Infering from text: Some factors influencing which inferences will be made∗

Two studies are reported in which the importance of an event in a story was systematically varied to test whether this would affect the likelihood that inferences about the event would be made. Eight versions of each of six 500‐word stories, identical except for the sections varied in order to manipulate the importance of the event and the salience of material which cued the target inference, shared the same target inference. Importance was varied without changing the plausibility of the inference. In half of the passage versions, the “inference” was explicitly stated in the text. High school students served as subjects. In the first experiment, inference was tested using multiple‐choice recognition, while a cued recall test was employed in the second experiment. The results of both studies confirmed that people are more likely to make an important rather than unimportant inference. Highly salient premises were also found to increase the probability that an inference would be made.