Re: PSPP-BUG: Different results in PSPP and SPSS

[John Darrington]

Thanks for the report.

Can you send a copy of the syntax (including the dataset) which you produces 
the differing results.
And can you also send a copy of the results which R and SPSS produces.

(If it's just the sign that is difference to what you expect, then I guess its 
merely a question of
the groups being transposed so I don't think its a question of being "right" or 
"wrong".  However
I would like PSPP to be consistent with other software in any case.)

Thank you

J'

On Fri, Mar 11, 2022 at 06:46:47PM +0000, Maricel Knechtel wrote:
     Hi,
     I 'am considering to switch over to PSPP, but my first test did not go 
well.
     When running a logistic analysis with the same data I got different 
results in SPSS and PSPP. I know the SPSS results are right because I got the 
same results in R)
     
     Looking at the results I found following.
     The negative B, became a positive B in PSPP.
     E.g. Right result in SPSS  for B =  - . 84 , in PSPP =.84
     This led to exp(b) being completely wrong in pspp
     
     Looking further into the output viewer, I found that what is coded as 1 in 
my file, was recoded to 0 (for the independent variables) and viceversa.
     
     The frequencies and -2 likehood and Nagelkerke R square are otherwise 
right.
     
     Is there anything I am doing wrong?
     
     
     Kindly
     Maricel
     

     _______________________________________________
     Bug-gnu-pspp mailing list
     Bug-gnu-pspp@gnu.org
     https://lists.gnu.org/mailman/listinfo/bug-gnu-pspp