How Accurate are Aptitude Tests?
17th July 2015 | Jim Barrett
Aptitude tests are increasingly used by organizations to select the best staff for their teams and develop those already employed. Designed to measure your competence and potential for achievement, these tests demand a serious approach. Ultimate Aptitude Tests, by Jim Barrett and Tom Barrett, will prepare you for any aptitude test eventuality.
Predicting human performance is extremely complicated, much more difficult that predicting what will happen to machines. This is why predictions based on tests, even those that are well researched, commonly fall well short of a perfect 100 per cent. This may be for two reasons:
- There may be circumstances related to the test itself, including its administration or interpretation, that undermine its predictive value
- There may be circumstances around the person, or subject, who has taken the test that alter the chance of the test being predictive, such as altered social or emotional circumstances.
Among selection devices, graphology, astrology and similar methods are no better than chance. In contrast, aptitude tests are:
- an efficient way of collecting information;
- objective, as the information is difficult to obtain by any other method; for example, where an interview question might be ‘How good is your maths?’ a test score can say precisely how good a person is;
- comparative: individuals can be compared directly with a relevant group;
- a better way of predicting success or satisfaction at various jobs than other selection devices.
It is worth remembering that tests of one kind or another are being used all the time, as we constantly judge people against our experience. Sometimes we get it right, sometimes we don’t. Tests should help us to get it right more often, though nothing can ever be asserted with complete finality.
What it is possible to do with psychometric tests is to assert a probability, for example the probability that a particular event will happen is less than one in a hundred or one in a thousand, and so on. This is the nature of both statistics and human potential, where we are dealing with probabilities, not certainties.
Therefore, we have to try to estimate how certain, or uncertain, we are when we use tests. For example, just because someone has not obtained a certain level on a test, it does not mean that they cannot succeed in a particular job. So, how likely is it that an individual might succeed or fail? If it can be shown that only 5 per cent of people who obtain a particular score are likely to be successful, then an organization may be justified in giving the job to someone who has a 95 per cent chance of being successful.
Start by testing your core intelligence on the webpage set up by the authors of Ultimate Aptitude Tests, Jim Barrett and Tom Barrett