Re: Error in computation of Mann-Whitney-U Test

[John Darrington]

On Sat, Mar 04, 2017 at 11:14:33PM +0100, Dr. Walter Statistics wrote:
     I cannot confirm your objection, John. I checked it again, now with a
     published data set of 3 - three - different groups. I ran three Mann
     Whitney U tests (group 1 vs 2 excluding the cases of group 3, group 1 vs
     3 excluding the cases of group 2, group 2 vs 3 excluding the cases of
     group 1). There were no differences. PSPP seems to be working properly. I
     did not find anything of importance. There seems to be no problem at all.


Thanks Walter.  In that case perhaps it is indeed a different problem.
     
     Prof Nuszbaum has to be more specific in describing her problem.
     

Yes.  Perhaps she could give us an example including:

1. The dataset used.
2. The syntax which provokes the problem.
3. The exact version number of PSPP which is problematic.
4. The output produced - and any contradictory results from elsewhere.


J'

     

-- 
Avoid eavesdropping.  Send strong encrypted email.
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://sks-keyservers.net or any PGP keyserver for public key.

signature.asc
Description: Digital signature