Count data regression with survey data - goodness of fit / model comparison

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Count data regression with survey data - goodness of fit / model comparison

Gineer
This post has NOT been accepted by the mailing list yet.
Hi Statalistors,
 
I am working on a thesis about count data regression using NHIS survey data. I am trying to compare different models and select the best model for mydata.
But I am running into some problems below.

First, I found that fitstat and estat are not allowed after svy: poisson/zip, so I cannot obtain AIC, BIC etc to test the model fit.  Do you know how to test model after svy: poisson/zip/zinb?

Second, the "vuong" option are not allowed after -svy:possion/nbreg/zip-. Do you know how to compare svy:zip model to other non-nested models using STATA?

Thanks! I am looking forward to your responses.

Jin Peng
 
2013 MS Candidate | CPH Student Ambassador
The Ohio State University | College of Public Health