Аннотация:Estimates of the relative influence of peers and parents on adolescents' drug use and other forms of deviance have inflated the importance of peers and underestimated the influence of parents. Following a brief review of major findings in research on parental-peer linkages, sources of distortion and overestimation in peer effects are identified: reliance on cross-sectional designs, which confound selection and socialization effects; reliance on perceptual reports of friends' behaviors, which reflect projection and attribution; failure to take into account parental contributions to children's peer selection; and failure to consider genetic contributions to observed parental effects. Selected empirical studies that have estimated peer and parental effects on drug use and delinquency from relational and longitudinal designs are utilized to develop correction factors. These are incorporated in equations designed to estimate biases in peer estimates and the effects of parental contributions to peer selection. Within the limitations of the available data, I conclude that peer effects based on cross-sectional data and perceptions of peer behavior are overestimated at least by a factor of five.
