R.
Assuming the missing is random. What I usually do is create a nmissvar
(COMPUTE nmissvar = NMISS(items in regression)).
Once that is created, I could sort nmissvar and delete cases not used in the
regression (note. I would have backed up the original); I could TEMPORARY
SELECT IF (nmissvar EQ 0) or just SELECT IF (nmissvar EQ 0).
I hope this helps,
Steve
-----Original Message-----
From: SPSSX(r) Discussion [mailto:
[hidden email]] On Behalf Of
Rayman Mohamed
Sent: Saturday, May 05, 2007 1:12 PM
To:
[hidden email]
Subject: Capturing Records Used in Regression
Hello all,
I have about 1,000 observations on which I am doing regressions. Some
records have missing data for which there were no responses (this is survey
data). Thus, SPSS performs the regressions only using those observations for
which there are data for the dependent variable and the independent
variables (included in the regressions).
How do I isolate the records that SPSS actually used such that I perform
additional analyses of the observations actually used in the regression.
I hope that this is clear. I am accustomed to using the GUIs in SPSS.
Many thanks for your help.
R.
Locked
