Why Personality Tests Feel So Accurate

Bertram Forer was a psychologist in Los Angeles who also taught a psychology class. In 1948 he gave his thirty-nine students a personality test, collected their answers, and a week later handed each of them a sheet with their individual profile on it. He asked them to rate how well it fitted, from zero to five.

The average rating was 4.26. Most of the class gave it a four or a five.

Then he told them that every single profile was identical. He had assembled it out of lines from a newsstand astrology book.

What the profile said

Here are some of the sentences Forer used. Read them slowly, and notice what happens.

You have a great need for other people to like and admire you. You have a tendency to be critical of yourself. You have a great deal of unused capacity which you have not turned to your advantage. While you have some personality weaknesses, you are generally able to compensate for them. At times you have serious doubts as to whether you have made the right decision or done the right thing.

Almost everybody reads that and recognises themselves. That recognition is the effect, and it has carried Forer's name ever since - along with that of the showman P. T. Barnum, after the psychologist Paul Meehl coined the term "Barnum effect" in 1956.

Why it works

Three things are happening at once, and they reinforce each other.

The first is that the statements are double-headed. Each one contains a claim and its opposite, arranged so that whichever is true of you is the part you notice. "At times you are outgoing, while at other times you are reserved" cannot be wrong about anybody, because everybody has been both.

The second is that the statements are flattering in a specific direction. Nobody wants to be told they are boring or cruel. They are quite happy to be told they are self-critical, that they have untapped potential, that they think more deeply than people realise. Forer's profile is essentially a list of compliments phrased as insight, and we are considerably less rigorous about checking things we would like to be true.

The third, and the one that does most of the work, is active participation. You are not reading the profile neutrally. You are searching your own memory for evidence, and a memory of forty years contains evidence for almost anything. Told you avoid conflict, you find the three times you did. Told you confront it head-on, you find the three times you did that instead. The profile is a prompt, and you are supplying the content.

Where this leaves personality tests

It would be easy to conclude that all of this is nonsense. That is too strong, and the more interesting answer is in the middle.

The Barnum effect explains why a bad test can feel excellent. It does not mean a good test is impossible. What it means is that feeling accurate is not evidence of being accurate - the two are separate properties, and the first one is much easier to produce.

This is why psychologists do not judge an instrument by whether people like their results. They judge it on two boring technical questions. Does it give you roughly the same answer if you take it twice a month apart? That is reliability. Does it predict anything outside itself - job performance, relationship satisfaction, how somebody behaves in a laboratory task? That is validity.

Measured that way, the picture separates fast. The Big Five does well on both. Type-based systems that sort you into one of sixteen boxes do considerably worse on the first one: in some studies around half of people get a different type when they retake the test five weeks later, which is a problem if the type is supposed to be who you are.

How to read your own results

There is a simple test you can run on any profile, including ours.

• Could the opposite have been written? "You are energised by other people" is a real claim, because "you find groups draining" is a different one. "You value both connection and independence" is not a claim at all.
• Does it say anything you would not want to hear? A result that is entirely complimentary is marketing.
• Would it be wrong about somebody? If you cannot think of a person it does not describe, it is not describing you either.
• Does it predict anything? A useful result should let you guess how you will behave in a situation you have not been in yet.

The honest position

We run a site full of personality tests, so this is worth saying plainly: some of ours are built on published psychological research and some are for fun, and we label which is which on every page. The fun ones are subject to everything described above and we do not pretend otherwise.

Even for the science-based ones, the right way to hold a result is loosely. A short test is a mirror with a slightly uneven surface. It can show you something you had not looked at directly, which is genuinely useful. It cannot tell you who you are, and any test that claims otherwise is selling you Forer's paragraph with better design.