When repeatedly testing abilities such as memory, processing speed and reaction time, we recommend using parallel or adaptive tests.
For many people, applying for a job is a stressful, strenuous event. Consequently, the job search often involves more trial and error than error. Writing cover letters, the long wait for a response, the fear of a negative answer… This is why it is important that the selection process for candidates is as efficient and enjoyable as possible each time. As a consultant, how can you answer questions such as Do I have to take the same test again? or Can’t my old test results be used?
Repeated testing can bias test performance in several ways. For example, for some skills tests, there is a chance that performance on the second test-taking is better than on the first test because candidates know how to approach the test better or because they can remember certain questions.
Also, candidates may have trained themselves to solve the types of questions asked in the test between the two takeings. This may result in a “learning effect” or even: a better score if the same test has to be taken again.
Another possible problem we see in questionnaires that gauge personality or motivations. If such a questionnaire is offered in a job application context, where the stakes are high for the participants, there is a good chance that the answers will be influenced by the image a person wants to paint of himself.
Thus, our advice around retesting candidates should be formulated according to the type of test being administered. The advice below is based on our own extensive empirical database of data around retesting candidates.
Cebir researchers’ advice for retesting soft skills
This research shows that personality, attitude and value tests have little sensitivity to training and learning effects. Test scores for that type of questionnaire are thus fairly stable over time, especially when both test-taking occasions occur within the two years and the job for which a person is applying remains similar in content and/or hierarchy. A new test is therefore not necessary as long as the existing test is less than two years old. When both test moments do lie more than two years apart, the chance of changes as a result of (positive or negative) life experiences increases. It is also best to have a soft skills questionnaire completed again when there are major content and/or hierarchical differences in the positions for which someone is applying.
Cebir researchers’ advice for retesting hard skills
Next, we looked at the influence of retesting on scores in proficiency tests. This showed that results from skill tests for reasoning, tests for visual information processing, and knowledge tests were little sensitive to learning effects. In contrast, for memory tests and tests on processing speed and reaction speed, a learning effect may occur temporarily (one month after the first test-taking) when retested. This can result both from changed test conditions (less alert candidate, quieter test environment …) and from better task comprehension (as a result of previous experience with this test). Therefore, when repeatedly measuring these skills, we recommend using parallel tests. Another solution is the use of adaptive tests, as the content adapts to the candidate’s level during test-taking. Candidates are thus never given exactly the same questions in adaptive tests. This reduces the risk of learning or training effects within a selection or coaching context.
Would you like to read more on this topic? As a customer, you can download this and other articles.