| Abstract: | Describes an experiment in which 144 undergraduates read 28 sentences, each 6 words in length. The last word was a homograph for which either of 2 meanings could be biased by the context of the sentence. 6 of the homographs appeared twice, with 1, 6, or 11 sentences intervening between the 1st and 2nd presentation. For 72 of the Ss, the same meaning was biased on both occasions (condition SC), while the remaining Ss were given 2 different meanings (condition DC). In free recall of the homographs, probability of recall was an increasing function of the number of intervening sentences (lag) for condition SC, whereas this function was invariant with lag for condition DC. For both conditions, the probability of an early output of a homograph was also an increasing function of lag. Findings are interpreted within an encoding variability framework. (French summary) (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
