I have a dataset with a small amount of missing data (8% on one variable
or about 160 missing value points and <1% on another; total sample size =
1968). In computing the multiple imputation, I have included a large
number of other variables in the model. The multiply imputed datasets,
however, have only imputed about one-half of what is missing. There is
some missing data on the predictors included for the multiple imputation
model, but there still should be enough valid data to compute the
imputation. Can anyone offer advice on this?
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
