Abstract
Decreases in old-new recognition performance may be due to confusability of items in memory. Hue-dependent mask-induced interference has been shown to affect stored representations of colour stimuli within short-term perceptual memory by reducing the fidelity, or sensitivity, with which information is stored in memory. We manipulated sensitivity using a delayed hue-discrimination paradigm in a recognition memory task. We examined old-new recognition judgments for cardinal stimuli that are embedded in a continuous, multidimensional space. Cardinal colours typically do not have learned colour names and are perceived as desaturated and more perceptually mixed than monochromatic stimuli. Five colour-normal subjects completed the experiment. Stimuli were 8 degree centrally-located patches from the subjectively equiluminant plane of cardinal colour space, presented for 1 sec (500ms half-width) in a raised-cosine temporal-envelope where the inter-stimulus interval between the reference and test stimuli ranged from 0.5ms to 5s. The patches consisted of 14 colours drawn from equally spaced intervals within the ranges 350° to 11° and 34° to 56°. To measure sensitivity, the observers' task was to make same-different judgments for the pair of stimuli. Three types of masks were examined: (1) a mask that was either similar in hue to the study and test items, (2) a mask that was drawn from the opposite region of colour space, and (3) an achromatic (luminance) mask. In the control condition, no mask was presented. The data show the masking conditions had virtually no impact upon similarity-based recognition judgments. These findings are inconsistent with previous studies demonstrating mask-induced interference, particularly when the mask is identical or of a similar hue to the study and test item. These findings suggest that short-term memory for cardinal colours is robust despite the presence of internal (the delay) or external (the mask) noise.
Meeting abstract presented at VSS 2017