Area activation: a computational model of saccadic selectivity in visual search

Abstract The Area Activation Model (Pomplun, Reingold, Shen, & Williams, 2000) is a computational model predicting the statistical distribution of saccadic endpoints in visual search tasks. Its basic assumption is that saccades in visual search tend to foveate display areas that provide a maximum amount of task‐relevant information for processing during the subsequent fixation. In the present study, a counterintuitive prediction by the model is empirically tested, namely that saccadic selectivity towards stimulus features depends on the spatial arrangement of search items. We find good correspondence between simulated and empirically observed selectivity patterns, providing strong support for the Area Activation Model.