• Description

We analyze peer effects in sleeping behavior using a representative sample of U.S.teenagers from the National Longitudinal Survey of Adolescent Health. The sampling design of the survey causes the conventional 2SLS estimator to be inconsistent. We extend the NLS estimator in Wang and Lee (2013a) to estimate network models with sampled observations on the dependent variable. When accounting for sampling, we find that the sleeping behavior of the friends is important to shape own sleeping behavior, besides the impact of individual, family and friend characteristics.