I’m guessing that you would not expect me to write about discontinuing the use of a pre-employment or pre-promotional assessment. But, there are instances when it is appropriate to do so.
For instance, the National Football League (NFL) has decided to stop using an intelligence test that they had been using for years to evaluate new players. I have written about the league’s use of the test before, so I won’t rehash the arguments about it here. However, its reasoning for not using it any more really comes down to:
- They did not feel it was predictive.
- It led to a poor candidate experience (which, to the NFL means bad publicity).
- And those are two very good reasons not to use a test.
Another reason to discontinue the use of a test is when knowledge, skills, abilities, or personal characteristics (KSAPs) required of a job change. At some point, administrative assistants stopped typing pages of documents, so a test of how quickly someone could manipulate a keyboard no longer made sense. Changes in customer dynamics can impact KSAPs as well. When working with a call center client, our validation data showed that personality tests that predicted performance for those taking phone calls were not effective for those who took customer inquiries via e-mail or chat. This led to a change to how the tests were scored depending on the open position.
This does not mean you should automatically drop using assessments because a job changes or has converted to WFH from an office position. However, knowing that for many people WFH is the new normal, it may be time to see if the work has really changed and the if that impacts the KSAPs. If the status quo has held, you have your answer. If there are some changes, then another validation study is likely in order.
The use of assessments, like many HR procedures, tends to take on a life of its own. Once they are in place, there is a lot inertia (we have always done it this way) keeping them there. It does not have to be that way. A good job analysis and validation study can help you modify your testing tools so that you get high value from them.