Wegman plagiarism case: GMU jury out to permanent lunch

I “crib” from Wikipedia (well, with a link) for blog comments. If only I’d known that I could have been writing scientific papers instead.

Seems like this creates the need for an external group maybe of publishers to investigate plagiarism and misconduct.

“The big question is how long GMU can keep on getting away with doing nothing”

As long as Koch money pours in.

Even if Wegman retires, it won’t “render the case moot”. At GMU, there would remain issues with several doctoral theses done under his supervision, as well as the question of whether any of the questionable articles or questionable supervision was performed under the sponsorship of federal grants. Over at Wiley, there remains the questionable editorial process at WIREs: Computational Statistics, for which Wegman and Said constitute 2/3 of the editors-in-chief. And finally, and most important of all, there remains the report and testimony provided by Wegman and Said (with Scott) to Congress. Wegman’s employment is just a small part of what has to be addressed. Also worth asking who else might have helped or enabled all this …

Adding on to Ted’s point, it’s clear that the administration and board of regents of GMU should be replaced. The top level of that university is corrupt (even compared with top leaders in general).

“Yes, and if you plagiarize your thesis you can be defense minister of Germany.”

I’m glad you mentioned this because the cases of both Guttenberg and Koch-Mehrin (and there was at least one other case but the name eludes me) show that GMU’s cavalier attitude to plagiarism doesn’t have to be the norm. Interestingly, US media found it somewhat bizarre that Germans make such a fuss about politicians’ academic misconduct.

Mpowell @ 10:

What you’re saying is literally true — research on Global Warming has become more definitive over time. It leaves out something important, though, about the response of so-called ‘skeptics.’ If an estimate as unbiased, even if it has a relatively large error range, you should be considering the high and low scenarios. There’s really no defense of just looking at one side of the distribution and planning around that.

Of course, there’s an Everest-high mound of bullshit claiming that the studies are biased due to corrupt researchers or the ‘protest industry’ or whatever. But back in the real world, current measurements are coming in worse than predicted by earlier models. We seem to be living through the exact scenario that ‘skeptics’ told us to ignore.

1) Of the three deadly sins of FFP (falsification, fabrication, plagiarism), the FFs are far harder to detect, prove and explain to a non-expert audience. P can be hard to detect, but once found, is trivial to display, especially after we added the color to the side-by-side format. (I conjecture that readers check the identical-cyan blocks, see that they are identical, and then ignore them to focus on the minor changes (yellow). There are probably some papers to be done by cognitive psychologists. I do know that I’ve found errors and problems only after doing the coloring, because the real changes then leapt out.)

For those unfamiliar with FFP, see PDF @ SFWR, which gives some of the simplest examples of FF from the Wegman Report, highlighted in red.

p.4 Wegman Report text
p.5 WR with cyan&yellow for plagiarism
p.6 WR with cyan&yellow, then adding red for errors/falsification

It is sometimes hard to tell accidental FF from serious incompetence, whereas massive plagiarism is different. Hence SSWR covered the massive numbers of errors, distortions, etc in the WR (of 91 pages, only a few were problem-free), and in fact, only about 25% of the pages were on plagiarism, and those mainly to show the side-by-sides. Most of the discussion was on other topics, but ti was complex. It took more months to do SFWR, but plagiarism helped show the FF. When it is clear someone copied text, their changes leap off the page.

2) As for duration, see SIGMU.
GMU broke their own rules fairly thoroughly.

Rice U handled the *same* complaint in *9 days*. Now, that was simpler, because Scott quickly showed that Wegman had taken responsibility – Scott had written only the 3-page math appendix. But it does show that a credible university acts quickly for something so obvious. This had to have gone from President to VP Research within a day or two, because plagiarism was not in question, only who did it. I applaud Rice for acting with honor, integrity and alacrity.

In the US, there is an entity called the Office of Research Integrity. It has teeth.
It debars (health-related) researchers, typically for 3 years, and Yasmin Said ACk’d a health contract on the CS&DA paper.
Debar = no Federal research contracts, period, not just health.

Both Said & Wegman Ack’d Army contracts.

ORI says:

“Who can be debarred?

Both individuals and entities may be subject to debarment. In the area of grant and cooperative agreement supported research, this includes anyone who participates in the research: the principal investigators, researchers, contractors, students, and technical and support staff. To date, all ORI debarments have involved individuals, not institutions or other entities.”

Institutions have to follow certain rules in reporting to ORI. Outside, one cannot know if GMU has done so. If they haven’t, at some point they will be in trouble. No one messes with ORI. Academic friends shudder at the idea. Of course, debarring GMU is an institutional death sentence, so that likely wouldn’t happen, but something short of that, who knows? SIGMU p.17 lists people who must have been involved in this.

jim 10.07.11 at 9:56 pm

” The President of GMU has announced his retirement. There’s a current search running for his replacement. It would not be unlikely that the Provost would want to at least discuss the Wegman case with the incoming President before taking action.”

Jim, good point about GMU and the GOP – it’s another data point to show that the right is incapable of *not* trashing institutions.

But as has been poined out, if GMU had followed it’s own rules, the investigation would have been over months ago – and that’s if it had taken the full allowed time. And it shouldn’t have taken the full time, because there’s a core of easily proven plagiarism. Other stuff might have taken longer, but GMU should have fired Wegman a while back. Instead GMU stalled it, with no good reason given. And even now, if they are waiting on consultation with the incoming president on an open-and-shut case of plagiarism, that’s not a valid reason.

At this point, any assumption of good faith on the part of climate “sceptics” is unwarranted.

This doesn’t seem like a scientific approach.

@Bill Benzon…

The concern was probably a bit excessive in your case, but editors and academic publishers do often care about “self plagiarism”, i.e. publishing the same thing in two different places. There are two concerns here:

a) Academics padding their CV’s by publishing the same thing twice under two different titles, to give the impression that they’ve done twice as much research.

b) Copyright. Typically, when a commercial publisher like Springer publishes an article, they sign a contact with the author that limits where the content can be republished (in the particular case of Springer, they’re reasonably open to negotiating contracts that let you, for example, republish on your web site, but the principle still holds). So if the same content ends up published in two venues, the concern is that the author is breaking the terms of their contract with one or both publishers. (And, that the second publisher isn’t going to get the sales they expect because potentially interested readers have already read it).

Prior publication in The Value looks OK on both grounds, but I can quite understand Questions Being Asked.

I think jourals are starting to use more automatic plagiarism detection, because it’s becoming needed.

Turnitin has its own issues: see here. Producing algorithms that cannot be relatively easily gamed is very difficult when the incentives line up this way.

Plagiarism casts doubt on Wegman’s scholarship. If he is willing to ignore one ethical standard, then it makes it more likely that he will ignore others. If he can’t be trusted not to plagiarise his sources, how can he be trusted not to falsify them?

Plagiarism also makes it easier to falsify a source. An author who quotes a source must indicate elisions, and distinguish his own conclusions from those in the source. Plagiarising enables elisions to be hidden, and the source to be altered. John Mashey’s “Strange Falsifications in the Wegman Report” documents how Wegman plagiarised a section on tree rings from Bradley in such a way as to undermine the conclusions of the original. Thus Bradley’s “If an equation can be developed that accurately describes instrumentally observed climatic variability in terms of tree growth over the same interval, then paleoclimatic reconstructions can be made using only the tree-ring data” is replaced by Wegman’s “Thus tree ring proxy data alone is not sufficient to determine past climate variables” with no indication from Wegman that he is contradicting his source.

The big question is how long GMU can keep on getting away with doing nothing.

Until the rising sea level floods Fairfax County, Virginia.

JQ:
“Although it’s hard to prove (there is some evidence but not conclusive) it’s obvious to anyone who knows how these things work that the entire report was a put-up job, and that Wegman was recruited by one of his friends at GMU (the initials FS come to mind) as someone who wasn’t (at the time – he’s come out openly since) publicly identified as a denialist but could be trust to come up with the right answer.”

1) We have strong evidence that Wegman was recruited via Jerry Coffeey, a VA Tea Party friend of his fond of Singer, Michaels, etc. See CCC p.115. Yasmin Said revealed this in that foolish 2007 seminar she gave, whose file suddenly disappeared summer 2010, and whose mention in the seminar list was edited away.

2) I think the evidence for a put-up job is by now overpowering, which is why SSWR grew to 250 pages. But, in particular, see pp25-32, that go through the chronology and people involved.

Also, see pp.89-95 on Modified and disappeared files.

Finally, see Strange Tales and Emails p.17, which tracks how a raw PPT of a McInytre+McKitrick talk for George Marshall Institute got to Wegman via Joe Barton staffer Peter Spencer. That came out of FOIAs after SSWR.

“So is the whole issue with Wegman that he took credit for someone else’s work, or is there something that makes his results unreliable?”

If previous work is being plagiarized rather than new work being done, no actual expansion is being made to the field of inquiry but cover is provided to those who claim that it has been expanded. This is akin to the “teach the controversy” people who are themselves creating the only controversy in evidence.

For example, one of the earlier cases was a “correction” to a bad analysis that claimed to refute global climate change. The “correction” seem to support the conclusion of the original work… except there was no “correction”. Not only was there plagiarism, it was plagiarism of wrong answers.

Wilfred asked: What, EXACTLY, does it mean to spend years mastering a relatively esoteric academic area only to find out that someone can ‘google’ it and cut out the hours your spent prowling around libraries or counting your pennies in some used bookstore, angsting about whether that book was worth the expense, or if anyone would notice you ripping out the odd page or two?

If you spend years doing the actual research in the stacks, it means you actually know what you’re talking about and can back it up with hard evidence and detailed citations from peer-reviewed journals.

People who cruise Wikipedia get what they pay for: zero. There is no assurance that any given Wikipedia article contains any actual verifiable knowledge, and there’s a lot of evidence of systematic bias in most (all?) Wikipedia articles introduced by facts edited out by a biased administrator and citations removed to slant the presentation.

Consider, just by way of example, the Wikipedia article on the Chicago School of Economics. This philosophy of economics has made predictions which have been comprehensively disconfirmed. This philosophy of economics has been so thoroughly debunked that it is no longer possible to describe it as a legitimate branch of economics: it’s now something closer to Dianetics. Yet look at the Wikipedia article of the Chicago School of Economics and you won’t find a word about how completely the Chilean experiment falsified the predictions of the Chicago School. You won’t find a word about how the predictions of Chicago-School-influenced economists leading up to and following the 2008 global financial meltdown were completely utterly 100% wrong. Most of all, you won’t find a single word detailing the many instances in which Chicago-School-influenced economists, having found embarrassing their false predictions of sky-high interest rates due to the Fed’s expansion of the money supply after the 2008 financial collapse, now cast about for arcane and outlandish second-order effects to explain with absurdly convoluted complexity what Keynes’ and Hicks’ IS-LM model simply and easily explains — and which Keynes’ and Hicks’ model predicted correctly in the immediate aftermath of the 2008 crisis.

The world is dividing into lazy people who trawl the net and come up with junk science and economic just-so stories and foolishly parrot this nonsense, to their discredit…and hard-working people who hit the library stacks and do the actual research and come up with valid knowledge that actually works in the real world. Wikipedia is just one of many flash points in this schism.

@John #34:
@mclaren:

Clearly there’s a difference between standards for an academic/graduate student and an undergraduate student. Having slogged through endless cut-and-paste jobs from students it seems to me that we are experiencing a kind of electronic epistemology, where process is trumped by product and superficial ‘knowing about’ something has replaced what an older generation called knowledge.

Some, most, of my colleagues march about harrumphing about student laziness, vapidity, etc., but maybe the real challenge is for us to find a way to accomodate this new epistemology. So the statues don’t have eyeballs, what now?

Wikipedia is not fundamentally different from relying on Cliff Notes or the like to write a book report.

The basic divide is between reading the primary sources and reading the interpretation of the primary sources provided by others. There is a role for the latter, but true expertise requires the former.

I should clarify that when I referenced “online journals” above I meant not just true e-journals but, more importantly, electronic access to traditional print journals.

It is also insane to demand that all public and citizen discourse on important issues of the political economy be of peer-reviewed original-source academic quality. I do not need to study reports of steel production in Chile in the 80s to hate the “Chicago Boys.”

Go to hell.

46: I think the academic web needs to determine who and what it is for. Discrediting Wegman among yourselves sort of misses the point. Wegman wasn’t writing with you as the intended audience. You need to write for Wegman’s intended audience in such a way that does not insult them. You need to gain some kind of authority and legitimacy with them, for they will not be able to sift the evidence, weigh the arguments, compare the credentials of competing references in the way you seem to expect them too.

Any discourse analysis that dismisses Wikipedia outright has really removed itself from the useful and influential discourse. Sorry.

Or y’all can just talk among yourselves. Sorry for trolling.

I’ll second bob at 44. wikipedia is an astonishing resource and elitist disdain for the lack of academic quality is misplaced. As bob points out, wikipedia will almost always provide valuable links where the person searching for information can go deeper. Secondly, it usefully highlights controversies even if the account given may be biased. And, its plain factual statements (often on obscure topics that would be difficult to find elsewhere) are usually reliable. The bottom line of course is that schools and colleges are called on to educate students in the good use of online sources. I don’t think, speaking from my limited experience, that they are currently doing a good job. I found to my surprise that most students do not know how to look for information online, not even wikipedia. In my view there should be dedicated courses on that subject.

I personally make heavy use of online sources for teaching, including wikipedia. Part of the reason is that I can use current information where in earlier times, you’d have had to rely on text book statistics from ten years ago. I try to enable students to seek out up-to-date and accurate information about the state of the world and to use time honored scientific tools for interpreting these data.

The Wegman case isn’t about elitist disdain for Wikipedia. Wikipedia is great, and by reading it you can quickly get a basic level of understanding of a subject area. But Wegman didn’t represent himself as someone with a basic level of understanding of climate science. He represented himself as someone capable of writing an expert Congressional report evaluating the work of Mann, Bradley and Hughes. The fact that large parts of the report were plagiarised (and not only from Wikipedia) undermines that claim to expertise.

Bob et al, adding to what Gareth Rees said — John Mashey’s awesome SSWR (linked by JM above and by me here) spends approximately nil time pooh-poohing Wikipedia as a resource for the hoi polloi, and gets straight to the more important point–from the SSWR’s executive summary:

Of 91 pages [in Wegman’s report], 35 are mostly plagiarized text, but often injected with errors, bias and changes of meaning.

Three issues here.

Plagiarism is an acknowledgement of an upper bound on one’s own expertise. Setting aside questions of mere laziness, which do not obtain here, a plagiarist is implicitly confessing that they are not capable of synthesizing the information they are presenting as expertly as the source they are stealing from.^1^ The maxim “if they could have done better, they would have done better” applies. To plagiarize Wikipedia is therefore to admit that one is not capable of synthesizing the information one presents as well as a cursory explanation in a crowdsourced layperson-friendly encyclopedia. That encyclopedia might well be super awesome (and it is!), but we certainly would like for an allegedly expert report to achieve a greater upper bound of mastery, in depth and breadth of synthesized understanding, than a cursory summary in an encyclopedia contains.

This is why people are pooh-poohing Wegman’s use of Wikipedia. Plagiarizing from an expert source’s tremendously difficult original research would at least be implicitly acknowledging a very high ceiling over one’s expertise. “Can’t navigate my way through that argument as well as this person… but who could?” Plagiarizing from Wikipedia (or any entry-level resource) is effectively confessing to a very low level of expertise.

Plagiarism is evidence of dishonest intent. Setting aside the plagiarism concern, there’s still copious evidence that the Wegman report contains enough false statements and incorrect interpretations of data to completely compromise its theses. There are two plausible explanations for a report that fails this badly: either Wegman et al are well-meaning but terribly error-prone, or Wegman et al are deviously dishonest.^2^ The fact that Wegman et al plagiarized about a third of their material establishes one of those two explanations as likely, allowing us to place confidence in assertions like “Wegman intentionally misrepresented the information available to him in order to mislead Congress.”

The less complex a plagiarized source is, the greater freedom it provides the plagiarist to redirect arguments and manipulate conclusions.

Because the less-complex source speaks more broadly and with fewer technical caveats, it will inevitably elide some of the nuance details that support each conclusion it communicates. A plagiarist can exploit the ambiguity, with selective rephrasing and invisible elision, in order to produce conclusions that would be plainly at variance with the more complex (and therefore more technically specific and rigorous) expert source. Plagiarizing from a high-level original research source may not provide one with nearly so many opportunities to engage in that kind of manipulation, as the statements and inferences will be constructed more precisely, making it harder to systematically redirect the flow of the argument.

Even a well-intentioned plagiarist might accidentally produce a wildly divergent line of reasoning from a less complex source, accidentally eliding or obfuscating important technicalities, hedges, and qualifications, because the critical importance of the more technical caveats is inherently obscured by the simplicity of the source text. Therefore, plagiarizing from a less complex source is more worrisome, because of greater likelihood that the plagiarist’s text diverges meaningfully from the assertions of its source.

^1^Self-plagiarism being the exception that proves the rule, in the classical sense of that phrase: you plagiarize your own work when you can’t communicate the idea at hand any better now than you did back when you first wrote it.

^2^These are not disjoint categories, but it helps to consider them separately.

Just to circle the square, it turns out that denial is well practiced in the arts of self plagiarism, publishing the same article multiple times.

On another subject, it was interesting reading various claims for Wegman in the 2006 comments. Some have held up, others, not so muck. Eli is less inclined to blame folk like TCO, because the entirety of the fraud was not obvious then, but it is an object lesson in whom to trust.

Tom Atchison@50
“easy for average citizens to make up their minds as to who is telling them the truth about global warming, because (to summarize crudely) a Google search on the term leads initially to a quite reliable Wikipedia article laden with references to authoritative sources, whereas the skeptical (or ‘denialist’) sites that come up are pretty obviously industry-funded and/or right wing think tanks.”
There is an assumption that everyday people use information the way scholars do. I don’t think so. For average people hearsay weighs more than anything on the untrustworthy “Internet”. I have a cousin that tells me that scientists around the would bury dinosaur bones and then dig them back up to prove evolution. She is not a moron. Its just that in her circle of friends their say weighs more than some paleontologist’s boring dissertation – from “The Ivory Tower”.

Isn’t GMU an accredited university?

I’m acquainted with someone who teaches there in a non-controversial area, and from her descriptions it seemed like a fairly normal kind of university.

I would think that having faculty who are ethically challenged would raise a few flags the next time that accreditation came around.

Or are there areas that are such a political hot potato that accreditation committees don’t want to touch them?

If so, it’s a sad day for American academia. Tenure should protect people who pursue unpopular research–just think about someone who was researching developments in the USSR during the 1950s. But that’s different from committing fraud.

maidhc:
Yes, GMU is accredited.
Nature raised the issue in its editorial “copy-and-paste”:

“Perhaps it should fall to accreditation agencies to push for speedy investigations. Tom Benberg, vice-president of the Commission on Colleges of the Southern Association of Colleges and Schools — the agency that accredits George Mason University — says that his agency might investigate if the university repeatedly ignored its own policies on the timing of misconduct inquiries. To get the ball rolling, he says, someone would have to file a well-documented complaint.”

As it happens, GMU’s usual scheduled accreditation review has been going on this year, with results due in December. Hopefully, they will perform better than GMU.

I find Wikipedia most useful for checking basic assertions that lie well outside of my field of understanding, and where necessary giving me links to sources. If someone makes a basic claim about a field I’m unfamiliar with, and the wikipedia entry seems to strongly disagree with the claim, then I’m on fairly solid ground investigating it further.

Wikipedia pages also often alert one to the existence of a controversy within a field of knowledge. Iris Chang’s Rape of Nanking is a good example. Not only does the wikipedia page point out dubious scholarship that it would take an expert on the topic to notice (e.g. the mis-captioned photos) but it also gives a reasonably good insight into the nature of academic debate about what happened in Nanking. I’ve got no experience of the study of history, so being able to check up on the contents and conclusions of a book like that is really useful. In fact, Wegman is already on the wikpedia page about the hockey stick, so it would be hard for someone checking his assertions to not notice what he’s been up to.