Abstract
Researchers often study visual working memory (VWM) using spatially organized arrays of objects (e.g., differently colored dots at various locations). The availability of the originally encoded spatial information can aid retrieval at test. For example, change detection in a single item after retention can benefit from preserved locations of other items (Olson & Jiang, 2000). In another influential paradigm, continuous report, observers memorize a set of items and report one target cued by location using a continuous color wheel. Although the colors of non-target items are not shown at retrieval, is it possible that some spatial information still available at continuous report? We tested a hypothesis that the spatial adjacency of the tested item and a particular color on a test color wheel can affect observers choices. Our observers memorized three color dots and, after a 1-second retention interval, reported the color of a cued dot. The rotational position of the color wheel around the dots changed across trials. Unbeknownst to observers, the correct target color was rotated by 0°, 30°, 60°, 90°, 128°, or 158° away from the target. We analyzed the distribution of angular errors using a method of calculating the discriminability index d’ from continuous error distributions (Schurgin, Wixted, & Brady, 2020). We found that the d’ gradually declined as a function of distance between the target location and the correct answer on the color wheel. Further research should focus on whether it is a genuine effect on retrieval (like retro-cue) or response strategy (e.g., choosing the closest response). At this point, we conclude that the spatial relationship between the tested target and test alternatives (such as in color wheel) can have an effect on the results of VWM measurement that is typically overlooked in VWM research.