Abstract
Sustained attention is critical to daily functioning but the relationship between perceptual processing during sustained attention and the ability to remain focused is incompletely characterized. To explore this issue, we administered the gradual onset continuous performance task (gradCPT), a sensitive sustained attention task, to a large web-based sample (N>10,000, testmybrain.org). The gradCPT requires participants to respond to the majority of stimuli (cities-90%) and withhold to rare target images (mountains-10%). The images gradually transition from one to the next every 800ms, which eliminates the exogenous effects of abrupt stimulus onsets. By computing the pairwise pixel-image correlations of the stimuli across category (cities and mountains), we were able to assess which city exemplars were most "mountain-like" and which mountain exemplars were most "city-like". If perceptual representations influenced performance, we predicted that RTs for city images that were more "mountain-like" would be slower and "city-like" mountain images would be more prone to erroneous button presses. Consistent with our predictions, we found that the degree of cross-category similarity significantly correlated with RTs for cities (r=0.37; p< 0.0001) and erroneous button presses for mountains (r=0.44; p< 0.05). Furthermore, by computing the visual similarity of the top and bottom halves of the images separately, we revealed a bottom-half bias for cities and a top-half bias for mountains. Notably, individual differences in sensitivity to stimulus similarity were correlated with measures of ability (d-prime r=0.36; RT variability r=-0.44) and only weakly correlated with measures of strategy. In other words, individuals who displayed greater sustained attention ability were more likely to be influenced by stimulus similarity, suggesting enhanced perceptual processing is associated with greater sustained attention ability. Ultimately, stimulus similarity provides a new means to explore perceptual processing during sustained attention and could provide a novel behavioral marker for disorders of attention.
Meeting abstract presented at VSS 2016