The effects of time-of-day, task and effort on resource allocation

Performance on a range of cognitive tasks is known to vary across the day. The precise effects of time-of-day depend on individual differences, the nature of the task and physiological factors. Several event-related potential (ERP) studies have found significant time-of-day effects on ERP components, including a reduction in P3 amplitude in the afternoon/evening compared to the morning (Higuchi, Liu, Yuasa, Maeda, & Motohashi, 2000; Wesensten, Badia, & Harsh, 1990) and increased P2 amplitude in the afternoon compared to morning (Wesensten & Badia, 1992). In contrast, other studies have reported minimal (or no) time-of-day effects (Broughton, Aguirre, & Dunham, 1988). Geisler and Polich (1990) suggest ERPs are only indirectly affected by time-of-day, with factors which vary across the day, such as food intake and body temperature, having more direct effects. With the exception of Higuchi et al., all published ERP investigations of time-of-day have used single-task paradigms, which may be less sensitive to time-of-day effects than dual-task paradigms. In addition, no reported studies to date have investigated effects of time of day on P3a. The primary aim of the current series of seven experiments was to determine the impact of time-of-day on ERPs during a series of dual-task experiments and to determine if a dual-task paradigm would be more sensitive to time-of-day effects than a single-task paradigm. As voluntary levels of effort invested in task performance have been reported to increase during periods of low arousal (Hockey, 1997), it is possible fluctuations in effort mask effects of time-of-day on resource availability. Therefore, an additional aim of this series of experiments was to determine whether voluntary effort invested in tasks varies across the day and if so, the impact of this on ERP components. During Experiment 1, 12 participants completed a visual oddball braking task and a concurrent visual tracking task with varying priority, at 0900 and 1300. When priority was given to the oddball task the effect of time-of-day was significant, with amplitudes of P2 and P3b greater at 0900 than at 1300. Fourteen participants completed a single oddball braking task and easy and difficult dual-tasks (braking and tracking) during Experiment 2. P2, N2 and P3b amplitude were greater during the single-task than both dual-tasks and there was no evidence of P3a during any of the task conditions. Modification of the braking task stimuli during Experiment 3 resulted in significantly greater P3a amplitude during the single-task compared to the dual tasks and lower P2 and P3b amplitude during the dual difficult task than the dual easy task. During Experiment 4, 16 participants completed single, dual easy and dual difficult tasks at 0900, 1300 and 1700. The effects of time-of-day on P3b were not significant during the single-task however during the dual-tasks significant time-of-day effects occurred with larger P3b amplitude at 1700 than at 1300. At 1300 P3a was consistent across the tasks, whereas at 0900 and 1700, P3a amplitude was greater during the single-task than the dual-tasks. Experiment 5 aimed to determine if effort modulated ERP effects. Sixteen participants completed easy and hard versions of a cued visual RT task, at 0900, 1300 or 1700, in which they were instructed to used standard or 'extra' effort. Contingent negative variation (CNV) was significantly greater following effort cues than standard cues; however the effects of time-of-day were not significant following either cue type indicating effort was consistent across the day. During Experiments 6 and 7, 16 participants completed the tasks used in Experiments 4 and 5, in caffeine and placebo conditions, to determine whether previous significant findings were consistent with effects of increased arousal associated with caffeine. The results of Experiment 6 revealed a trend towards greater P3b amplitude in the caffeine condition than the placebo condition. During Experiment 7 the effect of caffeine on CNV was not significant suggesting levels of voluntary effort were consistent across both conditions. Overall, the results indicated time-of-day has significant effects on several ERP components including P2, P3a and P3b; however it appears the effects on P2 and P3b are only evident during tasks which are sufficiently difficult to require total capacity of attentional resources. Dual-tasks appear to be more sensitive to time-of-day effects than single-tasks, with dual-tasks that engage the same supply of attentional resources being more sensitive than those tasks which place demands on wider resources. Although subjective effort varied across the day, changes in arousal due to time-of-day and caffeine did not significantly affect levels of effort invested in task performance, as measured by CNV amplitude. Together, the findings suggest time-of-day variations in arousal affect ERP measures of resource allocation, but not effort, during tasks which place high demands on resources.