Question for BRFSS data

yyuan — Mon, 22 Aug 2016 20:17:06 GMT

Hi all,

I'm a new user of Proc Glimmix and I would like to specify a multi-level logistic model by using this procedure.

The Data set I use is BRFSS. (Cluster sampling weight) From 2011 - 2014.

The dependent variable is either "Ever been tested" or "Tested in the past 12 month" . Also these two variables are dummy variables.

The data set only contain one global weight for each observation.(_LLCPWT)

Years(2011, 2012, 2013, 2014) are level 2 variables in my model.

Currently my codes are as follows:

proc glimmix data=xxx.test method=quadrature(qpoints=7) empirical=classical;
Class year;
model "Dependent variables" = "Here will be the independent variables" /

dist=binomial link=logit obsweight = _LLCPWT solution CL ODDSRATIO (DIFF=FIRST LABEL);
random intercept / subject=year cl;
run;

Just wondering is there any suggetion of my code? Thank you very much!

Re: Question for BRFSS data

Ksharp — Tue, 23 Aug 2016 03:02:30 GMT

Or you could also check PROC GEE .

topic Re: Question for BRFSS data in Statistical Procedures

Question for BRFSS data

Re: Question for BRFSS data