>Well, looks like they're lining up!! You've got a discussion going, Lachlan!
Yes, this answers Lachlan's proposal for replacing the Rietveld mailing list :-)
>True to your request, Alan, I will lead you through it step by step off-line
Thanks :-) I asked that you do the refinement both ways and demonstrate that:
>>you get the same estimates of the parameters and standard deviations.
You reply: >as a happy
>coincidence, the combined use of Pawley and Rietveld methods gives
>the sort of esds increases that Dave Cox majors on and we all know about.
OK, now you are saying that they DON'T give the the same results after all !
We agree that the extracted integrated intensity method gives larger esds,
but you think this is good.
I hope Dave will forgive me for interpreting what he, and others say about standard deviations. That is that the calculated Rietveld standard deviations
are correct - PROVIDED that the model is capable of completely describing the
data (which of course is never really true :-). The esds calculated from
statistics are always underestimated due to systematic errors in the data and
model (but perhaps not by much if you look at Rod Hill's intercomparison). And
you can't really address the question of systematic errors by simply changing
the statistics or refinement method - you have to improve the model/data.
Of course you know that, and that is why you say "as a happy co-incidence".
Alan Hewat, ILL Grenoble, FRANCE <hewat@ill.fr> tel (33) 4.76.20.72.13
ftp://ftp.ill.fr/pub/dif fax (33) 4.76.48.39.06 http://www.ill.fr/dif/