LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (September 2002, week 4)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Mon, 23 Sep 2002 14:09:55 -0700
Reply-To:     Dale McLerran <stringplayer_2@YAHOO.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Dale McLerran <stringplayer_2@YAHOO.COM>
Subject:      Re: heteroscedasticity robust t test
Comments: To: Xuesong Hu <xhu@USC.EDU>
In-Reply-To:  <amgn40$orq$1@usc.edu>
Content-Type: text/plain; charset=us-ascii

Here is what the PROC REG documentation has to say:

When you specify the SPEC option, tests listed in the TEST statement are performed with both the usual covariance matrix and the heteroscedasticity consistent covariance matrix. Tests performed with the consistent covariance matrix are asymptotic. For more information, refer to White (1980).

It would appear to me that what is reported from the fit of the MODEL statement would not be heteroscedasticity consistent, but functions of the parameters coded in a TEST statement would be heteroscedasticity consistent. Still, I would prefer the direct approach of modeling the variance structure employing the MIXED procedure rather than employing the sandwich estimator. With the MIXED procedure there are all manner of modeling the variance including 1) specifying that the variance is a power-of-the-mean, and 2) that the variance differs according to some group variable. I believe that you would know more about your data, and the tests obtained when the variance structure is modeled would not rely on asymptotics to the same degree as tests obtained from PROC REG employing the SPEC option.

Dale

--- Xuesong Hu <xhu@USC.EDU> wrote: > Can SAS conduct heteroscedasticity robust t test? I check the manual > for REG > proc, it has a SPEC option to test the heteroscedasticity, however, > it does > not mention whether it can produce heteroscedasticity consistent t > statistics. > > Anybody know how to produce a heteroscedasticity robust t statistics? > thanks > a lot!

===== --------------------------------------- Dale McLerran Fred Hutchinson Cancer Research Center mailto: dmclerra@fhcrc.org Ph: (206) 667-2926 Fax: (206) 667-5977 ---------------------------------------

__________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com


Back to: Top of message | Previous page | Main SAS-L page