|
|
Missing data for a twolevel model |
|
Message/Author |
|
|
Hello, I am running a multilevel model (children nested within families) and have missing data (<10%) on predictors, covariates and the outcome variable. I have see in past threads (2008) that use of auxiliary variables in FIML is not possible using the twolevel command. Has this changed? If not, is the best alternative multiple imputation? |
|
|
No change. Multiple imputation can be used, or ML where predictors/covariates are brought into the model by mentioning their means (or variances). |
|
Back to top |
|
|