I am requesting to get an idea on the use of flag variable in restricting my sample to those adults who had records during NHIS 2011-2015. Usually, analysts do
drop age if age<18.
However, I am worried about not to miss those respondents who are below 18 years and still consider as adults in some states. Therefore, I would like to consider “ASTATFLG” var to select my adult sample size. Would you please have any suggestions?