search

AaronGullickson / panethnicity_intermar

Data for "Patterns of Panethnic Intermarriage in the United States, 1980-2018" forthcoming in Demography
MIT License
0 stars 0 forks source link

Generational Coding by Pan-ethnic groups and for 1980 #21

Closed AaronGullickson closed 3 years ago

AaronGullickson commented 3 years ago

Currently I only test the different generational coding possibilities (Birthplace, USA, Both) in the ACS data for all respondents. I would like to test the fit:

  1. In 1980.
  2. Separately for the three pan-ethnic groups (E/SE Asian, South Asian, Latino).
AaronGullickson commented 3 years ago

OK, I think (2) above is difficult conceptually and also computationally. Birthplace endogamy is not a feature of individual partners but rather the combination of their birthplaces - so there would be lots of issues about how to code particular combinations of husband's and wife's race. Given that issue and the number of possible ways to code it generally, the number of cases needing coding and testing quickly becomes intractable. So, I think the parsimonious approach I am using here is best.

However, I can do (1). I am currently running the models on 1980 data and will incorporate this into the analysis and the paper once it is finished.

AaronGullickson commented 3 years ago

Ok, I have run the models for 1980 and they prefer an all first generation coding (all birthplace), which is interesting. The preference is not particularly strong, but I think it makes sense to run the model with the preferred coding. I am now re-running those models on the server. They should be ready tomorrow.

I have made changes to the analysis.html and I have changed the table in the paper, but I need to also change the corresponding description there.

AaronGullickson commented 3 years ago

Ok, going to hold off on actual rerunning these models until I deal with #20 because that is also going to require re-running the models. That will produce the deviance tables again, which could possibly shift due to the different coding of marriage markets. Its not likely but is possible. If everything comes up the same there, then I can safely close this one.