Subconscious police

Intrusive ‘anti-bias retraining’ is baloney science

Silicon Valley tech companies are implementing mandatory “anti-bias retraining” programs, in a collective effort to combat the prejudices supposedly hidden in their employees’ subconscious. If that sounds Orwellian, it is because it is.

Anti-bias retraining is a top-down intervention into people’s thought lives, based on discredited science. And tech is far from the only industry to adopt such initiatives.

If there were any evidence that unconscious bias influences people’s actions, then perhaps a case for such measures might exist. But the behavioral effects of unconscious bias have been conclusively shown to be “slight” at most. Moreover, the test used to measure a person’s unconscious bias — the Implicit Association Test — is neither reliable nor valid.

The IAT is actually a word association game that purports to identify a person’s level of unconscious bias against various group identities not the test taker’s own (whether race, gender, sexual orientation, etc.). It is pseudo-science.

For a diagnostic test to be considered reliable, the test taker must get the same result twice. Takers of the IAT get differing results with each attempt. On these reliability grounds alone the IAT and its findings could be dismissed, but the test has another fundamental flaw: the behaviors it predicts rarely appear.

In January of this year, the Chronicle of Higher Education reported that a group of researchers (among them one of the IAT’s two creators) analyzed the results of hundreds of studies of the test. Their work revealed that the correlation between implicit bias and discriminatory behavior is far feebler than originally hypothesized. From the Chronicle: “Everyone agrees that the statistical effect linking bias to behavior is slight. They only disagree about how slight.”

And yet, in the same article, the author gives us this astonishing additional information: “Since it first went online in 1998, millions have visited Harvard’s Project Implicit website, and the test’s results have been cited in thousands of peer-reviewed papers. No other measure has been as influential in the conversation about unconscious bias.” The writer goes onto to recall how the concept “circulated widely after Hillary Clinton mentioned it during the presidential campaign.”