Why Personality Tests Can Be Wrong (And When to Trust Them)

You take a personality test and it describes you perfectly. You take it again a few months later and get a different result that also describes you perfectly. Both can’t be right. Here’s what’s going on, and how to tell a test worth trusting from one that just feels good.

Two things a test has to get right

Psychologists judge a test on two things.

• Reliability: does it give the same answer twice? If you retake it in a few weeks, without having changed, you should get roughly the same result.
• Validity: does it measure what it says it does? A valid extraversion scale should relate to how sociable people actually are, as reported by friends or observed in real life.

A test can be reliable without being valid: a scale that always tells you you’re 3 kg lighter is consistent, and consistently wrong. It can’t be valid without being reliable. Most online quizzes have never been checked for either.

Why vague descriptions feel so accurate

In 1948, the psychologist Bertram Forer gave his students a personality test, then handed each of them a personal profile. On average they rated its accuracy 4.3 out of 5. Every student had received exactly the same text, assembled from horoscopes. It said things like “you have a great deal of unused capacity which you have not turned to your advantage” and “at times you are extraverted, affable, sociable, while at other times you are introverted, wary, reserved”.

This is the Forer effect, sometimes called the Barnum effect: statements vague and flattering enough to fit almost anyone feel personal. Many test results work this way. If a description would fit most of your friends too, it isn’t telling you much about you.

Why you get different results

You’re near the middle. Most people score in the middle on most traits. If a test turns scores into types, a small change in your answers can flip you into a different box. This is the biggest reason people get a different Myers-Briggs type on retest.

Your mood leaks in. If you take a test on a bad day, you’ll likely score higher on neuroticism. The effect is modest, but real.

You answer as who you’d like to be. It’s easy to tick “I’m always prepared” because you mean to be. Answer for what you actually do.

Context changes the answer. “I enjoy being the centre of attention” might be true with close friends and false at work. When a question doesn’t say which, you answer for whichever situation comes to mind.

You really have changed. Personality does shift over years, especially after big life changes. See Can Personality Change?

Which tests have science behind them?

Roughly, from strongest to weakest:

• The Big Five and HEXACO. Decades of research across cultures, stable scores, and real predictive power. The HEXACO adds a sixth trait, honesty-humility.
• Adult attachment, measured on its two scales (anxiety and avoidance). Well researched, especially in relationships.
• Conflict and communication styles. Useful frameworks for learning skills; they describe habits more than fixed traits.
• The Myers-Briggs and other 16-type tests. They measure four real traits, but forcing them into types makes results unstable, and there’s weak evidence they predict anything.
• DISC. Popular in workplaces, with little independent research, and its styles overlap with the Big Five.
• The Enneagram and love languages. Popular and often useful for reflection, but with little research support for their central claims.

Red flags

Be sceptical of a test when:

• Every result sounds good. Real personality has trade-offs. A test with no unflattering results is built to be shared, not to inform.
• It puts you in exactly one box. People are mixes. A good result shows how strongly you lean, not just which way.
• It predicts your perfect job or partner. Personality shifts the odds a little; it doesn’t decide your life.
• It won’t say where its questions came from. Legitimate tests can point to published research.
• It’s used to hire people. Most personality tests, the Myers-Briggs included, aren’t designed for selection, and their own publishers often say so.

How to use a test well

1. Treat it as a hypothesis. “The test says I’m low on conscientiousness” is a prompt to check: does that match how you actually handle deadlines?
2. Look at the scores, not just the label. A 52 and a 98 can get the same type. The number tells you more.
3. Compare frameworks. Where a Big Five result and an attachment result point the same way, you’re probably onto something.
4. Ask people who know you. Friends’ ratings of your personality often predict your behaviour as well as your own.
5. Never use it as an excuse. A result describes a tendency. It doesn’t decide what you do next.

Try the tests

Every free test on Cognosc is marked with how much the science behind it holds up, and shows you your scores rather than just a label. Start with the Big Five, the best-supported of them.