Conflict and control in error monitoring during sustained visual discrimination

Increases in response errors across time-on-task are ubiquitous in vigilance research, particularly for tasks that require visual operations. An important aspect of vigilance tasks is response monitoring, particularly when a prepotent response must be inhibited. Neural indices of performance monitoring, such as the error-related negativity (ERN), provide strong evidence for changes in response conflict across time during vigilance tasks; however, it is unclear whether the allocation of additional cognitive resources (i.e., working memory) during such tasks leads to differences in response errors and performance monitoring. To examine this relationship, we employed a visual discrimination task in which participants must compare pairs of lines either to each other (simultaneous discrimination) or to a template held in working memory (successive discrimination) and decide whether to withhold or provide a response across 4 blocks of trials. We predicted lower ERN amplitudes and greater false alarm rates during visual discrimination trials that employ working memory (successive discrimination) compared to those that do not (simultaneous discrimination). While ERN peak amplitude was significantly greater during simultaneous compared to successive discrimination for incorrect responses, participants showed significantly higher false alarm rates during simultaneous discrimination, contrary to our hypotheses. To further investigate these results, we developed and examined outputs from computational models of the two experimental tasks based on a parallel-distributed processing account of cognitive control in which working memory (during successive discrimination), response conflict, and effortful control modulate continuous task performance. Task simulations suggest that pre-stimulus processing of a working memory template during successive discrimination helps to lower false alarms and alleviate post-stimulus error monitoring. These results are interpreted in the context of resource allocation and compensatory control with respect to the vigilance decrement.