The Effects of a 'Compound Set' on Learning Efficiency
Peter K. Janke · The American Journal of Psychology · 1962
A recent experiment by DiVesta and Blake reported evidence for differential effects on learning of three combinations of (a) verbal reinforcement, and (b) knowledge that some principle was the basis for reinforcement.l The sets induced by instruction were presumed to permit verbalizations which mediated adaptations to the problem in the form of hypotheses. As might be expected, they found that a set to 'not look for a principle' resulted in making a task involving a principle signiScantly more diEcult to learn than did a set to look for a principle, aIthough similar reinforcement was given in both cases. In short, a set of a positive nature proved more beneficial for learning than did a set of a negative nature. Following closely the experimental design used by DiVesta and Blake, the present experiment attempts to investigate further the effects of set on learning efiiciency. Whereas previous experiments in this Seld have dealt with relatively simple 'homogeneous' sets, however, the present study deals primarily with a set combining both positive and negative elements-a socalledtheterogeneous' or 'compound' set. It was hypothesized that the negative component of the compound set would become less effective with time (as a result of non-reinforcement), allowing the positive component to be reflected in an increase in performance. Aus, since the compound set is assumed to produce a decrement in performance not shared by the positive set, and an increase not shared by the negative set, the prediction was made that it would produce a learning curve which would fall somewhat between the cunes produced by the positive and negative sets.