Control of Fixation Duration and Saccade Direction during Search
To investigate the interaction between saccade preparation and foveal analysis on multi-fixation search, we designed an experiment in which the result of the foveal analysis is relevant for programming the direction of the next saccade. We examined whether fixation duration is adjusted to the duration of the analysis process or not. Subjects were asked to find a circle (diameter 2.1 deg) among 35 circles (gap size 0.3 deg, diameter 2.1 deg) that were placed on an invisible hexagonal grid (35 deg × 35 deg). Eye movements of the right eye were measured. The gaps of the circles indicated the direction in which the circle could be found. Subjects were instructed to make eye movements in the direction that was indicated by the gap of the fixated circle. The proportion of eye movements made in the indicated directions ranged from 0.54 to 0.72 (0.25 is chance level). Thus direction information is indeed, but improperly, used for the programming of eye movements. Durations of fixations preceding eye movements made in non-indicated directions were smaller than durations of fixations preceding eye movements made in indicated directions. From these results we conclude that saccade preparation only uses the result of the analysis if it is available in time and that there is no direct control of fixation duration by a mechanism that monitors the analysis process.