Welch versus Satterthwaith (PR#2111)
Wed, 13 Nov 2002 14:31:56 +0100 (MET)
>>>>> "TL" == Thomas Lumley <firstname.lastname@example.org>
>>>>> on Sun, 6 Oct 2002 09:19:27 -0700 (PDT) writes:
TL> On Sat, 5 Oct 2002 email@example.com wrote:
>> This is not a bug report but didn't see another way to
>> ask a question.
TL> Well, you could try the r-help or r-devel mailing lists
>> For the approximate t-test assuming unequal variances,
>> the R docs cite Welch's method for the df of the
>> approximating distribution. I have several methods
>> books, and they all uses Satterthwaite's method. Why
>> does R use Welch's method where can I learn about Welch's
TL> It's the same method.
TL> The t-statistic is the difference in means divided by a
TL> more-or-less unbiased estimate of its standard
TL> deviation. There's only one natural way to do this and
TL> in sufficiently large samples that's all you need.
TL> The more difficult part is the sampling distribution,
TL> which is approximated by a t-distribution with degrees
TL> of freedom from some complicated formula involving the
TL> two sample sizes and variances. It was indepedently
TL> derived by Welch and Satterthwaite.
I've added "Satterthwaite" to the help page now.
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: firstname.lastname@example.org