The only thing we can be sure of about the future is that it will be absolutely fantastic. So if what I say now seems to you to be very reasonable, then I have failed completely. Only if what I tell you appears absolutely unbelievable, have we any chance of visualizing the future as it really will happen.

- Arthur C. Clarke (h/t Brin)

Friday, October 8, 2010

Can the Press be Awakened from its Slumber?

Well, a story making Wonk Room isn't necessarily a story making the mainstream, but this is more than a little bit interesting:
Insurance Commissioner Mike Geeslin testified that his office was receiving questions about “coverage for dependents and children,” “pre-existing conditions,” “the new federal risk pool or the pre-existing condition insurance plan, and then they want information about the effective days more among the lines of what I need to do and when.” This surprised Rep. Susan King (R), who wanted to know if Texans were complaining about the new law:
KING: On the complaints…what does that mean? There has not been a single complaint?

GEESLIN: That’s not to say you can’t get 10 in the office today.

KING: But you’ve had nobody really calling and being concerned? That’s really amazing. I just didn’t know….
And then there's this from USAToday, which is MSM however you look at it:
In 2006, GMU statistics professor, Edward Wegman, spearheaded a Congressional committee report critical of scientists' reconstructions of past climate conditions -- notably the 1999 "hockey stick" paper in Nature, which concluded that the 20th Century was the warmest one in a millennium. A National Research Council report later that year largely validated the 1999 paper's research, but the "Wegman" report has knocked around in public debate over climate ever since.

GMU spokesman Daniel Walsch confirms that the university, located in Fairfax, Va., is now investigating allegations that the
Wegman report was partly plagiarized and contains fabrications. Last month, a 250-page report on the Deep Climate website written by computer scientist John Mashey of Portola Valley, Calif., raised some of these concerns. Mashey says his analysis shows that 35 of the 91 pages in the 2006 Wegman report are plagiarized (with some of the text taken from a book, Paleoclimatology: Reconstructing Climates of the Quaternary, by Raymond Bradley of the University of Massachusetts) and contain erroneous citations of data, as well.

30 comments:

King of the Road said...

Call me thick if you will, but I'm missing your point here. Is it that King should know that that type of question or protest or complaint would be extremely unlikely to come to that office? Or is your point that the press has not reported that the public is (arguably) not generally unhappy with the ACA? Or something else I'm missing entirely?

Michael Tobis said...

The short answer is that the Republican party lives in a bubble.

I think the idea that there has ever been a groundswell of opposition to Obama's actual health care reform from any significant group is simply a delusion paid for by billionaire money and propagated by a gullible press.

Very much like "climategate", there is a republican reality and a real reality. The influence of this confused though well-intentioned worldview should be challenged on its points of confusion. That is the job of the press.

I think it likely that most Texans are not unhappy with ACA, even though most Texans hate liberal democrats for shallow reasons. I think it likely that if that is the case, the republicans in the house would be unaware of it.

Steve Scolnik said...
This comment has been removed by the author.
Steve Scolnik said...

But the WaPo will give you all the info you need about 80-lb. parmigiana cheese wheels:
WaPo, WaPo, Wherefore Art Thou? Press Sleeps While Climate Report Fraud Festers

Hank Roberts said...

I'm looking for a good clear writing guide --- something to show news people what computer people are taught and expected to know about plagiarism -- best I've found so far is this bit I excerpted:
http://www.realclimate.org/index.php/archives/2010/10/cuccinelli-goes-fishing-again/comment-page-5/#comment-188297

Has anyone found clear guidance actually taught at GMU?

David B. Benson said...

Hank Roberts --- Strunk and White's Elements of Style remains top grade. Tarcher/Penguin sent me a free copy of Now Write! Nonfiction: Memoir, Journalism, and Creative Nonfiction Exercies From Today's Best Writers and Teachers.

Rattus Norvegicus said...

Hank,

Deep Climate had an example from the GMU website which provided guidance similar to that you cited on RC.

I tried the example and it wasn't easy to avoid plagiarizing. Probably easier to quote and cite. You could use it in a paper titled "Bears: Threat or Menace" (h/t to Colbert).

King of the Road said...

With respect to your desmog blog link re.: Wegman, I suppose you're aware of Mr. Fullers two part post on the subject at WUWT.

here and here

"Move along folks, nothing to see here."

Michael Tobis said...

As is normal, Fuller misses several key points, even in my casual reading of the Wegman story.

First of all, I've met John Mashey and take him very seriously.

Secondly, "The leaked emails clearly showed bad and bullying behaviour that left a stench in any honest reader’s nostrils." is true if and only if you believe that the people writing the denial papers are not charlatans or fools. It certainly leaves no such impression with me. But it's not relevant to Wegman.

Thirdly, the Schneider and Prall matter is totally irrelevant, and Fuller's position on it is bizarre.

Fourthly the words "carefully concocted hockey stick chart" beg the question. There is nothing concocted about it, as all subsequent work in the same area has pretty much confirmed the broad outlines of that result.

Fifthly, a quick glance at the Mashey piece shows that " The text is not identical, but to be an accurate summary would have to be similar." But Mashey's point is that the quotiations garble and lose meaning, i.e., that the summary was plagiarized in the same way that a weak student would plagiarize a term paper; lifting text without understanding it, in the hopes that the reader would not have the time and resources to recognize this. That, in this case, the grader was the US Congress is more than a little striking.

Michael Tobis said...

In the second Fuller piece, Fuller begins by setting up Bishop Hill's ludicrous straw man:


Wegman et al are guilty of plagiarism; short-centred principal components analysis is biased and can produce hockey sticks from red noise

Wegman et al are not guilty of plagiarism; short-centred principal components analysis is biased and can produce hockey sticks from red noise.


Clearly, even if these were the only two issues at stake there are two other possibilities.

But it has been shown time and again that the statistical flaws with MBH are not crucial to reaching the conclusions. The lack of collaboration between climatologists and statisticians is something that a great deal could and should be written about in the context of the institutional incapacity of American science to support interdisciplinarry collaboration, but this is just a tiny instance. Once better methods were applied, and better data were collected, the same results emerged, showing that the intuitions of the original researchers were sound. Twelve years have passed. A methodological flaw in a single 1998 paper is not a matter for a congressional investigation. This question has not deserved any attention.

I am very amused at the end to see Fuller hemming and hawing about the trivial nature of the actual transgressions involved. This is as may be, but what is good for the goose is good for the gander.

I have not followed the matter in enough detail to vouch for John's version. But I will with confidence stand up for John Mashey as a sane, intelligent, moderate, perceptive, competent and responsible human being. His accusations carry some considerable weight with me.

On the other hand I have followed the matter closely enough that it only takes a few minutes to see Fuller's position for what it is; mostly defensiveness, diversionary tactics and bilge.

Eli suggests that Wegman's university is having a very hard time with this. We shall see.

I have always believed that the Wegman report was biased, but I didn't expect it was so shabby a piece of work as to allow substantive accusations in the other direction. I can't say that the prospect makes me unhappy, though one hopes in future that people like Joe Barton keep their fingers out of judging science.

Alexander Ac said...

OT but interesting,

peak ET? Nature reports:

"Recent decline in the global land evapotranspiration trend due to limited moisture supply"

http://www.nature.com/nature/journal/vaop/ncurrent/full/nature09396.html

Alex

Michael Tobis said...

KOTR, I am afraid you yourself have been plagiarized.

Steve Bloom said...

Thanks, Alexander. AFAICT, having just seen summaries of both, last week's paper on increased river flow as an indicator of a global acceleration of the hydrological cycle would seem to fill in the observational gap noted in the first paper. As Michael has observed, our understanding of the climate tends to become more coherent as we learn more about it.

John Mashey said...

MT: thanks for the kind words!
Hopefully, my report stands on its own merits.

Many people have gotten confused about the plagiarism, and while cringing at the thought of another -gate, this is actually a better match than most.

The Watergate breakin by itself was only a small thing, but it was absolutely clear, and it ended up leading to bigger things.

This one is easier, but some people want to act like plagiarism is no big deal or there's nothing else.

1) As I wrote, 35 of 91 pages are mostly plagiarism, and every such page is shown either there, or via URLs to DC's.

2) W.11.1 cites 4 different sources on plagiarism, including both Rice and GMU, and even quotes GMU. I must admit amusement at the fanciful opinions I've seen in the last few days on what plagiarism is not. {Of course, I checked out more university sites, asked various academic experts, and at least one involved in court cases, i.e., the source of the "striking similarity" term.

3) There *are* 35 pages that are mostly obvious plagiarisms, although often mixed with errors of various kinds, some so ludicrous undergrad should be ashamed. People with experience take one look at this and say "yes," others express opinions far from reality.

4) But so what? Well, this was not presented as a term paper by an undergrad, it was presented to Congress, repeatedly an "independent,impartial, expert" report by a team of "eminent" statisticians. Scott only wrote the 3-page Appendix A, straightforward math.

More than half the report was thus almost certainly written by Yasmin said, less than a year post-PhD, plus some help from grad student(s). It is not even clear that Wegman read many of the papers Summarized or even the Summaries or Literature Review, as his testimony contradicts them. It is very unclear whether anyone did any serious proofreading of the Summmaries or Bibliography.

5) So, we have a 9-month effort constructed to mislead Congress. There is plenty of evidence that shows that it was neither independent nor impartial, and the real authorship simply shows it was not an eminent team. The plagiarism alone shows that it was not expert.

6) Hence, I think there is enough evidence for an investigation into 18USC1001, 18USC371 on the part of a bunch of people ... and that's pretty serious, up to 10 years. If that happens, 18USC4 comes into play, and a whole bunch of people who likely knew about it get to be asked hard questions and would be living in fear that the wrong emails appear. That's limited to 3 years. Finally, if 18USC1001 happens, and somebody decides certain files are evidence, deleting them in August may have been a mistake, 18USC1512, 1519. I'm no lawyer, and the law can be fuzzy, but I talked to a very savvy Washington-experienced lawyer about this, and even on what we knew in February, he explicitly said this wasn't fantasy.

Making it happen is something else, but people should not dismiss it out of hand as unthinkable.

So, plagiarism is merely one little piece, but it's just the easiest, like the breakin.

Even without that, many people have yet to realize how bad an idea it was to cite 3 government contracts on a paper that had nothing to do with those funders, contained clear plagiarism, and was junk anyway. HHS ORI has serious teeth, and messing with your biggest research funder is not a good idea for a university.

John Mashey said...

MT: thanks for the kind words.

Hopefully, my report stands on its own merits.

In any case, many people have gotten confused about the plagiarism, and while cringing at the though of another -gate, this is actually a better match than most.

The Watergate breakin by itself was only a small thing, but it was absolutely clear, and it ended up leading to bigger things.

This one is easier, but some people want to act like plagiarism is no big deal or there's nothing else.

1) As I wrote, 35 of 91 pages are mostly plagiarism, and every such page is shown either there, or via URLs to DC's.

2) W.11.1 cites 4 different sources on plagiarism, including both Rice and GMU, and even quotes GMU. I must admit amusement at the fanciful opinions I've seen in the last few days on what plagiarism is not. {Of course, I checked out more university sites, asked various academic experts, and at least one involved in court cases, i.e., the source of the "striking similarity" term.

3) There are 35 pages that are mostly obvious plagiarisms, although often mixed with errors of various kinds, some so ludicrous undergrad should be ashamed.

4) But so what? Well, this was not presented as a term paper by an undergrad, it was presented to Congress, repeatedly an "independent,impartial, expert" report by a team of "eminent" statisticians. Scott only wrote the 3-page Appendix A, straightforward math.

More than half the report was almost certainly written by Yasmin said, less than a year post-PhD, plus some help from grad student(s). It is not even clear that Wegman read many of the papers Summarized or even the Summaries or Literature Review, as his testimony contradicts them. It is very unclear whether anyone did any serious proofreading of the Summmaries or Bibliography.

5) So, we have a 9-month effort constructed to mislead Congress. The plagiarism alone shows that it was not expert. Of course, there is plenty of evidence that shows that it was neither independent nor impartial, and the real authorship simply shows it was not an eminent team.

6) Hence, I think there is enough evidence for an investigation into 18USC1001, 18USC371 on the part of a bunch of people ... and that's pretty serious, up toe 10 years. If that happens, 18USC4 comes into play, and a whole bunch of people who likely about it get to be asked questions. Well, that's limited to 3 years. Finally, if 18USC1001 happens, and somebody decides certain files are evidence, deleting them in August may have been a mistake, 18USC1512, 1519. I'm no lawyer, and the law can be fuzzy, but I talked to a very savvy Washington-experienced lawyer about this, and even on what we knew in February, he explicitly said this wasn't fantasy.

Making it happen is something else, but people should not dismiss it out of hand as unthinkable.

So, plagiarism is merely one little piece, but it's just the easiest, like the breakin.

Steve Bloom said...

Great work, John, and thanks.

Michael Tobis said...

Fuller: "I do not intend to present myself as an unbiased reporter of fact."

Cool. A point of agreement, then. I don't think Fuller is an unbiased reporter of fact either. Can we make it unanimous?

The rest of the discussion at Kloor's is revealing. I think NewYorkJ is on point when discussing what the comparison of text (the foolishly-maligned color coding) reveals:

"From what I’ve seen, Wegman didn’t simply copy and paste text without attribution. He made an attempt to cover it up by adding or swapping words, in a rather comical fashion that renders the text incoherent in some cases, and changing the meaning in other case. It’s forgivable with school children, not a professor who is dubbed “world-renowned” (albeit by political hacks). Even so, a child would receive an F on their paper and a severe scolding if caught."

Fuller misses this point entirely, but that is because Fuller has demonstrated a very limited capacity to read substantive text.

John Mashey said...

Just to remind people, I think it is virtually certain that Yasmin Said did the actual mechanics, given the plagiarism style seen in her PHD, and unlikelihood of an ASA Fellow actually doing this work. Of course, Wegman was responsible:
1) lead author
2) only senior person really involved.
3) and a 3rd reason I am not free to say.

For reasons given in SSWR, it seems unlikely that Wegman ever read the Summaries carefully.
There us certainly no evidence of careful proofreading, given the various gaffes.

There are 2 separate aspects:
A) the straightforward plagiarism, as in the SNA section, where the occasional gaffes ate obviously oversights.

B) the pervasive biases and meaning changes in the Bradley tree-ring section and the summaries.

All evidence points to Said for A).
The interesting question is who did the various elements of B). I suppose it is possible that a new PhD might have done this on her own, pulling the wool over Wegman's eyes.... But I think that is a pretty hard sell.

Ted Kirkpatrick said...

Michael:

"But it has been shown time and again that the statistical flaws with MBH are not crucial to reaching the conclusions."

And no such analysis was ever attempted by the Wegman team. This is John's most striking conclusion (p. 135):

"Despite all the graphics and technical discussion, the WR fails to do the one thing to be expected of statisticians criticizing a method‘s results: reanalyze with the 'right' method (centered PCA and correct number of PCs) and see if it makes any difference. The WR [Wegman Report] contains no actual new statistical analysis of MBH itself, just reworks of MM material." (bold italics in the original, denoting an opinion held by Mashey)

All the furor over plagiarism is justified but it's obscuring the more important point: The Wegman committee does not appear to have even attempted to fulfill the mandate given them by Congress. The slipshod scholarship is not a sign of a rushed job, it is instead a sign that the actual job was quite different from what the public was told.

The plagiarism isn't the real issue. It's just pointing us to a much bigger problem.

dhogaza said...

"Fuller: "I do not intend to present myself as an unbiased reporter of fact."

Cool. A point of agreement, then. I don't think Fuller is an unbiased reporter of fact either. Can we make it unanimous?"

Except, of course, he does present himself as an unbiased reporter of fact, for instance in his role as co-author with Mosher of the hit piece on Climategate.

But let's keep this little gem around, just like his statement regarding men of honor not reading other people's stolen e-mails, made shortly before embarking on the aforementioned book project about climategate based on a dishonorable midreading of stolen e-mails ...

Neven said...

The plagiarism isn't the real issue. It's just pointing us to a much bigger problem.

That's exactly it. I'm trying to get that point across on some of the lukewarmer blogs.

Marco said...

Ted, you are touching on an interesting point here, as the Wegman report lists Von Storch & Zorita as well as Huybers (2005), both comments on M&M in the bibliography. Neither is discussed in the Wegman report. Methodological flaws in M&M found, articles known to Wegman et al, and yet no comment in the report itself. Interesting, no?

Hank Roberts said...

Wait -- when did Fuller quit claiming to act as a journalist?

thingsbreak:
"Tom Fuller | September 11, 2010 at 12:55 pm Yes. And as a journalist, I look for trusted sources with expertise that exceeds mine...."

RC:
"Tom Fuller says: 26 June 2009 at 12:07 PM ... as a journalist I’m trying to be fair to both sides ...."

Tom said...

Thought better of you, Tobis. Expected that crap from dhosed, but thought better of you. Okay, you're both the same.

Michael Tobis said...

Huh. Tom, "better" means what?

I think you are inadvertently toxic, and most of my friends think you are deliberately toxic. I think you are lazy and irresponsible, and have thought so ever since I've known about you. Others think you are evil. If that makes me your friend you have problems.

So here I am noting your latest irresponsibility. I saw nothing you said that was convincing, and a lot you said that needed shooting down. I did my part. What else would you have me do?

If you have anything to say in your defense besides that I hurt your feelings, please do so.

dhogaza said...

"Thought better of you, Tobis. Expected that crap from dhosed, but thought better of you. Okay, you're both the same."

This is Tom's response to Michael's repeating a direct quote he's made?

Ted Kirkpatrick said...

Marco: the von Storch reference is an excellent example.

This is the point we need to be making constantly. The poor scholarship doesn't show that the report was a sincere attempt that was executed carelessly, it instead shows that they weren't even trying to understand the material they listed. They only wanted the appearance of a summary that seemed to support their claims about MM and social networks. Appearances were good enough.

Michael Tobis said...

Ted, that seems like a good way to put it. But that is a difficult story for the press to tell, and I fear they won't bother.

Florifulgurator said...

So what was the bibliography for? I guess not to list their sources. Perhaps those unused papers are a Potemkin village?

Hey, the bibliography is the second thing I look at after abstract/intro. (So, Wegman has some serious stuff in his bibliography, which hints his report might also contain serious stuff.)

Do I smell a rat?

John Mashey said...

Well, I did think of using a Potemkin village metaphor, but then decided it might be obscure. At one point, I thought of a graphic that had a wall of blue/green, but wit ha few red bricks on top, but never got around to it.