Andrea Yaoyun Cui¹, Simona Buetti¹, Pengfei Yu¹, Alejandro Lleras¹; ¹University of Illinois at Urbana-Champaign

This study builds upon the findings of Cui et al. (2022) and Cui et al. (in revision), which proposed that the difficulty in conjunction search arises from a distractor rejection mechanism operating in a parallel-sequential manner, whereby participants first reject distractors based on one feature dimension, then reject distractors based on the second feature dimension. Here, we employed eye-tracking to investigate how participants process and reject color-differentiated and shape-differentiated distractors in conjunction search tasks. We recorded eye movements as participants searched for a target (a red triangle) among two types of distractors: orange triangles (color-differentiated) and red circles (shape-differentiated). Stimuli were arranged in quadrants, each containing only one distractor type to maintain local feature homogeneity. We analyzed fixation counts, durations, and sequences to investigate attentional allocation to different distractor types. Our results revealed a strong bias toward fixating on shape-differentiated distractors as early as the second fixation, suggesting that color-differentiated distractors were efficiently rejected via peripheral vision. Participants made significantly more fixations and had longer total fixation durations within shape-differentiated regions, indicating higher attentional costs associated with rejecting these distractors. Temporal dynamics showed that while initial fixations were unbiased, participants quickly adjusted their strategy to focus on the more challenging shape-differentiated distractors. These findings support the parallel-sequential mechanism, demonstrating that participants employ a strategic allocation of attention during conjunction search—prioritizing the rejection of easily distinguishable distractors via peripheral vision before focusing on more difficult ones. Our study underscores the importance of eye movement analysis in understanding visual search mechanisms and suggests that attentional strategies are dynamically adjusted based on distractor characteristics. In two follow-up experiments, we varied the discriminability along each feature dimension to evaluate the extent to which the parallel sequential rejection mechanism is sensitive to the relative processing ease of one feature discrimination over the other.