Development in causal reasoning: information sampling and judgment rule

Abstract

The present study investigated development in the ability to sample and use data about probabilistic relationships in order to test causal hypotheses. Third-grade, seventh-grade, and college students were asked to sample cards representing observations of two potentially related events and then to judge the causal relationship between those events. Overall accuracy of causal judgment was low (67-70% correct), with more errors on the noncontingent than on the two contingent relationships. Causal judgment accuracy improved with age across problem types. Biased information sampling was associated with poor accuracy of causal judgment in a pattern which suggested that subjects judged causal relationships according to the rate of the target outcome when the possible cause was present, with little attention to the same proportions when that cause was absent. © 1989.