Hi, I have a question about logistic regression contrasts. I would
appreciate whoever can give me a quick answer.
I have collected four personality types for smokers and nonsmokers and want
to compare risks of smoking among the personality types. I used deviation
coding scheme for personality type. as you all know only three of the four
categories are presented in equation and, therefore, i got odds ratio for
them. How about the forth one in terms of its odds when compared to average
effect which is what deviation contrast method does? I would like to provide
more information if i am not clear here. thank you in advance.
kelly
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|