Testing Times: Online psychometric testing and age discrimination
Research has shown that correctly applied psychometrics are one of the most effective techniques for predicting future job performance. The increasing use of the internet in recruitment over the last few years has led to a demand for online tests, and test publishers have been quick to produce online versions of their paper and pencil tests. Online tests have the advantage that they less resource intensive, don't require a face-to-face meeting with the candidate and can be used at the early screening stage of the process.
However, like any tool, online psychometrics are only useful if they are the right instrument for the task and are correctly used. One crucial consideration is whether tests used for selection are free from bias. Bias exists if the testing procedure is unfair to a group of individuals who can be defined in some way. The use of a test that is biased against candidates on the basis of their race, sex, or disability is already outlawed in the UK. And from October 2006 when the Employment Equality (Age) Regulations come into force candidates will also be protected against discrimination on the basis of age.
So what about age and online psychometric testing? Surely online tests are no more or less biased against different age groups than their paper-based equivalents? Recent research carried out by Denise Potosky and Philip Bobko compared paper-and-pencil and internet administration of two timed cognitive ability tests. The research examined whether there was a relationship between age and scores on the tests. What they found was quite surprising. On one of the tests older people scored worse than younger people in the online version although they did just as well as younger people in the paper-based version of the same test. At this point we might look to our stereotypes about older people and wonder whether the older people just weren't as familiar with computers, but Potosky and Bobko found that ability with computers was unrelated to test scores. There are clear implications if this online test had been used in selection: older candidates would have scored lower overall than younger candidates regardless of ability, and would have been less likely to be invited to the next stage of the selection process. This would be bias and could lead to litigious action under the new age discrimination laws.
Of course, this is just one study and this difference was found for only one of the two tests used. However it should lead us to think carefully about how we choose psychometric tests and there are a number of key questions that need to be asked of selection tools to ensure that all candidates are treated fairly and objectively:
-
Has the psychometric tool been checked to ensure it does not contain items that could bias on the grounds of race, gender, disability or age?
-
Does the psychometric display a high degree of face validity? That is, are the items acceptable to the candidate?
-
Does the tool have good content validity? That is, does the candidate perceive it to be a fair assessment of the job itself?
-
Does the tool have good predictive validity? That is, can the organisation be confident that a high score will predict success in the role?
-
Importantly, if you are using online selection tools, are the online versions proven to be equivalent to the paper versions and have they been designed to ensure all respondent groups (particularly older respondents) have an equal and fair opportunity to succeed?
How can Human Assets help?
There are hundreds of psychometric tests on the market and it is not always easy to choose the best one for your purpose. Test publishers are keen to sell their products and without specialist knowledge of how psychometrics are designed and the statistics behind their validation and reliability checks it can be difficult to know which psychometric tests do what they are supposed to do and are fair. Having an independent opinion to help answer the questions listed above can be very helpful in making the right choice. Human Assets is a firm of qualified business psychologists and we are independent of test publishers. We can recommend the right instrument for a particular purpose and have no financial interest in recommending any test publisher or particular tests. We take a keen interest in research such as that described above and work hard to make sure that our knowledge in this area is up to date. We can help you choose the test that gives you the information you need and meets the very important requirement that it is fair and unbiased.
If you would like to find out more, please contact our consultants on +44(0)20 7434 2122 or by email at enquiries@humanassets.co.uk
Reference
Potosky, D. & Bobko, P. (2004). Selection testing via the internet: Practical considerations and exploratory empirical findings. Personnel Psychology, 57, 1003 - 1034.
Further reading
Rust, J. & Golombok, S. (1999). Modern Psychometrics: The Science of Psychological Assessment. (2nd ed.). London: Routledge.
Newsletter: October 2005