Uniqueness of behavioral effects in causal attribution1

Most research on causal inferences has dealt with the conditions under which an observed behavior is attributed to a disposition of the actor rather than to some external factor (e.g., Steiner & Field, 1960; Jones, Davis, & Gergen, 1961; Frieze & Weiner, 1970; Ajzen, 1971; McArthur, 1972). Much less is known about causal attribution to an actor whose behavior may reflect any of several different internal factors. That is, even when external determinants can be ruled out, it is often possible to invoke more than one disposition to account for the observed act. Jones & Davis (1965) suggested that any outcome or effect produced by the actor's behavior can provide the basis for a causal attribution. Suppose that an actor is observed buying a package tour of the following European capitals: Oslo, London, Paris, and Rome. His desire to visit any one of these cities could provide an explanation for his behavior. According to Jones & Davis, a behavior tends to be attributed to its noncommon or unique effects; i.e., to outcomes that would not be produced by alternative behaviors.^ By implication, the more uniquely a given effect is associated with the actor's chosen alternative, the more likely it is that a dispositional attribution will be based on that effect. Imagine that the actor in the above example had a choice among four different package tours of Ein^opean capitals. If only the chosen tour included Oslo, an observer would be more