Using data from the past 25 years of the National Longitudinal Survey of Youth, this proposal will be the first to test causal pathways from adolescent and young adult marijuana use to adult labor market outcomes comparing biostatistical and econometric modeling techniques. Surprisingly, much literature has reported associations between marijuana use and positive labor market outcomes, but their reliance on cross- sectional data, and their inability to properly control for differences in baseline covariates (heterogeneity) between marijuana users and non-users brings into question the validity of their conclusions. This study aims to 1) utilize biostatistical (propensity score) and econometric (instrumental variable) modeling techniques to test for causal associations between adolescent and young adult marijuana use and adult labor market outcomes, and 2) to adjust for moderating and mediating factors on the association of marijuana use and labor market outcomes by building on the previous propensity score and instrumental variables models. Identification and proper control for baseline covariates, understanding moderating and mediating pathways, and validating results will strengthen our understanding of the negative outcomes of marijuana use. ? ? ?
| Harder, Valerie S; Stuart, Elizabeth A; Anthony, James C (2010) Propensity score techniques and the assessment of measured covariate balance to test causal associations in psychological research. Psychol Methods 15:234-49 |
| Harder, Valerie S; Stuart, Elizabeth A; Anthony, James C (2008) Adolescent cannabis problems and young adult depression: male-female stratified propensity score analyses. Am J Epidemiol 168:592-601 |
