logistic response

#2
Check out logistic regression using google. Usually you have a response (y) which takes values of 0 or 1 (fail or success). The model of logistic regression is

p=exp(b0+b1*x)/(1+exp(b0+b1*x)) if you have one predictor (x). And p is thr probability of success.


Leo
 
#5
Hi Nelson,

If you want a really thorough, really readable, and applied book on logistic regression, I recommend Paul Allison's "Logistic Regression using the SAS System." It's even worth it if you don't use SAS.

Scott Menard also has one called "Applied Linear Regression Analysis." It's one of those little green Sage books. It's also very good, but won't go through doing it in SAS the way Allison's book does. Slightly less applied, but also less expensive--it's only about $15.

Karen