vary depending on the cultural context. For example, both theory and empirical research suggest that a sense of competence is key to positive human development. However, the exact domains in which one “should” feel competent are likely to be quite culturally specific; that is, a positive sense of competence will depend on competence in those areas valued by the cultural group of which one is a member (Ogbu, 1994; Shweder et al., 1998).

The committee concluded that this debate could be reconciled to some extent by carefully considering the level of analysis. On one hand, we agreed that there are some universals at the most abstract level of consideration—universals such as the need to feel competent, to be socially connected, to feel valued and respected, to actually be making a difference in one’s social group, to feel that one has some control over one’s own behaviors and experiences, and to have one’s physical and emotional needs met. We also agreed that the failure to have these needs met has a negative effect on development (see evidence later in this and the next chapter). On the other hand, we agreed that there would be extensive cultural specificity when one moves to a more specific level of analysis. What this means is that one must take the local cultural context into account as programs are designed and evaluated.

The committee also spent a great deal of time discussing how to select a list of indicators of well-being, particularly given the cultural issues just outlined. We decided to rely on three sources: theory, practical wisdom, and empirical research. Within the theory category, we drew on developmental theories from psychology, anthropology, and sociology. Within the category of empirical research, we reviewed three types of evidence: (1) evidence that particular characteristics are either positively related concurrently to other indicators of well-being or negatively related concurrently to indicators of problematic development; (2) evidence that particular characteristics predict positive indicators of adult well-being and of a “successful”³ transition to normative adult statuses; and (3) evidence that the experimental manipulation or training of particular characteristics produces changes on other indicators of either