Which Personality Test Is Most Accurate? Big Five, MBTI and the Rest

For anyone whose four letters changed on a retest and who wants to know which version is real.

Your friend retook the Myers-Briggs at a work offsite and came back a different type. INFP had sat in her bio for three years, and now she's asking the group chat which one is really her.

Which personality test is most accurate? Look for one built on the Big Five traits (openness, conscientiousness, extraversion, agreeableness and neuroticism) that publishes evidence it's reliable and valid, such as the BFI-2. Myers-Briggs assigns a four-letter type, and many people get different letters when they retake it.

We make MirrorMii, a game with a just-for-fun personality quiz on its homepage. A scientific assessment makes a different promise.

What are the main types of personality tests?

The tests discussed here use either traits or types. Trait tests place you somewhere along a line for each trait, the way a ruler shows height. Type tests sort you into a box, like INFP or a Type 4.

The Big Five is a trait model. HEXACO is another, with six dimensions including honesty-humility.

A model is a way of describing personality; a questionnaire is one attempt to measure it, and each questionnaire has to earn its own evidence.

In a 2019 replication project, Christopher Soto of Colby College retested 78 published links between Big Five traits and life outcomes with large, preregistered samples. About 87 percent held up in the same direction, though usually a bit weaker than first reported.

What should you look for in a free personality test?

A reliable test gives consistent results, and a valid one measures what it says it measures. Look for a Big Five test that names its questionnaire, then check whether that questionnaire has published evidence for both.

The BFI-2 is one. It's a 60-question Big Five inventory developed by Christopher Soto at Colby College and Oliver John at UC Berkeley. Their 2017 validation studies found it reliable and valid. It's free for non-commercial research use, and the Berkeley Personality Lab's BFI-2 page links to an online version with instant feedback.

For more free questions, the IPIP is a public-domain bank of more than 3,000 personality questions that anyone can use without asking permission or paying a fee. Psychologist Lewis Goldberg launched it in 1998, and it includes ready-made Big Five scales.

Is the Myers-Briggs test real?

It measures something real, just not the way its letters suggest. In a 1989 study, Robert McCrae and Paul Costa at the National Institute on Aging compared MBTI results from 468 adults with Big Five self-reports and ratings from people who knew them.

The four MBTI scales lined up with four of the five Big Five traits, but the data didn't support 16 truly separate types. The trouble starts when an answer near the middle gets rounded into a hard letter.

Its publisher, The Myers-Briggs Company, says the assessment is designed for development and for understanding personality differences, and isn't intended for picking job candidates.

Why do you get a different result every time?

In a 1993 article, psychologist David Pittenger of Marietta College pointed out that MBTI results don't split into two neat camps; most people land between the two ends. He also cited studies where as many as half of people were sorted into a different type on a retest only five weeks later.

The surprising part is that the publisher's own figure looks similar. The Myers-Briggs Company says about half of people get the same four letters when they retake it, and about 90 percent keep at least three. That's well above what random sorting would give. When a letter flips, it's usually the one the person was least clear on.

If your answers sit near the dividing line, a small change can flip the letter.

Are fun personality quizzes worth anything?

Yes, if you know what they are. Pittenger noted that type descriptions tend to be flattering and vague enough that most people accept them as true of themselves, which psychologists call the Barnum effect.

That doesn't make a quiz useless. You read your result aloud, your best friend laughs at line three, and suddenly you're arguing about who really plans the group trip.

For personality as pure play, try our guide to Tomodachi Life personalities.

Where does MirrorMii's homepage quiz fit?

In the fun row of that table, on purpose. The quiz is hosted by Genii, a glass teardrop slime who also looks after the island in the MirrorMii game. You tap through 40 quick cards, then Genii locks in eight guesses about you, and you find out which ones missed. It's built for laughing with friends, and it isn't a validated personality test.

For play beyond a four-letter result, MirrorMii has Miia, a reflection of you who lives her own life. She writes a diary you can read but never rewrite.

FAQ

Is 16Personalities the same as Myers-Briggs?

No. 16Personalities borrows the four-letter format for convenience, but its model is reworked from the Big Five and leaves out the Jungian ideas behind Myers-Briggs. Its fifth letter, -A for Assertive or -T for Turbulent, reflects how confident you feel and how you handle stress.

Is the Enneagram scientific?

Partly. A 2021 review led by Joshua Hook at the University of North Texas looked at 104 samples and found mixed evidence. Some Enneagram measures relate to the Big Five in sensible ways. Statistical analyses usually found fewer than nine underlying dimensions, though, and ideas like wings have little research behind them.

Does a changed result prove you've changed?

Not on its own. In a 2000 review, Brent Roberts and Wendy DelVecchio at the University of Tulsa combined 152 long-term studies. People's relative positions on traits became more consistent with age, leveling off after 50, but never so consistent that change was ruled out. A different result can come from you, from the test, or from both.

There's more to you than four letters. Meet Miia. She's a reflection of you with her own friends, drama and diary. Join MirrorMii free