With the ever-growing need for quick turn-around times on psychometric testing, online and unsupervised assessments are rapidly increasing in use. With the capacity to send assessments out to candidates at 4pm on Thursday, and have all of their results sitting waiting for you when you come in at 8am on Friday, the appeal of unsupervised testing is obvious, particularly when bulk numbers of candidates are involved. But speed isn’t everything. What could we be sacrificing by allowing candidates to complete their tests from home?
Unsupervised testing provides candidates with the opportunity to cheat. Whether this be at a smaller scale (e.g., using a calculator) or a greater scale (e.g., getting someone else to sit the assessment for them), the risk is there. People naturally want to put their best foot forward in a selection process, particularly if they really want the job. However, we also like to think that people have sufficient levels of integrity not to engage in cheating behaviour. But can we rely on this? If you were completing testing for a job and were able to do it from home, would you be tempted to engage in cheating of some form?
We need to accept that the risk is there. So, how do we maintain the advantages of unsupervised testing, without compromising on the validity of the results? There are a number of options for minimising the risk, including:
- Using adaptive testing – so that different test questions are administered to different candidates, thereby minimising the risk of questions being leaked or shared between individuals.
- Verifying test results – through either re-testing candidates in a supervised environment (where possible) and/or with information gathered from other parts of the selection process (e.g., referee checks).
- Engaging in some form of honesty contract with candidates.
In terms of point one, adaptive testing is certainly feasible, and is becoming more accessible as test providers develop further assessments using adaptive technology. In terms of point two, while asking candidates to re-sit their assessment in a supervised environment will verify their results, in many ways it reduces the advantages of unsupervised testing in the first place. On the other hand, using interviews, referee checks, and information from a candidate’s CV (such as educational achievement) to verify the accuracy of test results is more easily achieved. Regardless of how assessments are administered, incorporating and verifying test results with information from the wider selection process should be a standard part of the process. However, this method still doesn’t provide us with certainty that results are accurate and not the outcome of cheating or falsification. As for point three, asking candidates to be honest, or suggesting the possibility of a re-sit in a supervised environment, is in itself likely to be a strong de-motivator for engaging in cheating behaviour.
So, following these steps should allow us to enjoy the benefits of online and unsupervised testing, while minimising the risk of candidate’s falsifying their results. While these methods are likely to reduce the likelihood of cheating, the question of ruling the risk out all together remains unanswered.