Confirmation Bias

Abstract

Robert Zajonc once wrote that opinions of psychologists about people’s rationality tend to vary with every decade (Zajonc, 1989). The “irrational” New Look approach of the fifties was thus duly replaced by the rational “lay scientist metaphor” of the sixties, to be followed by the boom of interest in “nonrational” biases and errors of the seventies. Here, however, the regularity seems to end. While the next two decades witnessed the undiminished interest in the nature of biased cognition, manifested in growing number of monographs and collectively edited books (Evans, 1989; Hogarth, 1980; Kahneman, Slovic, & Tversky, 1982; Manktelow & Over, 1993; Nisbett & Ross, 1980), the preferred theoretical approaches appeared to lose their pendulumlike character and suggested some form of a compromise. This means that while nobody nowadays denies the existence of biases in lay cognition (it can hardly be done in the face of the amassed evidence), the focus of interest apparently has shifted from their evaluation (whether the biases testify or not to human rationality) to explanation.