When we conduct a logistic regression with a multiple number of levels and we want to find the coefficients of the fitted model I invoke dummy.coef() function to get all the coefficients. The first coefficient is always 0, for the first factor. Does it mean that the rest of the coefficients are being compared to the first one? That is, in the case of the logistic regression (logit link) a positive sign of a particular coefficient means that the factor has greater effect than the first factor, and lower effect otherwise?
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|