Statalist

st: hausman test

Classic

List

Threaded

11 messages Options

Marco Buur

st: hausman test

I run hausman test and got Prob>chi2 = 0.1668
According my undestanding this is confmation that I should use random
effect model. Am i right ?

Thanks a lot

xtreg A B C , fe robust
est store fixed
xtreg A B C , re robust
hausman fixed

Test: Ho: difference in coefficients not systematic

chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
= 11.67
Prob>chi2 = 0.1668
(V_b-V_B is not positive definite)

xtreg Public_cap_ln Prevalence_ln GNI_cap_ln reg4 reg8 Year_* if
Tag!=1 , fe robust

est store fixed

xtreg Public_cap_ln Prevalence_ln GNI_cap_ln reg4 reg8 Year_* if
Tag!=1 , re robust

hausman fixed
*
* 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/

archsteve

Re: st: hausman test

The positive definite error is concerning. I would suggest using
xtoverid as an alternative test.

Schaffer, M.E., Stillman, S. 2006. xtoverid: Stata module to
calculate tests of overidentifying restrictions after xtreg, xtivreg,
xtivreg2 and xthtaylor http://ideas.repec.org/c/boc/bocode/s456779.html

On Thu, Aug 13, 2009 at 8:38 AM, Marco Buur<[hidden email]> wrote:

> I run hausman test and got Prob>chi2 = 0.1668
> According my undestanding this is confmation that I should use random
> effect model. Am i right ?
>
> Thanks a lot
>
> xtreg A B C , fe robust
> est store fixed
> xtreg A B C , re robust
> hausman fixed
>
>
> Test: Ho: difference in coefficients not systematic
>
> chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
> = 11.67
> Prob>chi2 = 0.1668
> (V_b-V_B is not positive definite)
>
>
>
>
> xtreg Public_cap_ln Prevalence_ln GNI_cap_ln reg4 reg8 Year_* if
> Tag!=1 , fe robust
>
> est store fixed
>
> xtreg Public_cap_ln Prevalence_ln GNI_cap_ln reg4 reg8 Year_* if
> Tag!=1 , re robust
>
> hausman fixed
> *
> * 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/

Schaffer, Mark E

RE: st: hausman test

Marco,

-hausman- isn't valid with -robust- (see Hayashi, "Econometric", 2000, p. 234, note 18).

Steve's advice to use -xtoverid- is the right way to go. It implements a Hausman-like GMM test that extends to robust VCV estimators.

--Mark

> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf Of
> Steven Archambault
> Sent: 13 August 2009 16:24
> To: [hidden email]
> Subject: Re: st: hausman test
>
> The positive definite error is concerning. I would suggest
> using xtoverid as an alternative test.
>
>
> Schaffer, M.E., Stillman, S. 2006. xtoverid: Stata
> module to calculate tests of overidentifying restrictions
> after xtreg, xtivreg,
> xtivreg2 and xthtaylor
> http://ideas.repec.org/c/boc/bocode/s456779.html
>
>
>
>
> On Thu, Aug 13, 2009 at 8:38 AM, Marco
> Buur<[hidden email]> wrote:
> > I run hausman test and got Prob>chi2 = 0.1668 According my
> > undestanding this is confmation that I should use random
> effect model.
> > Am i right ?
> >
> > Thanks a lot
> >
> > xtreg A B C , fe robust
> > est store fixed
> > xtreg A B C , re robust
> > hausman fixed
> >
> >
> > Test: Ho: difference in coefficients not systematic
> >
> > chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
> > = 11.67
> > Prob>chi2 = 0.1668
> > (V_b-V_B is not positive definite)
> >
> >
> >
> >
> > xtreg Public_cap_ln Prevalence_ln GNI_cap_ln reg4 reg8 Year_* if
> > Tag!=1 , fe robust
> >
> > est store fixed
> >
> > xtreg Public_cap_ln Prevalence_ln GNI_cap_ln reg4 reg8 Year_* if
> > Tag!=1 , re robust
> >
> > hausman fixed
> > *
> > * 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/
>

--
Heriot-Watt University is a Scottish charity
registered under charity number SC000278.

*
* 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/

Marco Buur

Re: st: hausman test

In reply to this post by Marco Buur

Thanks Mark and Steve

I tried to use

> xtreg A B C , fe robust
> est store fixed
> xtreg A B C , re robust
> xtoverid

but Stata says "Error - must have ivreg2 version 2.1.15 or greater installed"
I installed ivreg2 it but there is no any progress.
Please advice

On Thu, Aug 13, 2009 at 4:38 PM, Marco Buur<[hidden email]> wrote:

*
* 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/

Martin Weiss-5

AW: st: hausman test

<>

How did you install the files? Through the usual -ssc inst ivreg2-. If so,
you should get the latest version. What does -which ivreg2- give you?

HTH
Martin

-----Ursprüngliche Nachricht-----
Von: [hidden email]
[mailto:[hidden email]] Im Auftrag von Marco Buur
Gesendet: Freitag, 14. August 2009 12:19
An: [hidden email]
Betreff: Re: st: hausman test

Thanks Mark and Steve

I tried to use

> xtreg A B C , fe robust
> est store fixed
> xtreg A B C , re robust
> xtoverid

but Stata says "Error - must have ivreg2 version 2.1.15 or greater
installed"
I installed ivreg2 it but there is no any progress.
Please advice

On Thu, Aug 13, 2009 at 4:38 PM, Marco Buur<[hidden email]> wrote:

*
* 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/

Siyam, Amani

st: Can Stata read .xml generated by Acrobate Desinger 7

Dear Stata-listers,

We have designed a short questionnaire using Acrobat 7 and the only
option we have for the respondents to email their entries is via an xml
data file.

I tried using xmluse but it does not seem to recognize the "doctype".

Many thanks for your help in advance.

Best regards,

Amani

*
* 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/

Marco Buur

Re: st: hausman test

In reply to this post by Marco Buur

Dear Martin

I have installed it from :

SJ-7-4 st0030_3 . . . . Enhanced routines for IV/GMM estimation and testing
. . . . . . . . . . . . . C. F. Baum, M. E. Schaffer, and S. Stillman
(help ivactest, ivendog, ivhettest, ivreg2, ivreset,
overid, ranktest if installed)
Q4/07 SJ 7(4):465--506
extension of IV and GMM estimation addressing hetero-
skedasticity- and autocorrelation-consistent standard
errors, weak instruments, LIML and k-class estimation,
tests for endogeneity and Ramsey's regression
specification-error test, and autocorrelation tests
for IV estimates and panel-data IV estimates

If i run ssc inst ivreg2 i got

*****************ssc inst ivreg2**********************
checking ivreg2 consistency and verifying not already installed...

the following files already exist and are different:
c:\ado\plus\i\ivreg2.ado
c:\ado\plus\i\ivreg2.hlp
c:\ado\plus\i\ivreg2_p.ado

no files installed or copied
(no action taken)
r(602);

****************which ivreg2**************

c:\ado\plus\i\ivreg2.ado
*! ivreg2 2.0.04 14Apr2004
*! authors cfb & mes
*! cloned from official ivreg version 5.0.9 19Dec2001

Please advice

On Fri, Aug 14, 2009 at 12:21 PM, Martin Weiss<[hidden email]> wrote:

>
> <>
>
> How did you install the files? Through the usual -ssc inst ivreg2-. If so,
> you should get the latest version. What does -which ivreg2- give you?
>
>
>
> HTH
> Martin
>
>
> -----Ursprüngliche Nachricht-----
> Von: [hidden email]
> [mailto:[hidden email]] Im Auftrag von Marco Buur
> Gesendet: Freitag, 14. August 2009 12:19
> An: [hidden email]
> Betreff: Re: st: hausman test
>
> Thanks Mark and Steve
>
> I tried to use
>
>> xtreg A B C , fe robust
>> est store fixed
>> xtreg A B C , re robust
>> xtoverid
>
> but Stata says "Error - must have ivreg2 version 2.1.15 or greater
> installed"
> I installed ivreg2 it but there is no any progress.
> Please advice
>
>
>
>
> On Thu, Aug 13, 2009 at 4:38 PM, Marco Buur<[hidden email]> wrote:
>> I run hausman test and got Prob>chi2 = 0.1668
>> According my undestanding this is confmation that I should use random
>> effect model. Am i right ?
>>
>> Thanks a lot
>>
>> xtreg A B C , fe robust
>> est store fixed
>> xtreg A B C , re robust
>> hausman fixed
>>
>>
>> Test: Ho: difference in coefficients not systematic
>>
>> chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
>> = 11.67
>> Prob>chi2 = 0.1668
>> (V_b-V_B is not positive definite)
>>
>>
>>
>
> *
> * 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/

Martin Weiss-5

AW: st: hausman test

<>

The -which- output should be:

. which ivreg2
c:\ado\plus\i\ivreg2.ado
*! ivreg2 2.2.09 17jul2008
*! authors cfb & mes
*! see end of file for version comments

So go to your -c:\ado\plus\i\- folder, kill everything beginning with
-ivreg2- , and then type -ssc inst ivreg2- in Stata. After that, -which-
again and see whether your output matches mine...

HTH
Martin

-----Ursprüngliche Nachricht-----
Von: [hidden email]
[mailto:[hidden email]] Im Auftrag von Marco Buur
Gesendet: Freitag, 14. August 2009 13:48
An: [hidden email]
Betreff: Re: st: hausman test

Dear Martin

I have installed it from :

SJ-7-4 st0030_3 . . . . Enhanced routines for IV/GMM estimation and
testing
. . . . . . . . . . . . . C. F. Baum, M. E. Schaffer, and S.
Stillman
(help ivactest, ivendog, ivhettest, ivreg2, ivreset,
overid, ranktest if installed)
Q4/07 SJ 7(4):465--506
extension of IV and GMM estimation addressing hetero-
skedasticity- and autocorrelation-consistent standard
errors, weak instruments, LIML and k-class estimation,
tests for endogeneity and Ramsey's regression
specification-error test, and autocorrelation tests
for IV estimates and panel-data IV estimates

If i run ssc inst ivreg2 i got

*****************ssc inst ivreg2**********************
checking ivreg2 consistency and verifying not already installed...

the following files already exist and are different:
c:\ado\plus\i\ivreg2.ado
c:\ado\plus\i\ivreg2.hlp
c:\ado\plus\i\ivreg2_p.ado

no files installed or copied
(no action taken)
r(602);

****************which ivreg2**************

c:\ado\plus\i\ivreg2.ado
*! ivreg2 2.0.04 14Apr2004
*! authors cfb & mes
*! cloned from official ivreg version 5.0.9 19Dec2001

Please advice

On Fri, Aug 14, 2009 at 12:21 PM, Martin Weiss<[hidden email]> wrote:

Marco Buur

Re: st: hausman test

In reply to this post by Marco Buur

Thanks it works!!! I assume that based on this Sargan-Hansen test
below I need to reject RE model
is it the same Ho like in Hausman test?
Thank you in advance

xtreg A B C , fe robust
est store fixed
xtreg A B C , re robust
xtoverid, robust

Test of overidentifying restrictions: fixed vs random effects
Cross-section time-series model: xtreg re robust
Sargan-Hansen statistic 25.772 Chi-sq(8) P-value = 0.0011

On Fri, Aug 14, 2009 at 1:51 PM, Martin Weiss<[hidden email]> wrote:

>
> <>
>
> The -which- output should be:
>
>
>
> . which ivreg2
> c:\ado\plus\i\ivreg2.ado
> *! ivreg2 2.2.09 17jul2008
> *! authors cfb & mes
> *! see end of file for version comments
>
>
> So go to your -c:\ado\plus\i\- folder, kill everything beginning with
> -ivreg2- , and then type -ssc inst ivreg2- in Stata. After that, -which-
> again and see whether your output matches mine...
>
>
>
> HTH
> Martin
>
>
> -----Ursprüngliche Nachricht-----
> Von: [hidden email]
> [mailto:[hidden email]] Im Auftrag von Marco Buur
> Gesendet: Freitag, 14. August 2009 13:48
> An: [hidden email]
> Betreff: Re: st: hausman test
>
> Dear Martin
>
> I have installed it from :
>
> SJ-7-4 st0030_3 . . . . Enhanced routines for IV/GMM estimation and
> testing
> . . . . . . . . . . . . . C. F. Baum, M. E. Schaffer, and S.
> Stillman
> (help ivactest, ivendog, ivhettest, ivreg2, ivreset,
> overid, ranktest if installed)
> Q4/07 SJ 7(4):465--506
> extension of IV and GMM estimation addressing hetero-
> skedasticity- and autocorrelation-consistent standard
> errors, weak instruments, LIML and k-class estimation,
> tests for endogeneity and Ramsey's regression
> specification-error test, and autocorrelation tests
> for IV estimates and panel-data IV estimates
>
>
>
> If i run ssc inst ivreg2 i got
>
> *****************ssc inst ivreg2**********************
> checking ivreg2 consistency and verifying not already installed...
>
> the following files already exist and are different:
> c:\ado\plus\i\ivreg2.ado
> c:\ado\plus\i\ivreg2.hlp
> c:\ado\plus\i\ivreg2_p.ado
>
> no files installed or copied
> (no action taken)
> r(602);
>
> ****************which ivreg2**************
>
>
> c:\ado\plus\i\ivreg2.ado
> *! ivreg2 2.0.04 14Apr2004
> *! authors cfb & mes
> *! cloned from official ivreg version 5.0.9 19Dec2001
>
>
> Please advice
>
> On Fri, Aug 14, 2009 at 12:21 PM, Martin Weiss<[hidden email]> wrote:
>>
>> <>
>>
>> How did you install the files? Through the usual -ssc inst ivreg2-. If so,
>> you should get the latest version. What does -which ivreg2- give you?
>>
>>
>>
>> HTH
>> Martin
>>
>>
>> -----Ursprüngliche Nachricht-----
>> Von: [hidden email]
>> [mailto:[hidden email]] Im Auftrag von Marco Buur
>> Gesendet: Freitag, 14. August 2009 12:19
>> An: [hidden email]
>> Betreff: Re: st: hausman test
>>
>> Thanks Mark and Steve
>>
>> I tried to use
>>
>>> xtreg A B C , fe robust
>>> est store fixed
>>> xtreg A B C , re robust
>>> xtoverid
>>
>> but Stata says "Error - must have ivreg2 version 2.1.15 or greater
>> installed"
>> I installed ivreg2 it but there is no any progress.
>> Please advice
>>
>>
>>
>>
>> On Thu, Aug 13, 2009 at 4:38 PM, Marco Buur<[hidden email]> wrote:
>>> I run hausman test and got Prob>chi2 = 0.1668
>>> According my undestanding this is confmation that I should use random
>>> effect model. Am i right ?
>>>
>>> Thanks a lot
>>>
>>> xtreg A B C , fe robust
>>> est store fixed
>>> xtreg A B C , re robust
>>> hausman fixed
>>>
>>>
>>> Test: Ho: difference in coefficients not systematic
>>>
>>> chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
>>> = 11.67
>>> Prob>chi2 = 0.1668
>>> (V_b-V_B is not positive definite)
>>>
>>>
>>>
>>
>> *
>> * 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/
>
>
> *
> * 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/

Martin Weiss-5

AW: st: hausman test

<>

Well, all you have to do is check out

*************
help ivreg2,mark(s_overid)
*************

and note that you conclusively reject your null - which is stated there...

HTH
Martin

-----Ursprüngliche Nachricht-----
Von: [hidden email]
[mailto:[hidden email]] Im Auftrag von Marco Buur
Gesendet: Freitag, 14. August 2009 14:50
An: [hidden email]
Betreff: Re: st: hausman test

Thanks it works!!! I assume that based on this Sargan-Hansen test
below I need to reject RE model
is it the same Ho like in Hausman test?
Thank you in advance

xtreg A B C , fe robust
est store fixed
xtreg A B C , re robust
xtoverid, robust

Test of overidentifying restrictions: fixed vs random effects
Cross-section time-series model: xtreg re robust
Sargan-Hansen statistic 25.772 Chi-sq(8) P-value = 0.0011

On Fri, Aug 14, 2009 at 1:51 PM, Martin Weiss<[hidden email]> wrote:

wrote:
>>
>> <>
>>
>> How did you install the files? Through the usual -ssc inst ivreg2-. If
so,

>> you should get the latest version. What does -which ivreg2- give you?
>>
>>
>>
>> HTH
>> Martin
>>
>>
>> -----Ursprüngliche Nachricht-----
>> Von: [hidden email]
>> [mailto:[hidden email]] Im Auftrag von Marco Buur
>> Gesendet: Freitag, 14. August 2009 12:19
>> An: [hidden email]
>> Betreff: Re: st: hausman test
>>
>> Thanks Mark and Steve
>>
>> I tried to use
>>
>>> xtreg A B C , fe robust
>>> est store fixed
>>> xtreg A B C , re robust
>>> xtoverid
>>
>> but Stata says "Error - must have ivreg2 version 2.1.15 or greater
>> installed"
>> I installed ivreg2 it but there is no any progress.
>> Please advice
>>
>>
>>
>>
>> On Thu, Aug 13, 2009 at 4:38 PM, Marco Buur<[hidden email]> wrote:
>>> I run hausman test and got Prob>chi2 = 0.1668
>>> According my undestanding this is confmation that I should use random
>>> effect model. Am i right ?
>>>
>>> Thanks a lot
>>>
>>> xtreg A B C , fe robust
>>> est store fixed
>>> xtreg A B C , re robust
>>> hausman fixed
>>>
>>>
>>> Test: Ho: difference in coefficients not systematic
>>>
>>> chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
>>> = 11.67
>>> Prob>chi2 = 0.1668
>>> (V_b-V_B is not positive definite)
>>>
>>>
>>>
>>
>> *
>> * 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/
>
>
> *
> * 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/
>

archsteve

Re: st: hausman test

In reply to this post by Marco Buur

You need to install the program.

ssc install ivreg2

On Fri, Aug 14, 2009 at 4:18 AM, Marco Buur<[hidden email]> wrote:

> Thanks Mark and Steve
>
> I tried to use
>
>> xtreg A B C , fe robust
>> est store fixed
>> xtreg A B C , re robust
>> xtoverid
>
> but Stata says "Error - must have ivreg2 version 2.1.15 or greater installed"
> I installed ivreg2 it but there is no any progress.
> Please advice
>
>
>
>
> On Thu, Aug 13, 2009 at 4:38 PM, Marco Buur<[hidden email]> wrote:
>> I run hausman test and got Prob>chi2 = 0.1668
>> According my undestanding this is confmation that I should use random
>> effect model. Am i right ?
>>
>> Thanks a lot
>>
>> xtreg A B C , fe robust
>> est store fixed
>> xtreg A B C , re robust
>> hausman fixed
>>
>>
>> Test: Ho: difference in coefficients not systematic
>>
>> chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
>> = 11.67
>> Prob>chi2 = 0.1668
>> (V_b-V_B is not positive definite)
>>
>>
>>
>
> *
> * 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/