Longer Memory Delay Reveals Demands for Maintaining Multiple Features

[摘要] Interacting with real-word environments requires maintaining individual features and integrating them into a representation of a whole object. This requires encoding objects and their features as integrated wholes successfully into our visual short term memory (VSTM). Theories of VSTM suggest that remembering multiple features of an object does not require additional memory resources, but aging populations and Alzheimer's patients show specific deficits in maintaining bound objects. These processes could reflect impaired memory maintenance, but it is also possible that apparent memory deficits reflect perceptual and attention processes during encoding. To better understand the possible reasons for this discrepancy, we increased the maintenance demands on the VSTM system by increasing the memory delay in a change detection task. The increase in delay should indicate whether additional cognitive demands are needed to maintain two versus one feature objects. Subjects were run on a VSTM change detection task using both a 2000ms and 5000ms delay. On each trial, subjects were cued to detect a change in color, shape, or either. For the either cue, either color or shape could change. Our findings support the literature that indicates VSTM capacity is similar for single and two feature objects at shorter delays. However, we found a significant cost for remembering two features objects in the 5000ms delay. These results demonstrate that at longer memory delays VSTM for two-feature objects requires additional resources as compared to a single-feature condition. By taxing the maintenance demands of healthy individuals we can differentiate performance for single and two feature conditions. These findings offer a possible solution for the disparity of findings in healthy and clinical populations.

[学科分类] 眼科学