Strong decrease in predictive validity of IQ tests for work performance

As a psychology student in the 1980s, I learned that researchers Schmidt and Hunter indicated that IQ tests were strongly predictive of job performance, with correlations ranging from 0.51 to 0.65. However, recent research by Steel and Fariborzi (2025) suggests that this correlation may have decreased to as low as 0.16.

This decrease has two causes:

1. Statistical overcorrections in previous studies: Researchers applied corrections to compensate for “range restriction” – the fact that hired workers often have a narrower range of intelligence scores than all applicants. They often used national standards instead of data from actual applicants. National standards reflect a broader spread of intelligence than you actually see among job candidates, especially these days. This led to an overestimation of the relationship between intelligence and work performance. Recent research shows that these corrections are now often unnecessary, because applicant groups have become more homogeneous.
2. Changes in the labor market: There are now more highly educated people, partly due to the demands of the knowledge economy. Companies often use degrees as a quick indication of someone's capabilities ('credentialism'). As a result, applicant groups for many jobs have become more homogeneous in terms of intelligence. As a result, intelligence tests are less able to distinguish between candidates, which reduces their predictive value for job performance.

Relevance

This means that our ideas about the predictive value of intelligence tests may be outdated. Organizations should therefore reconsider how they use these tests when selecting staff.

PS: circular validation

In addition to the points mentioned in the above article, I think there is another reason why the predictive validity of IQ scores has often been overestimated in the past. This reason can be described as 'circular validation'. The frequent use of IQ tests in selection and promotion decisions can lead to a certain stratification within both society and organizations.

This process works on multiple levels. First, individuals with higher IQ scores are more likely to hold higher positions. In addition, knowledge of someone's high IQ score can lead to a kind of confirmation bias: managers may be inclined to perceive the performance of these individuals more positively, have higher expectations of them, and offer them more opportunities. These factors, in turn, can actually positively influence work performance.

Due to this complex process of circular validation, IQ scores and job performance assessments cannot be completely independent of each other. Both factors are partly influenced by the same initial selection criteria and subsequent expectations and opportunities. This interrelationship may give the impression that IQ tests have a higher predictive value for job performance than is actually the case. The circular validation may therefore reinforce the illusion of a stronger link between IQ and work performance than actually exists.