Abstract
Three experiments investigated factors affecting color precision with a fixed object load in visual working memory. The first experiment investigated color similarity and grouping across objects. Six colored-dots were presented every trial. In half the trials, three dots were the same color, and three dots were unique colors. In other trials all dots were unique colors. After a 500ms delay, participants selected a color from a wheel for a single probed dot. Color recall was significantly more accurate when the probed dot was identically-colored than uniquely-colored. The results could be due to the decrease in total quantity of colors to-be-remembered or due to the limited color range in trials with fewer colors. Experiment 2 tested the limited range hypothesis by presenting either a narrow or wide color range using three dots per trial. After a 1000ms delay, color memory for a single probed dot was significantly more accurate for a narrow than wide range of colors. Standard deviations and patterns of responses were comparable between both conditions, indicating that results were not due to guessing. However, in the narrow range condition, the test color wheel contained unprobed colors. These unprobed colors may have either allowed participants to rule out some color wheel options, or aided the recovery of the probed color by giving context. Experiment 3 used identical stimuli to Experiment 2 with participants choosing from six carefully selected options at response instead a color wheel. Probed colors were recalled much more accurately when unprobed colors were given as options than when decoys were all from the same color group (e.g. shades of blue). The results indicate that participants can use context at the recall phase to aid selection of the correct color, and show both encoding and retrieval manipulations influence color precision in working memory.
Meeting abstract presented at VSS 2014