Abstract
The triple conjunction effect (TCE) is characterized by faster response time (RT) when a target is defined by a conjunction of three features than when it is defined by two features. One possible, but unaccounted for, explanation of this finding is feature coactivation, in which information from multiple features combines to reach a response threshold more quickly than separate single features. The purpose of the current study was to determine if the TCE could be attributed to feature coactivation; in addition, we explored whether or not the TCE occurs in peripheral visual search. Participants searched for the presence of a target letter in 6 counterbalanced blocks, with 4 blocks of conjunction searches (2 of color and orientation, and 2 of form and orientation) and 2 blocks of triple conjunction searches (color, form, and orientation). Each trial contained 4 or 8 letters subtending 2° by 2° on an invisible circle 8° from the center of the display. Trials were terminated if participants moved their eyes more than 2.75° from the center or did not respond within 4 seconds. Each conjunction search had 4 types of distractors, and each triple conjunction search had 3 types of distractors. A second experiment had 3 distractor types in all search conditions, to rule out distractor homogeneity effects. In both experiments, RT was faster in triple conjunction (~206 ms) than conjunction search. The Townsend Bound, a race model prediction about the redundant target RT distribution, was violated at several quantiles (5-16 of 18 quantiles, depending on experiment, set size, and target), providing evidence for coactivation when RTs were averaged across participants. Anderson-Darling tests indicated that most participants individually violated the Townsend Bound, providing further evidence for coactivation. The results suggest that the peripheral TCE is at least partially due to coactivation of target-relevant features.
Meeting abstract presented at VSS 2015