Abstract
When observers perform an oddball feature search, their ability to select and discriminate a target is impaired if the feature that defines the target (e.g. “green”) also defined a field of homogeneous distractors on an immediately preceding target-absent trial. This phenomenon is commonly referred to as the distractor previewing effect (DPE), a robust inter-trial effect that reflects an inhibitory bias against selecting a recently rejected feature. In our previous studies, we have demonstrated that the DPE only occurs for features in the dimension that defines the target (search-relevant features), and no inhibition is observed along dimensions that never define the target (search-irrelevant features). In the present study, participants searched for an oddball that could occur either in the color or shape dimension (but never in both). On target-absent trials, all objects were identical (homogenous on shape and color). In addition, each stimulus contained a slanted line and participants were asked to report the orientation of the oddball's slanted line (if an oddball was present). Our results indicate that feature-based inhibition of attention (as indicated by a significantly positive DPE) was formed and maintained for both color and shape following a target-absent display. Furthermore, we find that inter-trial inhibition in a given dimension is partially modulated by recent experience with this dimension: if the dimension that defines a target (e.g. color) is repeated on trial N-2 and trial N (trial N-1 being a target-absent trial), the non-target dimension (e.g. shape, in this case) has no effect on the observed DPE. However, if the target dimension is switched from trial N-2 to trial N, inhibition along the non-target dimension on trial N magnifies the DPE along the target dimension. We conclude that inhibitory marking of features can simultaneously occur along multiple search-relevant dimensions, and is biased by recent experience.