Re: st: RE: Goodness of fit using Cox-snell residuals

[marco@vassar.edu]

Dear Emelda,

I am working on a similar problem. Most of the examples of this technique
use single observation per subject data.

If I'm not mistaken, you want to predict using ccsnell instead of csnell.
That way, you're using the cumulative cox-snell residuals.

Best,

Alan

Alan C. Marco
Department of Economics
Vassar College

--------------------------------------------------------------------------------