One additional historical source that I would suggest is the following which was written by Joan Fisher Box (yes, Fisher's daughter and wife of George. E. P. Box; she is also author of Sir Ronald's biography which I also recommend):
Fisher, J. F. (1987) Guinness, Gosset, Fisher and small samples. Statistical Science, 2(1), 45-52. For those of you with Jstor access, the stable URL is: http://links.jstor.org/sici?sici=0883-4237%28198702%292%3A1%3C45%3AGGFASS%3E2.0.CO%3B2-E or http://tinyurl.com/2o4v3y As for Fisher and Pearson's "feud", both, I think, engaged in the intellectual version of a "Greco-Roman Thumb to the Eye". Both could be quite harsh with each other and others with whom they disagreed (e.g., as editors of journals, they could reject papers with which they disagreed). -Mike Palij New York University [EMAIL PROTECTED] On Thu, 04 Oct 2007 11:44:30 -0700, G. Marc Turner wrote: > >Well, I guess to be fair we have to remember that Gosset >originally used z instead of t for his formulas in 1908. It was >later that Fisher came along and changed Gosset's z to t >(and added the df concept in I believe) because he wanted >to use z for something else... I believe it was Gosset though >that, when introducing the concept of his z (now t) in 1908 >referred to it as the difference between a sample mean and >the population mean divided by the standard deviation of the >sample, so that z was a measure in terms of standard deviations >(of the sample)... it could be that Fisher took this concept >and made z equal the mean difference divided by the SD of >the population instead of the SD of the sample while at the >same time arbitrarily renaming Gosset's z to t and changing it >to be divided by the DF instead of just the sample size. There >was a "feud" between Fisher and Pearson, whom Gosset >thanks a lot in his work, so it easily could be that Fisher chose >to use z just to irritate Pearson and Gosset... though he might >have given a reason somewhere later. I haven't had a chance >to go back and re-read everything related to this to figure out >the exact details, but most of these original works are available >on-the web, just not all in one spot... I'm not sure what it says >about me that I find the history of statistics so interesting... > >- Marc > >PS- For others who are (or might be) interested in the history >of statistics , I do recommend "The Lady Tasting Tea: How Statistics >Revolutionized Science in the Twentieth Century" by Salsburg if >you haven't read it already... and I welcome other recommendations >people might have on the topic :) ---
