Isolating retrospective priming processes using relatedness proportion manipulations
Date
1997
Authors
Bodner, Glen Edward
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Semantic priming is commonly thought to reflect the contributions of spreading activation and expectancy processes, both of which operate before target presentation. However, retrospective processes operating after target presentation can also contribute to priming effects. The contributions of these latter processes were assessed in the context of relatedness proportion (RP) manipulations in which the proportion of related-prime trials in the stimulus list was varied. In both lexical decision and word reading tasks, semantic priming is typically greater for subjects receiving a high-RP than for subjects receiving a low-RP, leading some researchers to claim that high-RP subjects are more likely to generate expectancies about the upcoming target. To test this claim, RP x Priming interactions were sought under conditions where their presence could be attributed to a retrospective mechanism but not to expectancy. In Experiments 1 and 2, the contributions of expectancy were controlled by a novel within-subject manipulation of RP in the lexical decision task. In Experiment 3, the effects of expectancy were limited by using a short 240-ms SOA in the word reading task. Finally, Experiment 4 examined the effect of the proportion of identity-prime trials on masked repetition priming in the lexical decision task, where masking eliminates expectancy. The results provide some support for an episodic retrieval account in which subjects retrospectively retrieve priming episodes to help guide target processing. Further experiments to help clarify these results are suggested.