Distinctiveness and Serial Position Effects in Recognition and Sentence Processing

Abstract Four experiments test the ability of Neath′s (1993a) dimensional distinctiveness model to account for serial position effects in recognition and sentence processing. Experiment 1 ruled out an account of primacy effects in recognition based on proactive interference. Experiment 2 determined that the model made accurate predictions for recognition of non-verbal stimuli in lists where the inter-item presentation interval varied: as the duration of the inter-item presentation interval was lengthened, the recency effect increased and the primacy effect decreased. Experiment 3 demonstrated the trade-off in primacy and recency effects using stimuli that could be verbally encoded, extending the range of stimuli covered by the model. Finally, Experiment 4 confirmed that the model predicted the switch in response-time advantage in verifying that either the first-mentioned or second-mentioned participant was in a sentence: with longer retention intervals, the advantage was for the first mentioned but with shorter intervals, the advantage was for the second mentioned. The model then successfully fit 17 out of 19 conditions reported by Gernsbacher and Hargreaves (1988). We argue that the same principles of distinctiveness, as described by the dimensional distinctiveness model, underlie all these diverse phenomena.