www.cmm.bristol.ac.uk/forum

Hi,

Having run a model using multiple imputation, is it reasonable to compare the variance components of the fitted model with the null model (which obviously didn't use multiple imputation)? i.e., I understand that the same number of cases should be used in the null model and the fitted model if calculating variance explained, but is it correct to think that the issue of listwise deletion doesn't apply once multiple imputation was employed?)

many thanks
Lorraine

Dear Lorraine
If the multiple imputation has done its job properly, i.e. te assumptions are reasonable (MAR) then you dont need to have the same no of records for null modle and imputed model in order to study the relative amounts ofvariance explained in both cases since you should be getting unbiased estimates of these components in each case. Anyway, in effect you can think of the null model as being carried out on the imputed dataset - which of course will include all cases.
Harvey