[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: xtlogit RE on Stata 7 vs Stata 9
At 09:22 PM 11/16/2006, Regina T. Riphahn wrote:
I don't really know, but according to -help version-, if you set
version to less than 9
I have a dataset and run xtlogit with RE in Stata 9. Then I save the
data as a Stata7 dataset using saveold and run xtlogit with RE under
Stata 7. The two xtlogit results differ.
The two data summaries are identical in Stata 7 and Stata 9. The
results of a simple logit (without xt) are also identical.
My questions are:
(1) why do the results differ and
(2) can I obtain the Stata 9 results with the Stata 7 software?
Commands xtintreg, xtlogit, xtprobit, xtcloglog, xtpoisson, and
xttobit revert to using nonadaptive Gauss-Hermite quadrature rather
than adaptive quadrature. In addition, option
quad() (modern name intpoints()) comes back to life.
So, in Stata 9, you might try something like version
7:xtlogit... and see if you get the same results as version 7. Or,
alternatively, see if version 7 lets you use nonadaptive
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
WWW (personal): http://www.nd.edu/~rwilliam
WWW (department): http://www.nd.edu/~soc
* For searches and help try: