Abstract
Introduction: previous research shows that observers can selectively attend to one of two fields of overlapping, transparent moving patterns. We were interested in the extent to which observers could divide attention between two overlapping patterns simultaneously. Methods: we presented two overlapping random dot kinemategrams that varied in both color and motion (e.g. red upward moving dots superimposed with green downward moving dots). Brief decreases in velocity or decreases in luminance could occur simultaneously on both surfaces with 50% probability for a given feature on a given surface. Observers made a yes-no detection judgment of a cued feature. There were three kinds of cues: cues to monitor a specific feature of a single surface (single-task control), cues to divide attention across the two features within the same surface (dual-task within surface), or cues to divide attention between features on different surfaces (dual-task between surfaces). The effect of divided attention was measured as the dual-task deficit: the decline in performance for a dual-task condition relative to the corresponding single-task control. Results: there was little dual-task deficit for the within surface conditions but a large deficit for the between surface conditions. The surprise was that there was also evidence of a failure of selective attention: responses were affected by changes in irrelevant features. This effect was present in all conditions but was most pronounced when attention was divided between the two surfaces. Conclusion: dividing attention between two surfaces therefore results in the failure of selective attention to specific features. Our interpretation is that one cannot simultaneously process two surfaces and must fall back on non-selective processing of the features.
Meeting abstract presented at VSS 2012