We review the evidence regarding the development of altruism and suggest that various forms ofperspective-taking" (perceptual, social, empathic, moral) might be the mediators of this development. Previous reviews in this area have generally concluded that the evidence relating perspective-taking and prosocial behavior is equivocal. Using a technique that allows the assignment of an exact probability to the results of a series of studies, we conclude that there are reliable relations between altruism and perceptual, social, and moral perspective-taking. The results concerning empathy and altruism are nonsignificant overall, but we suggest that a reliable association between empathy and altruism develops over time and is found in adults. The significant effects that were found cannot be attributed to artifactual effects of either age or other measures of perspective-taking. Nor, however, can these relations completely account for age-related increases in altruism. We conclude that the evidence for a causal influence of perspective-taking on altruism is strongest for studies of empathy and altruism among adults, but even there it is not totally conclusive.

The relationship between affect and altruism in 7-and 8-year-old middleclass white children was investigated. Children were asked to think of things that made them happy or sad, or they were assigned to control conditions. Subsequently they were given an opportunity to donate money to other children in the experimenter's absence. In accord with predictions, children who experienced positive affect gave more than control children while those who experienced negative affect gave less than controls. Girls gave more than boys overall. The relationship between self-reward and reward of others is discussed.Research into the determinants of altruistic behavior has increasingly implicated the role of affect as a setting condition. Several authors have investigated the effect of success on altruistic behavior and have speculated that their results are due to the affective component of the success experience. Berkowitz and Connor (1966), for example, found that adults who had experienced success proved more willing to help a beneficiary than did control subjects. These investigators interpreted their findings in terms of the feelings that arise from successful experience, terming them the "glow of goodwill." Isen (1970) obtained similar findings with regard to helpfulness and has suggested that the "warm glow of success" increases one's tendency to be helpful to others. Similar findings have recently been found among children by Isen, Horn, and Rosenhan (in press).These investigators and subsequent interpreters of their data (Bryan & London, 1970;Krebs, 1970; Rosenhan, in press) have attributed these findings to the positive affect that allegedly arises from the experi-

Children reminisced on matters that made them happy or sad. Subsequently, they were permitted to indulge themselves noncontingently with candies and to contribute money to other children, both in the absence of an experimenter. Both happy and sad children self-gratified more than the control group, but happy children contributed more than either the controls or unhappy ones. Among happy children, a strong positive correlation was obtained between self-gratification and altruism. Among unhappy children, that correlation was negative. Affect, therefore, moderates the relationship between self-gratification and altruism. Implications for a theory of altruism are discussed.

Previous research has shown that delay of gratification is decreased when children attend to the actual rewards in the contingency during the delay period. The present study investigated the effects of attention to symbolic presentations of the contingent rewards (in the form of slide-presented images) on children's ability to wait for the delayed reward. In sharp contrast to the effect of attention to the actual rewards, attention to the symbolic contingent rewards greatly increased the duration of the children's delay of gratification. This unexpected result was obtained both in conditions of waiting and working and both for continuous and intermittent presentation of the slides. The overall findings revealed the opposite effects of attention to real versus symbolic contingent rewards in the delay of gratification paradigm. Theoretical implications of different cognitive functions served by attention to real and symbolic reward stimuli were discussed.

No abstract