作者: Gary Chamberlain
DOI: 10.1016/0304-4076(77)90027-6
关键词:
摘要: Abstract The paper examines two approaches to the omitted variable problem. Both of them try correct for bias by specifying several equations in which unobservable appears. first approach assumes that common left out is only thing connecting residuals from these equations, making it possible extract this factor and control it. second relies on building a model unobservable, observable variables are causally related A combination methods applied 1964 CPS-NORC veterans sample order evaluate income- schooling regressions caused omission an initial ‘ability’ variable.