I currently have a data set with 7 binary variables, namely A,B,...,G. The data contains some structurally missing values and am trying to find the "saturated model" using the method from Rubin, 1974 (Characterizing the estimation of parameters in incomplete data problems).
Variables ABCD are always observed, E is observed more than G, but F in some cases is jointly observed with E, but G is not necessarily a subset of F.
I tried using this characterization [ABCD][EF][G] but it does not seem to work due to that F is not observed more than G.
I am not sure where I am missing it.
Thanks.
Vumani Dlamini Swaziland
_________________________________________________________________
Add photos to your e-mail with MSN 8. Get 2 months FREE*. http://join.msn.com/?page=features/featuredemail
