Over-Attribution Bias

views updated

Over-Attribution Bias


The over-attribution bias, also known as correspondence bias, occurs when people attribute human behavior to whichever causal factor is most available to them. Behavior often engulfs the field, and people draw dispositional inferences that correspond to the behavior. When a person freely expresses a certain attitude, others assume that the person believes it. The same inference is biased, however, when observers know that a powerful other asked the person to express that attitude. The bias is most striking when it is the observers themselves who constrain the respondents behavior.

The common interpretation of the correspondence bias is that it constitutes a fundamental attribution error (Ross 1977). This interpretation holds that people fail to fully discount the influence of a persons internal disposition as a cause of behavior. The error interpretation has been influential in social psychology because it implies that people are incapable of understanding the power of the typical social-psychological experiment, which is to demonstrate that subtle changes in a persons situation can dramatically change behavior.

Upon review, the idea that people fail to appreciate the power of social situations needs to be tempered. The correspondence bias reverses, for example, when people who know a persons disposition are asked to judge the strength of the situation. They continue to attribute behavior in part to the situation even when the behavior is freely chosen. Hence, the correspondence bias is generic rather than purely dispositional. People attribute behavior firstly to whichever causal factor they happen to be focused on, be it a property of the person or the situation, and then modulate this inference by considering the other, less salient causal factor. Because the former process is likely intuitive and automatic, whereas the latter is deliberate and controlled, the bias is larger when people are unmotivated or unable (e.g., because of distraction) to process all available information.

Most models of causal attribution are hydraulic in that they regard the total causal force directing behavior as a zero-sum quantity. As one causal factor is being favored, another one must yield. On this view, the correspondence bias reflects a failure to fully discount the primary and salient cause when the secondary cause is sufficient. For the explanation of everyday behavior, the hydraulic model is sometimes inadequate. For example, people often attribute aggressive behavior to an aggressive disposition. To do so, however, they require the presence of a facilitating stimulus, such as an insult or a threat. Whereas a hydraulic model suggests that inferences about an aggressive disposition should be stronger in the absence of provocation, an interactionist model recognizes that a situational cause (provocation) is necessary for a dispositional attribution. On this view, theories of personality that seek to capture individual differences by merely counting trait-related acts are likely contaminated by the researchers correspondence biases.

The common tendency of attributing correspondence bias to peoples dispositional failure to think logically may itself be an example of the very same bias. Correspondence biases are, after all, experimentally evoked when investigators limit the salience of the situational causes of behavior. Hence, it may be sufficient to attribute respondents preference for dispositional inferences to the nature of the experimental situation.

SEE ALSO Attribution


Gawronski, Bertram. 2004. Theory-Based Correction in Dispositional Inference: The Fundamental Attribution Error Is Dead, Long Live the Correspondence Bias. In European Review of Social Psychology, vol. 15, ed. Wolfgang Stroebe and Miles Hewstone. Chichester, U.K.: Wiley.

Jones, Edward, and Victor Harris. 1967. The Attribution of Attitudes. Journal of Experimental Social Psychology 3: 124.

Ross, Lee. 1977. The Intuitive Psychologist and His Shortcomings. In Advances in Experimental Social Psychology, Vol. 10, ed. L. Berkowitz, 174221. New York: Academic Press.

Joachim I. Krueger