Acquisition context and the use of causal rules

Judgmental asymmetries in using causal knowledge (e.g., for prediction or diagnosis) have been attributed to the inherent directionality of causal knowledge. The present study examines the effect of acquisition context--representations used for initial instruction, and the type of judgement required during acquisition--on judgments using causal rules. In contrast to traditional concept formation research, this paradigm examined the development of procedures for using rules, rather than rule induction. College-student subjects learned to use causal rules describing digital logic gates, receiving instruction with either verbal rules or truth tables, and practicing either predicting or verifying logic-gate outputs. After 200 trials of practice with each rule, subjects were transferred to the untrained judgment task. Transfer was strongly asymmetrical. Subjects trained to make prediction judgments were slowed substantially by transfer to the verification task, while subjects trained to make verification judgments had little difficulty with transfer to the prediction task. Truth-table representations resulted in superior performance, especially for verification judgments. Contrary to prediction, verification judgments always required more time. The results demonstrate that acquisition context may be partly responsible for judgmental asymmetries, and imply that examining conditions of acquisition is important for understanding how causal knowledge is used.