Page 1 of 1

Include level1 variables to impute level 2 missing variable

Posted: Wed Feb 11, 2015 10:35 am
by likestatistic
Dear All,

I am using REALCOM to impute missing variables. I have missing values on level 1 and level 2 variables. after running realcom on stata, the model in REALCOM doesn't include level1 variables in the imputation of level 2 variables. I have tried to include them directly on REALCOM using Add/Remove explanatory variables but I am getting the following message:

Bad Input! This is not a possible action and has been ignored.


Could someone help me?

Many thanks

L

Re: Include level1 variables to impute level 2 missing variable

Posted: Sat Feb 14, 2015 2:03 am
by Harvey Goldstein
When you specify a 2 level imputation model any level 1 variables that are declared responses at level 1 implicitly contribute through their level 2 random effects. However you cannot have a level 1 predictor for a level 2 response in any multilevel model so do not include level 1 variables as covariates for level 2 responses in the imputation model.

Harvey Goldstein