After 40 forced-choice trials to acquaint Ss with the reinforcement schedules, 120 children between the ages of 3.7 and 5.4 were presented with a two-choice, partial reinforcement situation to test the hypothesis that they would prefer predictable to unpredictable schedules. Choices of one of the alternatives were reinforced on either a 50% alternating or a 50% random schedule, while choices of the other alternative were reinforced on either a 50, 60, or 70% random reinforcement schedule. A significant preference for the predictable schedule was found, but only in groups for which both alternatives paid off with an equal percentage of reinforcement. The effects of a predictable schedule were apparently weak enough to be overcome by a discrepancy in percentage of reinforcement of 10%. The relevance of this finding for theories of cognitive development stressing the importance of predictability, or congruity, are discussed.
Read full abstract