‘Same’-‘different’ response times: A further test of a ‘counter and clock’ model☆

Abstract Three experiments were conducted for the purpose of evaluating further a model that was described in a previous paper ( Nicherson , 1969). The model was intended to be applied to situations in which ‘same’-‘different’ 2 judgments must be made with respect to stimuli that are allowed to differ only on a single perceptual dimension. In a previous experiment, the model had been tested in a situation in which the magnitude of the difference between different stimuli was constant over an experimental run. The purpose of experiments 1 and 2 of the present series was to extend the model to the situation in which differences of several magnitudes are used within the same session. The main implication of the model for this case is the following: Whereas ‘different’ response times (RTs) should decrease with increases in the magnitude of the difference between stimuli, (incorrect) ‘same’ RTs should be relatively insensitive to this variable. The results were generally consistent with this expectation. In experiment 3, a Donders' c-type task was used in the hope of simplifying the situation, thereby getting a somewhat purer measure of the time required to make the ‘same’ or ‘different’ decision. Contrary to expectations, response times were generally longer in this case than in the preceding experiments. It was suggested that the general increase in response times might have been due to the occurence of revised or ‘second-thought’ decisions, which the Donders' c-type task permits. This conjecture was made the more plausible by the fact that under both ‘respond-to-same-only’ and ‘respond-to-different-only’ conditions, errors of commission were about three times as frequent as were errors of omission. Although the possibility of the occurence of revised decisions makes the model less appropriate for the c- than for the b-type task, most (but not all) of the predictions that were developed with the latter task in mind were in fact obtained in this experiment