Abstract
Target/distractor similarity influences categorical search (Alexander & Zelinsky, 2011), as does the degree to which category members vary in appearance from one another (category variability). Searchers have better attentional guidance and shorter response times (RTs) to homogenous categories with low variability in appearance (Hout et al., 2015) likely due to template precision. We examined whether target/distractor similarity effects are moderated by category variability. We expected that the benefit of a precise search template for low variability categories (e.g., bananas) would only be present in search among dissimilar distractors compared to similar distractors because similar distractors would produce greater interference for low variability (LV) target categories than high variability (HV) categories (e.g., shoes). We conducted two experiments: one using color stimuli and another with greyscale stimuli. The participants were presented with the visual search task where they looked for a category target (e.g., butterfly) among 19 different distractors. We developed a novel approach to manipulating target/distractor similarity using multidimensional scaling (MDS). MDS was also used to measure category variability for each target category and identify categories that were LV or HV targets. Eye tracking was used to measure attentional guidance and category verification. Replicating previous studies, participants were slower to locate and identify targets among similar distractors. For color stimuli, there was an interaction between variability and target/distractor similarity in RTs. Participants were slower to find LV targets among similar distractors compared to HV, and faster to find LV categories among dissimilar distractors. While no effects on guidance were uncovered, for both experiments there were interactions in the verification measures; participants were slowest to verify LV categories among similar distractors but this effect was not present for dissimilar distractors. The results suggest that target/distractor similarity effects may not be uniform across categories and that variability is a moderating factor.