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

st: Re: specifying difference in H0 of ttest


From   "ignacio_cedes" <[email protected]>
To   [email protected]
Subject   st: Re: specifying difference in H0 of ttest
Date   Thu, 22 Jan 2004 15:26:10 -0000

Could someone tell me how can I write to the group from my mail and
recive your mails?

I'm trying to participate on the list but I don't know how is it.

Thanks a lot list.

Regards

Ignacio


--- In [email protected], Richard Williams
<Richard.A.Williams.5@n...> wrote:
> At 10:35 PM 1/18/2004 -0500, Michel Camus wrote:
> >In fact,  they do provide the same results.
> >I am not a statistician, but using the robust option with regress,
as
> >suggested by Christopher Baum, gave me the same result as teh
ttest with
> >unequal variance after subtracting the hypothesized difference.
> >I do not know why you obtain different results" Do you actually
have two
> >groups with different variances in you simulated example, Richard?
> >See my response (in 5 minutes from now... to Karen, simulating
data
> >similar to hers).
>
> In the various little experiments i have done, the differences
between
> regression with robust standard errors and the t-test with unequal
> variances have been very small.  My guess is the approaches aren't
100%
> equivalent (unlike the t-test with equal variances versus
regression) but
> that it would only make a difference in some incredibly borderline
> situation.  On the other hand, I have been surprised before, so it
might
> not hurt to do it both ways.
>
> I like your examples in your later message.  In terms of what
should be
> presented in the writeup, however Karen does it, she may just want
to give
> the final test statistic; my guess is the audience won't much care
how she
> computed it.  If I did have to explain the nitty-gritty of the
> computations, I suspect t-test routines are familiar to more people
than
> regression with robust standard errors.  But, the main thing is to
get the
> answer, and this discussion has come up with some good ways to do
that.
>
> *
> *   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/

*
*   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