Martin, A fun question. Looking back at my oldest books, Feller (1950) used chi-square. Then I walked down the hall to our little statistics library and looked at Johnson and Kotz, "Continous Univariate Distributions", since each chapter therein has comments about the history of the distribution.
a. They use 'chi-square' throughout their history section, tracing the distribution back to work in the 1800s. But, those earliest papers apparently didn't name their results as chi- whatever, so an "origin" story didn't pan out. b. They have 13 pages of references, and for fun I counted the occurence of variants. The majority of papers don't have the word in the title at all and the next most common is the Greek symbol. Here are the years of the others: chi-square: 73 43 65 80 86 73 82 73 69 69 78 64 64 86 65 86 82 82 76 82 88 81 74 77 87 86 93 69 60 88 88 80 77 41 59 79 31 chi-squared: 72 76 82 83 89 79 69 67 77 78 69 77 83 88 87 89 78 chi: 92 73 89 87 chi-squares: 77 83 chi-bar-square: 91 There doesn't look to be a trend over time. The 1922 Fisher reference uses the Greek symbol, by the way. Terry T [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.