Richard Perry wrote:
> ...This seems to me to be a bit unfair. SPM has to make some assumptions to get
> any statistics out at all.... I think that we can be reasonably confident that in
> developing SPM99, the authors haven't deliberately introduced an error that wasn't
> there in SPM96!...
I apologize to all (especially the SPM authors!) if my comments came across as
hostile or ad hominem. This was not intended! I have used SPM and SnPM myself, and I
have been impressed at the collegiality and remarkable efforts put forth by its
authors in making it available to the public. The idea of deliberately introducing
errors never crossed my mind and I hope it was not read that way by anyone else.
I do have genuine questions about the meaning of the p values. I am seeking expert
opinion on "fair" reporting of my own results and on how to approach papers I may
review if they used SPM96 or 99. I think it is fair to say that many publications in
fact used SPM96 or earlier versions to report on studies with less than 40-120
degrees of freedom, so I think the questions are germane. I have read the material on
the web site and a number of the original papers, plus Dr. Perry's comments, without
quite understanding these points. I do hope the authors could comment on my original
post?
Thanks,
Kevin Black
|