logistic regression ad nauseum

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

logistic regression ad nauseum
Message posted by Susan Daniels (via 128.173.184.227) on June 19, 2001 at 7:26 PM (ET)

Thanks again JG. I have a 0/1 dependent (specified descending), roughly half of which are 0's. I have three continuous independents, although all of them have less than 7 possible values. The variable of interest has a positive coefficient, the opposite direction to that expected. That's my big problem...the independents are not highly correlated and there are no points of high influence...

I may still be violating independent observations though...I wonder, can you use proc mixed for a binary response?

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: logistic regression ad nauseum
Message posted by JG (via 128.8.24.160) on June 20, 2001 at 12:14 AM (ET)

Your independent variables are not really continuous if they can have only 7 possible values. As long as there is some correlation of the independent variables the sign of single coefficient has no meaning, unless you regress against that variable alone and ignore the other independent variables.

Your $5 contribution helps cover part the $500 annual cost of keeping this site online.