The effect of retention interval on serial position curves for item recognition of visual patterns and faces.

In 9 experiments, participants were presented with a series of 4 novel items, followed by an immediate or delayed probe recognition test. Delaying the recognition test reduced performance on the final item, when the interpresentation intervals (IPIs) were kept constant (Experiments 1A-1G), and also when the IPIs were varied (Experiments 2A-2B). Only 1 experiment reported any evidence of increased primacy with delay when the IPIs were kept constant, but this result failed to replicate. Neither Experiment 2A nor 2B provided evidence of increased primacy with test delay or any effect of IPI on recency. However, Experiment 2A showed a deficit at the first serial position when followed by a short IPI. These results do not support the predictions of the dimensional distinctiveness model but are broadly compatible with established information-processing models of visual memory.