Item

Abstract

The multiple-object-tracking (MOT) paradigm is useful for studying how observers allocate their attentional resources over several moving targets. Despite claiming no effect of eye movements on performance, previous studies provided no clear evidence for this assumption. We investigated how eye movements affect performance in an MOT task under different viewing conditions (free eye movements vs fixation control vs instruction to fixate but without monitoring). Subjects (N=4) performed an MOT task (120 trials per condition) while eye movements were recorded (250 Hz, error < 0.4°). Allowing free eye movements led to significantly higher performance (93) than with proper fixation control (75) and without monitoring (83, all pairwise t-tests: p < 0.001). Thus, high performance in previous studies without fixation control could possibly be explained by fixation losses. On the basis of eye-movement data and subjective reports we suggest an alternative to t radition al multi focal at tention models. It predicts that observers group targets together and track a single "mental" object.