On Wed, 6 Aug 2008, Mike Fontaine wrote: Thanks very much, Mike. I see it's an ordered logit/probit you're talking about: the dependent variable has 11 distinct values (1 to 12, skipping 11). Anyway, you have exposed a bug: something has gone wrong with gretl's ordered probit (and so I'm cc'ing back to the users list). It's a pity Jack Lucchetti is offline at present, since he's the expert in this area. I've got a partial fix in place, in CVS and the Windows snapshot at http://ricardo.ecn.wfu.edu/pub/gretl/gretl_install.exe But there's a residual problem due to the fact that your dependent variable has a gap in its sequence of values. You can work around this for the moment by doing, e.g., genr y = (Rtg == 12)? 11 : Rtg probit y 0 2 3 4 5 That is, convert the 12s in Rtg to 11s and thereby create a replacement dependent variable, y, that contains the complete sequence 1 to 11. (We can fix this too, but I'd like to hear what Jack thinks before going ahead.) Allin Cottrell