[疑难杂症]Multilevel Modeling in SAS and Model Fit?

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

I am developing a two-level random intercept model with a dichotomous outcome in SAS. I would like to know what indices I may use to check if the model is on the right track. The outcome is administration of a certain drug across all hospitals in the data. The predictors include person level medical conditions and hospital level characteristics (size, beds, etc.)The ICC is only 0.04. Does that indicate that multilevel modeling is unnecessary? I'm also aware that ICC is not always reliable. Is there other index I may use to evaluate my model?

I'm new to multilevel modeling in SAS. Can anyone take a look at my syntax? I would greatly appreciate it!

Here's a simplified version (Person_condition is the within level variable, hospital_beds/size is the between level variable). My specific question is:should I use

ddfm=satterth, and tech=nrridg?

*proc* *glimmix* data= anti_LinkageFinal;

class hospitil person_condition1 person_condition2 hospital_size;

model drug  = person_condition1 person_condition2 hospital_beds
hospital_size

/ dist=binary link=logit ddfm=satterth oddsratio;

random intercept / subject=hospital solution;

nloptions tech=nrridg;

ods exclude solutionr;

run;

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：Multilevel Modeling model multi Level indicate include always person across

支持一下！！！！！！！！