Reward Contingencies and the Development of Children ' s Skills and Self - Efficacy

This experiment tested the hypothesis that rewards offered for performance attainments during competency development promote children's arithmetic skills and percepts of self-efficacy. Children received didactic instruction in division operations and were offered rewards contingent on their actual performance, rewards for simply participating, or no rewards. Results showed that performance-contingent rewards led to the highest levels of division skill and self-efficacy, as well as the most rapid problem solving during the training program. In contrast, offering rewards for participation resulted in no benefits compared with offering no rewards. The findings suggest caution in how rewards are distributed in educational contexts. Article: According to Bandura (1977b), reinforcing consequences inform and motivate. As persons engage in activities, they notice the differential consequences of their actions. They thereby learn which behaviors lead to desirable outcomes and which result in un-desirable ones. Such information guides future behavior. Further, the anticipation of attaining desirable outcomes motivates behavior. The capacity to represent valued future outcomes in thought promotes initial task engagement and persistence. The purpose of the present study was to determine the effects of tangible, extrinsic rewards offered in the context of arithmetic competency development on children's task mastery and percepts of self-efficacy. The conceptual focus was Bandura's theory of self-efficacy (Bandura, 1977a, 1981). According to this theory, different influences change behavior in part by strengthening percepts of self-efficacy. Self-efficacy is concerned with judgments about how well one can organize and implement courses of action in situations that may contain novel, unpredictable, or stressful elements. Percepts of efficacy can affect choice of activities, effort expended, and persistence in the face of difficulties. Efficacy information is conveyed through enactive attainments, socially comparative vicarious means, social persuasion, and indexes of physiological arousal. In this conception, the anticipation of a rewarding consequence should promote children's skill development and percepts of self-efficacy. The motivational effects of reward anticipation should boost task accomplishments. As children observe their progress they develop a heightened sense of efficacy, which should help sustain task involvement and lead to greater skill development. As compelling as these ideas sound, research has yielded mixed results on the effects of offering a reward for performance. Studies have shown that rewards facilitate performance, have a detrimental effect on performance, or have no effect. For example, offering a reward has been shown to boost performance on simple, unchanging tasks where performance is measured by rate or speed (McCullers, 1978), as well as on more complex tasks such as paired-associate learning (Goyen & Lyle, 1971), perceptual recognition (Glucksberg, 1962), and serial learning (Dornbush, 1965). Conversely, detrimental effects have been obtained with tasks requiring concept learning (McCullers & Martin, 1971), discrimination learning (Miller & Estes, 1961), and problem solving (Glucksberg, 1962). In an attempt to reconcile these contradictory findings, McGraw (1978) proposed that experimental tasks be distinguished by whether they require an algorithmic or heuristic solution and by whether they are initially attractive or aversive. When a task is initially attractive and requires a heuristic solution, offering a reward should have a detrimental effect on performance. For the other combinations, rewards may facilitate or have no effect on performance, depending on whether subjects initially view the task as aversive or attractive, respectively. In the present study, children who lacked division skills individually received instruction and opportunities to solve division problems, One group of children (performance-contingent reward) were told that they would receive points for each problem they completed and that they could exchange the points for prizes. A second group (task-contingent reward) were told that they would receive points for engaging in the task and that the points could be exchanged for prizes. A third group (unexpected reward) were not offered prizes but unexpectedly were allowed to choose prizes at the end of training. McGraw's (1978) analysis suggests that the present use of rewards should promote skill development and perceived efficacy: Division requires an algorithmic solution, and the present sample of children who displayed low mathematical achievement was not expected to view the task in an attractive light. In studying the effects of reward anticipation, however, it is important to differentiate between rewards used to regulate behavior and those employed to foster skill development (Bandura, 1981). Telling children that they can earn rewards based on enactive accomplishments conveys a sense of efficaciousness that can be actualized through effort. This sense of efficacy is subsequently validated as children observe their actual progress. Heightened efficacy should sustain task involvement and pro-mote skills. Efficacy is further validated on receipt of the reward since it symbolizes in concrete fashion the progress that children have made. In contrast, when rewards are offered merely for participating in a task, children should not experience a heightened sense of efficacy. Further, such rewards actually may convey negative efficacy information: Children may infer that they are not expected to accomplish much and that they do not possess the requisite efficaciousness to perform well. Subsequent task accomplishments and skill development should be lower than that obtained under a reward system tied to enactive attainments. In summary, it was hypothesized that offering rewards for performance attainments would be most effective in promoting children's skills and percepts of self-efficacy. In contrast, offering rewards for task participation was not expected to promote these outcomes over that expected as a function of merely providing training.