Attention and conceptual priming: Limits on the effects of divided attention in the category-exemplar production task

Prior research indicates that dividing attention during encoding reduces conceptual priming. The present study examines a limitation of this effect. Experiment la, in apparent contradiction to earlier research (Mulligan, 1997), found no effect of attentional load on later conceptual priming. Experiments 1b, 2, and 3 indicate that the discrepant results are not due to participant, materials, or power differences among studies, but rather to certain procedural differences. In particular, increased attentional load fails to reduce conceptual priming in the category-exemplar production task only when the categorical structure of the study list is salient (i.e., list items are blocked by category) and when attentional load is manipulated within categories, Experiment 4 indicates that processing earlier examples in the study block under full attention confers immunity to divided attention effects for later examples in the block. In contrast, if the first category members are in a divided-attention condition, then the usual negative effects of divided attention obtain. Experiment 5 verifies that this outcome is not due to within-block serial position effects, which occur for category-cued recall bur not category-exemplar production (rendering a new dissociation between matched conceptual explicit and implicit tests). The results are discussed in terms of item-specific and relational encoding processes. (C) 1999 Academic Press.