Closed AaronGullickson closed 3 years ago
I also want to expand this to think about the cases of Nepalese, Mongolian, and Bhutanese. These are very small groups and some question about where they fit in a South Asian/ESE Asian split. I mean thats part of what I am trying to highlight, but I can't really do anything with these groups specifically to highlight that because they are so small. So my initial thought is to drop them.
I definitely will remove Mongolians and Bhutanese for low sample size and some issues around identifying languages. I think I will drop Nepalese for similar reasons, although that group is bigger.
This is complete with commit 927cdd786fd0dc778e2fb7368d5f925400e81754.
The results for Vietnamese are clearly having estimation problems in 1980 because of the small size of this population. I think it is best to remove them from the 1980 data and the 1980 basis data in the ACS and then include them in the larger expanded ACS models.