RE: st: RE: RE: Robust Standard Errors in Small Sample Sizes

[Cameron McIntosh <cnm100@hotmail.com>]

With N=42, I might run to some exact (perhaps non-parametric) methods:
Oster, R.A., & Hilbe, J.M. (2008a). An Examination of Statistical Software Packages for Parametric and Nonparametric Data Analyses Using Exact Methods. The American Statistician, 62(1), 74-84.

Oster, R.A., & Hilbe, J.M. (2008b). Rejoinder to “An Examination of Statistical Software Packages for Parametric and Nonparametric Data Analyses Using Exact Methods”. The American Statistician, 62(2), 173-176.

Sawilowsky, S.S. (2011). Statistical Reanalysis of Jewish Priests’ and Non-Priests’ Haplotypes Using Exact Methods. SAGE Open, 1-3.http://sgo.sagepub.com/content/early/2011/06/10/2158244011413475.full.pdf

Cytel Software Corporation (2010). StatXact Version 9. Cambridge, MA: Cytel Software Corporation.

Cytel Software Corporation (2002). LogXact 5 user manual. Cambridge, MA: Cytel Software Corporation.

Cytel has a 30-day free trial of StatXact and LogXact, but you may have to ask them for a direct link to download them, as their website was recently upgraded and the demos haven't been migrated quite yet:

http://www.cytel.com/services

Cam

> Subject: Re: st: RE: RE: Robust Standard Errors in Small Sample Sizes
> From: william@williambuchanan.net
> Date: Thu, 21 Jun 2012 11:29:03 -0700
> To: statalist@hsphsun2.harvard.edu
> 
> Hi Quinn,
> 
> In addition to showing exactly what command you used, looking at the reference for the -vce()- option in the reference manuals will give you clearer guidance regarding which standard errors are more appropriate to use in smaller samples.
> 
> HTH,
> Billy
> 
> Sent from my iPhone
> 
> On Jun 21, 2012, at 11:20, "Swanquist, Quinn Thomas" <qswanqui@utk.edu> wrote:
> 
> > Sorry, yes the Breusch-Pagan test indicates heteroskedasticity. So if that's the case, am I in the clear to use robust standard errors regardless of sample size?
> > 
> > Quinn Swanquist
> > qswanqui@utk.edu
> > 
> > 
> > 
> > 
> > -----Original Message-----
> > From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Popick, Stephen J.
> > Sent: Thursday, June 21, 2012 2:12 PM
> > To: statalist@hsphsun2.harvard.edu
> > Subject: st: RE: Robust Standard Errors in Small Sample Sizes
> > 
> > Do you have an a priori reason for suspecting heteroskedasticity, or did you perform any such tests to check that returned statistically significant results?
> > 
> > 
> > -----Original Message-----
> > From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Swanquist, Quinn Thomas
> > Sent: Thursday, June 21, 2012 2:08 PM
> > To: statalist@hsphsun2.harvard.edu
> > Subject: st: Robust Standard Errors in Small Sample Sizes
> > 
> > I have a relatively small sample size (n=42) and find a statistically significant result using robust standard errors but no significance without robust standard errors. Is there a problem with using heteroskedasticity robust standard errors in small sample sizes?
> > 
> > Quinn Swanquist
> > qswanqui@utk.edu
> > 
> > 
> > 
> > 
> > 
> > *
> > *   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/