Abstract
The capacity to anticipate environmental demands and adjust spatial attention accordingly is a crucial aspect of attentional control. As an individual’s goals change, they must regulate their readiness to shift attention, such that the cost associated with shifting decreases as the likelihood of shifting increases (e.g., Sali et al., 2015). Moreover, the efficacy of sustained attention is subject to variability over time, with periods of high and low response time (RT) variability indicating unfocused and focused states, respectively (e.g., Esterman et al., 2013). In the current study, we measured the behavioral consequences and neural correlates of both forms of control fluctuations simultaneously. While inside an MRI scanner, participants completed a variant of the gradual continuous performance task (gradCPT) in which they detected frequent targets among distractors in one of two simultaneous visual streams, shifted or held attention in response to embedded visual cues, and made a digit categorization as a measure of shift readiness. We manipulated the shift likelihood across blocks of trials. As in previous studies, blocks with a high shift likelihood were associated with a reduction in response time shift costs relative to blocks with a low shift likelihood, reflecting learned shift readiness. We defined periods of high and low sustained attention efficacy, referred to as “in the zone” and “out of the zone,” respectively, based on ongoing changes in continuous performance task RT variability and found that commission errors were significantly more frequent when participants were “out of the zone” than when “in the zone.” Functional MRI analysis suggested that separate but interacting neural systems account for moment-by-moment changes in attentional control, with the frontoparietal cortex detecting and responding to violations of shift readiness and the default mode network regulating ongoing changes in sustained attention.