Purchase this article with an account.
Srivas Chennu, Patrick Craston, Brad Wyble, Howard Bowman; The influence of target discriminability on the time course of attentional selection. Journal of Vision 2009;9(8):196. doi: https://doi.org/10.1167/9.8.196.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
A wealth of neurophysiological data has demonstrated that visual attention can selectively enhance target representations early in the visual processing pathway. In conjunction, behavioural evidence tells us that the extent to which irrelevant items interfere with target processing depends on their featural similarity to the target. In this context, how does target discriminability influence temporal selection? We present results from an electrophysiology study that addresses this question by investigating the neural processes underlying the temporal dynamics of target selection.
Target distinctiveness is varied across a pair of blocked conditions: In the onset condition, letter targets could be discerned simply by their visual onset. In contrast, in the RSVP condition, they were distinguishable from temporally sequential digit distractors only based on semantic identity. The results suggest that, in line with previous findings, the relatively harder category discrimination task in the RSVP condition reduces the accuracy of target identification. Our electrophysiological data suggests that there are significant differences in the perceptual processing of the target in the two conditions, as indexed by early visual ERPs. Further, in the RSVP condition, we find an increase in the latency of the P3 ERP, which indexes target consolidation.
We ground this and previous empirical evidence within a theoretical framework for understanding the mechanism of attentional selection represented in the ST2 model, a neural network model of temporal attention and working memory. The model successfully explains a broad spectrum of behavioural and electrophysiological data relating to RSVP and the attentional blink. Here, by making theoretically justified changes to the model, we extend it to simulate the onset condition. We show that the model provides a convincing explanation of the pattern of experimental results, in addition to informing questions about early vs. late selection and the cognitive equivalence of target processing in masking and RSVP experiments.
This PDF is available to Subscribers Only