Re: not exactlyRe: [MORPHMET] PREPRINT: between-group principal components analysis (bgPCA)

[Una Vidarsdottir]

Thank you Andrea for clarifying this. You are one of the most honest and
modest people I know, and I am glad that your side of this story is now in
the open. You have my support, as always.
Una

On Wed, 15 May 2019, 06:33 andrea cardini,  wrote:

> I have to correct Fred on this:
> > we accelerated our writing. My paper was the first to be finished,
> > probably because it is a single-authored item by an emeritus with no
> > other obligations,
>
> No, WE did not accelerate the writing. We started a cooperation, after
> my small finding, and we were supposed to work all together on this. At
> some stage, we heard no more from Fred and I suggested to have two
> companion papers, but NEVER got an answer from Fred.
> Months later, Fred let us know he was presenting and discussing results
> (without ever asking me if I was OK with this). Finally, HE decided to
> go on on his own, submit and announce in this list (again letting me
> know after he was done). This is an accurate reconstruction of the
> events. The other one is not and Fred was not unaware that I wasn't OK:
> before the preprint he just announced, he (again without ever asking)
> had already done an informal presubmission to a journal and the journal
> has my written complaint about it.
>
> I let the morphometric community judge if this is the appropriate
> behaviour. Certainly it is not what I teach students, but possibly it is
> what a famous retired emeritus and one of the leader of a scientific
> community can do.
>
> All the best
>
> Andrea
>
> PS
> On a technical side, as I never thought that CVA was the source of all
> evil and BG-PCA a simple solution, here too I agree that the method has
> some problems but I am more than confident that it can still be WISELY
> applied in many cases. That small N (especially when one works with
> small differences) and large p (numbers of variables) are not desirable
> in very many types of analyses is written in all introductory textbook
> on multivariate stats (at least those written in simple non-mathematical
> language for non-numerically skilled people like me).
> In relation to this, there's a point I raised many times for years in
> this list and in some of my papers: one uses the specific landmarks
> required for her/his specific aim (I am in debt to Paul O'Higgins for
> teaching me this). Semilandmarks are a great tool but should be used
> when really needed and bearing in mind that almost inevitably p will
> become big and that might create problems. There are different views on
> this, including that having many points makes beautiful pictures: I
> agree but probably most of the time that is not the aim of a biologist.
> However, there might be cases when even with small N semilandmarks might
> be a huge step forward and possibly the best example I know it's the
> virtual reconstruction of fossils (further analysis of those data may
> then be harder, because of very big p and small N).
> I definitely share the frustration of many taxonomists and
> palaeontologists who have often very precious material and very small
> samples and want to get the most out of them. Regardless of p/N
> problems, estimates of means will be then inevitably inaccurate (and
> sometimes even biased, as the sample could be few and maybe related
> individuals of a rare species). Sometimes those means could be OKish
> (macroevolutionary analyses with very large differences?); most of the
> time they will be as accurate as trying to estimate the average body
> height of Italian men using a sample of 10 men from the same small
> region of Italy. Again, not my discovery: it's all in the introductory
> stats textbook, but I myself too often forget about it.
>
>
>
> --
>
> Dr. Andrea Cardini
> Researcher, Dipartimento di Scienze Chimiche e Geologiche, Università di
> Modena e Reggio Emilia, Via Campi, 103 - 41125 Modena - Italy
> tel. 0039 059 2058472
>
> Adjunct Associate Professor, Centre for Forensic Anthropology, The
> University of Western Australia, 35 Stirling Highway, Crawley WA 6009,
> Australia
>
> E-mail address: alcard...@gmail.com, andrea.card...@unimore.it
> WEBPAGE: https://sites.google.com/site/alcardini/home/main
>
> FREE Yellow BOOK on Geometric Morphometrics:
> https://tinyurl.com/2013-Yellow-Book
>
> ESTIMATE YOUR GLOBAL FOOTPRINT:
> http://www.footprintnetwork.org/en/index.php/GFN/page/calculators/
> SUPPORT: secondwarning.org
>
> --
> MORPHMET may be accessed via its webpage at http://www.morphometrics.org
> ---
> You received this message because you are subscribed to the Google Groups
> "MORPHMET" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to morphmet+unsubscr...@morphometrics.org.
>
>

-- 
MORPHMET may be accessed via its webpage at http://www.morphometrics.org
--- 
You received this message because you are subscribed to the Google Groups 
"MORPHMET" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to morphmet+unsubsc

Re: not exactlyRe: [MORPHMET] PREPRINT: between-group principal components analysis (bgPCA)

[Polly, P. David]

Wow.  Andrea invited me to collaborate with him, Paul O'Higgins, and Jim Rohlf 
on the BG-PCA problem back in July, 2018.  Andrea then graciously invited Fred 
a couple weeks later.  Andrea drafted a manuscript by early September, a long 
time before GMAustria19.  Through November he politely hounded me and Fred 
because we were not keeping up with discussion and comments.  Feeling more a 
hinderance than a help, I withdrew so they could get on with it.  It never 
occurred to me to quickly publish a paper of my own on the same subject
David


P. David Polly
Robert R. Shrock Professor
Earth and Atmospheric Sciences
(with affiliated appointments in Biology and Anthropology)
Indiana University
pdpo...@indiana.edu
https://pollylab.indiana.edu

On sabbatical leave 2018-19
Institute for Biospheric Studies
Yale University





On 15 May 2019, at 2:33 AM, andrea cardini 
mailto:alcard...@gmail.com>> wrote:

I have to correct Fred on this:
we accelerated our writing. My paper was the first to be finished, probably 
because it is a single-authored item by an emeritus with no other obligations,

No, WE did not accelerate the writing. We started a cooperation, after my small 
finding, and we were supposed to work all together on this. At some stage, we 
heard no more from Fred and I suggested to have two companion papers, but NEVER 
got an answer from Fred.
Months later, Fred let us know he was presenting and discussing results 
(without ever asking me if I was OK with this). Finally, HE decided to go on on 
his own, submit and announce in this list (again letting me know after he was 
done). This is an accurate reconstruction of the events. The other one is not 
and Fred was not unaware that I wasn't OK: before the preprint he just 
announced, he (again without ever asking) had already done an informal 
presubmission to a journal and the journal has my written complaint about it.

I let the morphometric community judge if this is the appropriate behaviour. 
Certainly it is not what I teach students, but possibly it is what a famous 
retired emeritus and one of the leader of a scientific community can do.

All the best

Andrea

PS
On a technical side, as I never thought that CVA was the source of all evil and 
BG-PCA a simple solution, here too I agree that the method has some problems 
but I am more than confident that it can still be WISELY applied in many cases. 
That small N (especially when one works with small differences) and large p 
(numbers of variables) are not desirable in very many types of analyses is 
written in all introductory textbook on multivariate stats (at least those 
written in simple non-mathematical language for non-numerically skilled people 
like me).
In relation to this, there's a point I raised many times for years in this list 
and in some of my papers: one uses the specific landmarks required for her/his 
specific aim (I am in debt to Paul O'Higgins for teaching me this). 
Semilandmarks are a great tool but should be used when really needed and 
bearing in mind that almost inevitably p will become big and that might create 
problems. There are different views on this, including that having many points 
makes beautiful pictures: I agree but probably most of the time that is not the 
aim of a biologist. However, there might be cases when even with small N 
semilandmarks might be a huge step forward and possibly the best example I know 
it's the virtual reconstruction of fossils (further analysis of those data may 
then be harder, because of very big p and small N).
I definitely share the frustration of many taxonomists and palaeontologists who 
have often very precious material and very small samples and want to get the 
most out of them. Regardless of p/N problems, estimates of means will be then 
inevitably inaccurate (and sometimes even biased, as the sample could be few 
and maybe related individuals of a rare species). Sometimes those means could 
be OKish (macroevolutionary analyses with very large differences?); most of the 
time they will be as accurate as trying to estimate the average body height of 
Italian men using a sample of 10 men from the same small region of Italy. 
Again, not my discovery: it's all in the introductory stats textbook, but I 
myself too often forget about it.



--

Dr. Andrea Cardini
Researcher, Dipartimento di Scienze Chimiche e Geologiche, Università di Modena 
e Reggio Emilia, Via Campi, 103 - 41125 Modena - Italy
tel. 0039 059 2058472

Adjunct Associate Professor, Centre for Forensic Anthropology, The University 
of Western Australia, 35 Stirling Highway, Crawley WA 6009, Australia

E-mail address: alcard...@gmail.com, 
andrea.card...@unimore.it
WEBPAGE: https://sites.google.com/site/alcardini/home/main

FREE Yellow BOOK on Geometric Morphometrics: 
https://tinyurl.com/2013-Yellow-Book

ESTIMATE YOUR GLOBAL FOOTPRINT: 
http://www.footprintnetwork.org/en/

not exactlyRe: [MORPHMET] PREPRINT: between-group principal components analysis (bgPCA)

[andrea cardini]

I have to correct Fred on this:
we accelerated our writing. My paper was the first to be finished, 
probably because it is a single-authored item by an emeritus with no 
other obligations,


No, WE did not accelerate the writing. We started a cooperation, after 
my small finding, and we were supposed to work all together on this. At 
some stage, we heard no more from Fred and I suggested to have two 
companion papers, but NEVER got an answer from Fred.
Months later, Fred let us know he was presenting and discussing results 
(without ever asking me if I was OK with this). Finally, HE decided to 
go on on his own, submit and announce in this list (again letting me 
know after he was done). This is an accurate reconstruction of the 
events. The other one is not and Fred was not unaware that I wasn't OK: 
before the preprint he just announced, he (again without ever asking) 
had already done an informal presubmission to a journal and the journal 
has my written complaint about it.


I let the morphometric community judge if this is the appropriate 
behaviour. Certainly it is not what I teach students, but possibly it is 
what a famous retired emeritus and one of the leader of a scientific 
community can do.


All the best

Andrea

PS
On a technical side, as I never thought that CVA was the source of all 
evil and BG-PCA a simple solution, here too I agree that the method has 
some problems but I am more than confident that it can still be WISELY 
applied in many cases. That small N (especially when one works with 
small differences) and large p (numbers of variables) are not desirable 
in very many types of analyses is written in all introductory textbook 
on multivariate stats (at least those written in simple non-mathematical 
language for non-numerically skilled people like me).
In relation to this, there's a point I raised many times for years in 
this list and in some of my papers: one uses the specific landmarks 
required for her/his specific aim (I am in debt to Paul O'Higgins for 
teaching me this). Semilandmarks are a great tool but should be used 
when really needed and bearing in mind that almost inevitably p will 
become big and that might create problems. There are different views on 
this, including that having many points makes beautiful pictures: I 
agree but probably most of the time that is not the aim of a biologist. 
However, there might be cases when even with small N semilandmarks might 
be a huge step forward and possibly the best example I know it's the 
virtual reconstruction of fossils (further analysis of those data may 
then be harder, because of very big p and small N).
I definitely share the frustration of many taxonomists and 
palaeontologists who have often very precious material and very small 
samples and want to get the most out of them. Regardless of p/N 
problems, estimates of means will be then inevitably inaccurate (and 
sometimes even biased, as the sample could be few and maybe related 
individuals of a rare species). Sometimes those means could be OKish 
(macroevolutionary analyses with very large differences?); most of the 
time they will be as accurate as trying to estimate the average body 
height of Italian men using a sample of 10 men from the same small 
region of Italy. Again, not my discovery: it's all in the introductory 
stats textbook, but I myself too often forget about it.




--

Dr. Andrea Cardini
Researcher, Dipartimento di Scienze Chimiche e Geologiche, Università di 
Modena e Reggio Emilia, Via Campi, 103 - 41125 Modena - Italy

tel. 0039 059 2058472

Adjunct Associate Professor, Centre for Forensic Anthropology, The 
University of Western Australia, 35 Stirling Highway, Crawley WA 6009, 
Australia


E-mail address: alcard...@gmail.com, andrea.card...@unimore.it
WEBPAGE: https://sites.google.com/site/alcardini/home/main

FREE Yellow BOOK on Geometric Morphometrics: 
https://tinyurl.com/2013-Yellow-Book


ESTIMATE YOUR GLOBAL FOOTPRINT: 
http://www.footprintnetwork.org/en/index.php/GFN/page/calculators/

SUPPORT: secondwarning.org

--
MORPHMET may be accessed via its webpage at http://www.morphometrics.org
--- 
You received this message because you are subscribed to the Google Groups "MORPHMET" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to morphmet+unsubscr...@morphometrics.org.

Re: [MORPHMET] PREPRINT: between-group principal components analysis (bgPCA)

[Ian Dworkin]

Fred

 I look forward to reading it. I am particularly interested since it was
probably your 2011 paper with Phillip (Mitteroecker and Bookstein, Evol
Biol 38:100-114) which advocated its use for ordination and has resulted in
its current widespread use, but which you address in an amusing way in this
current paper in figure 10 and with the term "less inappropriate" than CVA.

Ian

On Tue, 14 May 2019 at 13:32, MORPHMET  wrote:

> Dear MorphMetters,
>
> Some of you may have been in the auditorium in the Department of Botany,
> University of Vienna, back in March when Philipp Mitteroecker and I were
> the two scheduled discussants for the conference "GMAustria19" on
> applications of geometric morphometrics.  Several of the papers delivered
> there used between-group principal components analysis (bgPCA), and after
> each of those papers I mentioned in the course of my commentary that bgPCA
> was fatally flawed in applications to most GMM data sets and should NEVER
> be used here. In my keynote address, which closed the meeting, I had one
> cryptic slide about this assertion, with an example that flashed on the
> screen but was immediately replaced by the next slide.
>
> The typical response to both my own talk and my criticism of the talks of
> others, as far as bgPCA was concerned, was along the lines of "Hunh?" or
> sometimes "What are you blathering about this time? Isn't bgPCA in the
> standard toolkit?" I answered that the Bookstein paper they should read was
> just then being written, as one of a pair jointly arising from
> conversations with Andrea Cardini, Jim Rohlf, and Paul O'Higgins following
> an original hunch of Cardini's, and that my argument would be pretty
> convincing once it was actually written down.  The claim isn't that people
> are using bgPCA incorrectly. They're using it according to the published
> formulas, yes, but the method itself yields biological nonsense much too
> often.
>
> That was March.  In April, two different articles in Nature (one by
> Detroit et al., one by Chen et al.) buttressed claims about sister species
> of Homo sapiens using the bgPCA method, and so suddenly it became clear
> that we authors had to do something quickly lest this become an epidemic of
> bad biometrics. So we accelerated our writing. My paper was the first to be
> finished, probably because it is a single-authored item by an emeritus with
> no other obligations, and it seemed like a good idea to upload the final
> draft to https://www.biorxiv.org even before submitting the paper, so
> that any letter to the editors of Nature could include a link to  the
> argument as to exactly WHY bgPCA is nearly always unsound and its
> inferences invalid for applications in contemporary GMM.
>
> That is the draft that has just appeared as
>
> https://www.biorxiv.org/content/10.1101/627448v1
>
> For those of you who were at the March meeting, this is the argument
> (complete with formulas) defending my stern condemnation there. I won't try
> to summarize it in this morphmet note -- if you're interested, just read
> the abstract on page 1 of the link.  For those of you who have already
> published bgPCA analyses, you know who you are -- my paper argues strongly
> that you need to go back and revisit the inferences of those papers in a
> mood of much more intense multivariate skepticism.  For the rest of you,
> please consider this draft manuscript to be a wake-up call. A technique
> that has appeared in dozens of papers and that was, alas, specifically
> praised by Mitteroecker and Bookstein personally (back in 2011) could
> nevertheless, when examined closely (for the first time!), turn out to be
> algebraic garbage when applied to data sets where there are far more shape
> coordinates than specimens. But isn't that the usual situation in GMM these
> days?
>
> As always, I welcome all responses, both positive and negative. The
> biorxiv posting is permanent, but there is plenty of time for me to make
> changes before the paper is published (at present it has not yet even been
> submitted anywhere), so feel free to try to find the flaws in my argument.
> But I hope you will want to try some of these simulations on your own
> before you argue against mine. You will also want to study the companion
> piece by Cardini, O'Higgins, and Rohlf that should likewise be available
> for download before too long.
>
> Fred Bookstein
>
> --
> MORPHMET may be accessed via its webpage at http://www.morphometrics.org
> ---
> You received this message because you are subscribed to the Google Groups
> "MORPHMET" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to morphmet+unsubscr...@morphometrics.org.
>


-- 
Ian Dworkin
Department of Biology
McMaster University
Office phone 905 525 9140 ext. 21775
Lab phone 905 525 9140 ext. 20076
dwor...@mcmaster.ca

-- 
MORPHMET may be accessed via its webpage at http://www.morphometrics.org
--- 
You received this message because you are subscribed

[MORPHMET] PREPRINT: between-group principal components analysis (bgPCA)

[MORPHMET]

Dear MorphMetters,

Some of you may have been in the auditorium in the Department of Botany, 
University of Vienna, back in March when Philipp Mitteroecker and I were 
the two scheduled discussants for the conference "GMAustria19" on 
applications of geometric morphometrics.  Several of the papers delivered 
there used between-group principal components analysis (bgPCA), and after 
each of those papers I mentioned in the course of my commentary that bgPCA 
was fatally flawed in applications to most GMM data sets and should NEVER 
be used here. In my keynote address, which closed the meeting, I had one 
cryptic slide about this assertion, with an example that flashed on the 
screen but was immediately replaced by the next slide.
  
The typical response to both my own talk and my criticism of the talks of 
others, as far as bgPCA was concerned, was along the lines of "Hunh?" or 
sometimes "What are you blathering about this time? Isn't bgPCA in the 
standard toolkit?" I answered that the Bookstein paper they should read was 
just then being written, as one of a pair jointly arising from 
conversations with Andrea Cardini, Jim Rohlf, and Paul O'Higgins following 
an original hunch of Cardini's, and that my argument would be pretty 
convincing once it was actually written down.  The claim isn't that people 
are using bgPCA incorrectly. They're using it according to the published 
formulas, yes, but the method itself yields biological nonsense much too 
often.
 
That was March.  In April, two different articles in Nature (one by Detroit 
et al., one by Chen et al.) buttressed claims about sister species of Homo 
sapiens using the bgPCA method, and so suddenly it became clear that we 
authors had to do something quickly lest this become an epidemic of bad 
biometrics. So we accelerated our writing. My paper was the first to be 
finished, probably because it is a single-authored item by an emeritus with 
no other obligations, and it seemed like a good idea to upload the final 
draft to https://www.biorxiv.org even before submitting the paper, so that 
any letter to the editors of Nature could include a link to  the argument 
as to exactly WHY bgPCA is nearly always unsound and its inferences invalid 
for applications in contemporary GMM. 
 
That is the draft that has just appeared as 

https://www.biorxiv.org/content/10.1101/627448v1

For those of you who were at the March meeting, this is the argument 
(complete with formulas) defending my stern condemnation there. I won't try 
to summarize it in this morphmet note -- if you're interested, just read 
the abstract on page 1 of the link.  For those of you who have already 
published bgPCA analyses, you know who you are -- my paper argues strongly 
that you need to go back and revisit the inferences of those papers in a 
mood of much more intense multivariate skepticism.  For the rest of you, 
please consider this draft manuscript to be a wake-up call. A technique 
that has appeared in dozens of papers and that was, alas, specifically 
praised by Mitteroecker and Bookstein personally (back in 2011) could 
nevertheless, when examined closely (for the first time!), turn out to be 
algebraic garbage when applied to data sets where there are far more shape 
coordinates than specimens. But isn't that the usual situation in GMM these 
days?  
 
As always, I welcome all responses, both positive and negative. The biorxiv 
posting is permanent, but there is plenty of time for me to make changes 
before the paper is published (at present it has not yet even been 
submitted anywhere), so feel free to try to find the flaws in my argument.  
But I hope you will want to try some of these simulations on your own 
before you argue against mine. You will also want to study the companion 
piece by Cardini, O'Higgins, and Rohlf that should likewise be available 
for download before too long.
 
Fred Bookstein

-- 
MORPHMET may be accessed via its webpage at http://www.morphometrics.org
--- 
You received this message because you are subscribed to the Google Groups 
"MORPHMET" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to morphmet+unsubscr...@morphometrics.org.