Steven A. Sloman
Brown University, Providence, RI, USA
David E. Over
University of Sunderland, UK
One outcome of the study of judgement and decision making's heuristics and biases programme is that it has become conventional wisdom that people make systematic errors when judging probability. To take just one example: Judgements are sometimes guided by beliefs about causality, and these beliefs on occasion compete with sound probabilistic principles (Tversky & Kahneman, 1983). Some errors of this type can be reduced by changing the object of judgement. For example, Tversky and Kahneman (1983) showed that the incidence of the conjunction fallacy-cases in which the probability of a conjunction is judged greater than the probability of one of its constituents-can be reduced by asking people to imagine a finite set of individuals and to make a frequency judgement about that set rather than a probability judgement about a single individual. Since then, as we shall see below, parallel reductions induced by frequency frames have been observed for other fallacies and biases.
Some theorists have used this effect of frequency as the empirical basis for the argument that the heuristics and biases programme is deeply flawed because it fails to understand behaviour in its ecological context. The reason, on this view, that the programme has uncovered so much error is because it has primarily asked people to make judgements of single-event probabilities, that is, the probability of one-time occurrences. This is inappropriate, detractors have argued, because people did not evolve to make single-event probability judgements; they evolved to make judgements using natural frequencies. Ask people to judge the “natural frequency” of events and many errors disappear.