logistic regression include constant??

#1
Dear all,

I have been working on a binary logistic regression model. My dependent variable is death or no death, while the independent variables are a number of variables. When conducting a univariable regression analysis, SPSS asks me to include the constant or not. When I try both methods, the results differ very much. My question is, shold I include or exclude the constant for the univariable analysis, and why?

Thanks for your help,

best wishes.
 

obh

Well-Known Member
#3
It is the same as a regular linear regression, the logistic regression as the assumption of linearity between the logit of the outcome and the IVs. For a simple linear regression, it is recommended to include the constant.