In social psychology, fundamental attribution error (FAE), also known as correspondence bias or attribution effect, is a cognitive attribution bias where observers underemphasize situational and environmental factors for the behavior of an actor while overemphasizing dispositional or personality factors. This effect has been described as "the tendency to believe that what people do reflects who they are"; that is, to overattribute their behaviors to their personality and underattribute them to the situation or context. Although personality traits and predispositions are considered to be observable facts in psychology, the fundamental attribution error is an error because it misinterprets their effects.
As an example of the behavior which attribution error theory seeks to explain, consider the situation where Alice, a driver, is cut off in traffic by Bob. Alice attributes Bob's behavior to his fundamental personality; e.g., He thinks only of himself, he is selfish, he is an unskilled driver. She does not think it is situational; e.g., He is going to miss his flight, his wife is giving birth at the hospital, his daughter is convulsing at school. Alice might well make the opposite mistake and excuse herself by saying she was influenced by situational causes; e.g., I am late for my job interview, I must pick up my son for his dental appointment, rather than thinking she has a character flaw.
The phrase was coined by Lee Ross 10 years after an experiment by Edward E. Jones and Victor Harris in 1967. Ross argued in a popular paper that the fundamental attribution error forms the conceptual bedrock for the field of social psychology. Jones wrote that he found Ross's phrase "overly provocative and somewhat misleading", and also joked: "Furthermore, I'm angry that I didn't think of it first." Some psychologists, including Daniel Gilbert, have used the phrase "correspondence bias" for the fundamental attribution error. Other psychologists have argued that the fundamental attribution error and correspondence bias are related but independent phenomena, with the former being a common explanation for the latter.
Jones and Harris hypothesized, based on the correspondent inference theory, that people would attribute apparently freely chosen behaviors to disposition and apparently chance-directed behaviors to situation. The hypothesis was confounded by the fundamental attribution error.
Subjects in an experiment read essays for and against Fidel Castro. Then they were asked to rate the pro-Castro attitudes of the writers. When the subjects believed that the writers freely chose positions for or against Castro, they would normally rate the people who liked Castro as having a more positive attitude towards Castro. However, contradicting Jones and Harris' initial hypothesis, when the subjects were told that the writers' positions were determined by a coin toss, they still rated writers who spoke in favor of Castro as having, on average, a more positive attitude towards Castro than those who spoke against him. In other words, the subjects were unable to properly see the influence of the situational constraints placed upon the writers; they could not refrain from attributing sincere belief to the writers. The experimental group provided more internal attributions towards the writer.
The hypothesis that people systematically overattribute behavior to traits (at least for other people's behavior) is contested. A 1986 study tested whether subjects over-, under-, or correctly estimated the empirical correlation among behaviors. (ie traits, see trait theory) They found that estimates of correlations among behaviors correlated strongly with empirically-observed correlations among these behaviors. Subjects were sensitive to even very small correlations, and their confidence in the association tracked how far they were discrepant (i.e., if they knew when they did not know), and was higher for the strongest relations. Subjects also showed awareness of the effect of aggregation over occasions and used reasonable strategies to arrive at decisions. Epstein concluded that "Far from being inveterate trait believers, as has been previously suggested, [subjects'] intuitions paralleled psychometric principles in several important respects when assessing relations between real-life behaviors."
A 2006 meta-analysis found little support for a related bias, the actor-observer asymmetry, in which people attribute their own behavior more to the environment, but others' behavior to individual attributes. The implications for the fundamental attribution error, the author explained, were mixed. He explained that the fundamental attribution error has two versions:
The meta-analysis concluded that existing weight of evidence does not support the first form of the fundamental attribution error, but does support the second.
Several theories predict the fundamental attribution error, and thus both compete to explain it, and can be falsified if it does not occur. Some examples include:
The fundamental attribution error is commonly used interchangeably with "correspondence bias" (sometimes called "correspondence inference"), although this phrase refers to a judgment which does not necessarily constitute a bias, which arises when the inference drawn is incorrect, e.g. dispositional inference when the actual cause is situational). However, there has been debate about whether the two terms should be distinguished from each other. Three main differences between these two judgmental processes have been argued:
Based on the preceding differences between causal attribution and correspondence inference, some researchers argue that the fundamental attribution error should be considered as the tendency to make dispositional rather than situational explanations for behavior, whereas the correspondence bias should be considered as the tendency to draw correspondent dispositional inferences from behavior. With such distinct definitions between the two, some cross-cultural studies also found that cultural differences of correspondence bias are not equivalent to those of fundamental attribution error. While the latter has been found to be more prevalent in individualistic cultures than collectivistic cultures, correspondence bias occurs across cultures, suggesting differences between the two phrases. Further, disposition correspondent inferences made to explain the behavior of nonhuman actors (e.g., robots) do not necessarily constitute an attributional error because there is little meaningful distinction between the interior dispositions and observable actions of machine agents.