Adventures in social network analysis: approaching the finale

by John Q on May 17, 2011

A few years back as part of the attack on climate science (and in particular the famous ‘hockey stick’ graph) Congressman Joe Barton (R-TX) commissioned an assessment of the work of Michael Mann and others from Professor Edward Wegman of George Mason University, along with his former student Yasmin Said and some others. This included, not only Wegman’s supposedly independent assessment of the statistical methods used by Mann but a ‘social network analysis’ of the relationship between Mann and his co-authors, which purportedly showed that Mann’s network of co-authors dominated the climate science field. As I pointed out at the time, Wegman et al started the analysis with Mann at the centre, so the primary result was that Mann had written a paper with every one of his co-authors! Nevertheless, a version of the paper was published in Computational Statistics and Data Analysis, in which Wegman took this analysis to the startling conclusion that senior academics should not collaborate with each other, but should instead work only with their students. Wegman follows his own advice in this respect, and now we can see why.

It’s just been announced that the paper is to be retracted on the grounds that it contains extensive plagiarism, much but not all of it from Wikipedia. Wegman’s response, showing the wisdom of his research strategy, is to blame his graduate student, who was not, however credited as an author. USA Today, which has taken the lead in following the Wegman plagiarism story, asked an actual expert to look at the paper and her reaction was about the same as my amateur assessment (Wegman and Said are also newcomers to the field, which may explain their heavy reliance on Wikipedia as a reference source).

This kind of trouble seems to follow Wegman around, and to be contagious. Among those affected:

* George Mason University, which received a formal complaint about plagiarism over a year ago, and has yet to take any action. Amazingly, Wegman (or one of this graduate students) copied, without acknowledgement, the work of Raymond Bradley, one of the scientists he was attacking. As in the separate social networking case, USA Today asked experts about the case, and were told that this was an open-and-shut case of plagiarism.

* The editor of Computational Statistics and Data Analysis, Stanley Azen, who accepted the Wegman et al article a few days after its submission. Azen now says he would never do such a thing without having earlier sent the paper out for peer review. Sadly, all records of the review were lost in an office move. And, in accepting the article by email, Azen forgot to mention the rigorous review process, instead saying, in emails obtained under FOI by USA today “I personally reviewed your very interesting (and unique) manuscript.” In separate news, Azen’s pet dog has finally confessed to eating his homework in ninth grade.

* The journal WIREs Computational Statistics where Wegman and Said, along with co-author David Scott, were appointed as editors following their innovative work on climate science. There, Wegman and Said published an article on color design and theory, which, wouldn’t you know it, turned out to contain large slabs lifted from Wikipedia and other sources. The graduate student presumably responsible didn’t just cut and paste – there was extensive paraphrasing which often changed the meaning. The redoubtable blogger “Deep Climate” who’s done much of the work in documenting the extensive plagiarism that has plagued Wegman’s work has made some color innovations in the process, using color highlighting to document the various modes of unattributed borrowing that make up a typical Wegman piece. I should also mention John Mashey’s extensive work which has greatly helped our understanding of how delusions are manufactured and sold.

{ 47 comments }

1

roger 05.17.11 at 9:39 am

This is funny. Ah, that George Mason university.

2

JGabriel 05.17.11 at 12:03 pm

John Quiggan:

Wegman’s response, showing the wisdom of his research strategy, is to blame his graduate student, who was not, however credited as an author.

Hog all the credit, then blame the collaborator for the plagiarism?

Megan McArdle, this is why academics don’t want to work with conservatives.

3

EWI 05.17.11 at 12:50 pm

in which Wegman took this analysis to the startling conclusion that senior academics should not collaborate with each other, but should instead work only with their students. Wegman follows his own advice in this respect, and now we can see why.

I’ve seen this referred to in several places as the shining example of the “self-refuting paper”. I propose that this rare beast should be named after its leading exponent.

4

Kieran 05.17.11 at 1:11 pm

What a joke.

5

Sherman Dorn 05.17.11 at 1:17 pm

So George Mason has both Wegman and the George Mason Branch of the Koch Economics Departmentses to brag about? Unfortunately, this stellar reputation is being unjustly sullied by other entities such as the Roy Rosenzweig Center for History and New Media, which does awful things like digital humanities scholarship and infrastructure.

6

Bob 05.17.11 at 1:54 pm

Joe Barton is not a Senator. He is a member of the House.

Might want to correct that.

7

Satan Mayo 05.17.11 at 1:55 pm

This’ll teach Barton to stick to commissioning content-free studies from accountability-free “think tanks” instead of actual universities from now on.

8

SFAW 05.17.11 at 2:01 pm

This’ll teach Barton to stick to commissioning content-free studies from accountability-free “think tanks” instead of actual universities from now on.

That’ll happen right after Pantload announces he’s going to do all his own research, from now on.

9

Francis 05.17.11 at 2:23 pm

I hope that the next GMU student accused of plagiarism responds by requesting the Wegman treatment.

10

Absalon 05.17.11 at 2:23 pm

Climate skeptics and anti-carbon tax libertarians, being the vanguard against politicized, sectarian, sloppy and dubious climate science, will of course be all over this.

Aaaaany minute now.

Delingpole? Monckton? Watts? Cato institute? Helloooooo?

11

Steve LaBonne 05.17.11 at 2:23 pm

This’ll teach Barton to stick to commissioning content-free studies from accountability-free “think tanks” instead of actual universities from now on.

This assumes a fact not in evidence (that George Mason is a real university).

12

kth 05.17.11 at 2:24 pm

It’s the same Joe Barton (R-Waco, TX) who thought he stumped Energy Secretary/Nobel laureate Steven Chu with “if global warming is manmade, then how did all that oil end up in Alaska?” An utter buffoon, even by the standards of the region.

13

Bruce Baugh 05.17.11 at 3:17 pm

So this is a career dedicated to acting out the Buggles’ “Adventures in Modern Recording”.

14

Sev 05.17.11 at 4:19 pm

#3 the “self-refuting paper”

It’s a performance piece! A new conceptual artist is discovered!

15

psycholinguist 05.17.11 at 4:41 pm

For the few nonacademics that read this blog, plagiarism in the context of a supposedly peer reviewed journal submission is a really big deal – to the point of losing ones job, being completely cut out by the offended discipline – basically one’s “real” academic career should be over.
And, it would be very nice to know if there was any Fed money involved in any of this clown’s research. There is ample precedent that if you cheat when you disseminate your results, the fed is gonna come calling to get that money back, and they really mean it. We had a case here a number of years ago that was fairly innocuous compared to this one – a grad student was “loaned” some data his prof had collected, added it to the dissertation and defended – done deal. A disgruntled grad student turned them in, resulting in loss of degree, payback of funds, etc. Does this guy still have a job?

16

salazar 05.17.11 at 5:18 pm

How did Wegman win an academic appointment in the first place (even if he’s a newcomer to the field of climate change)?

17

Josh G. 05.17.11 at 6:05 pm

salazar @ 12: He’s most likely on wingnut welfare. A significant portion of the professorships at GMU are handed out not by academics, but by Charles and David Koch:
http://www.americablog.com/2011/05/it-isnt-just-florida-koch-brothers.html

We really need to start yanking the accreditation of universities that prostitute their hiring processes to wealthy outsiders.

18

EthylEster 05.17.11 at 6:14 pm

Someone asks: How did Wegman win an academic appointment in the first place?

As a former GMU faculty member, I must remind you that GMU is Virginia’s attempt to reproduce the stellar academic achievement of Liberty University in a public institution.

I’m sure there are intelligent and effective scholars at GMU but they must feel embarrassed by some of their colleagues.

19

Colin Danby 05.17.11 at 6:48 pm

There are lots of excellent people at GMU, though possibly more than the normal complement of loons.

But do I have this right … an article that alleges “peer review abuse” was accepted for publication six days after its submission?

20

Bloix 05.17.11 at 7:14 pm

Q: Will Virginia Attorney General Ken Cuccinelli investigate this clear case of academic fraud with the same vigor that he has exhibited in his pursuit of the allegation that Prof. Mann’s original “hockey stick” paper constituted fraud?
A: No.

This has been etc.

21

John Quiggin 05.17.11 at 7:17 pm

It appears there was federal funding, which creates all sorts of potential hurt for both Wegman and GMU.

@Colin Danby – you are right, and this kind of irony is typical of the case

22

Bob Munck 05.17.11 at 7:30 pm

Say what you will about Wegman, he has a great cheese selection.

(Come on, we were all thinking it.)

23

Warren Terra 05.17.11 at 7:57 pm

@ Bob Munck,#22
I wasn’t thinking that at all. I was thinking that he takes weird photos of his dogs.

24

Anderson 05.17.11 at 8:27 pm

So George Mason has both Wegman and the George Mason Branch of the Koch Economics Departmentses to brag about?

Not to mention the law-school luminaries David Bernstein, Ilya Somin, and Todd Zywicki.

25

DCA 05.17.11 at 8:27 pm

@12, 17: He had a long career in statistics (PhD in 1969), though largely outside academia–but so far as I know producing good stuff, or at least papers with a much higher intellectual content than the color theory paper. (Try “e wegman” in Google Scholar). Something happened, but it isn’t clear what.

26

eli 05.17.11 at 8:54 pm

Stan Azen is also, at the moment at least, “Assistant Dean of Research Integrity” at USC, according to his bio page.

27

Evil Bender 05.17.11 at 9:34 pm

I can’t get over Wegman’s strategy : “I didn’t plagiarize! It was that student! You know, the one who wrote part of the paper for me but who didn’t receive any credit for doing so!”

Is he plagiarizing Jonah Goldberg’s arguments now, too?

28

Ted Kirkpatrick 05.17.11 at 11:28 pm

@20, @24: Not to mention GMU law-school luminary Ken Cuccinelli.

@Colin Danby: The paper even argues that the peer review process is strengthened by having senior academics collaborate with their students, because “younger co-authors are generally not editors or associate editors”. But the first author of the paper (Said) had been appointed an associate editor just one year after her Ph.D—at the very journal that published this paper.

29

Dr. Hilarius 05.18.11 at 12:17 am

Is there money is this “university” business? I have a spare room in my basement. It won’t take me long to come up with some impressively named departments and start looking for faculty. Hey, Koch brothers, over here!

30

John 05.18.11 at 12:52 am

Wegman has a long and mildly distinguished record. Can’t say I ever met him myself, though. He received a “Founder’s Award” by the American Statistical Association (ASA) a few years ago, which is given out to people who do good stuff for the ASA, and only a few are given out in a year (three this year, for example). He is a former editor of Computational Statistics and Data Analysis (CSDA) – a well respected technical journal, just to make the point, despite this blow. The great bulk of the articles have titles like “Bayesian proportional hazards model for current status data with monotone splines”; the paper in question was somewhat of an anomaly, in more ways than one evidently.

As a professional statistician who’s been around a while, this is kind of a shock. This is the sort of thing that happens in other professions, not ours! Well, so much for that burst of professional chauvinism, and of course he wasn’t plagiarising a piece of statistical work… it just goes to show, every profession has its share of hacks and dishonest individuals, but still, he was fairly well respected… He may be innocent of the plagiarism, but having gone back and read the paper, it just doesn’t seem to fit in CSDA, although sometimes journals do publish outside their “typical zone.” Hard to believe Azen’s story, as I would imagine any other CSDA reviewers would have not recommended it for publication in CSDA, and it would be a miracle if he got other reviewer’s comments back in less than five days from the submission of the paper! And of course Azen’s email hardly supports the idea that anyone else was asked to see it.

It’s a big no-no to publish a paper with significant text contributed by someone who’s not listed as an author, so blaming the non-author grad student for this would still be a big hit to his reputation, over and above the plagiarism itself, even if it proves correct. A whole lot like plagiarising from your grad student, who in turn plagiarised from someone else, then blaming the grad student for the whole mess. After all, if the stuff the grad student wrote wasn’t meant for publication, but was just a document summarising some things for the authors, the standards for “plagiarism” may still apply but it’s hardly a capital offence in the circumstance. (Not that it’s no offence, just not the same level of offence.)

31

John 05.18.11 at 12:58 am

To clarify my last sentence – what the graduate student has done is hardly a capital offense…

32

Substance McGravitas 05.18.11 at 1:09 am

what the graduate student has done is hardly a capital offense…

Who is that dastardly grad student?

33

Tony Lynch 05.18.11 at 2:29 am

That wikipedia was plaigarised is indicative surely – given that wikipedia is hardly abstruse or difficult to access – that the offending grad student (if they exist) lacks the concept of plaigarism in the first place.

In fact, I do not find this all that unusual in my more recent teaching experience. It seems to me that when everyone is deidicated to “going forward” at all costs – or as President Obama put it possesses “a belief that we need to look forwards as opposed to looking backwards” – then plaigarism definitately is off the conceptual map.

34

John Quiggin 05.18.11 at 3:03 am

My understanding is that the graduate student was asked to prepare an overview with no indication that it would be included in a published document. In that context, lifting the material from Wikipedia is a bit slack, but not an academic offence.

35

roger 05.18.11 at 9:51 am

Wegman’s story is an embarrassment – but it will prove a very minor bump for the denialist industry. Over at the Atlantic, that hotspot of Meganomics, there is a reference to it that is almost perfect spin:
http://www.theatlanticwire.com/national/2011/05/climate-change-scientists-face-inconvenient-truths/37824/

It is worth looking at because it confects a denialist position while remaining “rational”. Thus, in its final sentence, it concedes the human effect on climate, although, gosh, how this is happening is beyond the human cognitive capacity to understand – and has nothing to do with CO2! However, the beginning paragraph somehow blames Al Gore, of all people, for Wegman’s plagiarism. Al Gore, it turns out, drove Wegman crazy! And then it find a fair and balanced solution by spotlighting a scientist who, as it happens, made no mistakes whatever on his research, but who, somehow, is to blame for the way Gore presented his work in his film. I loved that film when I saw it, and I love it all the more that it is targeted, whenever possible, by rightwing driveby-ers.
The reader of the Atlantic blog piece will have learned absolutely nothing by reading the whole piece, which is where it really is good spin. It doesn’t deny – it obfuscates to the point where whatever happened is simply unintelligible, and thus probably unimportant. And so this little piece of academic malfeasance will be swallowed in the pipeline, to make no difference whatsoever. However, it might be interesting to work on the William Easterly, the economist who was so impressed with “climategate’ that he decided climate science was a fraud and climate change a hoax. For a little fun with powerful but truly dumb academics, I’d urge people to read this post: http://aidwatchers.com/2009/12/copenhagen-special-climategate-and-the-tragic-consequences-of-breaching-scientific-trust/. As it happens, it is a piece of post-spin – while good spin leaves you feeling empty, post-spin waxes indignant at what hasn’t happened, in order to draw large ideological conclusions that are not entailed by any argument whatsoever. The cycle of spin to post-spin is now standard procedure in the mechanism for tilting politics to the right. It deserves study.

36

John Quiggin 05.18.11 at 10:32 am

@Roger I read the appalling piece of tripe in the Atlantic. I hadn’t caught up with Easterly being sucked in. Emeritus disease again, I guess.

37

Barry 05.18.11 at 12:05 pm

Considering the number of hacks at The Atlantic, it isn’t ’emeritus disease’; it’s deliberate. As Megan shows, one doesn’t need to be emeritus status there to be pretty bad. In addition, when ’emeritus disease’ seems to run in one political direction, it should be called ‘partisan hackery’.

38

Rob 05.18.11 at 3:50 pm

@Steve LaBonne

George Mason is most certainly a real university. Their basketball team made it into the NCAA tournament. QED

39

John Mashey 05.18.11 at 7:59 pm

I believe I can shed some light on various questions, and I note some Australian academics helped a lot, specifically Stephan Lewandowsky for introducing me to SNA researcher Garry Robins at U of Melbourne, among others.
(But I don’t know why previews show weird font sizes.)

0) If you want to see the detailed analysis, see SSWR – Strange Scholarship in the Wegman Report.

pp.118-129 shows the side-by-side-by-side of Said, et al(2008), the Wegman report (WR) and the original antecedents that DC found. you can see the flow and compare, in the color-highlight scheme we’ve adopted.
The new information does clarify the necessary uncertainty on p.118 over who do did what. Most rational people need spend only a few minutes to decide whether or not this is plagiarism.

1) pp.143-152, 159 discusses SNA and its roots, offers evidence of the Wegman team’s unfamiliarity with SNA, and spends a few pages on CSDA journal. It has a table of all the papers in that issue, and graph by Receive-Accept interval. Median = 204 days.

2) p.151 has an analysis by Garry Robins, who panned it.
I later sent the paper to a very distinguished SNA researcher, got the same sorts of comments, but didn’t include them, as I thought the horse was already dead. The researcher was quite happy yesterday when this finally came out.

Vergano got (expert) Kathleen Carley to quickly review it; she panned it. The 3 experts thought it would not pass peer review in a relevant journal, that the SNA methodology was flawed, and it leapt to conclusions. Ted Kirkpatrick had earlier pointed out the scholarship problems apart from SNA.
Dan Vergano is clever. Where did he find Carley’s name? Why, it was in Wegman’s email to Elsevier, where he says:

“Denise (Reeves) worked (and still works) for Mitre Corporation. Her company sent her to take a short course on social network analysis from Kathleen Carley, a professor at Carnegie-Mellon University. Dr. Carley is an internationally recognized expert on social network analysis. When Denise returned from her short course at Carnegie-Mellon, I took her to be the most knowledgeable among us on social network analysis…”

MITRE pays for short course for Reeves, the WR lists her as MITRE, not as “GMU” (student). I wonder why? She’s taken a short course, so they base one of the two key memes of the WR on this, because Wegman thinks she knows the most.

2) George Mason University has interesting characteristics. They have some good people (I know some), but I would certainly look closely at certain departments. SSWR called the WR a facade to deliver the messages. It’s possible that GMU’s key mission is one thing (Kochs), but one needs a bunch of credible departments.

3) I think I coined the self-refuting phrase, p.150, but I may have used it at DC.

4) Yes, there is Fed money: 3 contracts, p84. Worse, one of them (NIAAA) is covered by the Office of Research Integrity, who takes a dim view of such. Two are from US ARMY. ORI *debars* people, but also, it is not obvious why these agencies would fund bad SNA to attack climate peer review.

5) I know some fine statisticians, who thought Wegman was once pretty good. They are mostly sad that something has happened to him. Some are pretty peeved, and were, years ago. The earliest plagiarisms we’ve seen are fairly small in 1996, and fairly substantive in 2002 (that’s the Wiley posts @ DC.) It’s too bad: in my SGI days, I might even have attended his Interface conference. SGI even sponsored one.
Another statistician told me “I’m glad this is coming out … but why did it have to be *my* profession.” I think most statisticians are good guys.

See also p.83, Odd loose ends at Wiley. Said has been claiming to be a Professor @ Oklahoma State U. She isn’t and never was. This is in addition to any plagiarisms and junk papers they write/edit. Much of WIRES:CS looks reasonable, but…

6) The Cuccinelli attack on U VA and Mann references, among other pieces, the material in WR equivalent to that being retracted.

7) GMU has not handled this well. See
Strange inquiries at GMU.
You can see the timeline of GMU policy versus what happened. Ray Bradley sent complaints over a year ago, have yet to send him the simple inquiry report, much less a real investigation. Rice did theirs (For Scott) in 9 days.
See p.13, where Donald Rapp forwards email from Wegman to Elsevier (Ray’s publisher).
We learn that
“The deepclimate website is full of crackpot conspiracy theories. I avoid reading it in order to keep a semblance of normalcy. I believe the GMU inquiry will vindicate me and my co-authors. (Rice University has already concluded that the charges against David Scott have no merit.)”
Yes, Rice concluded that, but apparently Scott didn’t tell Wegman why…
I’d have loved to hear the conversation when that happened.
p.9: See Wegman’s August Facebook post, “Want to know a bad week?”
He has sometimes written papers on security.

8) No one could make up all the details we already know, they’d be laughed away….
And note: people write lots of junk, but mis-use of funds is bad, and out in the distance, is oen more possibility: misleading Congress is a *felony* 18USC1001, and conspiracy to committ felony is one also (18USC371) and there are more.

People can write junk and mislead, and there’s not much you can do, unless they get into libel and you can catch them, or if you do it in court, or for Congress … but those are *actionable*.
See the presentation in B.C., especially delightful because of the libel efforts udner way there, p.38. McCOnchie “wrote the book” on Canadian libel. Tim Ball = toast.
“You ain’t seen nuthin yet.”

40

herr doktor bimler 05.18.11 at 11:56 pm

Wait, what. “Senior academics should not collaborate with each other, but should instead work only with their students”, specifically to avoid the unfair opportunities to bypass peer review that occur when academics collaborate with other academics…
All in an article that was submitted to a journal where the author had previously been editor-in-chief, and using collegial connections with the current editor to bypass peer review?

And the Wegman-Said plagiarism on colour vision, published in a journal edited by Wegman & Said?

As well as the self-refuting paper, there is the self-refuting editorial policy. The term ‘Wegman’ should include “This bad thing I’m doing, other people should go to great lengths to avoid doing it”.

41

herr doktor bimler 05.19.11 at 1:48 am

the graduate student was asked to prepare an overview with no indication that it would be included in a published document. In that context, lifting the material from Wikipedia is a bit slack, but not an academic offence.

Unlike the two contributors to the paper who were listed as authors, Reeves has been called before a review committee to explain her behaviour, so no-one can accuse GMU of ignoring the ethical breaches on the part of the faculty.

42

John Mashey 05.19.11 at 1:53 am

“no one can accuse”: look at the chronology in SIGMU.

I’d coded Reeves as orange or green, not read in SSWR, because no one could figure out what she’d done. Clearly, GMU took some action by August, but it’s pretty clear they broke their own policies pretty badly.
I think there are multiple mis-uses of students here.

43

Pinko Punko 05.19.11 at 3:24 am

These people are so bad at what they do that it makes me feel like they couldn’t handle a massive conspiracy, which is a tiny solace, because they don’t even need one if they just keep playing the inaction game.

44

herr doktor bimler 05.19.11 at 5:55 am

“no one can accuse”: look at the chronology in SIGMU.

I will admit to a certain amount of sarcasm.
I will also admit to the possibility of misunderstanding Reeve’s statement that “she has met with a George Mason University misconduct committee”… I took that to mean she had been investigated for Wegman’s misuse of her literature search, but she could have been there to testify.

I see around the Googles that Wegman is also trying to shift the blame onto “an Asian student” for taking advantage of his trusting nature and inserting this recycled material into their various collaborations… hoping that everyone will think “Oh, Asians, they don’t share our attitude about intellectual property”. Who is this student, we wonder? Sharabati and Said are from the Middle East, which I suppose is Asia, but not really fitting the stereotype he’s trying to evoke.

Colour vision is my main research field, and reading how badly Wegman & Said distorted the issue of old-age-related tritanopic confusions made me want to throw things.

45

John Mashey 05.19.11 at 4:29 pm

1) Just to be clear, Wegman did *not* try to blame an Asian student.
That was confusion amongst others, actually referring to different cultural expectations. Faculty are often urged to take special care with foreign students and make US-style plagiarism rules clear.

2) The student was Denise Reeves, who is in her 40s or 50s and has worked at MITRE a while. I don’t know about Sharabati. Said grew up in US from 4th-grade onwards.

The Asian student idea is just a confusion, but not from Wegman. [he ah enough of his own.]

46

thomas 05.22.11 at 1:15 am

As John @30 said, Wegman used to be a real statistician. He did some good work on interactive graphics, back when computers were barely up to it. He also did a lot to get more academic recognition for statistical computing.

The statistics department at GMU, perhaps because it’s in the engineering school, has an atypical range of specialties, but certainly some of it is very good. In areas I know about there’s Dan Carr for graphics and William Rosenberger for adaptive clinical trial designs.

There’s an interesting contrast with the controversy on fine particulate air pollution. A lot of statisticians who went into that field were pretty confident that the results would disappear if you got the analysis right, but there we ended up being converted.

47

John Mashey 05.22.11 at 4:18 pm

As thomas says …
1) Wegman even did some OK work on SGI workstations…

2) Association alone is not evidence of guilt, and I know some others at George Mason University that are quire reasonable folks. However, given the heavy infouence of the Koch brothers, I sometimes wonder if:
a) The core mission of GMU is pursuing the public policy goals of the Kochs and their friends, via Mercatus Center (run by Koch exec.), the Institute for Humane Studies, other entities, and with some people in depts of public policy, economics, political science.

b) But for the credibility of a university, one needs a lot more, so that there are regular scholars and researchers doing normal work, but beign used as a “facade” perhaps, as per SSWR.

c) Hence, one has people getting normal Federal research grants at the same time that the Kochs&others pursue dismantlement of the Federal government and defunding of any research they don’t like.

Comments on this entry are closed.