A subset of a real data set was used to illustrate an ad hoc analysis
with missing data on multiple response variables. This strategy was in
itiated with a complete-case analysis to determine some variable(s) th
at may be deleted with no loss in effects of interest. By such a delet
ion, the number of complete observation vectors may very well increase
. Two straightforward imputed means analyses are also illustrated. All
illustrations are given in the context of predictive discriminant ana
lysis.