Gunesk,
I'm a little confused by your message as it seems like two different things
are being described. On one hand I think you are doing this:
Compute delta=dvT2-dvT1.
Regression dependent=delta/enter iv.
No moderation is possible. IV is going to predict the T1-T2 difference.
Actually a regression is not needed as a correlation will do the same thing.
On the other hand I think you are doing this:
Compute dvT1_iv=dvT1*iv. /* leave the centering business out.
Regression dependent=dvT2/enter iv dvT1 dvT1_iv.
This would be moderation or simply an interaction between iv and dvT1. But
note that this version does not predict the difference in development.
Gene Maguin
Hi,
I have a repeated measures design where I am looking at the development of a
dependent variable over time (T1 and T2). I want to predict the difference
in development (T2 - T1) with a continuous IV which I believe is a
moderator. So I want to look at an interaction between DV at T1 and the IV.
How can I do that in a multiple regression analysis? I have already centered
all the variables and created an interaction term but I don't know what I
put in SPSS. Someone help pleaseee!!!
--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/repeated-measures-multiple-regression-need-to-test-for-interaction-tp2653144p2653144.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.
=====================
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
=====================
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
