Abstract
A modified version of the masking paradigm first introduced by Averbach and Sperling (1961) was used to explore the effects of spatially-cued attention on metacontrast masking. Eight stimulus characters were presented in a circular array; following a variable delay (the target-prompt SOA), observers were prompted to classify the target as a digit or a letter; this basic procedure showed an expected decline in accuracy as a function of SOA, reflecting the decay of iconic memory. When the prompt was accompanied by a circle surrounding the target character, accuracy was a U-shaped function of SOA, showing a metacontrast masking effect. Both these results replicate Averbach and Sperling (1961). However, when the character array was preceded by a valid cue to the target location, accuracy was increased, and the cue eliminated the decline in accuracy with increasing target-prompt SOAs. After the effects of cuing on the availability of the target in iconic memory were accounted for, an additional effect of the spatial cue on masking remained: although the U-shaped metacontrast masking function remained in both cue condtions, validly cued targets were more accurately classified, and invalidly cued targets showed a smaller but significant decline in accuracy, relative to the control condition. The most relevant effect of the spatial cue for current theories of attention and perception is that directing attention to the target location decreased the target-mask SOA at which target classification accuracy reaches its minimum (that is, the masking function shifts to the left). This is important because previous results show that if display luminance and/or contrast are used to increase the salience of a masked target, the masking function is shifted to the right. This suggests that with respect to the metacontrast masking, stimulus intensity and attention operate in different ways.