Re: Pearson/Fisher Kurtosis

[Karl L. Wuensch]

I think you hit the nail on the head, Bob.  Thanks.  I got a similar answer locally, and I found in a discussion of skewness, kurtosis, and the normal curve, on [EMAIL PROTECTED] (archived at http://www.biostat.wustl.edu/archives/html/s-news/2002-06/msg00068.html) a post by Bill Venables:  �I prefer the definitions based on cumulants rather than on moments, and for these the normal has skewness and kurtosis BOTH zero.  Of course this is part of Fisher's angry revolt against the Pearsonian tradition, but he had a point: cumulants are much simpler to use for this kind of thing than moments.   Karl W. ----- Original Message ----- From: "Bob Wheeler" [EMAIL PROTECTED] I don't recall seeing the Pearson-Fisher names associated in this way; however, beta2=mu4/(mu2)^2 is one of the two parameters defining the Pearson curves and this could account for its ascription to him. Fisher discussed cumulants quit a bit, and since k4/k2^2=beta2-3, this may account for that ascription.