The study also finds that unobserved effects, sample selection bias and measurement errors play important role on results, and OLS really underestimates yield of schooling.
非观测效应，以及样本选择偏差和测量误差所引发的内生性问题对结果的影响很大，导致普通最小二乘法低估了我国的教育收益率。
No baseline data were found to be related to the prognosis due to a small sample size and selection bias.
因小样本及选择偏倚，无基线资料与预后相关。
应用推荐