DISCRIMINATION LEARNING IN CHILDREN AS A FUNCTION OF REINFORCEMENT VALUE

Abstract

The present study is concerned with the influence of reinforcement value on learning. Reinforcement value is defined as the individual S's relative preference for various types of reinforcers, when amount of reinforcement and probability of occurrence of reinforcement are held constant.2 The results of several studies show that, on the average, different types of reinforcers are differentially effective for learning (1, 5). For example, Terrell and Kennedy found that children learn faster when rewarded with candy rather than with praise (5). However, despite the fact that children as a group learn more effectively on candy reinforcement, it is hardly conceivable that any one type and amount of reinforcement has an exactly equal reinforcing effect for all Ss of any such group. Given one group of Ss and one type of reinforcement, there must still be some intragroup variability in effectiveness of that reinforcement as a function of variability in value of the reinforcer for the Ss. In other words, for different people, the same objective or external reinforcer may have different reinforcement values, and if so, the within-group variability in reinforcement value should be reflected statistically as a large standard deviation in number of trials to learning criterion. The present experiment used a discrimination learning situation to compare the relative effects of two experimental treatments: identity of reinforcement value vs. identity of external reinforcing object. Ss run under the first condition (group RV) were allowed to choose the most preferred one of three reinforcers: M & M candies, varicolored marbles, or varicolored plastic trinkets. Ss of the second experimental group (group R) were 1 The authors wish to express their thanks and gratitude for the cooperation and assist