On 18 Jun 2003 09:47:52 -0700, [EMAIL PROTECTED] (Robert J. MacG. Dawson) wrote:
> > > "Jane Roberts" <[EMAIL PROTECTED]> wrote in message > > news:[EMAIL PROTECTED] > > > This is a classical problem. No exact solution exists. See the works by > > > Satterthwise and Walsh. > ZHANG Yan wrote: > > > > Thanks for your information. I googled the keyword "Satterthwise" , but can > > not find one. Then tried "walsh" and find tons of links. > > Could that be "Satterthwaite"? > That seems likely to me. If that is the case, the other name might (or might not) be Welch. Here's a curious google sequence. < Satterthwaite Welch > reports 1040 hits. The names are together in the first half dozen, and the sites are talking about statistics. < Satterthwaite Walsh > reports 1030 hits, despite the lack of any essential connection that I notice -- just two names in the same article. Welch and Walsh each produce a bit over a million hits separately, so that did not account for anything. Then I added "gamma" to each of the first two. S + Welch + gamma reports 60, which is like I expected. S + Walsh + gamma reports 69, which is far more than I expected. There's been a statistician Walsh who *has* written in the area. But I would not rule out Welch if the original citation missed it on Satterthwaite. -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html "Taxes are the price we pay for civilization." Justice Holmes. . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
