Running the rule over psychometric tests

The last decade has seen some disastrous bank bosses, but Paul Flowers stands out. After bringing his Co‐​operative Bank to its knees through an ill‐​considered merger, he was found guilty of possession of Class A drugs and later banned from the finance industry for life for using his work phone to call premium‐​rate sex lines.

Investigating the fiasco, a committee of MPs in 2014 asked one of the bank’s deputy chairmen how Mr Flowers had been appointed chairman despite having little knowledge of finance. The reply: “I was told afterwards that he had done very well in the psychometric tests.”

It’s not a great advertisement for the use of psychometric testing to cut business risk. Nevertheless, more firms are trying to use this kind of testing not only to identify talent but also to ensure that they hire employees who reduce – rather than raise – their risk profile.

In some cases – such as mining or construction staff who may be working unsupervised with dangerous equipment for long periods – specialist firms have been set up to try to assess candidates for risky behaviour. In other cases, tests may be part of a broader programme of risk reduction. Some firms that employ large numbers of drivers, for instance, use psychometric testing alongside more traditional assessments of vehicle control skills and report reduced accidents per mile as a result.

And, as Mr Flowers’ case shows, they may even be used for the highest roles in a company.

“Testing is applicable to all levels nowadays and while common among students, the sheer volume of candidates has made it the norm in everything from contact centres to shop floor retail staff too,” says James Hollingsworth, head of marketing at Saville Assessment, a specialist in psychometric assessments.

However, he cautions: “While profiling work is growing among C‐​suite candidates, nobody would pretend that psychometric testing alone will signal that somebody is a potential Harvey Weinstein.”

One reason for the increase is because it appears objective. MPs were told that the Co‐​operative Group had decided to rely heavily on testing because of guidance from the Financial Conduct Authority. Ironically, this was meant to reduce risk, as the regulator wanted to avoid the boardroom becoming a club for the chairman’s friends and falling into groupthink.

One issue with using tests to avoid hiring risky employees is that academics are still debating the nature of risk. Is a preference for risk a stable psychological trait that can be tested? Or is it situation‐​specific, in which case a job candidate could engage in a wide range of risky behaviour yet still, in the right company culture, be an excellent, risk‐​averse employee?

To find out, academics from the University of Basel and the Max Planck Institute in Berlin bombarded more than 1,500 test subjects with tests over six months. In a paper published in 2017, they came back with the answer: it’s about 50/​50. In other words, in any given situation about half of a person’s attitude to risk comes from a deep‐​seated psychological trait (“a general, stable component”, as described by the researchers), and around half from the specifics of the situation.

This is an important validation of the use of psychometric testing to reduce risk, although caveated by the fact that an employee’s behaviour at any given moment can still be shaped by other factors.

But could a high‐​risk employee simply game the tests to appear safe? MPs suggested that the disastrous Mr Flowers had perhaps done this. However, experts in the field say that presenting a false image in psychometric tests is not as easy as it sounds.

“With randomised sampling and a bank of more than 1,000 often intuitive questions, it is becoming very difficult to either rehearse for a test or mask the person you really are, particularly when the key questions are often repeated in a number of different ways,” says Matthew Davis, director at talent consultants Occupational Psychology Group. “Every day, I see test results from candidates who have attempted to outwit us by being too clever and in all cases, they crash and burn quite spectacularly, possibly because they have opted to do the assessment on a bus, rather than in a quiet room.”

Mr Davis says testing is on the rise, and that even those who are not specifically testing for risky attitudes could see gains in business resiliency.

“Lack of organisational fit around teamwork or openness to new ideas can be as much a de‐​railer to future business performance as deliberate corner‐​cutting and can cost a firm dear in terms of lost revenue and reputational damage,” he says. “But as we always tell our clients, relying purely on algorithms to make your hiring decisions is unwise.”