AS PART OF THE RESEARCH ON THE INFLUENCE OF RESPONSE-CONTINGENT REINFORCERS ON THE LEARNING AND PROBLEM-SOLVING BEHAVIOR OF CHILDREN, THE EFFECTS OF A LIMITED NUMBER OF VARIABLES ON THE PERFORMANCE OF CHILDREN OF THREE AGE LEVELS (4-5, 7-8, AND 10-11), SELECTED EQUALLY FROM MIDDLE- AND LOWER-CLASS BACKGROUNDS, WERE INVESTIGATED. THE EXPERIMENTAL DESIGN, A 2-ALTERNATIVE DISCRIMINATION TASK, WAS ONE IN WHICH THE REINFORCING EVENTS PROVIDED THE SUBJECTS WITH THEIR ONLY SOURCE OF INFORMATION ABOUT THE CORRECTNESS OR INCORRECTNESS OF THEIR RESPONSES. THE EXPERIMENT USED SIX TREATMENT GROUPS. THESE USED TWO TYPES OF REINFORCERS IN EACH OF THREE REINFORCEMENT COMINBATIONS WHICH WERE REWARD ONLY, PUNISHMENT ONLY, AND REWARD OR PUNISHMENT. THE TWO REWARD-PUNISHMENT SETS WERE (1) "RIGHT" OR "WRONG" SPOKEN BY THE EXPERIMENTER AND (2) CANDY OR A LOUD, RAUCOUS SOUND. EACH SUBJECT HAD 15 LEARNING TRIALS. RESULTS INDICATED THAT (1) NO BROAD GENERALIZATIONS ABOUT EFFECTS OR REINFORCERS COULD BE MADE SINCE REINFORCERS INTERACT, NOT ONLY WITH SUBJECT VARIABLES, BUT WITH CHARACTERISTICS OF THE TASK, INSTRUCTIONS, AND OTHER SITUATIONAL VARIABLES TO DETERMINE PERFORMANCE, AND (2) BOTH MIDDLE-CLASS AND LOWER-CLASS CHILDREN PERFORMED BETTER FOR THE SYMBOLIC RATHER THAN THE MATERIAL REWARDS. (AL)