Effective elaboration and inferential reasoning

The present study investigated the effects of several variables on subjects’ abilities to draw inferences from two related premises. Contiguous presentation of the related sentences led to significantly higher inference scores than did noncontiguous presentation. The nature of the initial premise was also manipulated. For some subjects, the initial premise was relatively consistent with subjects’ semantic knowledge, whereas other subjects received initial premises that were somewhat arbitrary with respect to semantic knowledge. Inference scores were significantly higher for subjects who received the more consistent initial premises. This effect persists even when differential memory for the initial premises has been accounted for. The results suggest certain constraints on the nature of the knowledge one must possess in order to make appropriate inferences from material.