Your ability split

You have both abilities on every trait; the numbers show how your answers divide between them. The label is the character that split describes.

What stands out about you

Compared with people in an open dataset of 249 responses. Traits close to the middle are left out — they're the ones where you use both sides.

How solid is this result?

Zodiac element · optional

Add your sun sign to see how its element (Fire, Earth, Air or Water) combines with your type. This is just for fun: astrology has no scientifically demonstrated link to personality.

1. How Hexacter analyses your answers

1. Dimensions, not boxes. Each of the six dimensions is measured as a continuous score from 0% to 100%. Your letters only summarise which side of 50% you fall on. Studies that tested whether personality types are real categories consistently found continuous dimensions instead^[1,2,8], so the percentages carry more information than the letters.
2. Balanced wording. Each sitting asks 10 statements per dimension: 5 worded toward one side and 5 toward the other, drawn at random from a bank of 20 (so two people rarely get the same set or order). People differ in how much they agree with statements in general (called acquiescence). Balanced wording cancels that out so it can't push you toward one side^[12,21]. Answers use a 5-point scale. Research finds 7 points a little more reliable, but the difference is small and 5 is quicker to answer, so we chose speed^[15].
3. From answers to scores. Each answer counts from −3 to +3 and is flipped for statements worded the other way. The answers are summed for each dimension and converted to a lean percentage: 50 + (sum ÷ largest possible sum) × 50. An exact 50% is a tie, which your most recent strong answer on that dimension breaks. The result tells you when that happened.
4. How strong is a lean? Strength labels follow the published "preference clarity" bands of the most widely used type questionnaire: slight < 60%, moderate 60–76%, clear 77–92%, very clear ≥ 93%^[18]. The soft band on your profile chart shows how far your score would realistically move if you retook the test, not where other people fall (see section 2b).
5. Answer check. Your result includes two standard quality checks from questionnaire research: agreement tendency (did you agree with most statements whatever they said?) and consistency (did your answers about the same trait hang together?)^[9,6]. We tested the consistency check on 3,000 simulated random and 3,000 simulated honest answer sets. It flagged about 3 in 4 random patterns and fewer than 1 in 100 honest but noisy ones.
6. The zodiac layer is separate. It's optional and never changes your scores. Large studies have found no link between birth date and personality^[7,4].

2. How researchers check whether a personality test works

Reliability: is it consistent?

• Internal consistency asks whether statements about the same trait agree with each other. It's usually measured with Cronbach's alpha, from 0 to 1: .70 or more is acceptable, and .80 or more is good^[5]. Type questionnaires typically reach .80–.92 per dimension. Thinking–Feeling is the weakest, with values reported as low as .61–.64^[3,16,18].
• Test–retest reliability asks whether people get similar scores weeks later. Continuous scores retest at about .57–.81^[18]. But because so many people sit near the midpoint, about half get a different four-letter type on retest^[13,14]. That's why Hexacter highlights your percentages and near-balanced traits.

Validity: does it measure what it claims?

• Convergent validity checks whether scores agree with established measures. The four type dimensions line up with four of the Big Five traits: E–I with Extraversion (r ≈ .7), S–N with Openness (≈ .7), T–F with Agreeableness (≈ .45) and J–P with Conscientiousness (≈ .5)^[8,18]. The Big Five's fifth trait, Neuroticism, barely relates to any of the four (r between −.11 and .07)^[18], which is why Hexacter adds a separate stress dimension (A/T).
• Factor analysis checks whether the statements group into the intended dimensions. For the main type questionnaire, they form the expected four factors^[18].

Types or continuous traits? Taxometric analysis tests whether a trait comes in real types or on a continuum, and distribution analysis checks whether scores split into two camps. Both point to continuous dimensions, with most people near the middle^[1,2,8].

Testing the theory itself. Some popular claims, such as "type dynamics" (the idea that each type has a dominant function that is its strongest preference), haven't held up when tested^[17,20].

2b. How results are shown, and why

• An ability split, not a grade. The hexagon chart and the letter grades have both gone. People read position along a common scale far more accurately than the angles and areas a radar chart depends on^[23], and in a controlled comparison the radial form came out "least effective and least liked"^[24]. Grades had a different problem: moderate preferences are the norm, so a graded scale handed most people a B, C or D. Each trait is now shown as both of its abilities, as two numbers that add to 100, with a plain-English character label for where you sit (for example "Lone wolf", "Ambivert" or "Crowd magnet"). The label changes at the published preference-clarity cut-offs^[18], and the middle has a name of its own, because a near-even split means range, not a low score.
• Left-brain / right-brain, as a metaphor only. Results include a "brain-style mix" because people find the image intuitive, but it is not neuroscience: resting-state brain scans of over 1,000 people found no evidence that individuals are left- or right-brained^[30]. On this site it is simply the average of your S, T and J scores (the popular "left" list: detail, logic, structure) against N, F and P (the popular "right" list: imagination, feeling, spontaneity).
• Direction, not just strength. This is the change that matters most. The old chart plotted how strongly you lean, which is almost identical for all 16 types — so every type drew nearly the same shape. What actually separates the types is which side they fall on. Measured on the open dataset, plotting direction spreads the 16 types across 26–39 points per trait where plotting strength alone spread them across only 2.6–11.6 — so the shapes are now genuinely different, because the underlying numbers always were.
• Uncertainty as a soft band. The range around each score is drawn as a faded band rather than hard whiskers, because error bars are widely misread as firm cut-offs, while density-style shading produces better-calibrated judgements^[28].
• Ranges, not false precision. Each score shows the range it would realistically move within if you retook the test (one standard error of measurement), which is standard practice in test feedback^[18].
• How solid each letter is. Letters near the middle flip easily. We work out the chance of each letter changing from its distance to the middle and the published test–retest correlations^[18]. Applied to the open dataset, the model predicts 51% of people keep all four letters, close to the ~50% reported in research^[13]. It is shown as "if you retook it 20 times", because people reason about chances far more accurately as natural frequencies than as percentages^[31].
• What stands out, not everything at once. The scores worth discussing are the ones furthest from the middle, which is how bipolar profiles such as the 16PF are read^[25].
• Contrast instead of flattery. Type descriptions are easy to write so that they fit nearly everyone — the "Barnum effect", which has been demonstrated specifically for MBTI-style descriptions^[26],[27]. That's why each type page shows how it differs from its one-letter neighbours, how common it is, and what each letter is actually correlated with, rather than only a flattering summary. The correlations shown there, and the figure for how common each type is, were read directly off the published tables^[29] rather than copied from secondary sites — several widely repeated "population percentage" tables could not be traced to a checkable source, so they are not used here.

3. Where the stats on this site come from

• Most people are in the middle. Type scores form one central peak, not two separate camps^[2]. In an open-test sample, 60–82% of people scored in the two middle bands on each dimension^[11,22]. We estimate that about 3 in 4 people are near-balanced on at least one classic dimension, so being an all-rounder is normal.
• The typical preference is "moderate". In the official Form M data (about 8,800 people), average clarity was moderate on every dimension. E–I and J–P were a little clearer than S–N and T–F, and 27% of people changed at least one letter after reviewing their type^[18].
• Typical values and bands. We converted those published averages to lean percentages. Because scores form one central peak, the spread on each side follows a folded bell curve, which gives the "typical" value (the median) and the middle-half band. For example, E–I has a typical value of 70% and most people fall between 60% and 85%.
• Per-type figures come from real responses. The "All 64 types" page averages an open dataset of OEJTS responses (OpenJung, CC BY 4.0, de-identified and consented; 249 responses as of 20 September 2026, 5–29 per type). Because those per-type samples are small, each average is pulled toward the overall average by how much data backs it (empirical-Bayes shrinkage), so a thin sample can't invent a difference. After that correction, types differ by about 12 percentage points on Extraversion–Introversion — introverted types lean more strongly in this self-selected online sample — and by only 3–4 points on the other three dimensions. A/T and D/H aren't in the dataset, so they keep the published-research estimate.
• No signature strength for each type. Because type-dynamics claims haven't held up^[17,20], every type's typical strength is shown the same way. Types differ in which side they lean to and in how common they are.
• How common each type is comes from the US national sample in the MBTI Manual (1998, N = 3,009)^[10].

4. Limitations: what Hexacter hasn't validated yet

• Hexacter's own statements haven't yet been tested for reliability or validity with real respondents. The norms above come from other questionnaires. With opt-in, anonymous data, the next steps would be alpha for each dimension, test–retest reliability, a factor analysis and recalibrated norms.
• A/T and D/H are Hexacter's own dimensions and have no published norms, so they're shown as estimates (marked *).
• Children's self-reports are less consistent than adults', and younger children agree with statements more often whatever they say^[19]. The kids' edition uses fewer, simpler statements and treats results as conversation starters.
• Type frequencies are US figures from 1998. Other countries and other questionnaires differ.
• Results can change with mood, life stage and context, and retaking later is normal and informative. Don't use this test for hiring, clinical or other high-stakes decisions.
• The function stacks shown with each type (e.g. Ni · Te · Fi · Se) follow the Jungian convention. They're descriptive, not measured.

References

1. Arnau, R. C., Green, B. A., Rosen, D. H., Gleaves, D. H., & Melancon, J. G. (2003). Are Jungian preferences really categorical? An empirical investigation using taxometric analysis. Personality and Individual Differences, 34(2), 233–251. Link
2. Bess, T. L., & Harvey, R. J. (2002). Bimodal score distributions and the Myers-Briggs Type Indicator: Fact or artifact? Journal of Personality Assessment, 78(1). Link
3. 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. Link
4. Carlson, S. (1985). A double-blind test of astrology. Nature, 318, 419–425. (A later reanalysis disputes parts of the conclusion.) Link
5. Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16(3), 297–334.
6. Dunn, A. M., Heggestad, E. D., Shanock, L. R., & Theilgard, N. (2018). Intra-individual response variability as an indicator of insufficient effort responding: Comparison to other indicators and relationships with individual differences. Journal of Business and Psychology. Link
7. Hartmann, P., Reuter, M., & Nyborg, H. (2006). The relationship between date of birth and individual differences in personality and general intelligence: A large-scale study. Personality and Individual Differences, 40(7), 1349–1362. Link
8. McCrae, R. R., & Costa, P. T., Jr. (1989). Reinterpreting the Myers-Briggs Type Indicator from the perspective of the five-factor model of personality. Journal of Personality, 57(1), 17–40. Link
9. Pearson Assessments. Interpretation of MMPI-2 validity scales (VRIN and TRIN response-inconsistency scales). Link
10. Myers, I. B., McCaulley, M. H., Quenk, N. L., & Hammer, A. L. (1998). MBTI manual: A guide to the development and use of the Myers-Briggs Type Indicator (3rd ed.). CPP. (US national representative sample, N = 3,009.)
11. OpenJung. (2026). Research data and statistics API (OEJTS-based; CC BY 4.0). Retrieved 19 September 2026. Link
12. Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of personality and social psychological attitudes (pp. 17–59). Academic Press.
13. Pittenger, D. J. (1993). Measuring the MBTI … and coming up short. Journal of Career Planning and Employment, 54(1), 48–53. Link
14. Pittenger, D. J. (2005). Cautionary comments regarding the Myers-Briggs Type Indicator. Consulting Psychology Journal: Practice and Research, 57(3), 210–221. Link
15. Preston, C. C., & Colman, A. M. (2000). Optimal number of response categories in rating scales: Reliability, validity, discriminating power, and respondent preferences. Acta Psychologica, 104(1), 1–15. Link
16. Randall, K., Isaacson, M., & Ciro, C. (2017). Validity and reliability of the Myers-Briggs Personality Type Indicator: A systematic review and meta-analysis. Journal of Best Practices in Health Professions Diversity, 10(1), 1–27. Link
17. Reynierse, J. H. (2009). The case against type dynamics. Journal of Psychological Type, 69(1), 1–24. Link
18. Schaubhut, N. A., Herk, N. A., & Thompson, R. C. (2009). MBTI Form M manual supplement. CPP. (Tables 1–7, 11, 17–20.) Link
19. Soto, C. J., John, O. P., Gosling, S. D., & Potter, J. (2008). The developmental psychometrics of Big Five self-reports: Acquiescence, factor structure, coherence, and differentiation from ages 10 to 20. Journal of Personality and Social Psychology, 94(4), 718–737. Link
20. Stein, R., & Swan, A. B. (2019). Evaluating the validity of Myers-Briggs Type Indicator theory: A teaching tool and window into intuitive psychology. Social and Personality Psychology Compass, 13(2), e12434. Link
21. Weijters, B., & Baumgartner, H. (2012). Misresponse to reversed and negated items in surveys: A review. Journal of Marketing Research, 49(5), 737–747. Link
22. Cleveland, W. S., & McGill, R. (1984). Graphical perception: Theory, experimentation, and application to the development of graphical methods. Journal of the American Statistical Association, 79(387), 531–554. Link
23. Albo, Y., Lanir, J., Bak, P., & Rafaeli, S. (2016). Off the radar: Comparative evaluation of radial visualization solutions for composite indicators. IEEE Transactions on Visualization and Computer Graphics, 22(1), 569–578. Link
24. Cattell, H. E. P., & Mead, A. D. (2008). The Sixteen Personality Factor Questionnaire (16PF). In The SAGE Handbook of Personality Theory and Assessment (Vol. 2). Bipolar sten profiles, where the most distinctive traits are the ones furthest from the mean.
25. Forer, B. R. (1949). The fallacy of personal validation: A classroom demonstration of gullibility. Journal of Abnormal and Social Psychology, 44(1), 118–123.
26. Dickson, D. H., & Kelly, I. W. (1985). The "Barnum effect" in personality assessment: A review of the literature. Psychological Reports, 57(1), 367–382. Link
27. Correll, M., & Gleicher, M. (2014). Error bars considered harmful: Exploring alternate encodings for mean and error. IEEE Transactions on Visualization and Computer Graphics, 20(12), 2142–2151. Link
28. Schaubhut, N. A., & Herk, N. A. (2009). MBTI Form M Manual Supplement. CPP, Inc. Table 8 (CPI 260, N = 1,460), Table 9 (FIRO-B, N = 1,900), Table 10 (Adjective Check List, N = 185), Table 11 (Big Five via ACL), Table 12 (Strong Personal Style Scales, N = 6,000), Table 13 (Thomas-Kilmann, N = 2,600), Table 15 (type distribution, best-fit type, N = 8,836), Table 19 (preference clarity ranges). PDF
29. Nielsen, J. A., Zielinski, B. A., Ferguson, M. A., Lainhart, J. E., & Anderson, J. S. (2013). An evaluation of the left-brain vs. right-brain hypothesis with resting state functional connectivity magnetic resonance imaging. PLOS ONE, 8(8), e71275. Link
30. Gigerenzer, G., & Hoffrage, U. (1995). How to improve Bayesian reasoning without instruction: Frequency formats. Psychological Review, 102(4), 684–704.
31. Jorgenson, E. (2015). Open Extended Jungian Type Scales 1.2. Open Psychometrics (CC BY-NC-SA 4.0). Link

Figures were checked against these sources on 19 September 2026. The full working, including a verification log, is in the project's research notes.

Privacy and data waiver

• Stored in this browser only: quiz answers (so you can resume), your optional zodiac sign (never your birthday), the date you accepted the waiver, and your theme setting.
• Not collected: names, email addresses, accounts, IP-based analytics, cookies or ads. The site makes no network requests with your data.
• Share links contain your type code, your percentages and, optionally, your zodiac sign. Anyone you send a link to can see those.