Subjective randomness and the capacity to generate information

Literature concerning randomization tasks contains contradictory results since both increases and decreases of nonrandomness have been reported as a function of the response rate. The present paper proposes that the contradictory results are attributable to differences in experimental procedure and the different mathematical definitions of ‘objective’ randomness. In an experiment Ss randomized 2, 3, 4, 6 or 8 alternatives with speeds of 4, 2, 1 and 0.5 sec per response. The results showed that there was no significant effect of response rate. There appeared to be substantial individual differences for first- and second-order nonrandomness.