Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
MANOVA but multicollinearity!
4 views
Skip to first unread message
girl...@hotmail.com
Sep 17, 2005, 9:59:59 AM9/17/05
to
Can someone help please? I was giong to do a MANOVA but my variables
have turned out to all be correlated. I was gionhg to do a multiple
regression instead but my IV (I only have one) is dichotomous. Does
anyone know a test I can use instead?
have turned out to all be correlated. I was gionhg to do a multiple
regression instead but my IV (I only have one) is dichotomous. Does
anyone know a test I can use instead?
Øyvind Langsrud
Oct 3, 2005, 7:31:13 AM10/3/05
to
Hello,
A method called 50-50 MANOVA (Multivariate Analysis of Variance for
Collinear Responses) is available at
http://www.matforsk.no/ola/ffmanova.htm
Best regards,
Øyvind Langsrud
MATFORSK, Norwegian Food Research Institute.
http://www.matforsk.no/ola
0 new messages