrieved when: O3 >= 70
Seah, T. H. S., & Coifman, K. G. (2022). Emotion differentiation and behavioral dysregulation in clinical and nonclinical samples: A meta-analysis. Emotion, 22(7), 1686-1697.
seah-coifman-2022 · doi:10.1037/emo0000968
Sample. Meta-analysis of 17 studies of negative emotion differentiation and maladaptive behaviour, across clinical and non-clinical samples.
Findings. Negative emotion differentiation was negatively associated with maladaptive behaviours, r = -.15. The effect did not differ between clinical (k = 7, r = -.15) and non-clinical (k = 10, r = -.16) samples. The association held after controlling for mean negative affect (k = 11, r = -.09), and did not depend on the level of negative affect.
Licenses. That the ability to distinguish negative emotions is associated with less behavioural dysregulation, independent of how much negative affect a person feels. That this holds outside clinical samples, so it is not a finding about disorder.
Does not license. Treating the effect as large. It is r = -.15, falling to -.09 once negative affect is partialled out, which is the honest number for the unique contribution. Outcomes beyond behavioural dysregulation. Facet-level claims about O3, for the same reason as kashdan-2015: differentiation is an experience-sampling index, not a self-report of emotional awareness. Strong conclusions from 17 studies with acknowledged methodological heterogeneity.
Retrieved when: O3 >= 70
Rasmussen, K. R., Stackhouse, M., Boon, S. D., Comstock, K., & Ross, R. (2019). Meta-analytic connections between forgiveness and health: The moderating effects of forgiveness-related distinctions. Psychology & Health, 34(5), 515-534.
rasmussen-2019 · doi:10.1080/08870446.2018.1545906
Sample. 103 independent samples, 606 correlations, N = 26,043, drawn from 17 countries and including students, older adults, divorced mothers and combat veterans.
Findings. A reliable overall association between forgiveness and health, rho = .19 in published and .20 in unpublished samples, with overlapping intervals indicating no evidence of publication bias. Psychological health was much more strongly associated (k = 91, N = 23,446, rho = .22) than physical health (k = 57, N = 15,514, rho = .11). Among physical markers only some held up: somatization rho = .17, heart rate rho = .24, blood pressure rho = .26. Cortisol, substance use, bodily pain and infections were non-significant and near zero. Substantial heterogeneity across studies (Qw = 586.66), with a credibility interval from -.06 to .45.
Licenses. That a forgiving disposition is associated with better psychological health, and with cardiovascular indicators specifically. A positive framing of slowness to resent, as something associated with health rather than merely the absence of a problem.
Does not license. That forgiveness improves physical health generally. Most physical markers were near zero; only cardiovascular ones and somatization held. Equating forgiveness with low Anger. Forgiveness includes an active reconciliation component that rarely feeling angry does not, and the paper measures forgiveness scales, not N2. Causal language. These are correlations, and the authors say the causal direction is plausible rather than established. Presenting the association as consistent. The credibility interval crosses zero.
Retrieved when: N2 <= 30
Fehr, R., Gelfand, M. J., & Nag, M. (2010). The road to forgiveness: A meta-analytic synthesis of its situational and dispositional correlates. Psychological Bulletin, 136(5), 894-914.
fehr-2010 · doi:10.1037/a0019993
Sample. Meta-analysis of 175 studies, N = 26,006, covering 22 constructs as correlates of interpersonal forgiveness.
Findings. The largest correlates were state empathy (r = .51), perceived intent (r = -.49), apology (r = .42) and state anger (r = -.41). Trait forgiveness correlated with state forgiveness at r = .30 across 30 studies. Trait perspective taking correlated with forgiveness at r = .19. Situational constructs accounted for more variance in forgiveness than victim dispositions did. Mean and median correlations for situational cognitions were .37 and .35, about 12 to 14 per cent of variance.
Licenses. That dispositional forgiveness sits in a coherent, quantified network of situational and trait correlates. The point that whether someone forgives depends more on what happened than on who they are, which is a useful corrective in a report built entirely on traits.
Does not license. Overstating the dispositional side. This paper mainly shows that the situation dominates; the trait-to-state correlation is .30. Facet-level claims about Anger. Forgiveness is not the inverse of trait anger. Quoting the original Table 2 and Table 3 correlations without checking the 2011 erratum, which revised them.
Caveat. The largest paper in this batch. Held behind rasmussen-2019 for the same pole, so on a tight budget the smaller and more directly health-relevant source loads first.
Retrieved when: N2 <= 30
Hu, T., Zhang, D., & Wang, J. (2015). A meta-analysis of the trait resilience and mental health. Personality and Individual Differences, 76, 18-27.
hu-2015 · doi:10.1016/j.paid.2014.11.039
Sample. 60 studies, 111 effect sizes.
Findings. Trait resilience correlated -.361 with negative indicators of mental health (k = 76) and .503 with positive indicators (k = 35). Age moderated the association with negative indicators, stronger in adults than in children and adolescents (children and adolescents r = -.273); it did not moderate the positive indicators. Effect sizes were significantly stronger among people facing adversity than among those who were not. As the proportion of male participants rose, the effect weakened.
Licenses. That a stress-resistant disposition is associated with better mental health, and more strongly so for adults and under adversity. A positive framing of composure under pressure, as something with its own correlates rather than the absence of vulnerability.
Does not license. Any Big Five claim. This paper does not report correlations with Neuroticism, Extraversion or any other trait, and does not mention the five-factor model. A published proposal attributed a set of Big Five correlations to it that are not in the text. Facet-level claims about Vulnerability. Trait resilience is a separate self-report construct that overlaps heavily with low Neuroticism as a whole. Causal or protective language. These are cross-sectional correlations, and resilience scales and mental-health scales share method and content.
Retrieved when: N6 <= 30
Eschleman, K. J., Bowling, N. A., & Alarcon, G. M. (2010). A meta-analytic examination of hardiness. International Journal of Stress Management, 17(4), 277-307.
eschleman-2010 · doi:10.1037/a0020476
Sample. Meta-analysis of hardiness and its three subfacets (commitment, control, challenge) against stressors, strains, social support, coping and performance.
Findings. Hardiness was positively associated with self-esteem, optimism, extraversion, sense of coherence and self-efficacy, and negatively with neuroticism, negative affectivity, trait anxiety and trait anger. Hardiness correlated -.45 with negative affectivity (k = 6, N = 3,115). Hardiness was not significantly related to Conscientiousness (rho = -.06, k = 4, N = 1,428), and challenge was negatively associated with it. Hardiness was negatively related to a wide range of stressors and strains, most strongly to supervisor conflict and role ambiguity (rho = -.47, k = 2, N = 412).
Licenses. That commitment, control and a view of change as challenge form a stress-resistant pattern with its own correlates. The observation that this pattern is largely independent of Conscientiousness, so composure under stress is not a by-product of being organised.
Does not license. Facet-level claims about Vulnerability. Hardiness is a three-component construct with its own scales, further from N6 than trait resilience is. The incremental-validity-over-negative-affectivity claim as settled. It has been contested, and some included studies did not in fact control for negative affectivity. Treating the estimates as robust where they rest on very few samples. Several of the strongest are k = 2. Any causal reading; the authors note nearly all primary studies were self-report, so common-method variance is unaddressed.
Retrieved when: N6 <= 30
Cruwys, T., Dingle, G. A., Haslam, C., Haslam, S. A., Jetten, J., & Morton, T. A. (2013). Social group memberships protect against future depression, alleviate depression symptoms and prevent depression relapse. Social Science & Medicine, 98, 179-186.
cruwys-2013 · doi:10.1016/j.socscimed.2013.09.013
Sample. Two longitudinal community samples: proximal effects across 2 years (N = 5,055) and distal effects across 4 years (N = 4,087).
Findings. Depressed respondents with no group memberships who joined one group reduced their risk of relapse by 24 per cent; joining three groups reduced it by 63 per cent. Group memberships predicted lower depression 2 and 4 years later, so the association is not only concurrent.
Licenses. That participating in groups is associated with better mental health over years rather than days. A positive framing of enjoying company, as connected to something with longitudinal evidence behind it.
Does not license. Equating group membership with Gregariousness. The paper measures how many groups a person belongs to and how strongly they identify with them; E2 measures whether they find crowds and company stimulating. A reader can score high on E2 and belong to nothing. Any claim about depression risk for an individual reader. The relapse figures come from a sample already depressed at baseline and are not a personal forecast. Causal language. These are longitudinal observational data, so joining groups was not assigned. Clinical framing of any kind on the strength of this paper.
Retrieved when: E2 >= 70
Postmes, T., Wichmann, L. J., van Valkengoed, A. M., & van der Hoef, H. (2019). Social identification and depression: A meta-analysis. European Journal of Social Psychology, 49(1), 110-126.
postmes-2018 · doi:10.1002/ejsp.2508
Sample. Meta-analysis of 76 studies from 59 papers, N = 31,016.
Findings. Higher social identification was associated with less depression, average r = -.15, 95 per cent CI [-.20, -.11]. Heterogeneity was very large. The 95 per cent prediction interval for a future study runs from r = -.50 to .19, so the next study could plausibly find the opposite. Correcting for possible publication bias by trim-and-fill reduced the estimate to r = -.08, still significant. The authors conclude it would be premature to say the results uniformly support the social identity approach.
Licenses. That identifying with groups is associated with lower depression on average. The heterogeneity itself, which is the honest headline: this is a real but unreliable effect whose size depends heavily on circumstances the meta-analysis could not pin down.
Does not license. Presenting the association as dependable. A prediction interval spanning -.50 to .19 is the paper own reason for caution, and the corrected estimate is r = -.08. Equating identification with Gregariousness, for the same reason as cruwys-2013. Any individual-level or clinical claim.
Caveat. Held behind cruwys-2013 for this pole: larger, and its central finding is mostly a warning about variability.
Retrieved when: E2 >= 70
Deliberately excluded
Considered and rejected, with the reason in each case.
Roberts, Kuncel, Shiner, Caspi & Goldberg (2007), "The power of personality"
Its only citable claim is that traits predict life outcomes about as well as SES or IQ — a comparison between populations, in a report about one person, where it also forbids individual-level prediction. So it licensed almost no sentence here while supplying the most quotable overclaim in the corpus, next to the two papers whose whole job is discounting effect sizes.
MacCallum, Zhang, Preacher & Rucker (2002), on dichotomizing continuous variables
A statistics methods paper. It says itself that it licenses no claim about the instrument, and the rule it argues for — keep percentiles continuous, never cut them into high and low — is stated flatly in the composer's constraints, where it binds without needing a citation.
Schmidt & Hunter (1998) uncorrected validity estimates
Superseded by Sackett et al. 2022. Their r = .51 for GMA and structured interviews are exactly the figures Sackett corrected down. Cited only as the superseded baseline.
Duckworth et al. (2007), the original grit paper
Superseded by Credé et al. 2017; the incremental-validity claim does not hold. Only Credé is included.
DeNeve & Cooper (1998) and Steel, Schmidt & Shultz (2008) subjective-wellbeing meta-analyses
Both superseded by Anglim et al. 2020, which uses IPIP-NEO facets. Distinct from the included Steel (2007) procrastination meta-analysis, which is not superseded.
Carter & Weber (2010), "higher generalized trust predicts lie detection ability"
N = 29, and a 2025 registered replication (Levine et al., SPPS, DOI 10.1177/19485506241298340) failed to replicate. Excluded rather than carried with a caveat.
Ego-depletion / willpower-as-resource self-control papers (Baumeister and colleagues)
The depletion mechanism failed large-scale replication. Galla & Duckworth 2015 (habits) is used instead.
Power-posing, facial-feedback and priming-based studies
Excluded on replication grounds.
MBTI and other type-based instruments
Categorical, psychometrically inferior, and directly contradicted by the dimensional evidence in Gerlach et al. 2018.
Primary Bandura self-efficacy theory pieces
The Stajkovic & Luthans meta-analytic estimate is preferred over the primary theory.
MacLaren et al. (2011) as a sensation-seeking-to-gambling citation
It found essentially no effect (d ≈ .04); it cannot license a gambling-risk claim.
Non-English IPIP-NEO validation studies
This deployment administers the instrument in English against English-language norms, so a translated validation would not apply to any result here. Recorded so the gap is not silently filled later if the instrument is ever translated.
Where the research runs out
Thirty facets have two ends each, so there are sixty places a result can land. 5 of them have no paper behind them at all: the low end of Depression, the low end of Imagination, the low end of Artistic Interests, the low end of Trust, the low end of Achievement-Striving. Where your result lands in one of those, the report says the research does not extend to it rather than reaching for a tangential paper.
These are silences in the literature rather than gaps in the collection. Nobody has studied a preference for the concrete over the imaginative as a strength in its own right, so there is nothing to hold. Low Trust is the instructive one: the obvious candidate is the research on gullibility, and it turns out gullibility and trust are empirically unrelated — the trait that protects people from being exploited is social intelligence, not suspicion. Citing it would license a claim the evidence contradicts, so the cell stays empty.
The other fifty-five are covered, but not all equally. A good number are served by papers measuring the whole domain rather than the single facet, and each of those says so in its own does not license list.