I would like to run Full Information Maximum Likelihood estimation in my multilevel logistic model (under PROC GLIMMIX). The only documentation I have found for FIML is in PROC CALIS which only appears to allow linear modeling. Does anyone know how to do FIML in GLIMMIX?
Also, my code currently isn't running for my GLIMMIX procedure and I'm not sure why! I'm very new to SAS, so it may be an obvious mistake.
Proc glimmix data=work.smhse method=rspl noclprint;
class x_idschool x_classid x_studentid male prob need;
model intent= male age int ext prob need assets familyincome schoolsize/CL dist=binary link-logit solution oddsratio cl corrb covb;
reference=first event=last;
lsmeans adjust=bon;
random intercept/subject=x_idschool s cl type=vc;
random intercept/subject=s_classid(x_idschool) S CL type=vc;
covtest/wald;
run;
Any advice is greatly appreciated!