Psychometric tests… How to choose

Psychometric tests are a popular tool to help training professionals identify the learning and development needs of staff. Used effectively and appropriately, they can offer valuable insights into subjects' personalities, traits and even skills. But how do you decide which test is appropriate?
There are dozens of tests all claiming to offer fresh insights, so the first step is to decide on the ones you need.

Caroline Dunk, principle of Leeds-based HR consultancy CDA, says an effective test should cover one or more of three key areas: personality; emotional intelligence; and team type. "To decide which test is most appropriate, you need to work out the answers you are looking for," she says.

She recommends tests that subjects feel comfortable sharing the results of, so that they can compare them with their colleagues. "Tests such as Myers Briggs are great for learning and development needs as they value different behaviour."

Different insights

Other tests that Dunk says are appropriate in an occupational setting are OPQ, a personality test that has been in use for more than 20 years, and her own organisation's CDAQ. "This is a questionnaire based on neuro-linguistic programming, which analyses how we handle data and interact with others. It can offer some really different insights," she claims.

Another test recommended by Dunk recommends is the NEO personality test, which is similar to Myers Briggs inasmuch as it measures personality traits. "I believe this could become a very significant test for the future," she says.

Belbin, which assesses given traits and preferences in a team setting, is also in Dunk's top five.

Recommendations are certainly one way to sort the wheat from the chaff. But there are more formal ways to find out if you are getting a test that provides reliable and valid results.

George Sik is an occupational psychologist at Norfolk-based Eras, which develops psychometric tests. He says that all test providers should provide evidence that the scores obtained are consistent and precise.

Dunk agrees that the easiest way to check validity is to look at the statistics. "All reputable tests come with a set of statistics, which should show consistency of results displayed in technical tables," she says.

Sik says that employers should be wary of providers that promise to offer a one-size-fits-all solution. "Validity depends on context. Beware of any suppliers who say their tests are valid and leave it at that. Valid for what exactly?" he asks.

Keep it simple

Catherine Hick, managing director of York-based test provider TMS Development International, says the best profiling tools are reliable in two ways.

"First, they have good face validity - people find them easy to understand and quickly see a benefit in the knowledge they gain. And second, the tool has been subjected to rigorous research, which is available for you to see, clearly demonstrating that the internal structure of the questionnaires is reliable and that the questions measure what they are supposed to measure."

Case studies can provide more evidence. Hick says providers should let you see how clients use their tools. "They can also put you in touch with people who are willing to talk about how the tools have been used," she says.

One clear indicator of reliability is the British Psychology Society's (BPS) stamp of approval. In October 2005, the BPS psychological testing centre launched 'test registration', which is based on a good practice model. It provides quality marks on those tests that meet the relevant criteria. The BPS's psychometrics website (www.psychtesting.org.uk) lists tests that carry its kitemark, with reviews.

Dunk says that 'test fatigue' has been known to skew results so try to focus on newer tests. "There are a relatively small number of personality profiles so candidates can get too familiar with them. Tests need to energise and offer new insight if they are to be effective."

Finally, keep your eyes open to marketing ploys so you are not fooled by a glossy website. "Employers can be seduced by marketing literature and don't actually look at the reliability and validity data," says Dunk.

Online psychometric tests - can you trust them?

Online tests are now common - in fact, most tests are now available in an online format, helping to cut administration costs. Like all tests, reliability can be checked by investigating the reputation of the supplier and the evidence they provide with the test. This should show scores are consistent.

But George Sik, occupational psychologist at Eras, says there is a danger with online tests that applicants could get someone else to complete them.

"Despite sophisticated systems by some test suppliers to randomise the questions, one can never be sure the person who did the test is the one genuinely going through the development process," he says.

To overcome this, he advises employers to talk to staff about their results, so they can discuss - and possibly defend - the content.