Measuring Visual Attention Across Concurrent and Sequential Visual Perception Tasks

Abstract

Eye tracking provides a noninvasive method for measuring visual attention during human-computer interaction. Identifying which oculometric measures are most sensitive to different demands is important for the development of attention-aware systems. This pilot study examined whether common oculometric measures could distinguish between global, selective, and sustained visual attention tasks during low-demand viewing of static images. Ten participants (two males and eight females, 27.5 ± 4.7 years) completed descriptive, comparison-based, and memory-based visual tasks while wearing an eye tracker. Measures included task duration, fixation count, fixation percentage, blink count, blink percentage, and pupil diameter. Repeated-measures analyses showed a marginal effect of task condition on fixation count, with fixation counts descriptively highest during the global attention task and lowest during the selective attention task. Fixation percentage, blink measures, pupil diameter, and task duration did not significantly differ across conditions. These preliminary findings suggest that fixation count may be more sensitive to task-related attentional differences during low-demand visual interaction than blink- or pupil-based measures, supporting its potential use in future attention-aware and multimodal human-computer interaction systems.