Abstract
The change detection paradigm has been used to implicate perception, memory and decision in effects of divided attention for simple features such as orientation or contrast (e.g. Mayo & Maunsell, 2016; Pestilli, Carrasco, Heeger, & Gardner, 2011; Scott-Brown & Orbach, 1998). We ask whether the effects of divided attention are due to perceptual limits, memory encoding or storage, memory retrieval and/or decision. We use three tasks – simple change detection, post-cued change detection and retro-cued change detection – using the same stimuli and procedure to determine the source of the limits in change detection. Subjects were presented with two intervals of dynamic 1/f noise patches on the left and right sides of fixation and were pre-cued to attend to either one side (cue-one) or both sides (cue-both). For the simple change detection task, Gabors were embedded in the noise on both sides, in both intervals, with a 50% chance of a change occurring between intervals somewhere. For the post-cued change detection task, stimuli were the same (including cue-one and cue-both conditions) but now the chance of change was 50% independent on each side with a post-to indicate the relevant side. For the retro-cued change detection task, stimuli were the same but an additional cue was added between the intervals matching the post-cue. For the simple change detection task, subjects performed worse in the cue-both condition compared to cue-one. This effect was reduced but still reliable for the post-cued change detection task. However, for the retro-cued task where only one memory retrieval, comparison and decision must be made, there was no difference between the cue-one and cue-both conditions. Results are consistent with divided attention effects in this change detection task being due to the processing of memory retrieval, comparison and decision. They are inconsistent with accounts based on perception, memory encoding and storage.
Meeting abstract presented at VSS 2018