Statalist The Stata Listserver


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

Re: st: robust


From   Scott Cunningham <scunning@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: robust
Date   Sat, 13 May 2006 23:04:46 -0400

Debbie,

xtreg supports a robust option. I wonder if this is because you are not telling Stata the identifying variable for the panel? For instance, assume that each observation is delineated by the variable - id. So, like this:

id lnCA lnC1 lnC10 date lnincome
1 100 200 300 2000 25000
1 101 201 301 2001 28000
1 102 202 302 2002 40500
2 200 300 400 2000 14000
2 ....

This would be estimated as:

. xi:xtreg lnincome lnCA lnC1 lnC10 lnC15 i.date, fe i(id) robust

In other words, you need to tell Stata the variable that marks observations from a common "family", so to speak. Try that. If that doesn't work, then I don't know why this isn't working, as xtreg, robust works fine for me.

sc



On May 13, 2006, at 10:24 PM, Susie Enders wrote:


Hi,

I am running panel data regressions with 2-way fixed
effects, between companies and between years. I need
to obtain robust standard errors and t-statistics. The
problem is that when I try to use the robust command I
get an error message.

When I run the regression as follows it works fine:
xi:xtreg lnincome lnCA lnC1 lnC10 lnC15 i.date, fe

However when I add "robust" as follows:
. xi:xtreg lnincome lnCA lnC1 lnC10 lnC15 i.date, fe
robust

I get the following output (in the first line it looks
as if it will give the regression output, but then
instead gives an error message that the robust option
is not allowed. The part about naturally coded,
_Idate_136 omitted can be ignored as it gives this
anyway without robust):

i.date _Idate_136-183 (naturally
coded; _Idate_136 omitted)
option robust not allowed
r(198);

I have tried studying the help files on robust, xtreg,
xtdata etc, but still do not understand the problem.
Is it impossible theoretically to get robust t stats
and standard errors with such a regression, or is it a
programming issue? (I did also read about clustering
but I was wondering if it is possible to get the
robust stats without that).

I would really appreciate any help, thanks!

Debbie


Send instant messages to your online friends http:// uk.messenger.yahoo.com
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index