Take care when testing staff

Paul Barrett

Psychometric tests have become hot items with HR managers, but for companies looking to gain any benefit from these tools, there is a lot to consider.

Tests include verbal, numerical and abstract reasoning exercises as well as personality questionnaires. But Paul Barrett, adjunct professor of psychometrics at the University of Auckland, says care should be taken when trying to use them purely to make improvements in the workplace.

"In terms of intellectual or cognitive performance of a certain kind, intellectual tests are very good," he says. "You can get a good measure of the reasoning ability. Whether or not that translates into job performance is not so clear."

Employers should consider many factors before deciding to use psychometric testing, including the type of job it would be used with.

"You can't just slot in an ability test where the job itself may not require very high ability," says Barrett. "In other words, the link between job performance and ability will be lost under those circumstances."

There are two broad general categories in which psychometric tests are used. Developmental use is for career counselling and training, while select/reject use is for screening job seekers. Barrett says the trend is leaning toward the developmental.

"It is tending to look toward the use for developmental more than select/reject. I think this is because there is also another aspect which I think people are more comfortable with. If you're going to start using these tests seriously for select/reject you've actually got to do a lot of groundwork on the validity issue."

If the validity can be established then that should, in some way, be communicated to the job candidates or employees who may not see the value in such an exercise.

"What kind of message does it send to your employees when you begin to use psychometric tests for hiring? Certainly within unionised industries, care has to be taken in how you begin to put this in.

"It can lead to a negative impact on your employees because of the way the tests are imposed, even for development, if there's no sense of credibility about what's being offered."

He says it is important to examine how you are going to use psychometric testing within a particular industry to maximise the opportunities for both the people working and also for the company.

"You have to be careful with psychometrics. There are ways of introducing this to at least try and show the people in an organisation that it may be beneficial in the end, that there may be some good things coming out of this."

Demonstrating any potential good from these tests for employees should be a priority when introducing them. The situation with older and more experienced workers is much more difficult than for younger workers or graduates.

"As you grow older, you actually develop a job history, you develop experience, you develop your CV. Which I think then says, 'look, here is my behaviour, this is the proof of what I can do, what I can't do and whatever'. The feeling is, 'why isn't that good enough?'."

Older workers may view the tests with increased scepticism. But younger workers have a much greater acceptance and are more likely to just take them for granted as being part of the employment process. For younger workers who have no track record, the tests may be a way of determining if they fit within the company.

If graduates have nothing to demonstrate that they would be good for a particular job, psychometric testing may be a good way of getting a better idea about them.

"If you look at the evidence with a rational view and a rational mind, you would have to be leaning toward wanting to use psychometric testing. If you can get this to work for you it will have a beneficial effect."

But if you're going to test for personality traits or attributes which you want your employees to have, you should sit down and figure out which specific attributes you are looking for and why. Which characteristics will maximise their job success within your organisation? Then it is a matter of quantifying the process at the end of the day. Barrett says you need to be specific about how you are to measure the outcome.

"How do you recognise that the initiative you've put in place has had an effect? What are you doing at the moment? What are your indicators of performance - the kinds of things you're really trying to improve with this testing.

"It's no good just saying, 'I'm going to get better staff'. What exactly does that mean?"

He says you should be able to determine a quantitative practical benefit or there is not much point in doing psychometric testing.

"You're trying to get a handle on the return on investment. It's a bit like any HR initiatives. It's easy to put in what looks to be a wonderfully enthusing and exciting initiatives, whether it's leadership stuff or development or whatever that - on evaluation - have very little impact. In other words, you've just spent a lot of money and nothing has really happened."

It can be expensive, which is why implementing a quantitative measurement of psychometric testing may not make sense for a small employer.

"It may, in some cases, cost more to figure out and to set up the evaluation process than it does to just implement a bit of psychometrics which is one reason why you only see large-scale serious psychometrics being done in big companies."

Small businesses don't have the bureaucracy of multi-level management and are in a better position to know if their employees are fitting into their roles and performing. But for anyone considering the tests, Barrett has some words of advice.

"The first port of call would be to a reputable test publisher or a company selling tests. There's probably a dozen or so who are high-level consultancies in New Zealand.

"There should be accountability for the performance of the tests and there should be a system to determine their performance.

He cautions that care should be taken before handing over large sums of money for tests, unless there is a way to quanitfy the results.

"How will you know that spending that money was a good idea?"

Barrett says psychometric testing should be more than just "psychobabble" and the resulting benefits should be measurable.

"That may go right through to shareholder value because most of these effects are going to be indirect.

"It's more of a diffusion process because you're having to work within market conditions for the company."

Like any HR initiative, psychometric testing should be well communicated to the employees so that it is not threatening and its uses are fully understood.

It should provide verifiable and quantitative results for employers and employers should be able to measure and demonstrate that their workers are well-qualified, well-suited, and high-performing employees.