Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: - xtlogit- and predictnl and -xtreg-


From   Christoph Engel <[email protected]>
To   [email protected]
Subject   Re: st: - xtlogit- and predictnl and -xtreg-
Date   Wed, 17 Aug 2011 23:29:02 +0200

Ricardo,

xtreg dv ivs period
margins, dydx(*) over(period)

might do what you want

Christoph Engel

Am 17.08.2011 22:39, schrieb Ricardo Ovaldia:
I wold like to get XB estimates from -xtreg- at fixed points in time. I used a command similar to the one that Maarten suggested for -xtlogit-:

predictnl or = exp(_b[1.group] +_b[1.group#c.year]*c.year), ci(lb ub)

of course, I do not want the eform and my group variable has 3 levels instead of two so I used:

predictnl xbeta = _b[1.group] +_b[1.group#c.year]*c.year +_b[2.group] +_b[2.group#c.year]*c.year, ci(lb ub)

I do not get the correct estimates.

Thank you,
Ricardo

Ricardo Ovaldia, MS
Statistician
Oklahoma City, OK


--- On Wed, 8/17/11, Ricardo Ovaldia<[email protected]>  wrote:

From: Ricardo Ovaldia<[email protected]>
Subject: Re: st: - xtlogit- and predictnl
To: [email protected]
Date: Wednesday, August 17, 2011, 9:05 AM
My mistake, you are correct Maarten,
they should be symmetric, and they are, in the or scale.

Okay, I think I got it. If I do not exp() the expression in
-predictnl-, I will get the corret p-values.

Thank you again,
Ricardo

Ricardo Ovaldia, MS
Statistician
Oklahoma City, OK

Ricardo Ovaldia, MS
Statistician
Oklahoma City, OK


--- On Wed, 8/17/11, Maarten Buis<[email protected]>
wrote:

From: Maarten Buis<[email protected]>
Subject: Re: st: - xtlogit- and predictnl
To: [email protected]
Date: Wednesday, August 17, 2011, 8:49 AM
On Wed, Aug 17, 2011 at 3:34 PM,
Ricardo Ovaldia wrote:
predictnl or = exp(_b[1.group] +
_b[1.group#c.year]*c.year), ci(lb ub) p(pval)
but included the extra optiop -p()- to obtain
p-values
My question is why do the pvalues and the CI do
not
agree?
It seems like the CI are incorrect. They are not
symmetric in the log-scale.

I guess that the null-hypothesis is that the odds
ratio
equals 0,
which does not make sense in this case, but how is
Stata
supposed to
know that our expression happens to be an odds ratio?
The
confidence
interval is symmetric in the or scale rather than the
log(or) scale
because -predictnl- just applies the delta method on
our
expression to
get the standard errors and than moves on to use these
to
compute
confidence intervals.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



--
_________________________________________________________________
Prof. Dr. Christoph Engel
Max-Planck-Institut zur
Erforschung von Gemeinschaftsgütern
Max Planck Institute for Research on Collective Goods
Kurt-Schumacher-Strasse 10
D 53113 Bonn
Tel. +49/228/91416-10
Fax +49/228/91416-11
e-mail: [email protected]
http://www.coll.mpg.de
http://www.coll.mpg.de/engel.html
http://ideas.repec.org/e/pen22.html
http://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=251559
_________________________________________________________________

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index