Personality Testing at Work

It has now become a bit of an annual event. Journalists declare open season on management psychologists and newspapers carry a repeat or reheat article on psychometric testing at work. And the best stories are of scandalous incompetence, injustice or greed in which ideal candidates are rejected; test questions exposed as ludicrous and, worse, there is evidence of illegal or immoral discrimination.

So the story is one of evil psychologists, naïve managers and badly treated applicants.

A good story or a fairy tale? A common occurrence or a rare event?

Recruiters, selectors and trainers ‘discovered’ psychometric tests about a quarter of a century ago although they are at least 100 years old.  Publishers and management consultants pushed, peddled and praised the tests so they soon became popular. There remained cynics, sceptics and traditionalists who never trusted them and rejoiced in their publicized failures.

But this periodic coverage is really very educative for those who know little about the subject. So here are 7 didactic questions with answers:

1. Why do people use psychometric tests in recruitment?

The answer is simple: to help make better people decisions.  People are complicated, ambiguous, capricious. They may be hard to read. They may attempt to conceal a lot. Recruiters are trying to gauge many different things: creativity, handling pressure, integrity, punctuality, team-working.  They need reliable data to help selection. Tests can provide that.

Of course some may use psychometric test because they are fashionable or they want to appear up-to-date or scientific. They may be persuaded by consultants without really understanding the tests..

2. Are tests cost-effective?

It depends, of course, on their accuracy and the ability (and courage) of selectors to apply their findings. The easiest way to think about cost-effectiveness is the cost of getting it wrong. Ever tried to get rid of a well-dug-in, incompetent staff member who was a bad selection decision right from the start? Ever seen someone you turned down years before now running a successful competitor company? If you see tests as a sensible prophylactic they can seem very cost effective.

3. Should tests be used to select in or select out?

Most recruitment (should) start with a job analysis followed by a (parsimonious, rank ordered) list of attributes required (both necessary and sufficient) for a competent, indeed excellent job holder.  Then through interviews and tests you look for evidence of these “competencies”. But what about looking for things you don’t want: arrogance, dishonesty, hypochondriasis? Tests may help considerably with the dark-side stuff that frequently derails people.

4. Is lying or faking on these tests frequent, easy and really a problem?

All people fake in interviews. They commit sins of omission and commission. They do self-presentation, impression-management.  They do the same with their CVs. And they can fake in tests. But if they all gave the “obvious” and desirable answer there would be two consequences.  First they would all give the same answer (which they patently do not). Second there would be no evidence of test validity, which there is. There are many ways to catch dissimulation (as psychometricians politely call dissembling) lie scales, forced choice (ipsative measure) using liar profiles. People are much more prone to lie in interviews, the most popular method of selection.

5. How do clients choose between tests?

There are well over 10,000 tests available yet the average (HR) manager can’t name more than ten.  Obviously they know what they are marketed by consultants or test publishers. Test usage is then a function of marketing not necessarily validity. Test peddlers of both valid and invalid tests know that clients do not know what questions to ask. People need a “Which Guide” with little stars on different criteria set by honest, disinterested experts. Why haven’t they done this? Litigation by rich test publishers? Little reward for them? Caveat emptor. So don’t blame the product if you don’t know how to choose it. You need to understand (at least a little about) psychometric qualities such reliability  validity and process and how to assess them.

6. How important is personality at work anyway?

A person’s work success at any level is a function of a number of things, but five are clear: their ability, their motivation, their personality, their colleagues and the organisation’s processes and procedure. You need to be bright enough for the job and motivated to do it (well). You need to have a functional ship-shape, well managed organisations. No “ideal” personality profile can compensate if the other features are missing.  So it’s as dangerous to believe personality is all important as to believe it is not at all important.

7. Does personality change over time?

Alas not much? Go to a school reunion for evidence. Most personality and ability characteristics are hard wired. We have data on people measured 50 to 70 years apart. There is always more evidence of continuity than change, of stability than variability, of consistency than inconsistency. Trauma, training and therapy can change people. But by the mid-twenties what you see is what you get. Introverts at 10 are introverts at 90 though they may have learnt to fake extraversion.

Of course personality is important at work. Of course there are more or less desirable profiles for particular jobs. The question remains: how you choose to find out about an applicant’s personality?

Adrian Furnham is Professor of Psychology at University College London and author or Personality and Intelligence at Work pubished by Psychology Press (2008)