Differences in SE between version 2.4 and 3.0

[sixtystrat]

I am running a 2-species model interaction model and am having trouble getting Presence 3.0 to give me the same output as 2.4. For example, my estimate of phi with 2.4 was 1.02 (SE 0.05) whereas the estimate with 3.0 is 1.02 (SE 19.0). The output says that the variance covariance matrix was not successful so I suspect a numerical (convergence) issue. Why did the estimation procedure in 2.4 (apparently) work? Is there some place I can download 2.4 (the Patuxent website has 3.0 even though it says 2.4). Thanks for the help!
Joe Clark

[sixtystrat]

By the way, I ran the 2.4 analysis in December 08 which was supposedly after the fix of this variance estimation problem (feb 08).

[sixtystrat]

Oh, one more thing. The results with 2.4 and 3.0 are the same unless I include a particular site covariate that has a wide range (distance from reintroduction site which ranges from 30m to 40000m). I tried to normalize it but that didn't help (still got high SE for phi). I can't see if the scaling button is doing anything... Thanks again.
Joe