Why Your MBTI Type Changes But Your OCEAN Profile Doesn't

You took the test in 2019 and came out an INFJ, then took it again last spring in a slightly different mood and came out INTJ. Nothing important about you changed in between, so either the test caught you becoming a different person, or it never measured you the way you assumed. It's the second one, and the reason is a single measurement mistake baked into the format, not a failure of self-knowledge on your part.

Studies that retest people on the MBTI a few weeks apart find that something like a third to a half come back with at least one letter flipped, which means a different four-letter type. Your OCEAN scores over the same interval barely move. One person, two instruments, wildly different stability. That gap is worth understanding, because it explains why one framework feels like a horoscope that keeps rewriting itself and the other feels like a ruler.

The coin on its edge

Most human traits are distributed like height: a big smooth bell curve with the majority of people clustered in the middle, not at the extremes. The MBTI takes a trait like that and draws a line down the center, calling everyone on one side "E" and everyone on the other "I." For the rare person way out at either end, the label is stable. But most people sit near the middle, and a person near the middle is a coin balanced on its edge. A slightly different day, a couple of ambiguous questions answered the other way, and it tips to the opposite letter. Nothing about them changed; the line they happened to be standing on didn't move, they just fell off the other side of it this time.

Multiply that by four independent dichotomies and the arithmetic gets brutal. If you sit near the midpoint on even two of the four axes, which is completely ordinary, your "type" is essentially a coin flipped twice, and it will keep landing differently on retest for reasons that have nothing to do with who you are.

Why OCEAN doesn't do this

The Big Five refuses to draw the line. Instead of sorting you into introvert or extrovert, it reports where you actually fall on the Extraversion dimension, as a percentile. If you land at the 52nd percentile, that's what it says, and next spring it will say 51 or 53 rather than flipping you into a different category, because there is no category to flip into. Measuring your position instead of which side of the line you're on is the whole difference, and it's why test-retest reliability for the Big Five stays consistently high while the MBTI's four-letter stability is poor.

This is also why the Big Five carries information the MBTI throws away. "I" tells you nothing about how introverted; a person one step past the midline and a genuine hermit get the identical letter. The percentile keeps the magnitude the dichotomy deletes, and magnitude is where almost all the useful prediction lives.

The 16 boxes hide the interesting parts

There's a deeper cost than instability. Cutting each dimension in two, then collapsing several separate traits into one letter, loses the internal structure that actually describes a person. The MBTI's single "T/F" split, for instance, blends together how much you feel other people's pain with how bluntly you tell the truth and how readily you back down in a conflict. Two "F" people can be opposite across those and still share the letter. The 16-box system can't see the difference, while a 30-facet reading puts each on its own dial, which is the whole argument of the MBTI-to-OCEAN translation and, more bluntly, the case that typology behaves like astrology.

Why the boxes feel so good anyway

None of this explains the MBTI's staying power, and the explanation isn't that people are foolish. A four-letter type is a tiny, shareable identity, and "INFJ" fits on a dating profile in a way "Extraversion 34th percentile, Agreeableness 71st" never will. The types also read as flattering, since each description is written as a set of strengths, and a compact affirming label you can trade with friends is a genuinely nice thing to have. The catch is that comfort and accuracy are separate products, and the retest problem is the receipt: a measurement that changes when you didn't is measuring the instrument rather than the person. Why an appealing but empty description feels so personally true is a documented effect, covered in the confirmation bias breakdown.

The practical upshot

If you've watched your own type wander over the years and quietly wondered which one is the real you, the honest answer is that the question is malformed. You were never one of two things on those axes; you were always a specific point on each, and the point has been fairly stable the whole time. The wandering was the format converting steadiness into noise. The 30-facet OCEAN personality test reports those points, in about 15 minutes, with domain results free, and if you retake it next year the numbers will look reassuringly like this year's, which is the entire idea. A reading you can repeat and trust is worth more than a label that reshuffles with your mood, however good the label looks on the shelf.