The Implicit Association Test (IAT) is the primary tool used by researchers to measure and quantify the existence of implicit bias in a person. Developed by psychologists at Washington University in 1995, it has been of fundamental importance to studies of implicit bias ever since.
Because of its importance, the test itself has come under great scrutiny: Is it scientifically valid? Does its data lead to falsifiable conclusions? What exactly is it measuring, anyway? This section will review some of the research around the IAT, and hopefully lead to a fuller understanding of the scholarship on implicit bias.
As a starting point: Take the test yourself!
The creator of the test, Dr. Anthony Greenwald, also has a remarkably useful webpage with many resources on the validity of the IAT.
Cognitive psychology is an imperfect science that relies on imperfect data. Many criticisms of the IAT are, as best I can tell, valid. However, the underlying methodology of the test, as measured again and again by scientists over the last twenty years, supports its use in producing knowledge about implicit bias. Just as importantly, there is no obvious extant or forthcoming test that would improve on the IAT. To the extent that the IAT is compromised, it is because it is premised on that most difficult to test subject: the human mind. Criticisms are a reason to refine and adjust the test, but it should not be ignored or discredited wholesale by the judicial system.
The continued utilization of the test in a myriad of contexts, spanning the academy and the business world, constitutes strong evidence as to its continued relevance and usefulness. Judges and attorneys should seek to understand the powers and limitations of the IAT in the various and specific contexts in which it is presented. They should not latch onto any of its particular weaknesses as a reason to throw it out any more than they should accept its conclusions with complete credulousness.