| Discussion: | All in Probability & Statistics |
| Topic: | Non-parametric statistics |
| Related Item: | http://mathforum.org/mathtools/tool/105/ |
| Post a new topic to the All in Probability & Statistics discussion |
| ||||||||
| Subject: | RE: Non-parametric statistics |
| Author: | Swizerland |
| Date: | Mar 22 2004 |
 |
|
In fact, my problem is to know if two samples are comparable. As far as I am
concern, I work on the meta-analyses on individual data. As I don't have
randomization lists, I only have the lists of the treatments assigned to the
patients and I dont't have any information on the manner which the randomization
list is done.
For that, I decided to assimilate the lists that I have with distributions of
pseudo-random numbers. And I propose to apply tests like the autocorrélation
test, the runs test, the poker test, the frequency test and the gap test (tests
of uniformity and independance).
Untill now, I used the R language for the kolmogorov-smirnov test.
For that, I'll use the SAS System.
that seems very interesting what you do with java !
If that does not disturb you, can you send to me a version of the test of
kolmogorov-smirnov made with java?
Think you.
Swizerland.
| |||||||
| Post a new topic to the All in Probability & Statistics discussion | |||||||
| Visit related
discussions: Measures tool Probability & Statistics | |||||||