Re: st: -robust- and -cluster- options in xtreg

["Narahari H.S." <naraharihs@yahoo.co.in>]

Thank you Austin. Apologies - I should have checked the archives before sending out my mail. Will definitely do so henceforth.




 

----- Original Message -----
From: Austin Nichols <austinnichols@gmail.com>
To: statalist@hsphsun2.harvard.edu
Cc: 
Sent: Saturday, 9 March 2013 2:00 AM
Subject: Re: st: -robust- and -cluster- options in xtreg

Narahari H.S. <naraharihs@yahoo.co.in>:
Both FAQs:
1)
http://www.stata.com/statalist/archive/2009-08/msg01094.html
2)
http://www.stata.com/help.cgi?j_robustsingular

On Fri, Mar 8, 2013 at 6:36 AM, Narahari H.S. <naraharihs@yahoo.co.in> wrote:
> Hello all,
>
> My question is about the difference between -robust- and -cluster- options when using xtreg.
>
> 1) The stata output seems to suggest that some clustering is done in the robust option also. What is the difference in the clustering between the 2 options?
> 2) Under the cluster option, stata gives a blank Wald chi2. What could be causing this?
>
> I have given the top portion of the stata output below
>
>
> *******************************Using cluster option*******************************
> Random-effects GLS regression          Number of obs      =      2351
> Group variable: idn                             Number of groups   =       468
>
> R-sq:  within  = 0.1830                         Obs per group: min =         1
>        between = 0.2300                         avg =       5.0
>        overall = 0.2089                            max =        18
>
> Random effects u_i ~ Gaussian              Wald chi2(31)      =         .
> corr(u_i, X)       = 0 (assumed)                Prob > chi2        =         .
>                                   (Std. Err. adjusted for 468 clusters in idn)
>
>
> *******************************Using robust option*******************************
>
> Random-effects GLS regression          Number of obs      =      2351
> Group variable: idn                             Number of groups   =       468
>
> R-sq:  within  = 0.1830                         Obs per group: min =         1
>        between = 0.2300                          avg =       5.0
>        overall = 0.2089                             max =        18
>
> Random effects u_i ~ Gaussian                   Wald chi2(32)      =    679.48
> corr(u_i, X)       = 0 (assumed)                Prob > chi2        =    0.0000
>                                     (Std. Err. adjusted for clustering on idn)
>
>
> ***************************************************************************************
>
>
> I have detailed my regressions model below. Please let me know if further information will be helpful
>
>
> I have an annual panel data of firms and am running the below regression model
>
> xtreg depvar var1 var2 var1*var2 var3 var4 var5 var6 industrydummies
>
> where
> var1 and var2 are dummy variables (var1*var2 is their interaction)
> depvar, var3-var6 are continuous variables
> industrydummies are dummy variables (included to control for industry fixed effects)
>
> Thanks for your time
> Hari
*
*   For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/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/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/