PSPP-BUG: [bug #61510] Defining categories as 1, 0 instead of 0, 1 in Kruskal, shows erroneous results

[elias tsolis]

URL:
  <https://savannah.gnu.org/bugs/?61510>

                 Summary: Defining categories as 1, 0 instead of 0, 1 in
Kruskal, shows erroneous results
                 Project: PSPP
            Submitted by: estatistics
            Submitted on: Fri 19 Nov 2021 08:43:07 PM UTC
                Category: Numerical Errors
                Severity: 5 - Average
                  Status: None
             Assigned to: None
             Open/Closed: Open
                 Release: None
         Discussion Lock: Any
                  Effort: 0.00

    _______________________________________________________

Details:

Defining categories as 1, 0 instead of 0, 1 in Kruskal, shows erroneous
results

eg.

NPAR TEST /KRUSKAL-WALLIS = followup BY Death (0, 1) .   
  --> (erroneous results)
NPAR TEST /KRUSKAL-WALLIS = followup BY Death (1, 0) .
  --> (right results)



What have not tested: 
@ If this true to other tests
@ if this can happen to other coding schemes eg. 1 & 2 / 2 & 1 etc.



    _______________________________________________________

File Attachments:


-------------------------------------------------------
Date: Fri 19 Nov 2021 08:43:07 PM UTC  Name: example.sav  Size: 808B   By:
estatistics
coding scheme results error results
<http://savannah.gnu.org/bugs/download.php?file_id=52298>

    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?61510>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/