Some online personality tests are useful and some are entertainment, and the two can look identical. The honest answer to “how accurate are they?” is that it depends on three things: the model behind the test, how the test was built, and what you plan to do with the result.
What “Accurate” Means For A Personality Test
A bathroom scale can be checked against a known weight. Personality has no such reference, so researchers ask two narrower questions instead.
- Reliability: do you get a similar result if you take the test again a few weeks later?
- Validity: does the score relate to things it should relate to in real life, such as how people who know you describe you?
A test can be reliable without being valid. A quiz that always tells you the same thing about your favourite colour is consistent, but that does not make it a measure of your character.
Why Almost Any Result Can Feel Right
In 1949 the psychologist Bertram Forer gave his students a “personal” description, then asked how well it fit them. Everyone had received the same text, assembled largely from a newsstand astrology book, and the class still rated it as highly accurate. This is now called the Forer or Barnum effect.
So a result feeling true is weak evidence. Statements such as “you can be outgoing, but you also need time alone” fit nearly everyone. A better sign is a result that could have been wrong: one that says you score low on something, and that a friend might recognise as well.
What The Evidence Supports
Trait-based models, especially the Big Five, have the strongest research base. A large review by Roberts and colleagues found that personality traits predicted mortality, divorce and occupational attainment about as well as socioeconomic status and cognitive ability did. That is about averages across many people, not a forecast of your own life.
Type-based models are more popular and less well supported. (We look at the reasons for that popularity in why MBTI is so popular.) Pittenger’s review of the Myers-Briggs Type Indicator raises concerns about how consistently people land in the same type on retest, and about splitting smooth scales into two boxes. Someone just either side of a cut-off is labelled as a different type, even though their scores are almost identical. That is why our 16 personality types test shows a score for each dimension as well as the four letters. We set out which parts of the type model hold up in is MBTI scientifically valid?.
Personality also moves. A meta-analysis of long-term studies found that average trait levels keep changing through adulthood, most of all in young adulthood. A result is a snapshot of how you see yourself now.
Four Checks You Can Run On Any Test
| Check | Good sign | Warning sign |
|---|---|---|
| Is the model named? | It says what it measures and where the idea comes from | “Our unique algorithm” with no detail |
| Is the scoring shown? | You can see how answers become a result | Only a label, with no scores |
| Could the result be unflattering? | Low scores and trade-offs appear | Everyone gets a compliment |
| Are the limits stated? | It says it is not a diagnosis | It promises to reveal your true self |
You can also run a fifth check yourself: take the test twice, a few weeks apart, and ask someone who knows you well whether the result sounds like you.
How To Use Your Result
- Treat it as a starting point. Use it to ask better questions about yourself, not to settle them.
- Read the scores, not just the label. Being near the middle of a scale is a real and common result.
- Do not use it to rule things out. A test should not decide your career, or whether a relationship can work.
- Keep fun quizzes as fun. A character quiz is entertainment, and that is fine as long as it is labelled that way.
- Talk to a professional about health concerns. If a result touches on your mental health, a qualified person can assess what an online test cannot.
How PrismaTest Handles This
Each of our tests names its model, explains the scoring and lists every possible result on the page. Fun quizzes are clearly labelled as fun. You can read the details in how we build and review our tests.
Frequently Asked Questions
Are free online personality tests less accurate than paid ones?
Price tells you little. Some strong measures, such as items from the International Personality Item Pool, are free to use, while some expensive tests have thin evidence. Check the model and the scoring instead of the price.
Why did I get a different type when I took the test again?
Your mood and how you read the questions both vary, and scores near a cut-off can tip either way. If your scores barely moved, the change in label means less than it looks. We worked through the numbers in why you might get a different type on a retest.
Can a personality test diagnose a mental health condition?
No. A self-report test cannot diagnose anything. Only a qualified professional can make that assessment, after looking at far more than your answers to a questionnaire.