Insensitivity to sample bias: Generalizing from atypical cases.

Abstract

Two experiments were conducted to determine whether subjects take into account the representativeness of a sample before generalizing from the sample to a population. Subjects were presented with vivid one-case samples of populations—a welfare recipient in one study and a prison guard in another. Subjects were then asked to rate the population (of welfare recipients or prison guards) on a number of dimensions. Exposure to the sample case influenced attitudes about the population whether subjects were told nothing about the typicality of the case, were told that the case was highly typical of the population, or were told that the case was highly atypical of the population. The results suggest that, at least when information about sample bias is pallid and information about the nature of the sample is vivid, people may make unwarranted generalizations from samples to populations.