Purchase this article with an account.
Bjorn Hubert-Wallander, Geoffrey M. Boynton; Domain specificity in integration of visual information across time. Journal of Vision 2014;14(10):1130. doi: https://doi.org/10.1167/14.10.1130.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Perceptual decisions often require integration of information across time. Even a static scene is effectively sampled over time via saccades. Here we consider how humans combine such serially presented information into an overall judgment, and whether the specific type of visual information (e.g., object location/size) affects that integration process. In Experiment 1, subjects viewed a sequence of ten serially presented, small white dots that varied in spatial position. After each trial, they were asked to indicate the perceived center of the sequence via mouse click. A set of ten weights was obtained for each subject that reflects the average influence that each of the ten dots had on the overall judgments. The results clearly show a much larger influence of early dots on the perceived center than later dots (a "primacy" effect), with the first item being weighted on average twice as much as each of the last two items. This primacy effect was replicated across manipulations of stimulus timing and spatial arrangement. In Experiment 2, subjects judged the mean size of eight serially presented discs that varied in radius. Unlike judging mean location, judging mean size across time shows stronger weighting for discs presented later in the sequence – a "recency" effect. We believe these apparently contradictory results can be accounted for by proposing that the dorsal and ventral visual pathways each have a unique method of integrating serially presented information, and that each emphasizes different epochs in the information stream. Specifically, judging mean location of a series may be supported by the dorsal 'where' pathway, while judging mean size may recruit the ventral 'what' pathway. Finally, we show that this novel hypothesis can also unify seemingly contradictory findings in the small number of previous studies that investigated similar effects across a variety of visual domains.
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only