Abstract
Perceptual aftereffects arising from adaptation have been widely used as a tool for exploring the neural representation of visual stimuli as simple as spots of color or as complex as faces. In most adaptation studies subjects make decisions about the appearance of a test stimulus presented alone or after viewing an adapting stimulus. Changes in decisions are commonly attributed to changes in sensory representation. Potential changes in the decision-making strategy, however, are rarely studied. Here, we utilize recent advances in our understanding of the decision-making process to disentangle representational and decisional effects of adaptation. A pure representational effect modifies the sensory evidence furnished by the test stimulus and is expected to cause equivalent shifts in the subject’s psychometric and chronometric functions. On the other hand, a change in the starting point or termination criterion of the evidence accumulation process can introduce asymmetries in the chronometric function. We tested these two hypotheses in a facial emotion adaptation task in which the subject reported whether a test face was happy or sad. The test stimuli were drawn from a spectrum of happy to sad faces of an individual and evaluated in three conditions: no adaptation, adaptation to a happy face, and adaptation to a sad face. As expected, adaptation shifted the perceived neutral face toward the adaptor. Surprisingly, reaction times for the two response categories (happy, sad) changed asymmetrically: for a fixed test stimulus responses opposite to the adapting category (adapt happy respond sad) were faster. A second experiment shows that our observation is invariant to adaptor size or location. These behavioral patterns are more compatible with a change in the decision-making process than a representational change. It is possible that the underlying neural mechanism for high-level face adaptation differs from low level adaptation, which may be best explained by representational changes.
Meeting abstract presented at VSS 2013