Dear Douglas,

thanks for the reply and the new bug descriptions regarding t-test groups and 
the regression command. 

I opened a bug in the bug database regarding the regression syntax:

http://savannah.gnu.org/bugs/?45667

The spss v23 syntax description 

ftp://public.dhe.ibm.com/software/analytics/spss/documentation/statistics/23.0/en/client/Manuals/IBM_SPSS_Statistics_Command_Syntax_Reference.pdf

says that the groups command should separate the values with comma, but I guess 
you produced the syntax via the spss GUI, right? 

Regards

Friedrich

Am 31.07.2015 um 07:04 schrieb Douglas Bonett <[email protected]>:

> I am using version 22 of SPSS and it does not complain if I use CIN.    
> 
> I also noticed that SPSS V22 uses slightly different syntax for an 
> independent samples t-test. SPSS produces:
> 
> T-TEST GROUPS=group(1 2)
>   /MISSING=ANALYSIS
>   /VARIABLES=score
>   /CRITERIA=CI(.95).
> 
> which PSPP does not like.  PSPP wants group(1,2) but SPSS does not complain 
> with group(1,2).
> 
> There seems to be more serious discrepancies between SPSS and PSPP syntax for 
> linear regression. For a simple linear model SPSS produces
> 
> 
> REGRESSION
>   /MISSING LISTWISE
>   /STATISTICS COEFF OUTS CI(95) R ANOVA
>   /CRITERIA=PIN(.05) POUT(.10)
>   /NOORIGIN
>   /DEPENDENT y
>   /METHOD=ENTER x.
> 
> while PSSS gives
> 
> REGRESSION
>     /VARIABLES= x
>     /DEPENDENT=     y
>     /STATISTICS=COEFF CI R ANOVA.
> 
> SPSS V22 does not accept the PSPP regression syntax. 
> 
> 
> 
> 
> On Thu, Jul 30, 2015 at 9:02 PM, Friedrich Beckmann 
> <[email protected]> wrote:
> Dear Douglas,
> 
> thank you for the bug report. I filed a bug in the bug database of pspp:
> 
> https://savannah.gnu.org/bugs/index.php?45664
> 
> As mentioned in the bug report the spss manual uses CIN in the examples. Does 
> your spss version complain if you use CIN instead of CI?
> 
> Friedrich
> 
> Am 30.07.2015 um 22:14 schrieb Douglas Bonett <[email protected]>:
> 
>>  PSPP uses CIN(.95) rather than CI(.95).  I hope this is a bug and not 
>> intentional. 
> 
> 
> 
> 
> 
> 
> 

--
Friedrich Beckmann, Bismarckstraße 11, 86159 Augsburg
Tel: 0821-44923693, Mobil: 017648576782




_______________________________________________
pspp-dev mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/pspp-dev

Reply via email to