When choosing employee selection tests to improve workforce quality, cognitive tests typically work better than various other methods, such as interviews. Although there are other useful research-backed methods, cognitive tests have the best ability to predict job performance.
Unfortunately, the use of such tests can lead to discriminatory hiring practices against minority groups, who often score below their white counterparts due to a variety of factors. Different strategies have been proposed to counteract this adverse impact in selection procedures in order to ensure a fairer hiring process and encourage greater diversity within the workplace. The research reviewed here investigated one such strategy.
WHAT IS ADVERSE IMPACT?
Adverse impact is a means of measuring this type of discrimination. It is calculated by dividing the selection ratio from the lower scoring group of applicants (a minority group) by the selection ratio of the higher scoring comparison group (historically, more privileged groups).
Adverse impact towards the minority group has occurred if the result of these calculations is less than 4/5ths. In such a case, it then becomes incumbent upon the organization to prove that the tests used in the selection process are all job-relevant and not easily substitutable with other measure that do not lead to adverse impact. This is a way of guarding against discriminatory selection practices and ensuring a more diverse and representative workforce.
When cognitive tests are used for selection procedures, it is perceived that the organization now has to make a trade-off between selection criteria related to work performance and selecting for diversity by adhering to the adverse impact ratio.
SEEKING FAIRNESS IN SELECTION METHODS
One strategy for overcoming adverse impact is to weigh cognitive and non-cognitive tests differently. The researchers investigated the use of this weighting strategy on cognitive sub-tests, which represented second-order components of cognitive ability.
Second-order cognitive abilities are not specific individual abilities, but rather a broader constellation of related abilities. Still, they are more refined than a measure of general cognitive ability (known as g). For example, measuring acquired knowledge in reading and writing (second order) would include relationships across vocabulary, reading comprehension, and analogy tests.
The researchers hypothesized that, although general cognitive ability may be a fairly good predictor of later performance, the second order abilities may be better predictors when a job requires that specific ability. By using a sophisticated weighing technique with varying values for specific abilities related to a job, the researchers found that this method could improve minority hiring, but not at the expense of selection quality if a test of general cognitive ability was used.
BIG PICTURE TAKEAWAYS
This research is particularly interesting for managers and recruiters because it provides a clear way forward in decreasing the possible adverse impact of company selection procedures, which helps to create a more diverse workforce.
Workplace diversity has been shown to have multiple benefits in terms of organizational outcomes. But you can also rest assured that using such weighted methods won’t decrease the quality of hires, as long as the abilities included on tests are job-relevant.
Wee, S., Newman, D. A., & Joseph, D. L. (2014). More than g: Selection quality and adverse impact implications of considering second-stratum cognitive abilities. Journal of Applied Psychology, 99(4), 547-563.