are. As we have seen, Hoffman and Vygotsky attach more importance than does Kohlberg to interactions between caregiver and child as processes by which a child arrives at an internal moral orientation.
This emphasis on social processes is consistent with evidence from one of the very few published studies in which the TRA has been used to analyze moral behavior ( Vallerand, Deshaies, Cuerrier, Pelletier, & Mongeau, 1992). Participants in Vallerand et al.'s study were athletes who answered questions about two hypothetical sports situations raising moral issues: whether or not to criticize an official for making a bad call that cost the athlete the event, and whether to inform an official of an undeserved outcome of one's own, if doing so would cost one the event. One finding to emerge from this study was that the overall fit of the model to the observed data was significantly improved by adding noncausal paths relating normative beliefs, supposedly a component of the subjective norm factor, to behavioral beliefs and outcome evaluations, the two components of attitude toward the behavior. In discussing this relationship, the authors suggested that it can be understood by positing the existence of a type of belief that underlies both behavioral and normative beliefs. This common belief, they suggested, might be "one's beliefs about what should or ought to be done" (p. 106)--that is, one's moral norm. A second, related finding was that the overall fit of the model to the observed data was significantly improved by adding a causal path from normative beliefs to attitudes toward the behavior. Adding this path improved the explained variance by no less than 35%. Thus the perceived expectations of others appeared to play a key role in shaping individuals' attitudes to the two moral behaviors under investigation.
Vallerand and his colleagues were led to conclude that the social or normative component of the TRA has an important bearing on individuals' intentions to engage or not to engage in these behaviors:
Normative beliefs, or the beliefs regarding what is perceived as appropriate in the eyes of important others, appear to loom large in the learning of moral beliefs, evaluations, attitudes, and subjective norms, and probably for good reasons. How do we learn that moral behaviors are to be valued (or devalued) if not through the influence of important others, such as parents, teachers, and peers? ( Vallerand et al., 1992, p. 108)
Where a measure of moral norm has been added to measures of the standard TRA and TPB constructs, there is rather consistent evidence that it adds significantly to the prediction of behavioral intentions. In only three of the studies reviewed above was there less than compelling support for the independent predictive utility of moral norm, and in these cases it is reasonably easy to generate post hoc explanations for fact that moral norm played