Personality and Individual Differences

Personality & Individual Differences

The evidence-based science of stable individual differences, meaning who a person reliably is across time and situations, applied to one practical job: tailoring how you communicate with and manage different stakeholders, grounded in evidence and avoiding the MBTI trap.

The single most important thing this skill enforces: use the dimensional trait model (Big Five / HEXACO), not type systems (MBTI, DISC, Enneagram, True Colors). The Big Five is the consensus, replicated, predictive model of personality structure in academic psychology. Type systems are popular, fun, and psychometrically weak. This distinction is the whole game. Get it wrong and every downstream inference is built on sand.

Hard ethical frame (read first). Traits are probabilistic, not deterministic; descriptive, not diagnostic. You observe behavior, you do not “diagnose” or label a person. Never use trait inferences for hiring, promotion, or selection decisions about a real individual — that is both scientifically unjustified and, in many jurisdictions, illegal. The legitimate use here is adapting your own communication to observed behavior, holding every inference loosely. See Ethics & Limits at the end — it is not optional.

1. The Big Five / Five-Factor Model (the evidence-based model)

Five broad, continuous dimensions describe the bulk of stable personality variance. The acronym is OCEAN (Wikipedia: Big Five):

Each dimension is a continuum, normally distributed. Most people are in the middle on any given trait; extreme scorers are rare. There are no “types”, which matters enormously for the MBTI critique below.

Facets matter. Each of the five domains splits into six narrower facets (30 total in the NEO-PI-R; Costa & McCrae 1992, 240 items) (Big Five history). Two people can share an Extraversion score but differ sharply on facets: one high on Assertiveness/low on Warmth (a driving, task-focused executive), another high on Warmth/low on Assertiveness (a relational, non-pushy collaborator). When a domain score doesn’t fit what you observe, drop to the facet level.

The lexical hypothesis — where the five came from

The Big Five was discovered, not invented, via the lexical hypothesis: the idea that the most important individual differences become encoded as single words in natural language (Wikipedia).

• Allport & Odbert (1936) extracted ~4,504 personality-descriptive adjectives from the dictionary.
• Raymond Cattell (1940s) reduced the list and built the 16PF.
• Tupes & Christal (1961) analyzing Air Force ratings, repeatedly recovered five factors (Surgency, Agreeableness, Dependability, Emotional Stability, Culture).
• Norman (1963) replicated it, relabeling Surgency → Extraversion and Dependability → Conscientiousness.
• Lewis Goldberg (1981) coined the term “Big Five” and built the marker adjective scales.
• Costa & McCrae operationalized the questionnaire tradition (NEO-PI → NEO-PI-R, 1992), giving the Five-Factor Model (FFM) and the OCEAN naming.

“Big Five” vs “Five-Factor Model” — a real but minor distinction. “Big Five” comes from the lexical/adjective tradition (Goldberg); “FFM” from the questionnaire tradition (Costa & McCrae). They converge on the same five dimensions; use them interchangeably in practice, but know they have slightly different historical roots and factor-recovery methods.

How good is it? (predictive validity, honestly stated)

Personality predicts real outcomes — modestly but reliably. Conscientiousness is the standout: it predicts job performance across virtually every occupation (Barrick & Mount 1991 meta-analysis; corrected ρ ≈ .19–.22) and academic performance even more strongly (≈ .28) (job-performance meta-analysis). Low Conscientiousness shows strong links to common mental disorders across a meta-analysis of n > 75,000. Other traits predict job-specifically: Extraversion for sales, Agreeableness for customer service. Calibrate expectations — these are correlations in the .1–.3 range, not destiny. Personality nudges the odds; it does not dictate the outcome of any single interaction.

Cross-cultural replication is good but imperfect. The five replicate across dozens of cultures, with documented exceptions: Openness is weakly supported in some Asian samples, and some languages don’t yield a clean single Agreeableness factor (Wikipedia).

2. HEXACO — the six-factor model and Honesty-Humility

Ashton & Lee (early 2000s) re-ran lexical studies across many languages and recovered six factors, not five. HEXACO = Honesty-Humility, Emotionality, eXtraversion, Agreeableness, Conscientiousness, Openness (SAPA Project; Psychology Today: HEXACO).

The new factor, Honesty-Humility (H), captures sincerity, fairness, greed-avoidance, and modesty. Broadly, it is the degree to which someone refrains from exploiting others when they could get away with it.

HEXACO is not “Big Five + 1.” It is a re-partitioning. Big Five Agreeableness and Neuroticism are carved up differently: the “even-tempered / not-angry” content that sits inside Big Five Agreeableness moves into HEXACO Agreeableness, the anxiety/sentimentality content moves into HEXACO Emotionality, and a distinct honesty/fairness/non-exploitation dimension separates out as H (HEXACO vs Big Five). HEXACO Extraversion, Conscientiousness, and Openness map closely to their Big Five counterparts; the other three differ substantially.

Why H matters for stakeholder work: Honesty-Humility is the best single-factor predictor of exploitative, manipulative, and unethical behavior, better than anything in the Big Five, because the Big Five lacks a dedicated factor for it. Low H is the trait core of the Dark Triad (§6). When a counterpart’s behavior is making your antennae twitch — over-claiming, subtle manipulation, win-at-any-cost framing — you are usually detecting low H, and the Big Five alone would have under-weighted it.

3. Stability vs change — and state vs trait

Traits are stable, but not frozen, and stability has two distinct meanings people constantly conflate:

• Rank-order stability — does a person keep their position relative to peers? (If you’re more conscientious than 80% of your cohort at 25, are you still at ~25?) Roberts & DelVecchio’s (2000) meta-analysis of 152 studies (3,217 correlations) found rank-order consistency rising with age: .31 in childhood → .54 in college → .64 at age 30 → a plateau around .74 at ages 50–70 (Roberts & DelVecchio 2000, PDF). This is the cumulative-continuity principle: personality gets more stable as we age, but is never perfectly fixed.
• Mean-level change — does the average level of a trait shift across the population with age? Roberts, Walton & Viechtbauer’s (2006) meta-analysis of 92 samples found it does, in a systematic direction now called the maturity principle: people rise in Conscientiousness, Emotional Stability, and social dominance (a facet of Extraversion) through young adulthood (20–40); Openness and social vitality rise in adolescence then decline in old age (Roberts, Walton & Viechtbauer 2006, PubMed). People, on average, become more dependable and even-keeled as they age.

State vs trait. A trait is your characteristic average; a state is how you are right now. A low-Extraversion person can be lively at their own product’s launch; a stable person can be visibly anxious during an outage. Fleeson’s Whole Trait Theory resolves this elegantly: a trait is a density distribution of states — each person varies enormously moment-to-moment, but the shape of their distribution (their average and spread) is highly stable over time (Whole Trait Theory, PMC). Practically: judge a stakeholder on their distribution (the pattern over many interactions), never on one data point. A single tense meeting tells you about a state, not a trait.

Heritability. Twin studies put Big Five heritability at ~40–60% (Jang et al. 1996: N 41%, E 53%, O 61%, A 41%, C 44%) (heritability twin study). The non-genetic variance is almost entirely the non-shared environment (unique individual experiences), not shared family upbringing. Siblings raised together are barely more alike in personality than strangers, once genes are accounted for.

4. The person-situation debate — behavior = trait × situation

In 1968, Walter Mischel’s Personality and Assessment nearly ended trait psychology. He pointed out that correlations between a trait measure and behavior in any single situation rarely exceed r ≈ .30 — the so-called “personality coefficient” — and argued broad cross-situational traits were therefore “untenable” (Wikipedia: person-situation debate). This is the consistency paradox: we intuit that people are consistent, yet single-situation correlations look weak.

The resolution — now the settled consensus — is interactionism:

• Epstein’s aggregation principle (1979): the .30 ceiling is an artifact of measuring one behavior on one occasion. Aggregate across many occasions and trait–behavior correlations climb to .60–.80. Traits predict patterns over time, not isolated acts. (This is the §3 “judge the distribution” point in another guise.)
• Situational strength: strong situations (a funeral, a tightly scripted compliance review, a hard contractual deadline) suppress trait expression — everyone behaves similarly. Weak situations (an open-ended discovery call, an unstructured workshop) let traits show. Personality is most visible, and most useful to read, in weak, ambiguous situations.
• Mischel & Shoda’s CAPS (1995): the Cognitive-Affective Personality System reframes consistency as stable if…then behavioral signatures — “if challenged in front of peers, then this person digs in; if challenged privately, then they concede.” The signature itself is the stable trait, even though raw behavior varies by situation.

The operator takeaway: “behavior = trait × situation.” Don’t over-attribute a stakeholder’s behavior to their personality when the situation (a tense escalation, an exec in the room, a renewal deadline) is doing the work. This is the Fundamental Attribution Error applied to reading people — see behavioral-decision-making and persuasion-and-influence-psychology for the attribution machinery; this skill owns the trait side of the equation.

5. Why MBTI and type systems are psychometrically weak (steer to dimensional models)

The Myers-Briggs Type Indicator (MBTI) sorts people into 16 types across four binary dichotomies (E/I, S/N, T/F, J/P), loosely derived from Jung. It is the world’s most popular personality instrument and the textbook example of what not to trust. Four fatal flaws:

1. Low test-retest reliability. Studies find 39–76% of people get a different four-letter type on retaking the instrument within weeks (MBTI validity, Truity). A measure that reclassifies up to half its takers in five weeks cannot be measuring something stable. (The publisher reports higher scale-level coefficients; the type-flipping problem is the damning one, and it follows directly from flaw 2.)
2. False dichotomies — the bimodality failure. MBTI forces a binary (you’re “an Extravert” or “an Introvert”), but every trait is continuous and normally distributed, with most people clustered in the middle. Someone at 49/100 is labeled the opposite type from someone at 51/100, despite being functionally identical. Cutting a normal distribution at the mean manufactures two “types” that don’t exist in nature (Truity).
3. Weak predictive validity. MBTI type does not reliably predict job performance or life outcomes. Even the Myers-Briggs Company states it is inappropriate and potentially illegal to use for hiring or selection.
4. The Forer / Barnum effect. Type descriptions feel uncannily accurate because they are vague enough to fit almost anyone. In Forer’s 1949 classic, students rated a “personalized” profile 4.3/5 for accuracy — it was 13 generic statements (many lifted from an astrology book) handed identically to everyone (Barnum effect, Wikipedia). “That’s so me!” is not evidence of validity — it is the failure mode to guard against.

DISC, the Enneagram, True Colors, Insights Discovery share flaws 1, 2, and 4: they impose discrete types/colors on continuous dimensions and lean on Barnum-style feedback. They can be fine as low-stakes team-bonding vocabulary, but never treat their categories as real, stable, or predictive.

What to do instead: when someone asks “what’s my type / their type,” redirect to dimensional thinking. Don’t ask “are they an introvert?”; ask “where do they sit on Extraversion, and on which facets?” The honest, useful, evidence-based model is the Big Five (or HEXACO). When in doubt, use the validated, free IPIP-NEO items rather than a branded type test.

6. The Dark Triad — recognizing difficult stakeholders (brief, non-clinical)

Paulhus & Williams (2002) grouped three subclinical, socially aversive but non-pathological traits that share a callous-manipulative core: the Dark Triad (Paulhus & Williams 2002, PDF).

• Narcissism — grandiosity, entitlement, need for admiration, thin skin to criticism.
• Machiavellianism — strategic manipulation, cynicism, ends-justify-means, low warmth.
• Psychopathy (subclinical) — callousness, impulsivity, low empathy, low anxiety, thrill-seeking.

A fourth, sadism (taking pleasure in others’ distress), extends it to the Dark Tetrad. Measured with the Dirty Dozen and the Short Dark Triad (SD3) (Jones & Paulhus 2014). Research debates whether the three are distinct or share a single “dark core” / D-factor (Moshagen et al. 2018) — an underlying tendency to maximize one’s own utility at others’ expense, with whatever justifying beliefs (Dark Triad, Wikipedia).

In trait terms, the Dark Triad maps cleanly onto very low Honesty-Humility (§2) and low Agreeableness — which is exactly why HEXACO detects this profile that the Big Five tends to under-weight.

Use this defensively, not diagnostically. You are not qualified to diagnose anyone, and you must not call a stakeholder a “narcissist” or “psychopath” — that’s defamatory and clinically meaningless from the outside. The only legitimate use: notice a pattern of manipulative, exploitative, or callous behavior over time, then adjust your own posture — get agreements in writing, reduce reliance on goodwill, document, set firm boundaries, and avoid being manipulated by flattery or false urgency. For the influence/manipulation tactics themselves and how to counter them, defer to persuasion-and-influence-psychology and trust-and-psychological-safety.

7. Applying traits ethically to communication (the practical payoff)

The legitimate workflow: observe behavior → infer a trait tendency (held loosely) → adapt your own communication → update on new evidence. You are tuning your style, not boxing them in. Below, each trait is paired with what to do when a stakeholder’s observed behavior sits high or low.

Conscientiousness

• High C (organized, deadline-driven, detail-oriented): lead with structure — agendas, clear next steps with owners and dates, thorough written follow-ups. Be punctual and reliable; sloppiness erodes trust fast. They reward precision.
• Low C (flexible, big-picture, spontaneous): don’t bury them in process. Lead with the headline and the “why,” keep documents short, offer to own the logistics so they don’t have to track details. Over-structuring reads as friction.

Extraversion

• High E (talkative, energetic, thinks out loud): give space for live discussion, calls over email, brainstorming, social warmth. They process externally — silence can read as disengagement.
• Low E (reserved, reflective, drained by big meetings): send materials in advance so they can process alone; favor written or small-group channels; don’t mistake quiet for disagreement or low engagement. Give them time to respond.

Openness

• High O (curious, idea-driven, likes novelty): engage the vision, the roadmap, the novel approach; analogies and possibilities land well. Rigid “this is the only way” framing bores them.
• Low O (practical, prefers the proven, concrete): lead with track record, references, standards, and the conventional path. Frame change as low-risk and incremental; “new and experimental” reads to them as risk, not appeal.

Agreeableness

• High A (cooperative, harmony-seeking, conflict-averse): they may agree outwardly while privately hesitant — explicitly invite dissent and surface concerns, or you’ll miss real objections. Don’t mistake politeness for buy-in.
• Low A (skeptical, blunt, hard-bargaining, competitive): expect challenge; don’t take pushback personally. Bring evidence, expect negotiation, be direct. They often respect a counterpart who holds their ground over one who folds.

Neuroticism (Emotional Stability)

• High N / reactive (anxious, stress-sensitive, especially in escalations): over-communicate, reduce uncertainty, give frequent reassuring status, avoid surprises, name and validate the concern. Calm, predictable cadence lowers the temperature.
• Low N / stable (calm under pressure): you can be more direct and matter-of-fact about risks and bad news without extensive cushioning.

Most stakeholders are mid-range on most traits. The high-payoff moves are at the clear extremes you can actually observe — a visibly detail-obsessed buyer, a visibly skeptical security lead, an anxious champion during an outage. Match their distribution, don’t manufacture a type.

Cross-references for the content you’re communicating

This skill owns who the stakeholder is (trait structure). For what you say and how to move them, defer:

• Persuasion / influence mechanism and tactics → persuasion-and-influence-psychology
• Building trust, rapport, psychological safety → trust-and-psychological-safety
• Framing decisions, anchoring, debiasing, attribution errors → behavioral-decision-making
• Driving product adoption via motivation/habit → behavior-change-psychology
• Reading emotion/mood and emotional intelligence in the moment → emotion-and-affect-psychology
• Motivation/grit/mindset/resilience as performance traits → performance-and-resilience-psychology

Ethics & Limits (not optional)

1. Probabilistic, not deterministic. A trait shifts the odds of a behavior; it never guarantees one. Treat every inference as a hypothesis to update, not a verdict.
2. Observe behavior; do not diagnose. You read what people do, over many interactions. You never assign clinical labels (“narcissist,” “neurotic case”) — you are not a clinician, and from the outside such labels are meaningless and harmful.
3. No selection misuse. Never use trait inferences to make hiring, promotion, firing, or selection decisions about a real person. It is scientifically unsupported (effect sizes too small for individual decisions) and frequently illegal.
4. It’s about adapting you, not boxing them. The whole legitimate purpose is tuning your own communication to be more effective and respectful — not building a dossier or a manipulation map.
5. Beware your own biases. Confirmation bias makes you “see” the trait you expect; the Barnum effect makes vague reads feel precise; the Fundamental Attribution Error makes you blame personality when the situation is the cause (§4). Hold every read loosely and seek disconfirming evidence.
6. Privacy and consent. Don’t administer personality tests to customers, don’t record speculative “personality profiles” of individuals in CRM/notes, and don’t share trait speculation about a named person as if it were fact.
7. Clinical scope is off-limits. Personality disorders, mental-health diagnosis, and therapy are out of scope — refer to a licensed professional.

References

• Wikipedia — Big Five personality traits (domains, lexical history, heritability, cross-cultural replication, criticisms)
• Roberts & DelVecchio (2000), The Rank-Order Consistency of Personality Traits (PDF) — rank-order stability by age
• Roberts, Walton & Viechtbauer (2006), Patterns of Mean-Level Change (PubMed) — the maturity principle
• Fleeson & Jayawickreme, Whole Trait Theory (PMC) — traits as density distributions of states
• Wikipedia — Person–situation debate (Mischel 1968, .30 coefficient, Epstein aggregation, CAPS, interactionism)
• SAPA Project — The HEXACO Model of Personality; Psychology Today — HEXACO
• high5test — Big Five vs HEXACO (re-partitioning, not addition)
• Truity — Validity of the MBTI: Is the MBTI scientific? (reliability, false-dichotomy/bimodality, predictive validity)
• Wikipedia — Barnum effect (Forer 1949, vague-statement validity illusion)
• Paulhus & Williams (2002), The Dark Triad of Personality (PDF); Wikipedia — Dark Triad (SD3, Dark Tetrad, D-factor / Moshagen et al. 2018)
• Jang et al. (1996) via Heritability of the Big Five (PMC) — twin-study heritability, non-shared environment
• Barrick & Mount (1991) via Big Five & Job Performance — Conscientiousness as cross-occupational predictor (ρ ≈ .19–.22)