Choice Reaction Times and the Theory of Intermittency in Human Performance

Abstract

One visual signal, drawn from two equi-probable alternatives, was followed at a variable interval by a similar signal, also drawn from two equi-probable alternatives. The relationship between reaction time to the second signal and the interval between signals was studied under the following conditions: (1) Subjects were trained from the beginning to regard the first signal as irrelevant. (2) Subjects were asked to report on the nature of the first signal after they had responded to the second. The interval between signals was one of the five values, 50, 100, 150, 200, 250 millisec. In the regular interval situation the same interval was used over a block of 20 trials. In the random interval situation the values of the interval were randomly arranged, with equal frequency within each block of 20 trials. The results showed: (1) In the random interval situation there was a definite disturbing effect of the first signal on the reaction time to the second, despite the fact that subjects were trained to disregard the first signal. However, the delays in reaction time were significant only at the shortest interval, 50 millisec. (2) In the regular interval situation this disturbing effect was not evident. (3) The additional requirement of having to report on the nature of the first signal produced no consistent increase in reaction time to the second signal. Some implications of these results for the conception of the human operator as a channel of limited capacity are discussed.