I once knew a maintenance mechanic who had a completely organized toolbox. Literally blindfolded, he could reach into the box and find the right tool for the job. I wish we could all be that organized. I thought about the tools that are available to us for selecting people for a job and how to tell the “right” tools from the “wrong” ones.
PREDICTORS IN SELECTION
The term we use for selection tools is predictors and there are a lot of them. For example, there are cognitive ability tests, job tryouts, biographical inventories, structured board interviews, reference checks, grade point averages, job-related experience, one-on-one interviews, ratings of training, years of education, and preferences. Jack Hunter and Frank Schmidt studied all these predictors in a landmark study and what they found was very interesting.
Hunter and Schmidt studied the predictors below on the basis of their validities -- the strength of the relationship between scores on the predictors and measures of job performance.
- A perfect relationship has a validity of 1.0.
- A random relationship has validity of 0.0.
In terms of the selection tools studied, those with a validity of 1.0 would be correct 100% of the time and would have a “hit rate” of 100%. Those with a validity of 0.0 would be correct half or 50% of the time. Here’s what the researchers found:
THE STRONG PREDICTORS
Please note that the most powerful tool we have at our disposal is a battery of cognitive ability tests. Its expected validity is .53 -- far from perfect, but the most powerful tool we have! (A perfect scheme probably does not exist.) Another powerful tool is the 15dots™ Structured Board Interview Procedure at an expected value of .35. Job tryouts are also rated high, but are unrealistic in situations where a “failure” is costly in terms of money or danger. Biographical Inventories are also reasonably good, but they are very industry-specific.
THE WEAK PREDICTORS
As attention-grabbing as it is to see the most powerful predictors and their validities, it’s even more eye-catching to see how low the validities are for some of the most popular predictors, namely assessments of experience, one-on-one interviews, ratings of training, interest, and number of years of education. All of these predictors enjoy widespread use even in the face of research which shows that they are relatively weak in predicting future job performance. (The question of why organizations persistent in the use of these weak predictors as selection tools is the subject of another blog. Hang in there with me!)
I just described the right selection tools to find highly motivated, highly able people. Next time, my partner Mike Quinn is going to focus on why organizations should invest considerable effort in a rigorous employee selection process. The point is that people are an organization’s greatest asset.