[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

normality test

From: Toke Reichstein
Subject: normality test
Date: Thu, 20 Feb 2003 00:05:26 +0100 (MET)


A while ago I wrote and asked about combined plotting of histograms and
normal distributions (thank you for you help by the way). I was so
pleased with using octave that I was wondering if I should carry
out formal normality test in octave as well instead of using my normal
proprietary statistical package. But I just want to be certain that what I
am doing is right.

I searched the help file for a normality test. And I came by the
kolmogorov-smirnov test. The help file gives an example with a uniform
distribution. If I was to conduct the test using a normal distribution
would it be right to do it like this?

kolmogorov_smirnov_test([data], 'normal', [mean of data], [variance of

I have compared the results with my normal statistical package. But the
results are not the same. I am inclined to believe the octave
calculations. Because I have also conducted the tests with a earlier
version of the proprietary software that employed a shapiro-wilks test.
The conclusion of this test seem to concur with octave. So maybe I
cannot trust my latest proprietary software. It may look up the wrong
critical values in the new version in which it uses
the kolmogorov-smirnov test. Also the plotting of the data tends to point
toward the same conclusion as that of octave.

Thank you in advance for any comments.


Toke Reichstein                             Aalborg University
Cand. Oecon. (M.Sc. Economics)              Department of Business Studies
Ph.D. Student                               Fibigerstraede 4
e-mail  : address@hidden                    9220 Aalborg Oe
Homepage:   Denmark
Fax: +45 98156013                           Phone   : +45 96358241

Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:
How to fund new projects:
Subscription information:

reply via email to

[Prev in Thread] Current Thread [Next in Thread]