Researchers Uncover Hiring Method That Can Minimize the Effects of Unconscious Bias
New approach to personality-based selection minimizes adverse impact and unintentional bias while increasing the strength of the hiring process
BOSTON — April 14, 2014 — Will hiring managers soon be automating the first steps in the employee screening process? Cangrade today announced research findings that may cause savvy HR professionals to move in this direction. Multivariate personality assessment—a method slowly gaining popularity in the selection community—not only can help professionals find better employees, but can also actually help companies build the diversity of their teams.
“Modern psychological science clearly shows that all people are prey to biases beneath our conscious knowledge and control, and that these biases undermine the quality and fairness of our decisions,” says Steve Lehr, Chief Science Officer at Cangrade. These issues are garnering international attention, with coverage in bestselling books such as Malcolm Gladwell’s Blink and Daniel Kahneman’s Thinking, Fast and Slow . Many legal scholars and HR professionals are taking notice. “More and more often, business leaders are actively looking for advice on how to minimize the effects of unconscious bias,” says Mahzarin Banaji, Richard Clarke Cabot Professor of Social Ethics at Harvard University and author of Blindspot: Hidden Biases of Good People . “Since all people hold these subtle biases, one solution is for organizations to explore more objective tools and processes that sidestep them.”
For those worried about unconscious bias creeping into their hiring processes, multivariate personality assessment may be a promising tool to make employee selection more objective. This technique assumes that each job candidate is complex, with many different aspects of their personality combining to form a unique “whole” person, and then uses predictive statistical analysis to estimate how likely each candidate is to perform well in a particular job.
Cangrade’s team spent years researching the effectiveness of different pre-employment selection methods using a statistical technique called meta-analysis. Their analysis combined the results from hundreds of published research studies over several decades, and statistically modeled both the effectiveness and the potential for bias in different job candidate screening methods.
It turns out that some criteria employers commonly use to screen applicants, such as College GPA and phone interviews, show only very small relationships with future job performance. Yet these same things inadvertently create disparate treatment of qualified applicants by factors such as race. Relative to the other methods, multivariate personality assessment models were better able to identify good candidates. Even more amazing, this method at the same time reduced disparate impact to nearly zero.
“These results are striking, and may explain the increased industry interest in modern personality testing and predictive analytics.” said Greg Willard, Senior VP at Cangrade and lecturer at Harvard University. “It goes beyond diversity initiatives or legal compliance. We can now clearly see that objective hiring tools can affect the bottom line. When used correctly, they can help employers to make better hiring decisions while at the same time increasing the diversity of their teams.”