Greetings from Texas,
I'm using the MULTIPLE IMPUTATION procedure for the first time and have a
question about using a weighting variable correctly when I do this.
I'm analyzing survey data that comes with a weight variable (used to match
the sample to U.S. Census defined national parameters for several
demographic variables). As best I can tell, I should have the weight
variable turned off in the data file when I run MULTIPLE IMPUTATION, and
instead specify the weight variable in the /ANALYSIS WEIGHT subcommand. But
when running subsequent analysis on the imputed data, I'm not sure if I then
need to specify WEIGHT=varweight or not. If the data were not imputed with
the weight variable already, I know I would need to specify
WEIGHT=varweight.
If you have any guidance, or can point me to a resource that will help me
with this puzzle, I would really appreciate it.
Thank you!
Mitzi
=====================
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
