Sample Selection Corrections
Non random sample selection generally occurs when the data is
truncated (see previous slide) or when respondents do not answer
to one or many questions.
Incidental truncation occurs when we do not observe y because of
Limited Dependent Variablesa
When the dependent variable is discrete and takes only a few
values, it doesnt make sense to treat it as normally distributed.
When fitting a linear probability model (LPM), the fitted
probabilities can be less than
When Omitted Variable Bias is present, it may not always be
possible to find a proxy.
log(wage) = + educ + where we do not have data on ability
(proxy: IQ) and therefore is a biased estimate of the effect of