Abstract
Two experiments demonstrate the effect of stimulus configuration and cognitive workload on patterns of initial fixation in a visual search task. The display was divided into four equal quadrants with the target (an L among Ts) placed at the center of one of the four. The density of one quadrant per trial varied from control (1.97 of visual angle), to low (0.97 ), to high (0.54 ). The control density was the same as the density of the remaining 3 quadrants. The location of the target was orthogonal to the location of the density-quadrant. We hypothesize that increasing levels of density “draws” initial fixations and that this effect increases with workload. In Experiment 1 the probability of initially fixating the density-quadrant increased from control, to low, to high. Post-hoc analyses revealed an effect of the target's previous location, History, and an additive effect from the Co-occurrence of History and Density. Experiment 2 added a secondary task, presented phonically, that taxed working memory. Participants performed both tasks simultaneously. Added workload doubled the effects of Density, History, and Co-occurrence relative to Experiment 1. We believe the secondary task taxed cognition to a point that it became difficult to impose conscious strategies, resulting in data-driven behavior. Results indicate an independent effect of Density and History. Strong densities have a greater effect than Moderate densities. All effects were doubled as a result of increasing cognitive workload. We believe these results to be examples of bottom-up processes peering through top-down strategies. That is, when attentional processes become sapped, saccadic selectivity becomes more susceptible to bottom-up, data-driven processes.