Original article

Scand J Work Environ Health 1998;24 suppl 3:109-114    pdf

Diurnal trends in mood and performance do not all parallel alertness

by Owens DS, Macdonald I, Tucker P, Sytnik N, Minors D, Waterhouse J, Totterdell P, Folkard S

Objectives This study examined the hypothesis that alertness can be used to predict time-of-day effects on performance.

Methods For 6 or 7 days the volunteers (24, highly preacticed young women) were required to retire to bed at 0000 and were awakened at 0800. A battery of mood and performance tests was completed every 2 hours while the women were awake; the result was 9 equally spaced measures per day. Measures of mood, serial reaction time, and memory scanning were recorded. Rectal temperature was recorded continuously.

Results After omitting the data from the first day to avoid any carry-over from the "first-night" effect on sleep, average time-of-day functions were calculated for each subject, for each variable, and were then z-transformed. Cross-correlations between the pooled time-of-day trends indicated that, while alertness was a reasonably good "predictor" of the simple perceptual-motor speed measures, it fared less well for some of the other measures. Two-way analyses of variance indicated that the time-of-day trend for all measures differed from that for alertness, although the magnitude of this difference varied substantially and, for some measures, was very largely due to the last reading of the day (0000).

Conclusion It is clear from these results that, while alertness may successfully "predict" variations in some measures of performance capability, and especially those of simple perceptual-motor speed, care should be exercised in extrapolating to other performance measures.