When covariation is not enough: The role of causal mechanism, sampling method, and sample size in causal reasoning.

College students and college-bound ninth and sixth graders read several story problems in which a problem solver tried to find out whether a target factor was causally related to an effect. Each story problem included information about possible mechanisms that could have mediated between factor and effect (mechanisms present vs. absent), sample size (large vs. small), sample method (direct intervention vs. natural occurrence), and results (target factor did vs. did not covary with the effect). The results suggest that adolescents hold a tacit theory of evidence in which the presence of covariation is accorded a kind of primacy so that the presence of covariation overrides other evidence that calls causation into question