For most of the twentieth century, HR decisions were made by feel. Who gets hired? The person the interviewer likes. Who gets promoted? The one who seems like leadership material. Who gets the high-visibility project? The manager's instinct. The problem was not that these decisions were necessarily wrong — some were excellent — but that they were made without systematic evidence, and without any reliable way to distinguish good instincts from self-serving bias, cultural homophily, or simple error.
Evidence-based HR is the attempt to change that. It applies the same commitment to data and rigor to people decisions that evidence-based medicine applies to clinical decisions: ask what the research actually shows, use validated instruments, measure outcomes, and update practices when the evidence demands it.
Among all the tools available to evidence-based HR practitioners, personality science — specifically the Big Five model of personality — has one of the strongest claims to a central role. Not because it is fashionable, but because it has been tested, replicated, and validated more rigorously than almost any other individual-difference measure in applied psychology. A general overview of the evidence-based management tradition is available at Wikipedia: Evidence-based management.
What Evidence-Based HR Actually Means — and Requires
The term "evidence-based HR" entered the management literature in part through Denise Rousseau and Eric Barends' 2011 paper in Human Resource Management (doi:10.1002/hrm.20409), which provided a working definition: the conscientious, explicit, and judicious use of the best available evidence to make decisions about the people practices most likely to achieve the organisation's goals.
Three elements of that definition deserve attention.
"Best available evidence" means not intuition, not industry convention, and not the most recent management bestseller — but research evidence that has been subjected to peer review, replicated across multiple samples and contexts, and made accessible to practitioners in a usable form.
"Conscientious and explicit" means that decisions are made through a deliberate process that can be examined and justified — not post-hoc rationalisation of an instinctive choice.
"Most likely to achieve the organisation's goals" means that evidence-based HR is not about measurement for its own sake, but about making better decisions that produce better outcomes: better hire quality, lower turnover, higher performance, stronger development.
Personality science meets all three criteria better than most alternatives in the HR toolkit. Understanding the evidence base is also essential for recognising what personality science cannot predict — the limits matter as much as the capabilities.
70 Years of Personality Research: What the Evidence Actually Shows
The Big Five / OCEAN model of personality — which structures personality into Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism — emerged from the convergence of multiple independent research traditions in the 1980s and has been the dominant scientific framework for personality research since then.
The evidence base is unusually deep.
Cross-cultural replication. The Big Five structure replicates in personality data from cultures across Europe, Asia, Africa, and the Americas. A landmark study by McCrae et al. (2005, doi:10.1037/0022-3514.89.3.407) found that the five-factor structure was recognisable in data from 50 cultures across six continents. This cross-cultural robustness is a significant advantage over many HR instruments that were validated on narrow, WEIRD (Western, Educated, Industrialised, Rich, Democratic) samples. The APA's guidance on psychological testing reinforces the importance of cross-cultural validation for any tool used in employment contexts: see the APA's standards for educational and psychological testing.
Predictive utility for job performance. Barrick and Mount's 1991 meta-analysis (doi:10.1111/j.1744-6570.1991.tb00688.x) established Conscientiousness as a consistent predictor of job performance across occupational groups. Subsequent meta-analyses confirmed the performance relationships for Extraversion (particularly in sales and management), Agreeableness (in team-based work), and Openness (in creative and knowledge work). Schmidt and Hunter's landmark 1998 review of personnel selection validity placed personality assessment — used alongside structured interviews and cognitive ability tests — among the highest-validity selection methods available.
Developmental utility. Beyond selection, personality data has demonstrated utility for leadership development, team composition, coaching, and career counselling. The trait measures are stable enough to be meaningful baselines, and sensitive enough to behavioural change that development programmes produce detectable shifts — particularly in observer-rated personality (the method Cèrcol uses in its Witness peer assessment). For a deeper exploration of what reliability and validity actually mean in practice, see what is reliability and validity in personality testing?.
What Big Five Personality Assessment Adds Beyond Interviews
The most common HR tool for assessing candidates is the unstructured employment interview. The interviewer meets the candidate, asks questions of their own choosing, and forms an impression. The problem is that the research on unstructured interview validity is discouraging. Schmidt and Hunter (1998) placed unstructured interviews at a validity coefficient of approximately 0.18 — modest predictive accuracy, and highly susceptible to interviewer bias, halo effects, and cultural match rather than actual competence.
Personality assessment adds three specific things that unstructured approaches miss.
Self-report bias reduction. Unstructured interviews elicit self-presentation, which is heavily shaped by social desirability and the candidate's theory of what the interviewer wants to hear. Validated personality assessments include item designs and scales that detect and partially correct for social desirability responding. Peer assessments — like Cèrcol's Witness instrument — bypass self-report entirely, measuring personality as it is actually expressed in behaviour. The problem of faking is explored in detail in can you fake a personality test?.
Structured comparison. A personality profile on a validated framework allows systematic comparison between candidates, between a candidate and a role profile, and between candidates and current high performers in equivalent roles. This structured comparison is impossible with unstructured interview data, which varies by interviewer and cannot be aggregated. To understand the 12 role archetypes Cèrcol maps onto Big Five profiles, see /roles.
Developmental continuity. Once a validated personality profile exists, it can support decisions beyond initial selection: development planning, team composition, promotion readiness, and succession planning. The same data that informs a hiring decision can inform an onboarding programme, a coaching relationship, and a development conversation three years later.
Where HR Goes Wrong: Unvalidated Tools and MBTI for Selection
Despite the depth of the evidence base for Big Five approaches, many organisations continue to use personality tools that lack scientific validity. The most prominent example is the Myers-Briggs Type Indicator (MBTI).
The MBTI was developed by Isabel Briggs Myers and her mother Katharine Cook Briggs, neither of whom were psychologists, based on their reading of Carl Jung's typological theory. The instrument classifies respondents into one of sixteen types along four dichotomies (Introvert/Extravert, Sensing/Intuition, Thinking/Feeling, Judging/Perceiving).
The psychometric problems with the MBTI are well-documented in the academic literature:
- Test-retest reliability is poor. Studies have found that a significant proportion of respondents (commonly cited as 35–50%) receive a different type classification when retested five weeks later.
- Dichotomous classification is arbitrary. The traits the MBTI measures are distributed continuously, not bimodally. Classifying a continuous trait as a type (Introvert vs Extravert) discards information and creates artificial categories.
- Predictive validity for job performance is low. Unlike Big Five measures, MBTI type classifications have not demonstrated consistent predictive validity for performance outcomes in peer-reviewed meta-analyses.
The MBTI is widely used for team communication and self-awareness exercises — for which its reliability is arguably sufficient — but using it for selection, promotion, or high-stakes people decisions is an evidence-based HR failure, regardless of how familiar or popular it is.
Other common unvalidated tools include DiSC (modest criterion validity), enneagram typologies (little psychometric validation), and various proprietary "culture fit" assessments that measure cultural similarity to current employees without examining whether those employees are high performers. For a comparison of open-source versus commercial tools, see personality testing: open-source vs commercial.
How to Build an Evidence-Based Personality Assessment Programme
An evidence-based personality programme has five components.
1. Choose validated instruments. Use tools based on established models (Big Five / IPIP, NEO) with published validity data, adequate test-retest reliability, and cross-cultural validation. Free IPIP-based tools — including Cèrcol — meet this standard. Many commercial tools do not publish their validity data; this should be treated as disqualifying. The IPIP — International Personality Item Pool is the public-domain foundation of validated personality items used in academic research worldwide.
2. Use personality data as one input, not the only input. Schmidt and Hunter's (1998) landmark review found that the highest-validity selection approaches combine multiple methods: structured interviews, cognitive ability tests, work samples, and personality assessments. No single instrument has sufficient predictive validity to be used alone in high-stakes decisions.
3. Train decision-makers in interpretation. Personality profiles are only as useful as the people reading them. Decision-makers should understand what the scores mean, what they do not mean, and how to use them in combination with other data.
4. Measure outcomes. An evidence-based programme tests whether its tools are predicting what they are supposed to predict. Track performance outcomes, development progress, and turnover against personality profiles over time. This allows the organisation to build internal validity data and detect tools that are failing to predict.
5. Apply ethical constraints. Personality assessment in selection raises legal and ethical questions in many jurisdictions. Valid instruments should be used only for the purposes they were validated for; results should be confidential; use of personality data in final selection decisions should be disclosed to candidates and, where legally required, subject to adverse impact analysis. For the full legal and ethical picture, see personality testing in hiring: what is legal and what is ethical.
Start with evidence-based assessment: Cèrcol is free
The case for evidence-based HR is also a case for using instruments built on transparent, peer-reviewed science rather than proprietary tools with opaque validation claims. Cèrcol implements the IPIP item pool — the same public-domain resource underlying academic Big Five research worldwide — and adds anonymous peer ratings through the Witness instrument. The result is an assessment that meets evidence-based standards without the commercial price tag. You can see how the science is implemented at /science, explore the 12 role profiles that map personality to team function, and run a full self-plus-peer assessment free at cercol.team. If your organisation is serious about replacing gut-feel people decisions with validated data, this is the most accessible entry point available.
Summary: How Personality Science Upgrades Each HR Practice
| HR practice | Evidence quality without personality science | Personality science contribution |
|---|---|---|
| Candidate selection | Low: unstructured interview validity ~0.18 | Incremental validity: adds ~0.10–0.15 when combined with structured interview |
| Leadership identification | Moderate: manager nomination with halo bias | Trait profile comparison with validated leadership correlates (Extraversion, Openness) |
| Team composition | Low: intuition and availability | Composition analysis: identify gaps and homogeneity risks before they cause problems |
| Development planning | Low: annual review with recency bias | Profile gaps and Witness data provide specific, evidence-based development focus |
| Succession planning | Low: informal sponsorship | Trait profiles support systematic comparison against role demands |
| Retention analysis | Low: exit interview data with hindsight bias | High Neuroticism and low-fit personality profiles predict early departure risk |
"Evidence-based management means translating principles based on best evidence into organisational practice." — Denise Rousseau & Eric Barends, Becoming an Evidence-Based HR Practitioner, 2011
Personality science is not the whole of evidence-based HR. But it is one of its most robust inputs — seven decades of cross-cultural replication, meta-analytic validation, and practical application. For organisations serious about moving from instinct to evidence in their people decisions, the Big Five is the natural starting point.
Further reading: Personality testing in hiring — what is legal and what is ethical · Personality testing: open-source vs commercial