CAUSAL ATTRIBUTION AS A SEARCH FOR UNDERLYING MECHANISMS - AN EXPLANATION OF THE CONJUNCTION FALLACY AND THE DISCOUNTING PRINCIPLE

We propose that causal attribution involves searching for underlying m echanism information (i.e., the processes underlying the relationship between the cause and the effect). This processing account can explain both the conjunction effect (i.e., conjunctive explanations being rat ed more probable than their components) and the discounting effect (i. e., the effect of one cause being discounted when another cause is alr eady known to be true). When two explanations cohere with respect to a single mechanism, they would be judged to be more likely than a singl e explanation which partly supports that mechanism. When the two expla nations imply two separate mechanisms, one would be discounted. In Exp eriment 1, both effects occurred with mechanism-based explanations but not with covariation-based explanations in which the cause-effect rel ationship was phrased in terms of statistical covariations without ref erring to mechanisms. In Experiments 2 and 3, the amount of the discou nting and conjunction effects varied depending on the relationships be tween specific mechanisms in the two given explanations. We discuss wh y the current results pose difficulties for previous attribution model s. (C) 1996 Academic Press, Inc.