Stroop-Type Interference in Random-Number Generation

Abstract

We describe a positive correlation between the extent of sequential counting in a random-number generation task and the magnitude of the interference effect in Stroop's color-naming task. This finding is compatible with the view that both counting and reading are highly automatized processes which constitute an inevitable source of interference in randomization and Stroop paradigms, respectively. On the other hand, cognitive psychological theories proposing that a generally biased concept of randomness would be responsible for human subjects' inability to generate true random sequences do not readily account for this correlation. Literature on repetition avoidance indicates this universal effect in random generation is likewise not explainable in terms of some “biased concept of randomness.” Repetition avoidance (“spontaneous alternation”) also occurs in lower invertebrates, is largely independent of mathematical sophistication in humans, dissipates with increasing time between consecutive responses, and is diminished by amnesia. We conclude that the failure of functionally intact organisms to display random behavior is due to basic neuropsychological limitations. In neglecting these biological foundations, “concept of randomness” theories of randomization behavior lack explanatory value.