ADJUSTING FOR ATTRITION IN SCHOOL-BASED SAMPLES - BIAS, PRECISION, AND COST TRADE-OFFS OF 3 METHODS

Attrition in longitudinal studies can introduce nonresponse bias when estimating parameters. Methods to correct for nonresponse include surv ey-based approaches (tracking) as well as analytically based, methods (weighting, sample selection modeling). Using data from a multiwave sc hool-based study of adolescents, substance use estimates are compared across methods. Methods are validated by simulating effects of attriti on at baseline, and the relative efficiency of each approach with resp ect to a known ''gold standard'' is calculated Results indicate that w eighting may provide sufficient adjustment for nonresponse in other si milar studies. Sample selection modeling requires assumptions that are not met in this setting, and severe bias results. The high casts asso ciated with full tracking efforts may be avoidable, as hem we find tha t tracking vias apr inefficient approach for bias reduction.