Abstract

Salient examples may bias human judgments about the probability or frequency of events, an effect known as the “availability heuristic”; or the “salience bias.”; Scientific work has not been immune to this bias; in particular, the existence of certain large‐scale trends in evolution, such as those in size, complexity, and fitness, is widely accepted among professionals within evolutionary biology and paleontology, as well as among people outside these fields, even though these trends are poorly documented. Often, what documentation exists consists mainly of long lists of cases exemplifying the trend, or detailed descriptions of a small number of salient cases. Here, it is argued that although these lists and salient cases are not good evidence that a trend is pervasive, they may convince both the trend researcher and his or her audience. The possibility is raised that the bias may be pervasive in science and everyday thought, and a strategy for avoiding it, namely the use of random samples, is offered.