Noise stress and circadian arousal in self-paced computation

Previous work(Frankenhaeuser & Lundberg, 1977) has shown that unpaced mental arithmetic is performed at slower rates in noise, despite unchanged catecholamine indices of arousal; only male subjects were used, tested early in the day. Since the times of testing entail arousal effects that interact with noise stress, and the sex of subject further modifies these interactions, a new experiment was designed to include these variables. Men and women were tested on the Norinder mental arithmetic task, in quiet or in noise, either in the morning, when arousal was low, or during the early evening, when temperature curves indicate that arousal should be high. Analysis of the number of problems attempted shows a significant drop for men in noise in the morning, but an interaction due to reversal of the noise effect in the evening; there is no main effect of noise when women's scores are included. The results are compatible with interpretations combining motivation and cognition, and they demonstrate the importance of the experimental variables in explaining noise stress effects.The authors are indebted to the late David A. Shipp for his assistance in running subjects.