Thanks for your help Jerry. That helps me tidy up my algorithm, but I'm still getting the same results.
I'm aware of existing applications Dennis, but I wanted to implement the algorithm for myself - make sure I really understand how it works. The fact it's not giving me the correct results implies I don't understand it fully yet. Could anyone confirm my understanding on the other two points? I'm assuming that I can use the normal standard linear regression techniques on the link funtion (i.e. DV=B0 + B1*IV1 + B2*IV2 +...) for coefficients, standard errors etc. Thanks, Darren. . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
