Location-based errors in change detection: A challenge for the slots model of visual working memory

The slots model of visual working memory, despite its simplicity, has provided an excellent account of data across a number of change detection experiments. In the current research, we provide a new test of the slots model by investigating its ability to account for the increased prevalence of errors when there is a potential for confusion about the location in which items are presented during study. We assume that such location errors in the slots model occur when the feature information for an item in one location is swapped with the feature information for an item in another location. We show that such a model predicts two factors that will influence the extent to which location errors occur: (1) whether the test item changes to an "external" item not presented at study, or to an "internal" item presented at another location during study, and (2) the number of items in the study array. We manipulate these factors in an experiment, and show that the slots model with location errors fails to provide a satisfactory account of the observed data.