[R] KS for goodness of fit

Frank E Harrell Jr fharrell at virginia.edu
Fri Nov 29 14:28:05 CET 2002

On Fri, 29 Nov 2002 12:20:01 +0100
Emmanuel Paradis <paradis at isem.univ-montp2.fr> wrote:

> >Dear All,
> >I have two distributions which I don't their nature. I want to check 
> >whether they come from the same distribution. I know that I can use KS 
> >test however the standart function ks.test applies only the ks test for 
> >testing the difference of two samples (non-parametric). By the way the 
> >distribution are of Euclidean distances. One of observed and the other of 
> >shuffled data.
> >Thanks,
> >Ron
> Ron,
> It seems to me that if your two distributions are not significantly 
> different, you can conclude that they come from the same (theoretical) 
> distribution. The KS test seems appropriate.
> EP
> Emmanuel Paradis

But beware.  Fisher said that failure to achieve significance means you need to acquire more data.  How do you know that the test has adequate power in any given situation, or that a test of a NULL hypothesis is appropriate?
Frank E Harrell Jr              Prof. of Biostatistics & Statistics
Div. of Biostatistics & Epidem. Dept. of Health Evaluation Sciences
U. Virginia School of Medicine  http://hesweb1.med.virginia.edu/biostat
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list