Diversity training shouldn’t be based on flawed implicit bias research | Opinion

Diversity trainings seem to be more popular than ever. Last week, Sephora closed all their U.S. stores for such a training, and in 2018, Starbucks required their employees to attend similar workshops.

But the effectiveness of these programs has been called into question. A 2016 study published in the Harvard Business Review, which analyzed 30 years of diversity training data from more than 800 U.S. companies, showed that on a whole, such programs not only fail, but can even decrease diversity.

“The positive effects of diversity training rarely last beyond a day or two, and a number of studies suggest that it can activate bias or spark a backlash,” the authors wrote.

Last year, the high school where I teach held a professional development aimed at improving school climate by helping educators recognize their unconscious bias. The facilitator from the Philadelphia School District’s Office of Student Rights and Responsibilities showed our staff national statistics on school discipline, and how black males were three times as likely to be suspended or expelled as their white counterparts. The facilitator suggested this was the result of implicit racial bias.

Although some of the younger teachers kept an open mind, many of the older and more experienced teachers like myself wondered: How could we be unknowingly discriminating against our students? We’d dedicated our lives to educating, coaching, and mentoring children from all backgrounds.

The answer, the facilitator explained, was in something called an implicit association test (IAT). For 20 years, the IAT has been used to measure the unconscious bias of teachers, police, businesspersons, and others; Hillary Clinton even mentioned the test in a debate against Donald Trump. The facilitator then pointed to data gathered by the highly touted Project Implicit, whose IAT data has been used to develop diversity training all over the country.

“These are the numbers,” she said, and told me that my resistance to her presentation was normal, that dialoguing my feelings was a good way to help me work through my own blind spots.

So the sacred IAT, the facilitator argued, suggested my colleagues and I were unconsciously racist. Yet a growing body of research shows that psychology’s favorite tool for measuring racism doesn’t live up to the job. In fact, a number of magazines and scientific journals have published articles in the last few years questioning both the validity and reliability of the IAT, showing that the results were unpredictable, and that they didn’t sufficiently correlate to actual behavior.

The Chronicle of Higher Education reported in 2017 that researchers had analyzed the results of hundreds of studies over 20 years of the IAT and found that “the correlation between implicit bias and discriminatory behavior appears weaker than previously thought.”

Amazingly, there’s even a disclaimer on Project Implicit’s website, which warns IAT test takers that they “make no claim for the validity of these suggested interpretations” from research done with these tests. In other words, the IAT might show evidence of racist beliefs where none exist.

So if conclusions based on the IAT are shaky, what can be done to help folks better act in a manner that is welcoming to all people?

For starters, we can dump the phrase “implicit bias,” which clearly has a negative connotation that can be offputting in training settings. “Conscious inclusion” is much more positive, and is now being used by some companies, such as Randstad USA. Studies also show making diversity training voluntary can increase its effectiveness, as can mentoring programs between managers and proteges, regular interaction with coworkers from diverse backgrounds, and encouraging social accountability. Hopefully such measures can get us to a place where we can simply treat people as people, and stop obsessing over the color of our skin.

Christopher Paslay teaches English at Swenson Arts and Technology High School in Northeast Philadelphia. cspaslay@yahoo.com