in article [EMAIL PROTECTED], christopher.mecklin at
[EMAIL PROTECTED] wrote on 8/29/00 8:05 AM:
> Ronny,
> Kurtosis is poorly defined in almost every elementary stat textbook around.
> "Tailedness" and "peakedness" are both components of kurtosis. It is
> impossible to adequately explain kurtosis with just one component. A good
> reference that discusses this in much more detail is:
> DeCarlo, L.T. (1997). "On the meaning and use of kurtosis". Psychological
> Methods, 2, 292-307.
I'm joining this thread late but i have to make a couple of comments:
1. the DeCarlo article is excellent; read it. why so many (including some
of this thread) is confused about kurtosis is because most textbooks
illustrate kurtosis with figures that vary not only kurtosis but also
variance. if you control the variance to be the same and only vary kurtosis,
the pictures look rather different and don't seem to have much to do with
the peak.
2. in any discussion of the comparison of means and medians i have to throw
in that wonderful fact that the mean and median can be at most one standard
deviation apart. one of those lovely results that makes theoretical
statistics so much fun.
gary
--
[EMAIL PROTECTED]
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
