Abstract
Recently, Glaholt & Reingold (in press) reported biases in looking behavior during eight-alternative forced-choice (8-AFC) decision tasks. We found that throughout the entire decision period, dwell duration (where a dwell is a run of consecutive fixations on an item) is longer on the chosen item than on not-chosen items. In the current study we investigated the ways in which this dwell duration bias depends on stimulus exposure. Eye movements were recorded while participants selected from a display of eight photographs the photo that they preferred the most, or the photo that was the most unusual (control task). In Experiment 1, we manipulated stimulus exposure by pre-exposing four of the eight photos, for one second each, prior to each 8-AFC decision. In Experiment 2, stimulus exposure was manipulated within-trial using a gaze-contingent methodology that limited viewing time during each dwell to either 200 ms or 400 ms. Experiment 1 revealed a larger dwell duration bias for items that were not pre-exposed compared to items that were, which suggests that the bias involves the selective encoding of task-relevant information. In Experiment 2 we found that a dwell duration bias is present even when stimulus exposure is equated between the chosen and not-chosen items, indicating that post-perceptual factors also contribute to the bias. In addition, we examined the component fixations within each dwell and discovered that the bias in dwell duration is the result of both an increase in the number of fixations per dwell as well as a lengthening of individual fixation durations. A similar pattern of results was present in both the preference and control tasks. Implications of the present findings are discussed in the context of the Gaze Cascade model of preference decision making (Shimojo, Simion, Scheier & Shimojo, 2003).