Choosing Psychometrics I. – Do I need psychometrics at all?

September 22, 2014

People assessment and talent intelligence is a $4 billion market, growing at a rate of 20% per year. It seems like everybody uses psychometric testing already: 80% of Fortune 500 and 75% of the Times Top 100 companies, UK schools, hospitals and universities all use psychometric tests.

And these are impressive numbers indeed. Yet, they do not really answer the question you are asking: Should I use them? Can they bring real added value to my organization?

digital brain 4

There are well over a thousand different kinds of psychometric tests and instruments on the market to assist you at all stages of workforce management. From selection, through engagement, to talent development and organisational planning there are several tools you can use to make more informed, more objective decisions. If you want to improve your talent management processes by starting to support your decisions with proven data-informed methods, the question is not whether you should be using psychometric tests or not, but rather: When and how to use them? What is the best tool for you?

Walking hand in hand

Picking the right tool goes hand in hand with picking the right test provider. Unless you want to develop your own tests, which would be quite a hefty investment to start with, you do not have to become a psychometrics expert yourself. However, you will need a basic understanding of how psychometrics work and knowing what to look for when selecting your partner in implementing psychometric assessment into your workforce management process. This guide was put together to help you with these challenges.

Measuring the mind – But what exactly?

Psychometrics means: measuring the mind. These tests are designed to measure complex psychological phenomena objectively. However, different test types measure different things. So, the first thing you have to ask yourself is: What is the problem you are trying to solve?

Transparency is a good indicator of quality here.

Say, you want to build a new creative team and want to predict how the members would work together. You‘d probably use a motivation test and mix and match profiles to get your A team, but may not need to back up the same decision with numerical reasoning test results. On the contrary, when selecting a new payroll officer, numerical reasoning scores can be very good indicators of candidates’ future work performance. A good tool ties closely to the problem you are trying to solve and yields relevant data to support your decision.

Am I measuring what I think I am measuring?

Another, more technical, aspect of the same question is the test’s validity. Validity shows that a test really measures what it is said to be measuring. A math test on addition, for example, is valid if it measures your ability to add up numbers, and not your reading or drawing skills.

It should be noted that there is no such thing as a 100% valid test. This is especially true when looking at complex phenomena. Even with the most basic maths test in our example, your reading skills will influence your results to some extent, because you will have to read the questions before you can start doing maths. Furthermore, there may well be other factors in play when the test taker has to use her mathematical skills, such as her motivation for example. However, in a well-designed test these effects are minimised, well understood and documented.

A trustworthy test provider has to be able to present you, amongst other technical data, with a series of validity studies. This should include information about how the test was developed, how many people it was tried on, and scientific evidence on how much you can trust the results and what they tell you. The best providers will even offer you advice and support in conducting a validity study in your own organisation. This step is always the ultimate test of the tool’s validity, as it controls for factors specific to your organisation, so you do not have to solely rely upon what another organisation has shown works in their setting.

An occupational psychologist should also be available to consult you anytime on choosing the ideal test suite for you. If something is not clear, go ahead and ask questions, look for client testimonials, case studies, try a demo test, or request a free trial if available. Transparency is a good indicator of quality here.

The second part of the guide on pitfalls to avoid and benefits to look for when applying psychometrics is coming next week. We’ll cover some technical and financial aspects to consider and discuss what the results can and cannot tell you.

*The number of test publishers is estimated to be in the thousands, but only a small number of them conduct proper validity studies. Hogan, R. (2005). In Defense of Personality Measurement: New Wine for Old Whiners. Human Performance. 18(4), 331-341.