Abstract
Conscious perception of a visual stimulus can be impaired both by decreasing the signal strength in the stimulus as well as by distracting attention from the stimulus. In the present study, we report a method that allows us to classify different types of psychophysical techniques for rendering visual stimuli invisible.
In experiment 1, we varied the visibility for a luminance blob by varying its contrast. Subjects were asked to report the presence or absence of the luminance blob together with their confidence rating: high/mid/low. In experiment 2, we maintain the contrast of luminance blob well above threshold but varied its visibility by manipulating the difficulty of a concurrent central visual search task. Observers detect the presence of blob with confidence rating (high/mid/low) as in experiment 1 in addition to their search result.
In both experiments, the target blob was presented only half of the trials. Between the two experiments, we identified conditions yielding a comparable d' and analyzed the confidence in “miss” trials in which subjects reported absence when the target was present. We found that subjects reported absence with high confidence when they missed a target due to low contrast, whereas they reported absence with little confidence when their attention was distracted. In the case of low-contrast stimuli, the confidence in reporting absence was as high as reporting absence in “correct rejection” trials in which no stimulus was presented. The distinct patterns in the confidence rating between the two types of blindness implies that blindness at a perceptual stage and attentional stage can be distinguished by objective measures based on a second-level signal detection framework.