Abstract
The reaction time (RT) to identify the orientation of a tilted-line target (two alternative responses) is longer when the target is flanked by incongruent distractors (whose identity is the competing target) than when flanked by neutral distractors. This delay is assumed to reflect distractor interference. Max and Tsal (2015) employed the mutations paradigm to assess the time-course of distractor processing: In each trial, while the target remained unchanged throughout the presentation, neutral distractors were replaced by incongruent distractors at a random time between 8 and 66 ms after onset. Results (Fig. 1): RTs decreased as the duration of the neutral display increased, up to 25 ms. When incongruent distractors appeared after 25 ms or later, RTs matched those in a neutral baseline. We consider two possible interpretations: (a) In neutral displays, the target may have been fully processed within 25 ms, which rendered incongruent distractors presented later than 25 ms ineffectual. (b) The distractors may have been suppressed within 25 ms, thus allowing target processing to continue unimpaired beyond 25 ms. We tested these two hypotheses by presenting the neutral display for periods varying between 8 and 91 ms, followed directly by a pattern mask (###). If targets are fully processed within 25 ms, the masking should impair target identification during the first 25 ms, but not beyond. In contrast, if distractors are suppressed within 25 ms, masking should remain effective beyond 25 ms because target processing would still be underway. Results (Fig. 2): Masking was effective for at least 74 ms after target onset, supporting the distractors-suppression option. Homologous patterns of results emerged with incongruent displays, albeit with a longer timecourse (58 and at least 91 ms for distractor suppression and target processing, respectively). These results have implications for perceptual load theory and for early-selection and late-selection theories.
Meeting abstract presented at VSS 2016