An exploration of conceptual similarity in multiple target visual search

Abstract

Visual search studies generally suggest that multiple target visual search is more difficult and less efficient than traditional single target search, with costs to both RT and accuracy. Typically, these studies use dual target trials and abstract stimuli. A study by Laderoute (2021) using varying target set sizes and naturalistic stimuli found better performance for many-target search, with per-item RTs that decreased as target set size increased. Interestingly, increasing the perceptual similarity of the targets had no effect on search performance. However, in that study, regardless of the perceptual similarity condition, all targets within a trial were always the same type of item. It is possible that this high degree of conceptual similarity between targets facilitated search performance. The present study further explores the many-target benefit and assesses the possibility of a conceptual similarity effect. Participants searched naturalistic scenes for 2, 6, or 18 targets. On each trial, targets were either variations of the same item, items belonging to the same specific category, or items belonging to the same general category (e.g., hair ties, hair accessories, toiletries). Performance was best in the same-item 18-target condition. Mean per-item RT decreased as target set size increased, with little to no effect on accuracy. Search performance was best for same-item targets, and worse for samecategory and similar-category targets, with very similar performance in the two “category” conditions. The effect of conceptual similarity decreased as target set size increased. This study provides additional information about the factors that enhance and impair multiple target visual search performance, and has relevance for both professional and non-professional searchers.