Re: st: ksmirnov and normality

[Maarten Buis <maartenlbuis@gmail.com>]

On Fri, Feb 8, 2013 at 12:42 PM, Nahla Betelmal wrote:
> I found a paper says that Kolmogorov-Smirnov test is used for
> normality test in SPSS and SAS, but not in STATA. In STATA ksmirnov
> command is used for the one or two samples Kolmogorov-Smirnov test.
> the paper http://faculty.ksu.edu.sa/hisham/Documents/Students/a_PHCL/Testing_Normality.pdf,
> page 2
>
> I have two questions, does this mean that we can not use ksmirnov
> command to test normality for one sample in STATA?
> Second, what is the meaning behind STATA ksmirnov command is used for
> the one sample Kolmogorov-Smirnov test, other than normality test?
> what is the point the author wants to make?

Technically speaking, the Kolmogorov-Smirnov test can be used to test
for normality in Stata, there is an example in the helpfile of
-ksmirnov- on how to do this. However, that helpfile also, quite
rightly, warns you not to do this. The properties of this test are
absolutely dismall, regardless of the statistical package used to
compute, so you should never even think about doing it. Here is a
simulation that illustrates this point:

<http://www.stata.com/statalist/archive/2013-01/msg00704.html>

-- Maarten


---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

http://www.maartenbuis.nl
---------------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/