Sentence memory prompted with inferential recall cues

Conducted 3 experiments with 410 college students to investigate cue effectiveness in sentence comprehension and recall. In Exp I, a list of sentence pairs was constructed so that members of a pair were identical except for the subject of the sentence, which was chosen to maximize the probability that the hearer would understand each sentence unambiguously. Ss were presented with 1 member of each pair. Recall cues named objects probably inferred from the event described in a sentence or its pair member. Cues relevant for sentences an S heard were more effective than cues related to pair members not presented. Exp II was an attempt to predict sentence recall from data on cue effectiveness for sentence subjects and predicates in isolation. An equation based on an associative model proved adequate for relevant cues but greatly overestimated recall with irrelevant cues. However, in Exp III, a simpler study with new sentences and only relevant cues, the associative predictions underestimated recall. (PsycINFO Database Record (c) 2006 APA, all rights reserved).