Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Re: specifying difference in H0 of ttest


From   Richard Williams <[email protected]>
To   [email protected]
Subject   Re: st: Re: specifying difference in H0 of ttest
Date   Fri, 16 Jan 2004 07:54:41 -0500

At 07:24 AM 1/16/2004 -0500, Christopher F Baum wrote:
On Jan 16, 2004, at 2:33 AM, Richard wrote:

Yet another more straightforward solution than I initially proposed; about
the only downside I see would be if you wanted to specify unequal variances
on the t-test.
Just use the robust option on the regress command to allow for unequal variances.
I thought about that, but the t-test with the unequal option and regression with the robust option do not produce identical results. In particular, the t-test with unequal variances generally produces these weird fractional degrees of freedom which sometimes differ quite a bit from what regression with robust does. In practice, does it make much difference? I'm not sure. Would either be correct or one considered more correct? Again I'm not sure, but I'm more familiar with the use of a t-test in the sort of situation described than I am with regression using robust standard errors.

Here is what the Stata 8 Reference Manual S-Z says in the ttest documentation, p. 299:

"The ttest with an unequal variance assumption does not lend itself to an easy representation in regression settings, and is not discussed here."


-------------------------------------------
Richard Williams, Associate Professor
OFFICE: (574)631-6668, (574)631-6463
FAX: (574)288-4373
HOME: (574)289-5227
EMAIL: [email protected]
WWW (personal): http://www.nd.edu/~rwilliam
WWW (department): http://www.nd.edu/~soc

*
* 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 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index