Interaction Effects Model

[hdilk0507]

1. Does Realcom-Impute impute missing values when the MOI (model Of Interest) consists of interaction terms?
2. Should MOI be a main effects model?
3. Does anyone has a clear idea about how MOI should be? Is it the best model for complete cases?

Please give your comments.

[Harvey Goldstein]

An interesting question. At the moment the MOI should not contain interaction terms. There are different approaches in the literature to incorporating interactions (including polynomials) - see e.g. van Buuren S (2007). Multiple imputation of discrete and continuous data by fully conditional specification" Statistical Methods in Medical Research, 16(3), 219-242. However, while this author recommends the JAV approach none of these existing methods works very well in many circumstances. A new approach that produces efficient and unbiased estimates is being developed in conjunction with James Carpenter and Bill Browne and is currently being reviewed for publication.
Harvey Goldstein

[hdilk0507]

Thank you very much sir. I would like to know from where I can find the new method.