The Effect of Superseding Signals

Abstract

Eight adult human subjects were given a step-tracking task in which an occasional second signal within 50, 70, 90, 120, or 240 millisec, called for curtailing or reversing the first command. It was found for inter-signal intervals through 120 millisec. that the shorter the interval the greater was the reduction in amplitude and duration of the majority of responses, with no delay in the effect of the second signal. Where a larger change of response was called for, reversal rather than curtailment, there was a greater effect. A second signal occurring at the 240 millisec. interval (in almost all cases after the start of the response), had no detectable effect. Since the over-all RT was about 180 millisec., it is evident that for at least the first two-thirds of the RT period the initial response is not typically impervious to the effect of a second signal. Contrary to the expectations of the uncommitted-period version of the hypothesis of substitutive grouping a reversing signal at the 50 millisec. interval did not yield many reversed responses. Moreover this view cannot accommodate the finding that for intervals through 120 millisec., relatively few distributions of response amplitude can be accounted for by the summation of instances of response to the first signal alone and to the second signal alone. It is concluded that for these intervals, there were generally either overlapping responses to the two signals or else unitary responses in which the two signals were grouped to produce a combined effect.