“Variables that Influence Random-Generation”: An Alternative Hypothesis

An alternative to Tune's memory hypothesis is proposed as the basis of the relative inability of Ss to generate a random sequence of responses. The memory hypothesis is inconsistent with results of studies showing decreased response stereotypy as a function of (a) time spent working on a task and (b) presence of potentially distracting extratask stimulation. Success at randomizing choices requires reduced awareness of previous responses and minimal sources of distraction to prevent Ss from “giving up the set.” Research implications of this attention-distraction hypothesis are discussed.