A Reexamination of the Role of Incentive in Children's Discrimination Learning

Abstract

2 studies are reported. The first examined the value of 12 typical incentive objects among 24 fourth-grade children by means of (a) rating scale, (b) paired comparisons, and (c) behavioral choice. Significant agreement was obtained between methods, and significant differences were obtained between objects. In the second study, 72 fourth graders reevaluated 4 objects from the first study and then performed in a discrimination learning task that manipulated: (a) type of reinforcement (S's most or least preferred object, or verbal feedback), (b) reinforcement contingency (for correct responses or errors), and (c) instructions (whether or not S was informed of the reinforcement contingency). Better performance was associated with higher-valued incentives, but high-value incentives given for errors produced as many correct responses as when given for correct responses. Best performance of all occurred when only verbal feedback was given. The results suggested that observed "reinforcement" effects may simply be the resultant effect of several variables, for example, motivation, information, and interference, acting in combination.