Posted by
Maurice Vergeer on
Feb 04, 2011; 4:11pm
URL: http://spssx-discussion.165.s1.nabble.com/Calculate-Interaction-Term-in-SPSS-tp3370327p3371271.html
there's a difference between regular collinearity and collinearity due to interaction terms. The first is really problematic, the second refers to inflated standard error because these are conditional ones (if I remember correctly).
There is a very good article on this matter by Friedrichs (1982) :
http://faculty.fuqua.duke.edu/~charlesw/s591/Methods/c09_Bennet/Friedrich_AJPS_82.pdf.
Maurice
On Fri, Feb 4, 2011 at 15:37, Gene Maguin
<[hidden email]> wrote:
R,
If A, B, and C are highly correlated, you have to assume that
you will have collinearity problems. So, you'll need to monitor that in
the analysis. That said, I'd center each variable by subtracting it's
mean and then simply multiply A*B*C. (As well as A*B, A*C, and B*C). I assume
you are doing a regression of some kind so you're final eqation will have seven
terms plus the constant (A, B, C, A*B, A*C, B*C, and A*B*C). I would use
the centered versions of A, B, and C rather than the raw score versions.
Gene Maguin
From: SPSSX(r) Discussion
[mailto:[hidden email]] On Behalf Of R Duke
Sent:
Thursday, February 03, 2011 5:46 PM
To:
[hidden email]
Subject: Calculate Interaction Term in
SPSS
Hello SPSS Experts,
I apologize for the novice question but I have 3 independent variables (A,
B, and C) that are highly correlated and I want to calculate the interaction
term for these 3 variables (one DV within). I am new to SPSS and I'm not sure
how to calculate the interaction between A*B*C. I have included some mock data
below and would appreciate any help.
Cheers,
R
|
0.78 |
0.62 |
0.39 |
|
0.80 |
0.98 |
0.83 |
|
0.51 |
0.42 |
0.39 |
|
0.45 |
0.60 |
0.20 |
|
0.35 |
0.53 |
0.37 |
|
0.12 |
0.58 |
0.12 |
|
0.40 |
0.43 |
0.28 |
|
0.57 |
0.84 |
0.55 |
|
0.86 |
0.68 |
0.42 |
|
0.45 |
0.21 |
0.14 |
|
0.57 |
0.53 |
0.45 |
|
0.29 |
0.37 |
0.17 |
|
0.48 |
0.81 |
0.45 |
|
0.76 |
0.44 |
0.24 |
|
0.54 |
0.62 |
0.30 |
|
0.07 |
0.06 |
0.06 |
|
0.42 |
0.92 |
0.66 |
|
0.92 |
0.91 |
0.92 |
|
0.70 |
0.76 |
0.44 |
|
1.00 |
0.25 |
0.09 |
|
0.45 |
0.45 |
0.23 |
|
0.81 |
1.00 |
1.00 |
|
0.02 |
0.53 |
0.25 |
|
0.49 |
0.91 |
0.56 |
|
0.25 |
0.58 |
0.35 |
|
0.00 |
0.00 |
0.00 |
--
___________________________________________________________________
Maurice Vergeer
Department of communication
Radboud University (
www.ru.nl)
PO Box 9104
NL-6500 HE Nijmegen
The Netherlands
Visiting Professor Yeungnam University, Gyeongsan, South Korea
contact:
E:
[hidden email]
T: +31 24 3612297 (direct)/ 3612372 (secretary) / maurice.vergeer (skype)
personal webpage:
www.mauricevergeer.nlblog:
http://blog.mauricevergeer.nl/
Journalism:
www.journalisteninhetdigitaletijdperk.nlRecent publications (see complete list at
www.mauricevergeer.nl/node/2):
-Vergeer, M., Hermans, L., & Sams, S. (accepted for publication). Online social networks and micro-blogging in political campaigning: The exploration of a new campaign tool and a new campaign style. Party Politics.
-Eisinga, R., Franses, Ph.H., & Vergeer, M. (2010). Weather conditions and daily television use in the Netherlands, 1996–2005. International Journal of Meteorology.
___________________________________________________________________