Bias in Stats Packages ?

[Warren Kovach]

> > Systat's 1500.565.  The resulting f-ratio was 364.1975 vs 364.300.  I 
> > think I would tend to believe Systat's results more :-)
> > 
> > Dr. Warren L. Kovach         Internet: warrenk at cix.compulink.co.uk
> > 
> Good point, but why?  Do they both give the method of calculation, and
> which method do you consider more accurate?  I always have this same
> concern
> with letting a computer do my dirty work, but prefer to base my bias on
> knowledge of the methods used rather than familiarity with the program.
> If you can share information about the method of calculation and your 
> reason for preference, that would be appreciated.  Thanks.
> .
> Leonard N. Bloksberg
> bloksber at pilot.msu.edu

The Systat manual states that the sums of squares are calculated in 
double precision using Spicer's one-pass provisional algorithm.  The 
Excel manuel says exactly nothing about the methods of calculation.  Like 
you, I prefer the known to the unknown.

--
Dr. Warren L. Kovach              Internet: warrenk at cix.compulink.co.uk
Kovach Computing Services                       CompuServe: 100016,2265
85 Nant-y-Felin                                tel./fax: +44-248-450414
Pentraeth, Anglesey, Wales LL75 8UY U.K.