One of the most interesting things about analyzing the power of pre-employment testing to predict on-the-job performance is the frequently non-intuitive nature of our findings. Recruiters and staffing professionals build screening criteria using thoughtful consideration of information they can easily learn about candidates. Unfortunately, very few companies take the time and conduct the research to see if the screening criteria actually make a difference and add value to the decision-making process.

A specific example is how companies value previous job tenure. When conducting statistical analyses of how candidate tenure in past jobs relates to performance in the new job, we often find that tenure is not a good predictor of success. In fact, length of tenure is often a negative predictor, so that more experience means a lower likelihood of success in the new job.

At face value, this does not seem to make a lot of sense. So, how can this be? Well, there are lots of possible reasons, such as:

• Deep experience in one organizational culture may bring with it certain rigidities that are hard to unlearn in a new organization
• Long tenure in the same role may imply a lack of motivation to pursue advancements, or worse, inability to get promoted
• Maybe those who leave after long tenure tend to do so for more negative reasons like burnout

To be sure, tenure, when used as part of an employee selection screening criteria, may well predict performance positively for some positions – especially sales jobs. But imagine the cost to our economy of hiring personnel constantly and indiscriminately looking for applicants with characteristics that are not associated with job success, or worse – are associated with failure. Modern scientific selection tools allow you to collect better data from candidates – data which you can analyze to see how well it predicts on the job performance. To see how we’ve done this for some of our clients, check out the case studies on our website.