|
|
Hello,
I am trying to run a cox regression for my choice based conjoint analysis. The subject of the research is the effect of different cause-related campaign messages on consumer choice. As predictor variables I have 3 product attributes: Price, Cause and Donation Message (DM). Each of the three variables has 3 categories, and I use the first one as a reference category.
When I run the regression, SPSS automatically codes the variables into dummies, so I am supposed to get two coefficients for the other two categories of each variable. However, I only get coefficients for one category and then on the bottom of the table I get this message: "Constant or Linearly Dependent Covariates S = Stratum effect". I checked for multicollinearity between the three variables and the VIF is around 2.5 which is not that high.
Do you have any idea what the problem is, and how can i fix it, or how to interpret it?
Any help will be very very welcome!
Thank you!
|
Locked
