Intercept changes when defining diff. reference categories

[jeanninekhan]

Hallo

I have a problem concerning the reference category of categorical variables.
Normally I take a low reference category to fit the model. I observed that the intercept changes in function of the reference category I putted to the model. I also observed that if I take the lowest reference category, I get an error message that the matrix has gone negative. But if I take the middle category of the same variable, the model calculates without any problem.
Could anyone explain me why this happens?

Best regards, Jeannine

[Lydia]

See my reply to your post in the topic http://www.cmm.bristol.ac.uk/forum/view ... ef9e6e36af- think that will answer this question too!