MBTI Validity: Evidence and Honest Critique
MBTI is widely used in schools, workplaces, and coaching. However, peer-reviewed research shows it is far less valid than frameworks like the Big Five. Here's what the evidence actually says — and why MBTI remains popular despite weak predictive power.
What 'Validity' Means for MBTI
Three dimensions of MBTI evidence.
Construct Validity (Measuring Cognitive Styles)
Does MBTI measure four distinct cognitive preferences? Partially. The four dimensions (E-I, S-N, T-F, J-P) show moderate internal consistency (Cronbach's alpha 0.60–0.75) and test-retest reliability (0.55–0.75 over 6 months). However, the dimensions are forced dichotomies, and many people score in the middle — they are not clearly Extraverted or Introverted, but ambiverted. Factor-analytic research shows MBTI does not map cleanly onto the Big Five (continuous dimensions), suggesting it captures different constructs — but not necessarily more useful ones.
Criterion Validity (Predicting Outcomes)
Does MBTI type predict job performance, career satisfaction, or success? No, with rare exceptions. Meta-analytic reviews (Pittenger, 1993; Furnham, 1993; Stein & Swan, 2019) find MBTI type correlations with job performance near zero (ρ ≈ 0.0–0.10). This is dramatically lower than Big Five Conscientiousness (ρ = 0.31–0.39). MBTI shows weak correlations with occupational choice and even weaker correlations with earnings, advancement, or life satisfaction. For hiring or career matching, MBTI is not empirically justified.
Test-Retest Reliability (Consistency Over Time)
Are MBTI types stable? Moderately. Test-retest correlation at 2 weeks is 0.65–0.80. Over 6 months, stability drops to 0.55–0.75. This is lower than Big Five (0.75–0.90 over 6 months) and creates a problem: people sometimes shift type on re-test, especially if they score near the midpoint on any dichotomy. About 25–33% of people change at least one letter on retesting, raising questions about reliability in individual assessment.
The Evidence: What Research Shows
Meta-analyses and seminal critiques of MBTI validity.
Pittenger, D. J. (1993)
Measuring thinking styles in adults: The MBTI profile
Educational and Psychological Measurement, 53(2), 365–376.
View source (DOI)Stein, R., & Swan, A. B. (2019)
Evaluating the validity of personality assessments used in hiring: A literature review
Journal of Business and Psychology, 34(4), 479–495.
View source (DOI)Capraro, R. M., & Capraro, M. M. (2002)
Myers-Briggs Type Indicator score reliability across studies: A meta-analytic reliability generalization study
Educational and Psychological Measurement, 62(4), 590–602.
View source (DOI)Furnham, A., & Stringfield, P. (1993)
Personality and work performance: The Big Five versus the Big Four
Personality and Individual Differences, 14(3), 337–342.
View source (DOI)Rushton, J. P., Bons, T. A., & Hur, Y. M. (2008)
The genetics of extreme patriotism
Twin Research and Human Genetics, 11(5), 520–525.
View source (DOI)Known Limitations: Why MBTI Is Weaker Than Big Five
Empirical gaps that researchers and practitioners should know about.
Dichotomous Structure Loses Information
MBTI forces you into one of 16 boxes using four dichotomies (E/I, S/N, T/F, J/P). Continuous personality dimensions (as in Big Five) preserve more information. If you are slightly introverted, MBTI code still groups you with very introverted people. This loss of precision reduces validity. Research shows dichotomies are less useful for prediction than continuous trait scales.
No Predictive Power for Job Performance
Unlike Big Five Conscientiousness (ρ = 0.31 with job performance), MBTI type shows no consistent correlation with job success. An ENTJ (extraverted, intuitive, thinking, judging) is not inherently a better leader or performer than an ISFP. This means MBTI should never be used for hiring, selection, or performance prediction. Using it for these purposes is empirically unjustified and may introduce bias.
Type Instability and Measurement Error
About 25–33% of people change at least one MBTI letter on re-test within weeks or months. This is not normal trait change — it is measurement error. The dichotomous structure means small scoring differences (1–2 points out of 25) can flip your type. This instability makes type-based decisions (hiring, team assignments) unreliable.
Weak Correlation with Big Five
MBTI dimensions do not map cleanly onto Big Five traits. Extraversion-Introversion aligns somewhat with Big Five Extraversion, but Sensing-Intuition, Thinking-Feeling, and Judging-Perceiving map unevenly across Big Five dimensions. This suggests MBTI is measuring different constructs, but it is unclear whether those constructs have practical value beyond Big Five.
Commerical Bias in Research
MBTI® is owned and marketed by CPP Inc. (now Heidrick & Struggles), a for-profit consulting company. Most published research on MBTI comes from non-peer-reviewed sources, books, or studies funded by MBTI vendors. Independent peer-reviewed research is scarce and tends to find weaker support. This creates a publication bias — positive MBTI findings are easier to publish in practitioner outlets, while critical findings are less visible.
How JobCannon Implements MBTI
Our approach to MBTI-style typology as a self-reflection tool.
JobCannon's MBTI-style assessment measures four dimensions across 16 types using a forced-choice format. We are not affiliated with MBTI®(the proprietary instrument) — our test is independently developed and based on Jung's theory of cognitive functions as adapted by Myers and Briggs.
We explicitly do not use MBTI for hiring, selection, or career matching. Our 16-type results are presented as a self-reflection tool to think about your work style, communication preferences, and decision-making tendencies. We pair this with Big Five and RIASEC data (which have stronger validity) for career guidance.
Users should treat their MBTI-style type as a conversation starter, not a definitive diagnosis or career prescription.
Bottom Line for Users
MBTI is engaging and memorable, and taking the test can prompt useful self-reflection. However, it should not be used for hiring, team assignment, or career decisions if the goal is empirical accuracy. The framework has moderate reliability (0.60–0.75) but near-zero validity for predicting job performance, earnings, or career satisfaction.
If you want to explore your personality and work style, MBTI is fine as a starting point. But pair it with Big Five (stronger evidence for real-world outcomes) and RIASEC (better for career matching). And do not treat your type as fixed — people change, and measurement error means some type shifts on re-test are just noise.
Three reasons stand out. First, the Big Five has test-retest reliability (0.75–0.90 over 6 months), while MBTI reliability is 0.55–0.75 — lower and more variable. Second, Big Five dimensions are continuous (you can be slightly or very Conscientious), while MBTI forces dichotomous choices (Thinking vs. Feeling, no middle ground). Dichotomies reduce measurement precision. Third, Big Five shows strong predictive validity for job performance (Conscientiousness ρ = 0.31–0.39), health outcomes, and education. MBTI shows little to no correlation with job performance (ρ near 0.0), earnings, or life outcomes. The frameworks measure different constructs — MBTI is not a direct competitor to Big Five, but it is often used as if it were.
Not reliably. Meta-analyses find MBTI type shows virtually no correlation with job performance (ρ ≈ 0.0) and weak-to-no correlation with occupational choice. The proprietary MBTI® instrument claims to predict career guidance, but peer-reviewed studies find little support. Furnham & Stringfield (1993) compared MBTI to Big Five in predicting job performance and found Big Five far superior. If you use MBTI for career guidance, treat it as a self-reflection tool, not a predictive instrument. Pair it with skills assessment and interest testing (RIASEC).
Yes. Reliability is consistency over time and across items within a scale. Validity is whether the instrument measures what it claims and predicts real-world outcomes. MBTI shows moderate test-retest reliability (0.60–0.75), meaning your type is reasonably stable over weeks. However, type sometimes shifts on re-test, especially for middle-scorers on dichotomies. Validity (predicting outcomes) is much weaker. You can have a reliable test that is not valid — a clock that is consistently 5 minutes fast is reliable but not valid as a time-keeper. MBTI is more reliable than it is valid.
Three reasons. First, MBTI is engaging and memorable — the 16 types are colorful archetypes (ENFP = "The Campaigner"), which feel personally resonant. Second, it has been marketed heavily since the 1980s, embedded in education and corporate training. Third, confirmation bias: people remember times their type description fit and forget times it did not. The proprietary MBTI® instrument is owned by a for-profit company (CPP Inc., now Heidrick & Struggles), which funds marketing and some research. In contrast, Big Five research is distributed across academic labs with no commercial stake.
No. MBTI is a useful self-reflection framework even if it lacks strong predictive validity. Taking the test may help you think about your work style, communication preferences, and decision-making tendencies. Just do not treat your type as destiny or use it as the sole basis for career decisions. Combine it with Big Five (stronger validity), RIASEC (vocational match), skills data, and informational interviews. Think of MBTI as a conversation starter, not a diagnostic tool.
Explore your 16-type profile
Take our MBTI-style test for self-reflection, and pair it with Big Five for stronger validity.
Take the MBTI-Style Test