Recenzat de Camille G., Founder of Loch Ness Paris · Ultima actualizare September 18, 2026 · 8 min read ·

Take an online MBTI-style test today. Take the same one again in a month, answering just as honestly. Research says there’s close to a coin-flip chance you’ll get a different four-letter type — not because you changed, but because of how the test is built.

That single fact — well documented, rarely acted on — is the reason we built True North around a different design principle: show the uncertainty instead of hiding it.

The reliability problem, in one number

In a widely cited review of the MBTI’s psychometric properties, researcher David Pittenger reported that between 35 and 50 percent or more of people receive a different four-letter type when they retake the standard instrument within just a few weeks. ^[2] Nothing about their personality changed in that time. What changed was which side of an arbitrary line their answers happened to fall on.

This isn’t a fringe criticism. It’s the reason personality researchers who study the trait structure of personality professionally tend to reach for the Big Five model instead of type-based instruments — the Big Five reports continuous scores that hold up far better on repeat testing.

Why forcing a binary choice breaks reliability

The mechanism is well understood. Traits like introversion-extraversion don’t come in two discrete flavors — they follow something close to a normal distribution, with most people clustered near the middle and few at the extremes. MBTI-style scoring takes that continuous scale and slices it into two categories at the midpoint.

For someone who scores near an extreme, that cutoff barely matters — their letter is stable test after test. But for the large number of people who land near the middle of a given dimension, a tiny shift in mood, wording interpretation, or context on test day is enough to tip the score to the other side of the line, flipping the letter entirely. Statistician Jacob Cohen demonstrated decades ago that converting a continuous score into a binary category this way discards a large share of the measurement’s usable information — modeling work has put the loss as high as roughly two-thirds of the shared variance with any external outcome the score is meant to predict. ^[3] A test built this way isn’t just occasionally unstable near the middle; it’s designed to be.

Ceea ce am construit în schimb

True North measures the same underlying territory — the empirically-derived Big Five domains — using the IPIP-50, a public-domain instrument developed by psychologist Lewis Goldberg. ^[1] We then translate those five continuous scores into a familiar 4-letter label, the same way most people already think about personality type.

The difference is what happens at the midpoint. Instead of silently resolving a 51% score into a confident-looking letter, we display the actual percentage and flag anything between 45% and 55% as a genuine toss-up — the same zone where independent research shows most MBTI-style instability actually comes from. A result that says “borderline” is a more honest answer than a clean box built on the same underlying uncertainty.

Two instruments, one honest report

True North offers two levels of depth, both drawn from the same public-domain research lineage:

• IPIP-50 (Goldberg, 1992) — the free screening form. Ten items per domain, five domain scores, a 4-letter type with confidence flags. ^[1]
• IPIP-NEO-120 (Johnson, 2014) — the full facet-level report. 120 items breaking each domain into its 6 underlying facets, 30 sub-scores in total, following the same structure used in the commercial NEO-PI-R. ^[4]

The full scoring methodology and design rationale behind this approach are documented in a published technical report. ^[6]

Both instruments are public domain, developed by academic psychologists, and freely documented in peer-reviewed literature — a different foundation from the many free MBTI-style clones online, most of which use small, unvalidated, in-house question sets with no published reliability data at all.

What this means for your result

A letter with a 78% score behind it and a letter with a 51% score behind it are not the same kind of information, even though a standard test would report them identically. True North’s report treats them differently on purpose. Where a dimension is clear, the test says so plainly. Where a dimension is genuinely close, it says that too — along with the reminder that Big Five research would predict exactly this kind of person to be the one most likely to get a different letter on a different day.

We also report a fifth dimension most type-based frameworks skip entirely: Emotional Stability. MBTI theory has no equivalent axis — a limitation researchers comparing MBTI dimensions to the empirically-validated Big Five structure have pointed to directly. ^[5]

Take the assessment →
Free IPIP-50 screening form, with an optional full IPIP-NEO-120 facet report.

Surse și lecturi suplimentare

1. Goldberg, L. R. (1992). The development of markers for the Big-Five factor structure. Psychological Assessment, 4(1), 26–42. https://doi.org/10.1037/1040-3590.4.1.26
2. Pittenger, D. J. (2005). Cautionary comments regarding the Myers-Briggs Type Indicator. Consulting Psychology Journal: Practice and Research, 57(3), 210–221. https://doi.org/10.1037/1065-9293.57.3.210
3. Cohen, J. (1983). The cost of dichotomization. Applied Psychological Measurement, 7(3), 249–253.
4. Johnson, J. A. (2014). Measuring thirty facets of the five factor model with a 120-item public domain inventory: Development of the IPIP-NEO-120. Journal of Research in Personality, 51, 78–89. https://doi.org/10.1016/j.jrp.2014.05.003
5. McCrae, R. R., & Costa, P. T. (1989). Reinterpreting the Myers-Briggs Type Indicator from the perspective of the five-factor model of personality. Journal of Personality, 57(1), 17–40.

Disclosure: True North is a free tool built by Loch Ness Paris. Unlocking the full IPIP-50 result requires an email address, used only to send that result. The IPIP-NEO-120 report is a paid, one-time purchase — see our notificare legală pentru detalii.