Predictive validity is a term used to describe the degree to which test results can be used to predict future work performance for newly hired individuals. Wouldn’t it be amazing if you could predict an employee’s performance before they started their new job?

A magic formula for this is still lacking, but we do know how to evaluate whether the non-magical predictive tests are doing their job.

“The Validity and Utility of Selection Methods in Personnel Psychology: Practical and Theoretical Implications of 100 Years” was the title of a meta-analysis published by Professor Franck L Schmidt in 2016.

In it, he analyzed 31 various assessment methods to identify which are of little benefit, which could harm the decision, and which assessments genuinely aid in the selection process. We were able to develop candidate selection approaches that have improved predictive validity as a result of these findings. Take a deeper look at what this lingo means:

What is Predictive Validity as it pertains to pre-employment?

In most cases, it is obtained through the correlation of various types of measured performance outputs (such as scores from technical or behavioral tests) with the criteria that the approach is evaluating (such as how well a candidate will do on that test). There is a correlation between these two variables known as r, which has a range of -1 to +1.

A correlation of +1, indicating 100 percent accuracy in performance prediction, is usually thought to be an impossibility. A correlation of -1, which means that the actual outcome is always the reverse of what was predicted, is not possible either.

Pre-employment exams and other general tests often have a score range of 0.30 to 0.70, with a minimum acceptable score of 0.30.

How does predictive validity work?

A test’s predictive validity evaluates how accurate it is at forecasting a candidate’s ability to perform on the job, as well as their ability to fit into the company’s culture. Predictive validity is frequently assessed using a scatter plot.

Overall, the graph shows test scores and work performance in relation to each other; a horizontal line represents a “ideal” test score for an individual’s ability to perform at their best. A candidate is considered successful if his or her score falls within that range or even goes higher.

The scatter plot effectively tells us the following:

• The level of difficulty of the exam
• Based on how closely one’s test scores correlate with one’s job performance, a test can either be considered acceptable or undesirable (weak positive correlation)
• Depending on the nature of the task at hand, a specific score line should be established.

What is the significance of predictive validity in the selection of candidates?

In today’s increasingly competitive business world, companies must find the appropriate people for the right jobs if they are to survive and thrive. There are critical vacancies that need to be filled, and the longer the recruitment funnel is, the more expensive it becomes.