The WRONG Questions to ask when HR is “Evaluating” Psychometric Tests
Is it a popular test?
A number of poorly developed internet tests are VERY popular but highly UNRELIABLE and not at all VALID!
Likewise, many tests on the market either have low validity overall, or lower validity than a competitor test but may have been adopted by HR for various reasons.
Tests which have been on the market for many years may have become very popular – but do you really want to use tests developed 20 or so years ago?
Can I see a list of clients who use this test?
This may break privacy and confidentiality agreements that test distributors hold with clients.
Some distributors may be happy to break confidence in order to get a sale – do you want to work with such people?
A new test will have a small client list and yet be the best test on the market!
Perhaps a very good test is rather pricey and only used by the most elite?
How long have you been selling the test?
If the test is new then the distributor may only have been selling it for a few months. This has no bearing upon the test’s validity.
Likewise, if a distributor has only just started to sell a well known, established test it has no bearing on the test’s validity.
Is it an easy test to understand?
This question may have some utility in that if the test is hard to understand, you may make selection mistakes. However, the easiest tests to understand usually are based on very simple models that may be unfair to use in predicting human workplace performance. It would be better to opt for more advanced tests that assess a broad range of human aptitudes, talents, motives and competencies and to commit to the necessary training to ensure you can use the test properly.
It’s expensive, perhaps I should go for a cheaper test?
Often, the more expensive tests are so for good reason! They have been developed well by the best in the field over a number of years and have been appropriately validated.
Always ask yourself about the ROI (return on investment). If you could pay $100 and get a 10% chance of choosing the right candidate or $1000 and have a 60% chance of getting the right candidate – which would you prefer? Most CEOs would opt for the $1000. Likewise, paying a little more for a better test will result in a higher likelihood of getting the right candidate as the test will be more reliable and valid.
Can I have a free trial?
Free trials don’t assist at all in evaluating the technical properties of a test.
Sometimes, when given free trials, clients reject a test because it is too long or does not have the questions they were expecting. This is subjective. In order to properly evaluate whether a test will predict the work competencies of importance, one needs to assess the test’s technical properties statistically.
Is it possible to “evaluate” without being trained to do so?
The clear answer is no!
To properly evaluate a test one needs to understand how to assess reliability and validity among other things. This is a difficult issue – in order to buy into a test and get trained on it you want to know it is a good test, yet you can’t fully evaluate a test until you’ve had training on how to evaluate it! You could have a look through PsyAsia’s blogs and psychometric training sites or attend a general psychometric course though.
More information
www.psyasia.com Knowledgebase
www.psychometricassessment.com/blog
www.psychometricassessment.com/training
http://dates.psyasia.com – all psychometric test training course dates in Singapore & Hong Kong as well as Online Training in Psychometrics
Hong Kong(+852) 8200 6005
Singapore (+65) 6521 3131
Malaysia (+60) 03 2782 6928
