Scientific Critiques
Addressing academic criticism of MBTI honestly and directly
最後更新: 2026年1月16日
Our Approach to Criticism
MBTI has been criticized by academic psychologists for decades. Rather than dismiss these criticisms or pretend they don't exist, we believe in engaging with them honestly.
Some criticisms are valid and important to understand. Others are overstated or based on misunderstandings. We address both below.
"MBTI doesn't predict job performance"
Criticism from: Morgeson et al. (2007), Pittenger (1993)
The Criticism
Meta-analyses show MBTI has weak to near-zero correlation with job performance, job satisfaction, and career success. Organizations shouldn't use it for hiring, promotion, or job placement decisions.
Our Response: This is valid.
We agree completely. MBTI should never be used for hiring decisions. The research is clear, and using personality tests to filter candidates is both ineffective and potentially discriminatory.
What we do: We explicitly state this in our guide and refuse to market OpenMBTI as a hiring tool.
"50% of people get different results when retaking"
Criticism from: Pittenger (1993), Hunsley et al. (2003)
The Criticism
Studies show that when people retake MBTI after a few weeks, about half get at least one letter different. A reliable test should give consistent results.
Our Response: Partially valid, but context matters.
The 50% figure is real, but it's largely a labeling problem rather than a measurement problem. Someone scoring 51% on Introversion one day might score 49% the next—their actual score barely changed, but their letter flipped.
The underlying dimension scores are more stable than the four-letter codes suggest. This is why we always show percentage scores, not just letters.
What we do: We display continuous percentages prominently and encourage users to focus on patterns across multiple tests via our history feature.
"Personality is continuous, not categorical"
Criticism from: McCrae & Costa (1989), Grant (2013)
The Criticism
MBTI treats personality as binary types (you're either E or I), but research shows personality traits are normally distributed on a continuum. Most people are near the middle, not clustered at extremes.
Our Response: This is valid.
This is probably the most fundamental criticism of MBTI, and the research clearly supports it. Personality dimensions follow bell curves, not bimodal distributions. The "type" boundaries are artificial.
However, categories can still be useful simplifications even if they're not technically accurate. We say "morning person" vs. "night owl" knowing it's a spectrum.
What we do: We always display scores as percentages and provide the Big Five calculator as a more continuous alternative.
"MBTI ignores Neuroticism, a key trait"
Criticism from: McCrae & Costa (1989)
The Criticism
The Big Five model includes Neuroticism (emotional stability), which predicts many important life outcomes. MBTI doesn't measure this dimension at all, making it incomplete as a personality model.
Our Response: Partially valid.
It's true that MBTI doesn't directly measure emotional stability. This was a deliberate choice by Myers and Briggs, who wanted to focus on "normal" personality variation without pathologizing any type.
Whether this is a flaw or a feature depends on your purpose. For clinical assessment, it's a limitation. For team building and self-reflection, avoiding judgmental dimensions may actually be beneficial.
What we do: We offer a Big Five estimator so users can explore how their MBTI might correlate with all five factors, including Neuroticism.
"Type descriptions are vague enough to fit anyone"
Criticism from: Forer (1949), general skeptic community
The Criticism
Like horoscopes, MBTI descriptions use "Barnum statements" that are vague enough to seem accurate for almost anyone. People accept flattering, general descriptions as uniquely applicable to themselves.
Our Response: Partially valid, but often overstated.
The Barnum effect is real and can inflate perceived accuracy. However, MBTI descriptions do contain specific, testable statements. An INFJ description saying "prefers one-on-one conversations over parties" is falsifiable—not everyone would agree with it.
The criticism applies more to pop-psych MBTI content than to careful type descriptions. We try to make our descriptions specific enough to be disagreeable.
What we do: We include specific behavioral examples and explicitly list potential weaknesses and blind spots—content that won't feel flattering or universally applicable.
"MBTI is a $2 billion industry with perverse incentives"
Criticism from: Grant (2013), Stromberg & Caswell (2015)
The Criticism
The MBTI industry makes billions selling certifications, workshops, and corporate training. This creates incentives to overstate validity and discourage critical examination.
Our Response: This is valid.
The commercial MBTI industry does have problematic incentives. Expensive certifications, proprietary instruments, and corporate contracts create pressure to oversell the tool's capabilities.
What we do: OpenMBTI is free, open source, and non-commercial. We have no financial incentive to overstate validity. We use OEJTS (a public-domain instrument) rather than the proprietary MBTI. Our business model is... we don't have one.
Summary: Which Criticisms Are Valid?
| Criticism | Validity | Our Response |
|---|---|---|
| Can't predict job performance | Valid | We explicitly advise against hiring use |
| Poor test-retest reliability | Partially valid | Show continuous scores, not just letters |
| Forced dichotomies | Valid | Display percentages; offer Big Five alternative |
| Missing Neuroticism | Partially valid | Provide Big Five estimator |
| Barnum effect | Partially valid | Include specific, falsifiable descriptions |
| Commercial exploitation | Valid | We're free, open source, non-commercial |