Psychological and Brain Sciences Brown Bag Seminar: Brad Stilwell & Brett Bahle
Psychological and Brain Sciences Departmental Brown Bag Seminar
Brad Stilwell, Psychological & Brain Sciences, The University of Iowa
Learned Distractor Rejection in the Face of Strong Target Guidance
Abstract: Most theories of visual attention posit a selective mechanism that allows relevant information to proceed through processing while filtering out irrelevant, distracting, stimuli. Visual search is more efficient when target-features are known in advance, through guidance with a target-template. Further, visual attention can learn to reject distractors through experience with statistical regularities, by establishing an experience-based template for rejection. The current study was aimed at determining whether visual attention could learn to reject distractors through feature-based experience in the presence of strong target guidance, and under which circumstances this learning occurred.
In three experiments, we presented individuals with two-color, spatially unsegregated visual search displays. Prior to each search display, individuals were presented with a word cue either indicating the target color or the word “neutral”. For each individual, the target never appeared in one color: the learned distractor color. In Experiment 1a, we presented neutral cues only, and observed learned distractor rejection, namely, faster mean RTs on trials with the learned distractor color present than absent. In Experiment 1b, to demonstrate learned distractor rejection with target guidance, we presented both types of cues, and observed a target cueing effect (faster mean RTs on trials following a target cue than a neutral cue), and critically, learned distractor rejection following both neutral cues and target cues. In Experiment 2, we asked whether learned distractor rejection required the neutral trials to establish learning, and found that it did not. Finally, in Experiment 3, we provided more direct target guidance, namely, the cue directed individuals to the only color-matching item in the display and we still observed learned distractor rejection during target guidance. Thus, we argue that visual attention can establish both target templates and templates for rejection and use both templates to guide visual search.
Brett Bahle, Psychological & Brain Sciences, The University of Iowa
Visual Search for Categorical Targets Is Biased Toward Recently Viewed Exemplars
Abstract: The guidance of attention toward task-relevant objects during visual search often relies on a target template representation. In addition, many searches are categorical, where the goal of search is to find any item belonging to a category. In the present study, we investigated the extent to which these categorical template representations are biased toward recently viewed category exemplars. Participants first completed an exposure task in which they viewed pictures of objects from common categories and classified them as natural or artifact. Critically, each exemplar from a given category always appeared in one color (e.g., all bears were brown). After completing the categorization task, participants performed four blocks of visual search. They saw a label specifying the target category and searched for the target picture within an array of objects. The target picture matched (e.g., a different brown bear) or mismatched (e.g., a black bear) the color of the previously viewed exemplars. Searches were more efficient for matching compared with mismatching color exemplars, and this effect reliably diminished across search blocks, as participants accrued more trials of search with exemplars of both colors. Moreover, eye-tracking results demonstrated that this effect was due to both more efficient guidance and categorization to/of the target, as evidenced by faster first fixations on the target object and faster/more accurate decisions when it matched compared with when it mismatched the previously viewed exemplar color. These results suggest that well-established category representations guiding visual search are constantly updated to reflect recent visual regularities within a category. The relationship between these findings and general theories of category structure will be discussed.