Olga Kozlova¹, Kirsten Adam¹; ¹Rice University

Visual working memory (VWM) is a limited-capacity storage that allows individuals to actively maintain information in mind. However, previous research conflicts on whether the capacity of VWM differs between simplistic stimuli (e.g., colored squares) and real-world objects. Here, we sought to 1) replicate and extend the VWM object benefit in a whole-report VWM task, and 2) test whether proactive interference may modulate the difference between simplistic stimuli and real-world objects. In two experiments (n=26 each), we compared VWM performance for colored squares versus real-world objects. On each trial, participants remembered 6 items across a blank delay, and reported the correct binding of each item to its location at test (i.e., participants made 6 memory responses in a “whole-report” paradigm). In Experiment 1, the stimulus pools were matched between colored squares (9 possible colors) and objects (9 possible objects). In Experiment 2, the stimulus pool was smaller for colors (9 possible colors) than for objects (hundreds of trial-unique objects). In both experiments, we found no evidence for an object benefit in VWM. Instead, we found that participants remembered significantly more color-location bindings compared to object-location bindings (p<.001). Was this “color benefit” driven by more lapses (0-1 correct reports) for real-world objects? A between-subjects repeated measures ANOVA (Experiment x Stimuli) revealed a main effect of the stimuli type on the proportion of lapses (p=.018), but no between-subjects effect of the experiment (p=.865), nor interaction effect (p=.110). Thus, participants experienced more lapses for objects than simplistic stimuli overall, and this was not related to using trial-unique or repeating objects. Altogether, our study failed to replicate a real-world object benefit for VWM, and instead found more failures in VWM performance for real-world objects than simplistic stimuli.