The right person for the job? Weeding out personality-test fakers isn’t easy

Job seekers always want to show their best side, so personality testing can be complicated. Vermin Inc

Most people have taken an online or magazine quiz promising to reveal information about their true personality, interests, or attitudes. These tests can be harmless fun.

But there is a serious side to personality tests. They are regularly used to make decisions in job applicant selection, legal judgments of diminished responsibility, compensation for trauma-induced conditions, diagnosis of clinical disorders, special consideration in education, and even in dating.

When the stakes are high, harmless fun can turn ugly. People can lie.

We have edited a book on faking and personality tests, which reveals that about one in four people will “fake” better scores on personality tests when the results matter to them.

The book brings together 35 academic researchers from around the world, and its conclusions represent the latest understanding in who fakes, when they fake, and whether faking can be detected, corrected or prevented.

One of its conclusions is a potential bombshell for organisational psychologists who use personality tests in their work, particularly when assessing job applicants.

It shows that the “lie scale”, which is the most commonly used methods to weed out fakers, simply does not work.

Lie scales and liars

Lie scales consist of items like “I always pick up my litter” or “I never swear”, which are mixed in with standard personality test items like “I worry about things” (which measures anxiety) or “I work hard” (which measures conscientiousness).

The logic behind lie scales is that at some point everyone has littered, just as everyone has at some point uttered a swear word.

So if respondents disagree with these lie-scale statements, it is assumed that they are lying, and that they have also lied on the real personality items.

However, research shows that these lie scale items can actually measure authentic personality traits.

Some people really are so responsible that they always pick up their litter, and some are so self-controlled that they honestly never swear.

That is, lie scales will identify Ned Flanders types, as well as the fakers.

There is, therefore, a false positive rate for identifying fakers, and some of the most deserving job applicants get caught out by lie scales.

For this reason, the book makes three recommendations to recruitment professionals:

  1. The results of lie scales should not be trusted.

  2. People identified as “fakers” should never be excluded from the job, but might be offered re-testing.

  3. It is a better idea to try to reduce faking rather than identify it after it has already happened.

Reducing the prevalence of faking

There are at least two of different tricks psychologists can use to reduce the amount and extent of faking.

First, test developers might try to make items sound less obviously positive or negative.

For example, the item “I work hard” might be replaced by the item “I am a perfectionist”. Both items measure the personality trait conscientiousness, but the second one does not sound so obviously positive.

Research by Swedish researchers indicates that making personality items sound more neutral will reduce the extent of faking.

Second, psychologists could include warnings about faking as part of the test or instructions.

Warnings can focus on the likelihood of getting caught, the ethics and fairness of faking, or the ultimate cost to the test-taker themselves if they fake – for example, they might end up with a job that is not suited to them.

Research to date has indicated that warnings may reduce faking, but may not stop it altogether or for all people.

Finally, it might be wise to use personality tests purely as a means of weeding out people with very low scores, rather than selecting from the top.

Personality tests are so interwoven into our lives these days that the thought of them being open to manipulation by liars and cheats can be scary.

Selecting a bully or psychopath to manage a work team or go out on a blind date is a worrying thought.

However, methods that reject the virtuous people along with the bad are not the solution.

Instead, a better idea would be to reduce faking through warnings and new test development methods.

New Perspectives on Faking in Personality Assessment is published by Oxford University Press, and is co-edited edited by Dr Matthias Ziegler (Humboldt University), Dr Carolyn MacCann (The University of Sydney), and Dr Richard D. Roberts (The Educational Testing Service).