Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Output from SKTEST
From
David Hoaglin <[email protected]>
To
[email protected]
Subject
Re: st: Output from SKTEST
Date
Mon, 2 Apr 2012 12:10:34 -0400
Maarten,
It's wonderful to have hanging rootograms available. For the present
purpose, though, a suspended rootogram (also available in the
-hangroot- package) should be even better. But even a suspended
rootogram may not show enough of the behavior in the tails, so a
normal probability plot would be worthwhile.
And if the data are highly skewed, either rootogram will mainly show
that a normal distribution fits poorly.
I strongly agree with your points that other aspects of the variable
are usually more important in a regression.
If the distribution shape of a variable is of interest, I suggest an
approach rooted in Exploratory Data Analysis. One discussion appears
in the chapter "Using Quantiles to Study Shape" in the book Exploring
Data Tables, Trends, and Shapes (DC Hoagiln, F Mosteller, and JW
Tukey, eds), Wiley, 1985.
David Hoaglin
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/