survey based multinomial regression

[stellapersis7]

Hi all,

I have a study with 3 categories dependent variable and 4 categorical independent variable and the data is survey based. I am using multinomial logistic regression with logit.

Can you please confirm if this is a correct procedure and code.

Thanks 

 proc surveylogistic data =  meps.sdm;

 weight poolwt;

 cluster VARPSU;

 strata VARSTR;

 class regionnew;

 model SDMSCORE =  regionnew / link =glogit;

 run;

[JosvanderVelden]

Just search for multinominal logistic regression in this community and you'll find threads with answers like for instance: https://communities.sas.com/t5/Statistical-Procedures/multinomial-logistic-regression/td-p/720924 & https://communities.sas.com/t5/Statistical-Procedures/SURVEY-LOGISTIC-and-multinomial-logitic-regres...