Dear Alan,
exactly, knowledge gaps and laziness do often cause blind trust in any
black box or statistical methods.
And, as AI is trained by open internet resources, the older bad
science rubbish will be probably perpetuated by AI. Bad perspectives
for science.
Terminology: We all know that automation in phase analysis
(qualitative and quantitative) has a tradition over decades, also in
applied mineralogy, and especially with XRPD methods. From this point
of view, the use of the term "Automated mineralogy" for any SEM/EDX
based methods sounds quite ignorant and arrogant to me.
Greetings
Reinhard
Zitat von Alan W Hewat :
Dear Rienhard
Yes everything is AI now, and people often prefer that to thinking for
themselves. I have nothing against automation as a human aid, but there is
a temptation with "black box" software to just click the "Refine
Everything" button and copy-paste the results. There is a problem when AI
is also responsible for reviewing and publishing this "science" :-)
Alan.
On Wed, 17 Jan 2024 at 09:37, Reinhard Kleeberg <
kleeb...@mineral.tu-freiberg.de> wrote:
Dear Alan,
I agree completely, especially with the evaluation of the "quality" of
the paper triggering the discussion.
Regarding misleading and capturing terminology in analytical methods,
I have an IMHO more serious, horrible example:
"Automated mineralogy"
https://en.wikipedia.org/wiki/Automated_mineralogy
what is very popular in a big community of geologists and aggressively
advertised, despite of nether any chemometric evaluation by
interlabory tests was published, phase abundances are given to second
to forth decimal digit of mass% and analytical errors resp. esd's are
typically not published. But even here we probably have already lost
the battle for precise terminology and quality of analytics, as people
can present such data and terminology in several journals of applied
geology and mineral processing without any problem.
Reinhard
Zitat von Alan W Hewat :
> Yes Stefan, it is a question of what you are refining and constraining.
> Before Rietveld it was of course known that the positions of the peaks
were
> determined by the unit cell. There were techniques like Search-Match that
> attempted to identify materials by the positions of the peaks. The
history
> of phase analysis is older than Rietveld. The Petten group were not
> interested in QPA, but instead in refining the crystal and magnetic
> structures of specific materials.
>
> If Rietveld had never existed, it is reasonable to assume that QPA would
> still have evolved much as it has today, refining the phase content from
> the whole profile, without trying to refine structures of mixed phases.
>
> And for those who worry about pedantry, "Rietveld Refinement" is just
> shorthand for the "Rietveld (method of) Refinement. It involves the
> refinement of the crystal structure, not Rietveld :-)
>
> Finally Kurt, perhaps as individuals we can't change the world, but if we
> are concerned that people are debasing our scientific techniques, we can
> expose them as Armel did. Or at least we can refuse to co-author such
> papers, and reject them as journal editors. As authors and referees we
can
> correct sloppy use of terminology. This paper is wrong on so many levels,
> apart from the apparent "misconduct" and the incorrect label of Rietveld
> Refinement.
>
> Dr Alan Hewat, NeutronOptics
> Grenoble, FRANCE (from phone)
> alan.he...@neutronoptics.com
> +33.476984168 VAT:FR79499450856
> http://NeutronOptics.com/hewat
> ___
>
>
> On Tue, 16 Jan 2024, 15:50 Stefan Seidlmayer,
wrote:
>
>> Dear all,
>>
>> I was following the discussion also with great interest, as terminology
is
>> important to distinguish properly between different items.
>>
>> To my understanding the Rietveld approach was new because it constrained
>> the fitting of a peak list generated "from a structure" with the
refinement
>> of the profile of the peaks themselves.
>>
>> Thus I would have the impression that everytime when we use a
>> constrainement of peak list which is generated from a structure and do
not
>> refine a list of "individual peaks" it is a Rietveld-type refinement.
>>
>> A Profile Refinement is/was in my current understanding, when the
>> peak/reflection position is not constrained by the structure parameters,
>> but can be refined in an arbitrary way, individual for each
peak/reflection.
>> From the positions refined in this way, one could then determine cell
>> parameter etc. But this would require a secondary step. First refine all
>> found peaks/reflection with a common profile. Then determine the lattice
>> parameters from the refined peak positions etc.
>> This is also troublesome as without prior structure "knowledge" it may
and
>> surely is that certain reflection which in fact are overlapping multiple
>> reflections are improperly identified as "one" reflection.
>> This is very the neatness of the
