ABSTRACT
In contrast to the sparse literature on the naive analysis of action, there is a considerable body of research on the methods people use to make causal attributions for behaviour. M. G. Lipe proposed that a general model of the causal attribution process has three stages: generating a hypothesis, assessing the strength of it, and considering alternative explanations. Tendencies in causal attribution that appear to be biases sometimes at least, therefore, emerge from rational use of a model and in a way that is predictable once one knows the individual’s background presumptions. The observer’s attributional focus serves to define a focal set, that is to say a set of cases within which the observer assesses causality by means of probabilistic contrasts. Beliefs about causal powers override information about empirical regularities in causal judgement. The length of the foregoing review gives a fair indication of the dominance of regularity-based approaches in the adult literature.
