Patterson said many applicants faced with an assessment will opt not to pursue an application any further. Blair said in a large company, which will often get 500 applications for a single position, tying an assessment to a position and using research to identify the top 30 percent reduces that number to 150 potential candidates, making it a more manageable group.

Blair said assessments are useful after the hire, when comparing later job performance to how an employee scored on a personality assessment for a given position. Successful employees who scored high and do well on the job mean that talent managers can predict similar success for those who score similarly in the future.

“One thing we try to do here is rather than talk about someone who performed better or didn’t, or who stayed with the company longer or didn’t, we turn it into a dollar figure,” Blair said. “Those who do better mean better revenue and longevity as well, and we don’t have to retrain them.”

He acknowledged that large companies like his — that make more than 2,500 hires a year — often have their own in-house data to use. But there is enough similarity in jobs, such as retail sales, that talent managers can pull data from a variety of companies, organizations and industries to predict what kinds of personality attributes they need to look for.

Given the current state of assessment instruments, their demonstrated accuracy and their ability to help predict performance, talent managers are wise to continue to apply them in hiring. A disciplined process supported by scientifically valid instruments is time consuming and expensive, but the alternative is far more costly.