The market in psychometric testing is huge. But where is the proof that a personality test really works in weeding out candidates?

I’m sure that many of us have participated in a personality test like the Myers Briggs Type Indicator (MBTI) in the past and subsequently labelled ourselves as ISTJ (Introverted Sensing Thinking Judging) or some other combination of personality types. 

MBTI was developed in 1921 by Katherine Cook Briggs and her daughter, Isabel Briggs Myers, who were drawn to the writings of Carl Jung. However, neither employed research to develop or test their concepts, relying instead on their observations, anecdotes and intuition. So, although you might feel attached to your MBTI label, it might have no valid empirical value.

In Australia, around 40 per cent of recruiters and employers still ask candidates or employees to participate in psychometric or personality testing and profiling.

Recruitment firms offer up a suite of aptitude tests including verbal reasoning, comprehension and grammar, spatial reasoning, information processing, problem solving and IQ.  They are based on the premise that organisations can gain greater confidence in knowing that a candidate will be skilled appropriately or a good fit for the organisation.

There is a sizeable market being generated around personality tests: A conservative estimate in the US puts the value at around $500m per annum. It’s not unreasonable to ask, therefore, does a personality test actually work?

Research would suggest not. Being generous, there is at best a tenuous link between the personality test and the competency being assessed, as it will not determine how the person will operate in their role in the organisation.

At times, the test scorers are not qualified in the science of psychology, and rarely are copies of the test results shared with the candidate, which does not lend reassurance that this is a transparent system. If the tests are valid there should be nothing to hide, so why not share those results?

Ninety-five per cent of all psychometric tests are created and tested on Anglo-Saxon people. I think we all know what that means in terms of skewing the results. Psychometric tests can kill your ability to create a diverse workforce and a healthy culture. If attempts are made to homogenise recruitment to get the right ‘fit’ and create a generic standard, then you will get what you desire – a generic organisation. That goal might also kill creativity and innovation, which are highly valued characteristics in an increasingly competitive business world.

The killer for me is one profile test offered by a company “designed to identify people who are most and least likely to engage in unethical and illegal behaviour within organisations”. The test supposedly identifies integrity, honesty, poor impulse control, stress tolerance and conscientiousness – and says that it will safeguard your workplace against fraud and misconduct.

No one profile test can determine on any day whether a person will engage in or be susceptible to any of those conditions. There is no absolute proof.

Despite the tenuous links between intent and outcome of these tests, the lack of proof and flaws in practice, candidates are screened out of opportunities on the basis of the results.

Of course, many organisations use these as just one part of a process of selection, but how sure can we be that the test results are the not the final deciding factor?

In several countries, an increasing number of civil rights groups and agencies look at instances of discrimination and bias that arise from applying a personality test of this type. By applying them to people who don’t conform, parameters are set that inevitably rule them out. We are not yet seeing the same trend in Australia, but I think we will.

If tests are to be used as a final decider in any process, we have to be sure that empirically and scientifically they are sound and valid.

Professor Petrina Coventry (FCPHR) is a speaker at AHRI’s National Convention 3-5 August 2016 in Brisbane. To check event details and register, click here.