Citation

Nelson JP. Int. J. Environ. Res. Public Health 2010; 7(3): 870-926.

Affiliation

Department of Economics, Pennsylvania State University, University Park, PA 16802, USA. jpn@psu.edu

Copyright

(Copyright © 2010, MDPI: Multidisciplinary Digital Publishing Institute)

DOI

10.3390/ijerph7030870

PMID

20617009

PMCID

PMC2872298

Abstract

This paper assesses the methodology employed in longitudinal studies of advertising and youth drinking and smoking behaviors. These studies often are given a causal interpretation in the psychology and public health literatures. Four issues are examined from the perspective of econometrics. First, specification and validation of empirical models. Second, empirical issues associated with measures of advertising receptivity and exposure. Third, potential endogeneity of receptivity and exposure variables. Fourth, sample selection bias in baseline and follow-up surveys. Longitudinal studies reviewed include 20 studies of youth drinking and 26 studies of youth smoking. Substantial shortcomings are found in the studies, which preclude a causal interpretation.

Language: en