Why personality tests don't work

It's time we stopped basing important decisions on them.

Sign Up

Get the New Statesman\'s Morning Call email.

There are 63 types of people in the world. Those who like the the Bee Gees but dislike mayonnaise, that’s one type. Those who make little colour-coded charts detailing whose turn it is to wash up, that’s another. And then there are those who believe in personality tests. That’s a third. I’m not going to list the rest here because of space issues. Plus the one I just listed was supposed to be the punchline. Point is, let’s not make this too laboured.

Right now the personality testing industry is big business: worth $500m and growing at ten per cent every year. IBM and Oracle have just bought firms with a personality testing arm, and Deloitte is rumoured to be planning a similar move. These tests are popular with recruitment because they’re cheap (at about $30 a candidate) – and a quick way of of procuring a shortlist from a pile of job applications.  You sit down and answer questions and after about half an hour are presented with your “personality” – usually summarised in four or five traits, like “extrovert” or “intuitive”. The company can then decide if you’d be a good cultural fit.

That’s how it’s supposed to go. But the process is a little murkier. First of all, the “right” answers are usually obvious. John Rust, director of Cambridge University’s Psychometrics centre, recently told the Economist that firms usually just end up “selecting the people who know what the right answers are”. Those who know the answers – and those who are willing to bend the truth to give them to you. Perhaps valuable, but not a proper measure of personality.

Get down to the science of personality and it’s likewise murky. A large portion of funding is funneled into defining and researching what makes people the way they are, but results are often vague or conflicting. One recent paper on the genetics underlying human personality simply came up empty. A large scale search, in over 5,000 Australian adults for genes associated with persistence, reward dependence and so on, and ... nothing. No associations at all.

There is at least one person this should please – the psychologist Walter Mischel. He spent much of his life persuading other scientists that personality is not a fixed thing – it changes with circumstance. You might be lazy one minute (e.g. when there is work to be done) and full of energy and purpose the next (e.g. when the pizza arrives).

He first came to this conclusion when employed by the Peace Corps as a personality consultant. They were fed up of complaints from people who, enthusiastic do-gooders safe at home, turned out to be impossibly squeamish out in the field. He conducted test after test but found little correlation between behaviour at home and abroad. 

He embarked on a large research programme to see if he could find any stable traits at all. One of his studies was on aggression in children. He found that some could take criticism from adults, but not from other kids. Others lashed out at authority, but got on well with their peers. Eventually he had his answer: personality doesn’t really exist. It’s entirely dependent on context. And we are not the same person in every situation.

A fixed personality would certainly make life easier – you could find it out early, match yourself to the right job and the right people, and then relax, safe in the knowledge you’ll never change. It’s a shame it’s not that simple.

Martha Gill is comment editor of the Evening Standard. She tweets as @Martha_Gill

Free trial CSS