Abstract
Humans are exceptionally good at resuming a search that has been momentarily interrupted; making accurate detections in far less than the half second normally needed to begin a search (Lleras et al., 2005). It is proposed that rapid resumption depends on an iterative hypothesis testing mechanism that compares information stored in short-term memory with the display. Three experiments investigated the role of eye movements in rapid resumption. Experiment 1 monitored eye position passively during interrupted search and found that rapid resumption was correlated with relatively short distances between eye position and target location. We reasoned that if eye position alone was predictive (near fixation hypothesis), we should be able to increase the proportion of rapid resumption responses by using a contingent gaze paradigm to present the target near fixation, regardless of where in the display the eyes were pointed. Experiments 2 and 3 used two different methods for presenting the target to the point of eye fixation on a subset of the trials. The results showed, contrary to the near fixation hypothesis, that the proportion of rapid responses decreased dramatically when the target was offered directly to the eye. However, when two successive fixations occurred in a single target location virtually all responses were rapid. We conclude that the contingent gaze procedure interferes with participants' ability to compare their perceptual hypothesis with the sensory information in the display and that rapid resumption depends on the confirmation of a hypothesis regarding objects in a stable world.