> > Does not multiple regression give you what you want ?
Thanks for your reply, Hmm. Do you mean the beta coeff for the income var? I see what you mean but is that the "best/most accurate" way? The multiple regression and the error term are derived from all the variables together, so might the income variable be swamped by the others? I can see that I might be wanting the best of both worlds, that is having to rely on the other variables yet wanting to exclude them also. Cheers . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
