Attribution Theory

Early Attribution Theories

Following Heider's lead, subsequent attribution researchers considered cognitive variables and processes. For example, suppose an observer wished to explain why Jill was impolite to Joe. Relevant considerations include whether Jill is impolite to others as well, whether others are impolite to Joe, and whether impoliteness arises when Jill and Joe get together. If Jill is impolite to others, this suggests that the behavior was due to Jill's being an impolite person. If others are impolite to Joe, this suggests that there is something about Joe that provokes impolite behavior. If Jill is not generally impolite, and Joe is not typically the recipient of impolite behavior, but impoliteness arises when they are together, this [Page 84]suggests that the problem resides in the interaction between Jill and Joe.

Other theorists have focused on social desirability. Because there is social pressure to perform socially desirable behaviors and avoid socially undesirable behaviors, socially desirable behaviors can be attributed to social pressure, whereas socially undesirable behaviors provide better clues to what the person is really like.

Attribution theorists also have considered “non-common” effects, which can be illustrated as follows. Suppose that Leslie chooses between two jobs; both jobs are near the ocean, near good restaurants, and pay equally well, but the first job is more interesting than the second job. Leslie's choice of the first job over the second one could confidently be attributed to the work being interesting rather than to the ocean, restaurants, or pay.

In addition to the foregoing “rational” attribution processes, early experiments also suggested a less rational process where observers overestimate the degree to which behaviors tell us about the traits of the people who perform them. Much early research seemed to indicate that observers give insufficient weight to the situations in which behaviors are performed, thereby leading to stronger trait inferences than are justified; this effect has been termed correspondence bias or fundamental attribution error. Later research, however, demonstrated that these effects are due, at least in part, to methodological artifacts. The use of different instructions or different experimental paradigms has been shown to eliminate or even reverse the tendency to underuse information about situations. At present, the pervasiveness of the phenomenon in the absence of method ological artifacts is unclear.

Later Cognitive Theories Pertaining to Ability and Morality

As attribution findings accumulated in the 1960s and 1970s, researchers began to suspect that perhaps the type of trait examined in particular experiments partially determines the results. To see why, consider ability and the premise that people without ability only can perform badly whereas people with ability can perform well or badly. It follows that a successful performance of a behavior indicates high ability (because a person without the ability could not have done it), whereas unsuccessful performance is ambiguous with respect to ability (because anyone can fail). In contrast, in the domain of morality, it is the negative behaviors that are diagnostic of traits because immoral behaviors only can be performed by immoral people, whereas moral behaviors can be performed by moral or immoral people.

...