GENERALIZATION OF THE EFFECTS OF TEACHER‐ AND SELF‐ADMINISTERED TOKEN REINFORCERS TO NONTREATED STUDENTS

Ten, black, second-grade boys served in a series of single-subject studies. They were from poor families, did not do well in arithmetic, were deficient in sustained attention, and presented behavior problems at school. One boy was a therapeutic confederate. Of the remaining nine nontreated students, three observed the confederate reinforced by a teacher, three observed the confederate self-reinforce without having an opportunity to use "self-reinforcement" themselves, and three observed self-reinforcement while having an opportunity to use "self-reinforcement." The target behavior was attending. Other measures of outcome were glancing, academic achievement, and accuracy of reinforcement. The basic experimental design consisted of an ABAB withdrawal applied to the confederate while the nontreated students remained on baseline. Generalization was expressed as a ratio (i.e., percent change in the generalization measure divided by percent change in the target behavior). Teacher-administered reinforcers to the confederate did not produce generalization of any kind. Both arrangements of self-administered reinforcers to the confederate produced across-subjects generalization and subject-response generalization. Additionally, the confederate manifested response generalization.