One of the Mann-Steyn lawsuit claims hits a rock

Steve McIntyre writes:

The Mann libel case has been attracting increasing commentary, including from people outside the climate community. Integral to Mann’s litigation are representations that he was “investigated” by 6-9 investigations, all of which supposedly gave him “exonerations” on wide-ranging counts, including “scientific misconduct”, “fraud”, “academic fraud”, “data falsification”, “statistical manipulation”, “manipulation of data” and even supposed findings that his work was “properly conducted an fairly presented”. Mann also represented that these investigations were widely covered in international and national media and thus known to Steyn and the other defendants.

In today’s post, I’ll look closely at the Oxburgh panel, one of the investigations cited in Mann’s pleadings. However, contrary to the claims in Mann’s litigation, not only did the Oxburgh panel not exonerate Mann, at their press conference, Oxburgh panelist David Hand, then President of the Royal Statistical Society, made very disparaging and critical comments about Mann’s work, describing it as based on “inappropriate” statistics that led to “exaggerated” results. These comments were widely reported in international media, later covered in a CEI article that, in turn, was reported by National Review. Moreover, information obtained from FOI in the UK a couple of years ago shows that Mann objected vehemently to criticism from an Oxburgh panelist, which he characterized as a “rogue opinion” and unsuccessfully sought a public apology.

Mann’s claim that the Oxburgh panel “exonerated” Mann on counts ranging from scientific misconduct to statistical manipulation to proper conduct and fair presentation of results has no more validity than his claim to have been awarded a Nobel prize for his supposedly seminal work “document[ing] the steady rise in surface temperatures during the 20th Century and the steep increase in measured temperatures since the 1950s.”

Read it all here:

Mann and the Oxburgh Panel

Get notified when a new post is published.
Subscribe today!
0 0 votes
Article Rating
139 Comments
Inline Feedbacks
View all comments
RomanM
February 18, 2014 12:45 pm

My comment seems to have disappeared into cyber-space (twice)…
[nothing pending in hold. mod]

February 18, 2014 12:48 pm

Gail Combs says February 18, 2014 at 6:27 am

It is too late we lost our free press in 1915 when …

You are too easily ‘bought’ when it comes to this kind of thing, “Ida” …
.

Snowshoe
February 18, 2014 12:53 pm

M.Mann to participate in an AMA at Reddit /r/science. February 21 1300 EST
Your thoughts will be moderated, don’t even think about posting….sad.

February 18, 2014 12:58 pm

iai mitchell says February 18, 2014 at 11:02 am
The proxy data has, without …

A “cheerleader” (literally: ‘a pretty face and an empty head’) and not a principal on this issue; your word therefore (in any kind of summary judgement or evaluation) means less than nothing, and is, as a matter of fact, a net ‘drain’ on the overall discussion of this issue (one on this side of the issue can quite figuratively feel one’s brain ‘leaking out one’s head and onto the floor’ having to ‘read’ your posts).
He!!, you can’t even be trusted to properly present, characterize or frame ‘your’ side of the argument!
.

RomanM
February 18, 2014 1:27 pm

jai mitchell says:
February 18, 2014 at 11:02 am

The proxy record is very clear, the instrumental record, when compared to the proxy record is devastating.

http://climateaudit.org/2013/03/13/marcott-mystery-1/#comment-404356
Unfortunately, Marcott himself admits that that particular portion of the reconstruction is “not robust” so any comparison of the two is spurious.
As far as the handle of his stick is concerned, you might also look here:
http://climateaudit.org/2013/04/04/marcott-monte-carlo/
Do you agree with my evaluation that the improperly applied methodology of the Marcott paper produces substantially underestimated error bounds is incorrect?
If not, perhaps you could indicate why I am wrong… 🙂

RomanM
February 18, 2014 1:30 pm

Oops, rewrote the end of the comment and screwed it up. It should read:
Do you agree with my evaluation that the improperly applied methodology of the Marcott paper produces substantially underestimated error bounds?

papiertigre
February 18, 2014 1:34 pm

Every American who cares about national security must.demand Kerry’s resignation. A delusional secretary of state is dangerous to our safety.
— Newt Gingrich (@newtgingrich) February 18, 2014
Gingrich calls for Sec Kerry to resign over global warming remarks
Whereas jai mitchell’s “devastating link” to Marcotte et al is pay walled into obscurity, Nature provides a link which is truly devastating to all of the sundry paleo tree studies.
Trees maintain their own internal temperature.
No paywall. Out in the open.
Measuring climate with a tree rings is akin to measuring room temperature with cats by rectal thermometer.

February 18, 2014 1:42 pm

The Emperor’s New Clothes:
http://www.itsnotclimatescience.com/0014.html

Joe Chang
February 18, 2014 1:43 pm

People, thanks for answering my questions. I agree that the Steyn accusation of fraud could be problematic. Serious error do occur periodically in scientific publications. An error however serious does not imply fraud. One possible explanation is general incompetence and there may be sufficient evidence to support such an assertion though I doubt the plaintiff would use this argument. I would suggest arguing that an incompetent but honest person could made any number of errors of no particular orientation. But for a series of errors all aligned to a specific objective should be sufficient to support a preponderance of evidence toward fraud and not incompetence.
On the McIntyre website, I asked a question that was not answered, needing a legal opinion. “Would it be necessary at some to provide evidence of damages? I would imagine that damages to a university professor’s reputation would include his ability to publish papers and get grants? has this been harmed? Also, given that the attack was made by a non-academic, would it carry weight within academic circles? Do other professors (or climate researcher now) avoid his company, or does he not get invited to parties since publication of the defamatory article?”

asybot
February 18, 2014 2:02 pm

Gail. 5.42,
I followed both links. I was a skeptic re the NWO then sitting on the fence for a few years , but I am truly scared after reading those, FYI when I tried getting on the roendaneducation .org I got the now more frightening 5-10 second delay before it connected and that happens only when I go to conservative sites. Adding to that what was said that the GOP in Washington is actually not standing in the way of the IRS etc to stop the Teaparty from various exemptions only adds to that (Pat Caddel on fox news I think it was Sunday’s FOX Report also talked about on Limbaugh on Monday).

Keith Sketchley
February 18, 2014 3:46 pm

Gail, some bullies go away when their victims fight back, others only go away when they are thrashed soundly enough often enough. Mann is the latter. Probably egged on by other alarmists, and maybe funded by some.
Some people don’t learn to stop and think, unfortunately Steyn fell into that category when the columnist he quoted retracted the ridiculous analogy but Steyn did not retract his reference to it. Steyn deserves huge praise for fighting the anti-speech hate laws in Canada, but he seems to have gotten too full of himself.

knr
February 18, 2014 4:18 pm

jai mitchell one small problem , the reviews themselves made it clear they did not look at the science, therefore than cannot be used to validate nor refute that which they did not consider .
Besides you can still be wrong and rubbish without being corrupt , with Mann who knows which two out these three three he was given he is always trying to achieve 3 out 3 .

ttfn
February 18, 2014 4:40 pm

Keith Sketchley says:
February 18, 2014 at 3:46 pm
“Steyn fell into that category when the columnist he quoted retracted the ridiculous analogy but Steyn did not retract his reference to it.”
Hasn’t the judge already ruled that the Sandusky/Mann analogy was protected free speech and that only Simberg’s accusation of Mann “manipulating data” and Steyn’s ‘fraudulent hockey-stick’ are potentially libelous? Isn’t that kind of ironic? The part that gets all the alarmists panties in a twist is the part that is “obviously protected free speech”. Steyn’s going to win. He knows he’ll either win the trial or win on appeal. The Supreme Court is not about to toss 200 years of first ammendment case law to protect Mann’s feelings. The ACLU and media attorneys will plead Steyn’s case before his lawyer even rises to address the court. If Mann takes this thing to trial, he’s going to have to get on the stand and answer hostile questions for days on end that will be discussed on the internet for decades. Mann’s climatologist friends will also get sucked into the fray. He’ll never do it. I predict Mann will offer to pay Steyn off to avoid going to trial after his attempts at delaying discovery fail (Styen will simply say, okay, we’ll do it with just the climategate emails). I also predict that Steyn won’t let Mann off the hook that easily.

HowSmart
February 18, 2014 6:36 pm

As long as there is one child on the planet who lacks food, shelter, education, medicine, clean water, the expenditure of even $1.00 on “climate change” or ‘global warming” must be seen as a crime against humanity.

hswiseman
February 18, 2014 6:49 pm

Maybe Mann will disclose and explain the “dirty laundry” in open court? The secret residuals probably make the fraud claim a slam dunk.

Wikus
February 18, 2014 7:12 pm

Joe Chang:
February 18, 2014 at 1:43 pm
Even if Mann honestly erred, the way that the hockey stick kept being promoted after its shortcomings had come to light certainly makes it “fraudulent”.

asybot
February 18, 2014 7:50 pm

@HowSmart, Talk about one sore nail head!

February 18, 2014 8:55 pm

RomanM says: February 18, 2014 at 12:12 pm
“jai mitchell says:
So are you saying that my evaluation that the improperly applied methodology of the Marcott paper produces substantially underestimated error bounds is wrong?
Perhaps you could indicate why I am wrong… :)”

It’s done here.

bushbunny
February 18, 2014 9:36 pm

I suspect libel and slander defamation suits are different around the world. But in Australia if a paper prints what can be sued on the grounds it is untrue, then the plaintiff can get some compensation. But is Mann a ‘real’ scientist? And more importantly has he made a substantial money from his dodgy research and data. There is something in academic circles that is called ‘Corrupting the data to suit the hypothesis’. When an academic gives a talk and lecture, and says ‘we haven’t compiled enough data to prove this’, the response is ‘why, are you just speculating at this point? Sounds to me like the dark ages, when persons presented information that went against the catholic church dogma, they got burned at the stake, and all their papers burned, as did the Nazis of course. Luckily this is America, the land of the free and not North Korea or China.

RomanM
February 19, 2014 4:56 am

This thread is not the place to continue this somewhat off-topic discussion, but the point should be made that the post at your link does not address the meat of the issues with the Marcott paper.
The methodology of the paper was intended to address all sources of possible error inherent in the reconstruction. These include the uncertainty as to how the proxy values relate to temperature, dating errors and the smoothing due to the interpolation to a 20 year grid of data measured at longer intervals. Monte carlo techniques using estimates of the variabilities of the first two components mentioned were applied to create a multitude of individual “reconstructions” which were then used to calculate a single reconstruction along with error bounds.
In my analysis I pointed out that the estimate of the variability of the proxy value to temperature relationship omitted the component due to the property that in a regression relating the two variables, even when the coefficients of the regression equation are known exactly, two proxies formed at the same temperature will generally differ from each other. Marcott estimated only the relatively minor uncertainty due to estimation of the coefficients and ignored the substantially greater one due to the fact that the relationship between the two variables is far from linearly exact.
Your link does not address any of these issues. You calculate a reconstruction using another technique (which can’t possibly include the uncertainties the monte carlo analysis attempts to address) and declare that the results look the same. In your post, you state:

In recent days there have been two issues raised at Climate Audit regarding the confidence intervals shown in the proxy reconstruction of Marcott et al. In the first, Romanm claimed that there was a major term missing from the confidence intervals, which could only be remedied by going back to the original calibration experiments and importing the variation of residuals as uncertainties in Marcott’s analysis.

Unfortunately, this shows a lack of understanding of regression models and how they are applied since the “variation of the residuals” is exactly what measures how much a proxy value may vary for a given fixed temperature value and it is the central part of the overall uncertainty structure of the proxies. Nick, I have been a statistics professor for almost forty six years. Would you not think that I might have learned something about simple regression during all that time? Enough with this side issue.

tom
February 19, 2014 7:03 am

As usual nick is only interested in trolling outside if his own blog.
Nick if you thought you had a good argument post it in the discussion at CA. Running off to yout own blog where yoi can control the debate is cowardly. You are a coward. Post your arguments at CA dont run off and post them at your own unread blog and act like romman has been refuted. That is tje same as running off and talking behind someones back. That is oure cowardice.

RomanM
February 19, 2014 7:26 am

I would not criticize Nick for doing a post on his own blog. It requires a lot of technical control to put together a complete analysis along with plots so this sort of material cannot be entered as comments on someone else post. Besides, he did put a link to it on my original post on CA:
http://climateaudit.org/2013/04/04/marcott-monte-carlo/#comment-412594
At the time I had arguing with him for over a week without success and didn’t feel like doing so further so I ignored it.
If nothing else, Nick is an honest guy – hard-headed and won’t admit being wrong 😉 – but certainly not sneaky in his behavior.

kim
February 19, 2014 8:08 am

Congrats to Sherry Moore, who’s been awarded the Purple Zamboni Medal for valor in battle.
=================

kim
February 19, 2014 8:09 am

A peculiar honesty, which cannot admit being wrong. Does he understand when he’s wrong? I think so.
============

jai mitchell
February 19, 2014 12:33 pm

RomanM
your analysis is focused entirely on .2% to .5% of the proxy record, over a period that is already covered by the instrument record.
Therefore, the entire concern that you have with the marcott record is a nit-pick and red-herring. It is not a valid criticism and does nothing to impact the overall gravity of the work. The curve verifies the Mann 1998 curve and the projections of future warming of 2-4C on a global scale within the next 85 years or so yields the following curve.
so remember, even if the .2% to .5% is not robust, that means that 99.5% to 99.8% of the proxy record IS,/b> robust!