Abstract
This paper reviews research on bias. We start by reviewing the New Look of the 1940s and heuristics and biases in judgment and decision making. We show how the waves of enthusiasm for some then-new forms of bias proved overwrought. We then present a “Goodness of Judgement Index” and show how it can extract information about unbiased responding from studies reporting only data on bias. The final section reviews some of the phenomena central to the newest wave of bias research – on various aspects of “social justice.” Core findings and limitations to stereotype threat, microaggressions, and implicit bias are discussed. We end by reviewing 1. research finding little or no biases against women in studies of workplace discrimination and peer review; and 2. “the discrimination paradox” – the seemingly conflicting results of high-quality studies seeming to provide evidence of both substantial and minimal racial discrimination. We resolve the paradox by showing that the findings are compatible. Across nearly 100 years of research on a wide variety of biases, some claims about biases were unjustified because they failed to rule out alternative explanations (unconscious influences on perception) or because advocates never tested core claims (such as that microaggressions can be unconscious). Others were outright false (such as Hastorf & Cantril’s (1954) claim that their participants saw entirely different realities). Yet others, when held up to critical scrutiny, were either not as powerful as originally claimed, or, sometimes, proved to serve people well in the real world (such as heuristics). In two cases, however, biases claimed to be substantial or important actually were: political biases and racial job discrimination.