gate increasingly well-specified models of how different types of program practices and processes may be instrumental in shaping consequential features of mentoring relationships and ultimately, the realization of particular desired outcomes for youth; and (d) establish a research registry to improve the quality and synthesis of available evidence regarding the effectiveness of youth mentoring as an intervention strategy.

These findings suggest a broad and multifaceted impact of mentoring relationships on adolescent health. However, mentoring relationships alone are not enough to meet the needs of at-risk youths and therefore should be incorporated into more comprehensive interventions.

This research investigated characteristics of natural mentoring relationships (mentor role, frequency of contact, closeness, duration) as predictors of adjustment outcomes among older adolescents and young adults (N = 2,053) in the Add Health study. Outcomes were assessed in the domains of education/work, problem behavior, psychological well-being, and physical health. Mentoring relationships with persons in roles outside of the family predicted greater likelihood of favorable outcomes in all domains except psychological well-being, relative to mentoring relationships with family members. Greater reported closeness in relationships was predictive of several favorable outcomes, particularly those in the domain of psychological well-being. These findings indicate that strategies to promote mentoring of adolescents may be more effective if particular categories of adults are targeted and an effort is made to cultivate relationships with strong emotional bonds. Editors' Strategic Implications: These data suggest that the cultivation of natural (especially non-familial) mentoring relationships during adolescence may be a promising strategy for prevention and health promotion. This study is impressive due to its large, nationally representative sample, the examination of relationship characteristics and multiple mentors, and the links to a variety of outcomes (controlling for earlier functioning). School officials and mentoring programs must consider how to capitalize on - and promote - naturally occurring mentor relationships.

Gender differences and similarities in the relations of key constructs in Eccles and colleagues' (Wigfield & Eccles, 2000) model of achievement were examined as predictors of math grades and enrollment intentions for Grade 9 boys (n = 263) and girls (n = 277). A number of gender similarities were found, particularly in the prediction of math grades. There were, however, two gender-specific paths: for girls, a direct path from competence beliefs to enrollment intentions, and for boys, a direct path from prior math grades to enrollment intentions. In addition, for boys, the path from utility value to enrollment intentions was stronger than it was for girls. These differential predictive patterns were found even though girls and boys reported similar levels of math utility and girls had lower math competence beliefs. For girls, competence beliefs were a significant predictor of both intentions and current math grades, which indicates the central role of competence beliefs.

No abstract