The Implicit Association Test at age 20: What is known and what is not known about implicit bias

[Version 3 (uploaded 21 April 2020) provides corrected list of co-authors and commenters; the ms. is otherwise unchanged from Versions 1 and 2.] Scientific interest in unintended discrimination that can result from implicit attitudes and stereotypes (implicit biases) has produced a large corpus of empirical findings. In addition to much evidence for validity and usefulness of Implicit Association Test (IAT) measures, there have been psychological critiques of empirical findings and theoretical disagreements about interpretation of IAT findings. Because of public attention drawn by the concept of implicit bias, commercial and other applications based on the concept of implicit bias have been developed by non-psychologists—some of these applications are not appropriately guided by the existing body of research findings. This article is in 5 parts: (1) review of best practices for research use of IAT measures, (2) summary of what has been confidently learned from empirical research using IAT measures, (3) accepted and controversial theoretical interpretations of IAT findings, (4) significant questions about the IAT and implicit bias that still await answer, and (5) questions arising in attempts to apply research findings to remedy unintended discrimination due to implicit biases.