Statalist The Stata Listserver

[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: svy:probit, margeff and lincom problem

From   Richard Williams <>
Subject   Re: st: svy:probit, margeff and lincom problem
Date   Thu, 10 Aug 2006 13:19:59 -0400

At 10:19 AM 8/10/2006, Chris Ruebeck wrote:
May I check on what problem mfx2 has with the svy: prefix?  Should I
expect mfx2 to work correctly after svy: probit?  I hope that the
email exchange quoted below indicates only that mfx2 makes problems
for lincom, not that svy: probit makes problems for mfx2 directly.

My coauthors and I have just started using mfx2 because (according to
the abstract on IDEAS) it works well with outreg2.  We were having
problems with svy: probit followed by mfx and then outreg2, but
replacing mfx with mfx2 appears to make things work.
You should be fine. If you ran svy:probit, followed by -mfx2, replace-, and then lincom, you would probably get a fatal error. This is just as well, because if it did give you an answer, it might be wrong! In the e(V) matrix created by -mfx2, replace-, all the off-diagonal elements are 0 (because mfx itself does not provide these values and mfx2 is mostly a shell for mfx).

As noted in the help for mfx2, "It is important to realize that, if mfx does not work for your problem, then mfx2 isn't going to work either. mfx2 simplifies the use of mfx and restructures mfx's output for easier use with table formatting programs. But, when it comes to actually estimating the marginal effects mfx2 has the same limitations that mfx does." That is, while mfx2 saves you some typing and makes it easier to format your results for output, it doesn't magically extend your abilities to do analysis, perform statistical tests, etc. beyond what mfx can do. The next time I update mfx2, I'll make the help even more explicit in saying that commands like test and lincom probably should not be used with the mfx2 estimates.

If someone really really really wants to use lincom with marginal effects, then they should see if margeff will meet their needs. mfx2 works with more programs than does margeff, which is why I wrote it; but when margeff does work, it is quicker and has abilities that go beyond mfx2.

Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX: (574)288-4373
HOME: (574)289-5227
EMAIL: Richard.A.Williams.5@ND.Edu
WWW (personal):
WWW (department):
* For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index