Inverse effects of time‐on‐task in task‐related and task‐unrelated theta activity

Abstract The phenomenon of mental fatigue has recently been investigated extensively by means of the EEG. Studies deploying spectral analysis consistently reported an increase of spectral power in the lower frequencies with increasing time‐on‐task, whereas event‐related studies observed decreases in various measures related to task engagement and attentional resources. The results from these two lines of research cannot be aligned easily. (Frontal) theta power has been linked to cognitive control and was found to increase with time‐on‐task. In contrast, theoretical frameworks on mental fatigue suggest a decline in task‐engagement as causal for the performance decline observed in mental fatigue. The goal of the present study was to investigate mental fatigue in time–frequency space using linear regression on single‐trial data in order to obtain a better understanding about how time‐on‐task affects theta oscillatory activity. A data‐driven analysis approach indicated an increase of alpha and theta power during the intertrial interval. In contrast, task‐related theta activity declined. This reduction of stimulus‐locked theta power may be interpreted as a reduction of task engagement with increasing mental fatigue. The increase of theta spectral power in the intertrial interval, moreover, could possibly be explained by an increased idling of cognitive control networks. Alternatively, it might be the case that the increase of theta power with time‐on‐task is a by‐product an alpha power increase. As alpha peak frequency systematically decreases with time‐on‐task, the theta band might be affected as well.