求助：用SAS逻辑回归计算出现问题 [推广有奖]

11楼

frankling 发表于 2010-2-22 12:49:03

Thanks,Bobguy and jingju11!
Can I understand in this way: because the "sex" is a catagorical variable, I can't simply calculate the P by putting numbers of variables into LnP/(1-P) = -11.1897 + 1.154X1 + 0.8777X2 - 5.7052X3, right? (x3=sex)

12楼

frankling 发表于 2010-2-22 12:59:55

10# jingju11
Yes, the default parameterization is Main Effect only in the menu of analyst. How to change the default set there? Thank you!

13楼

bobguy 发表于 2010-2-23 08:35:45

frankling 发表于 2010-2-22 12:59
10# jingju11
Yes, the default parameterization is Main Effect only in the menu of analyst. How to change the default set there? Thank you!

Check the manual to see what is the meaning of "param=glm".

proc logistic data=_PROJ_.DISCRIM;
class sex /param=glm;
model hired (event='1') = education experience sex ;
run;