Since I’ve made an appearance in a blog post of Judith’s about the road to scientific hell being paved with good intentions, I thought I might make a brief comment. It relates to a Twitter discussion I had with Judith about research integrity and whether to speak out or not
@curryja @gregladen @MichaelEMann I largely disagree. Your job is to have personal research integrity. Implying a lack in others, (1/2)
— There's Physics (@theresphysics) July 20, 2014
@curryja @gregladen @MichaelEMann suggests that you think you are somehow purer than others. I'd call that sitting on a high-horse (2/2).
— There's Physics (@theresphysics) July 20, 2014
Judith’s response in her blog post is
Because I talk about research integrity and try to defend it and point out problems when I see them, I am somehow dismissed as trying to present myself as purer than others.
Well, yes, precisely. What else does Judith expect? This isn’t some kind of debate about whether a particular scientific result is correct or not; this is a discussion in which Judith is implying nefarious behaviour in others. Accusing others of having questionable research integrity is an incredibly serious accusation. It’s not something someone should do lightly. It’s something for which – in my opinion – you should have almost watertight evidence. You don’t make such accusations just because you disagree with someone or because you dislike them. I don’t see why this isn’t patently obvious.
Judith also seems to think she should be applauded for highlighting possible improper behaviour. If by some chance Judith turns out to be correct, then this may well be what she would deserve. On the other hand, if Judith is wrong, then she would deserve no such thing. You don’t get applauded for making serious accusations that turn out to be wrong. If anything, making potentially career damaging accusations against others that turn out to be incorrect could – in itself – be career damaging.
I’ll make a few other comments. I do find those who pontificate about research integrity incredibly irritating. To me that we should behave with integrity is self-evidently true. I don’t explicitly say that I plan to behave with integrity, because I regard it as clearly my intention and would hope that this would be obvious. I have an immediate suspicion of those who choose to highlight research integrity. Additionally, those who do choose to highlight research integrity often do so in a way that implies that it’s obviously true of them, but not necessarily of others. Who made them the arbiters of scientific best practice? There’s a reason we often use the term research community. We don’t have those who police the behaviour of others.
To be clear, if someone is aware of genuine scientific malpractice, then speaking out is entirely the right thing to do and it should be applauded. However, advocating for some particular policy is not – in itself – evidence of research misconduct. An error in a paper, is not evidence for research misconduct. Someone being rude to you on Twitter is not evidence of research misconduct. Even refusing to acknowledge an error in a paper is not research misconduct. Of course, everyone should aim to be honest and act with integrity, but disagreements are unavoidable and just because you believe someone is wrong, doesn’t make them so.
Anyway, those are some of my thoughts on the whole research integrity issue. In my opinion, accusations of research misconduct should be made carefully and only when the evidence is compelling. Throwing such accusations around just because one works in a field where there are some strong disagreements is neither constructive nor – in my opinion – particularly good scientific practice. Of course, it’s a free world and if this is how someone chooses to behave, they’re largely free to do so. However, it’s not something for which they’re likely to be thanked, unless they happen to end up being correct.
I’ll end by giving my answer to the question Judith posits as the title to her post
Is the road to scientific hell paved with good moral intentions?
Yes, I rather think it is.
...and Then There's Physics 
Hmm maybe its time…
Anthony,
You do whatever you think is right, oh hold on……
This is tiresome. Not your post, but the discussion which led to it. For two reasons: 1) You can’t really be held responsible for someone else’s research and 2) There seems to be a little too much time spent on judging the integrity of others and not enough on the actual science.
Something I thought I might add to the post, but didn’t, was the relationship between advocacy and research integrity. It seems clear that simply because a scientist chooses to advocate, is not evidence for a lack of research integrity. Some may perceive it this way, but that doesn’t make it so. Research integrity applies only to your actual research. Arguing that scientists should not advocate because it may appear to others that they lack research integrity seems – to me at least – to be a suggestion that scientists shouldn’t always be honest. Basing what you say on how it might be perceived doesn’t appear to be a particularly honest strategy. I’d rather people said what they believed, and could be judged on that basis, than people decided not to say something simply so as to present an impression that may not be strictly true.
Rachel,
I agree. As I said to Judith on Twitter, you’re only actually responsible for your own integrity. Of course, if you do become aware of genuine research misconduct, speaking out would be the right thing to do. However, going around looking for signs of research misconduct is neither expected nor a particularly good practice.
==> “Hmm maybe its time”
Wow. So Anthony is finally going to reveal the identity of “Just the facts?”
I can’t wait. It’s about time that someone stood up to Just the facts and made his real name known. Because, you know, Anthony’s dedication to ethics dictates that he must absolutely flush out the anonymous cowards!!!
I mean think of integrity. Think of the starving children. And think of RealClimate moderation policies!!!11!!!1!!!!
“However, going around looking for signs of research misconduct is neither expected nor a particularly good practice.”
No, but insinuating research misconduct in those you disagree with can be politically expedient.
fyi You came up again here again in a post by Carrick:
This would make sense. Neither There’s Physics (“Anders”) nor Michael E Mann have ever demonstrated that they have a moral compass. In fact Anders seems to have suggest (in context with the ethical transgressions in Cook’s paper), that he (Anders) completely lacks one.
So it would makes sense that, rather than seeing a call for more ethical behavior on the part of scientists for what it is, both Anders and Mann would interpret it as moralizing. Scientific sociopathic behavior at its finest.
I can deal with a lack of competency a lot easier than I can people who themselves are so heavily invested in a particular socio-political outcome that they are completely morally and ethically compromised by their advocacy.
Physics doesn’t have a socio-political outcome.
Joseph,
Interesting, thanks. When I wrote this
and this
Carrick was one of the names that immediately sprang to mind. Good to see that he’s chosen not to disappoint.
But here’s the theme tune for the nonsense…
The only way to ensure research integrity is to establish a responsible regulatory body with enough non-scientists to avoid group-think, who are experienced with establishing rules and procedures, (lawyers perhaps). That body can then define a set of research methods that are permitted and protect the integrity of the research by delineating any proscribed behaviour. To ensure compliance all research, from planning to [publication will have to be fully recorded to allow independent auditing.
(sarc/off)
In practise there are two main causes of any lack of integrity in research. the personal, dishonest individual. Where research is anything from fraudulent, to merely over-optimistic and spread thinly across more papers than is justifiable. This is most often motivated by personal ambition. The competitive nature of research is embraced by a few who use, and abuse the research to further their careers and reputation. At the personal level research integrity is very rarely compromised to promote an external ideology or belief, it is nearly always for personal gain.
Research integrity can also be compromised by systemic factors in the field. The most obvious example present example is in the medical field where the preponderance of drug company R&D distorts things towards patentable chemical treatments for common or chronic conditions. There is also a problem with the skewing of published research from the omission of negative results and the constraints on publishing full methods because of commercial sensitivity.
Without real examples of personal dishonesty or a credible motive and means for biasing the money that funds research in a field, raising the issue of research integrity while discussing research you wish to disparage is dog whistle politics and a grubby rhetorical device.
ATTP – when I saw Anthony’s comment here and your response, all I could think of was how we deal with an old lady who keeps leaving threatening messages on our home phone. She calls about 10 times a week, usually in the middle of the night (our phones are where they won’t disturb our sleep). She has a number of conspiracy theories – the main one lately seems to be that somehow I arranged for people to use lasers in their cell phones to burn her leg. We don’t have the heart to involve the police – as far as we can tell this is all completely harmless, just a woman in her 70s with a mind not working properly. Anyway – “you do what you think is right” is exactly how I answered her last time we talked in person. Not much else you can do with the nutters of the world 🙂
In the future, government funded research integrity should be measured by units of ‘value’ and that value should not necessarily be measure by ‘dollars’.
I might have more respect for Curry, or indeed any at all, if she actually came out and made it clear exactly which actions by which scientists she considers unethical. Instead she makes insinuations on her blog but when challenged to name names or provide evidence she either comes up with some vague waffle or just doesn’t respond at all. I have no doubts about her integrity as it applies to her research activities but I’m not sure about how far that extends to her wider conduct.
Anders
I don’t suppose you want my opinion, but if you’re willing to publish it – here it is.
Those of us who debate anonymously on the internet have to tread a fine line.
If, like me, you’re just a name, with no identifiable back-story, assumed expertise or authority – your position is just that of the common man in the street, chucking rotten tomatoes at those who irritate him.
This sort of rough & tumble anonymous barracking of authority has been around for most of recorded human history.
You, however, have chosen a rather different path. You have carefully crafted a convincing back-story that you are a practising academic at a prestigious university with expertise in the areas where you choose to debate.
That’s fine, we can all choose to believe your story or not and weigh what you say according to our various prejudices.
It changes though when you involve yourself in highly personal disputes with known, named, fellow academics like Curry on issues which touch on honesty & integrity.
Basically, she has to risk her credibility and reputation when she debates with you – you risk nothing.
I think, if you want to get into ethics wars with opposing academics – you owe it to them to take Anthony’s advice and reveal who you really are.
FG,
Not strictly true, I have experience in physics which is relevant. No direct experience in climate science.
Well, Judith was happy to post our exchange, so I assume she doesn’t have an issue. Also, this isn’t really a personal dispute. I happen to think that one shouldn’t throw around accusations of research misconduct without very strong evidence. I’m not accusing Judith of doing anything that isn’t fairly obviously what she is openly doing and I’m not attributing any intent as to why she chooses to do this. I don’t have any personal issues with Judith, I just happen to disagree with her about many aspects of this topic. To be clear, I’m not accusing Judith of being dishonest, I’m suggesting that it might be better if she refrained from implying that about others (unless she was extremely certain).
As I say in the post, if she turns out the be right, she may well deserve plaudits. If not, then that would seem unwarranted.
This is still a blogosphere issue. Judith can take what I say with as big – or as small – a pinch of salt as she would like. If anything, I would be quite pleased if Judith chose to engage more constructively with those with whom she has in the past disagreed. Whether she choose to do so or not, is entirely up to her.
FWIW, I kind of agree with Foxgoose. At least he can’t say this blog is an echo chamber as he usually does.
Rachel,
To be fair, I can kind of see the point, but I need some sleep 🙂
It’s a bit different for me because I’ve met you and I know who you are and that was important for me. I know you’re not a dog. But no one else can be sure.
Why is all this meta-rationalization regarding scientific integrity, etc going on right now?
Probably because of this firestorm:
“National Review declares war against the nerds”
Guess who lost the war, but think they can still do battle?
Rachel, as a general principle, if you agree with Foxgoose, you’re wrong.
More specifically, Anders is not making any accusations. He is stating a general principle. Stating general principles on ethics anonymously is no different to discussing the validity of scientific theories anonymously. His comments should be assessed on their validity, and the reasoning and arguments provided, and accepted (or not) on that basis. It would be different if he were directly accusing Curry of lacking research integrity – something he explicitly disavows.
I will note that Foxgoose is being thoroughly hypocritical here in that he does repeatedly make accusations of academic misconduct against people whose careers could be potentially damaged by such accusations; and does not willingly abandon his anonymity on the internet when he does so.
Tom Curtis says:
August 4, 2014 at 10:18 pm
I will note that Foxgoose is being thoroughly hypocritical here in that he does repeatedly make accusations of academic misconduct against people whose careers could be potentially damaged by such accusations; and does not willingly abandon his anonymity on the internet when he does so.
Tom
As usual, you’ve got carried away with your own pomposity and magnificently missed the point.
Which was – that there is an important difference between anonymous criticism from a member of “the common herd” and the same criticism from someone who has partially outed themselves to build up their authority.
If that difference is too subtle for you I’m sorry – I can’t help you further.
“Is the road to scientific hell paved with good moral intentions?”
In Judy’s case, I’d say it’s paved with something else altogether. FG says Judy risks much by what she does, but in fact she risks little beyond reputation since her career is drawing to an end. As well, he elides that Judy is out in public doing what she does as a means of promoting her profit-making enterprise. IOW she *wants* to benefit from notoriety, something one doesn’t get by acting the part of a happy scientific cog.
Judy is a special case because of the financial benefit to her of damaging her own scientific reputation, i.e. it’s impossible to tell from a distance what her thoughts really are, but with Carrick and many others like him I think it boils down to having a very strong yet incorrect physical intuition about climate. Those who disagree must be judged either incompetent or unethical or both.
Re Carrick my favorite moment involving him was some months back, IIRC on Eli’s blog, when he sneeringly attacked Harold Brooks after the latter dared to dispute his wisdom on some aspects of tornado damage. Oops. And whaddayaknow, Carrick just vanished upon being informed (not by HB) as to HB’s credentials, not even pausng to apologize.
“Well, yes, precisely. What else does Judith expect? This isn’t some kind of debate about whether a particular scientific result is correct or not; this is a discussion in which Judith is implying nefarious behaviour in others. Accusing others of having questionable research integrity is an incredibly serious accusation. It’s not something someone should do lightly. It’s something for which – in my opinion – you should have almost watertight evidence. ”
I wonder if you hold yourself to the same standard of air tight evidence.
Here is a good test. Go back over all your posts and comments where you merely implied
that others were engaged in nefarious behavior. Ask yourself if you had air tight evidence.
I’ll wait. It doesnt take long to find examples.
or We can cue up the behavior around Ar4 chapter 6 .. or not.
But nobody wants to re hash climategate.
Two issues: Judith’s ideal of integrity is more strict than others who are ok with boys being boys.
Second, her standard of proof is lower than yours.
Watertight evidence is an interesting concept. Perhaps one should apply it to climate science rather than accusations of nefarious behavior.
“To be clear, if someone is aware of genuine scientific malpractice, then speaking out is entirely the right thing to do and it should be applauded. However, advocating for some particular policy is not – in itself – evidence of research misconduct. An error in a paper, is not evidence for research misconduct. Someone being rude to you on Twitter is not evidence of research misconduct. Even refusing to acknowledge an error in a paper is not research misconduct. Of course, everyone should aim to be honest and act with integrity, but disagreements are unavoidable and just because you believe someone is wrong, doesn’t make them so.”
Mistake number one is conflating lack of integrity with misconduct.
yes, misconduct is one way to lack integrity. Its not the only way.
So, you will find Judith is critical of certain behavior ( for example advocating) that falls short of misconduct, but still is related to the issue of integrity. That same would go for some of the conduct she saw in climategate. It didnt rise to the level of misconduct, but she felt that the integrity of science was at stake.
The way you’ve made your case here is by assuming that her arguments for integrity imply misconduct. They dont.
If you want to know what she means by integrity, read harder or ask her.
That way you could get some air tight evidence..
you know what that looks like right?
Nice tries, mosh. (OK, not really. But knowing Anders he’ll happily reply at great length.)
Even if Anders were making an accusation, “accusations” stand or fall on their own merits. The anonymity or lack thereof of the acccuser is irrelevant. FG’s argument amounts to an ad hom. He is using Anders’ anonymity as a cloak under which to attack Anders.
Allow me to quote my friend FG.
Now would the quality or value or veraciity of FG’s comment to me change one iota if we knew his real name? He hides behind a rhetorical ploy to call me pompous and fatuous and bereft of moral sense. Would his reasoning or moral standing change one iota if we knew his real name?
If I am not mistaken, carick’s last name is known on the Web – but would his pronouncements on Anders’ “moral compass” be materially affected one way or the other by whether I or anyone else have seen his last name or not? Of course not. I can see that Carick has climbed on his moral high horse whether or not I know his last name. The same is true for anyone else. I can evaluate Carick’s argument in support of his sense of moral superiority independently of whether i know his last name. Knowing his last name is irrelevant to my assessment of the strengtb of his reasoning. Would doing a search to find his last name make his characterization of Anders’ modality less vacuous?
Mosher:
1) If you want to make slurs about Ander’s consistency, first establish that he has accused anybody else of lack of academic integrity.
2) If Curry means “research integrity” in some vague general sense that apparently includes matters not (in theory) related to research, she needs first to establish that clear community standards exist as the basis of her accusation. Otherwise not only is she on a high horse, but she is describing behaviour that may be perfectly acceptable in abusive terms.
3) If Curry means “research integrity” in some vague sense that considers advocacy, sloppy research, or insufficient attempts to ensure your theory is supported by the evidence, then her pontificating about integrity takes on the appearance of a displacement device to draw attention away from her own poor practise.
Because it was cobblers the first time round.
Presumably ‘research integrity’ involves fair reporting of evidence related to research. I just had a quick look at the ATTP posts on Judith Curry’s Senate testimony and one striking point seems to have been left out of the discussions (I may have missed it in skimming and it may well have covered elsewhere of course).
In her testimony Curry says:
The IPCC AR5 notes the lack of surface warming since 1998:
“[T]he rate of warming over the past 15 years (1998–2012) [is] 0.05 [–0.05 to +0.15] °C per decade which is smaller than the rate calculated since 1951 (1951–2012) [of] 0.12 [0.08 to 0.14] °C per decade.” (AR5 SPM)
By contrast the full paragraph in the AR5 WG1 SPM p5 actually reads:
In addition to robust multi-decadal warming, global mean surface temperature exhibits substantial decadal and interannual variability (see Figure SPM.1). Due to natural variability, trends based on short records are very sensitive to the beginning and end dates and do not in general reflect long-term climate trends. As one example, the rate of warming over the past 15 years (1998–2012; 0.05 [–0.05 to 0.15] °C per decade), which begins with a strong El Niño, is smaller than the rate calculated since 1951 (1951–2012; 0.12 [0.08 to 0.14] °C per decade)5. {2.4}
Curry’s quotation not only omitted the crucial context of her AR5 quote but also fundamentally misrepresented the point of the IPCC’s “past 15 years” example, which was to illustrate that short term records based on potentially cherry picked start and end dates are likely to be highly misleading. Instead in supporting a short-term hiatus argument Curry then proceeded to make exactly the mistake the IPCC was cautioning against in the very quote she used. One has to ask why.
The most striking point here for ‘research integrity’ is that she entirely failed to make the crucial context and meaning of the quote clear to the US Senate. Curry may well be a competent climate scientist but I fail to see how this example supports her claim “to protect research integrity”. In public testimony regarding the IPCC report, making the exact mistake the IPCC warn against and also failing to fairly report the IPCC’s warning about the dangers of cherry picking short records (such as her own 15 year “hiatus” claim) looks a lot like misleading the government she claims to have a duty to as part of her job.
As Curry likes to have these problems pointed out, I’d be interested in her response.
Haha. That’s probably true 🙂
I do agree with AT’s post here so perhaps I’m not entirely in agreement with Foxgoose. My issue with anonymity is this: I am not a scientist and I haven’t spent years in the field taking observations and conducting experiments but I know that scientists do. I therefore trust that they have some expert knowledge to share in the same way I trust my dentist when she tells me I should be using a full-strength, fluoride-containing toothpaste on my kids’ teeth – something I have always done because this is what dentists recommend.
When I visit blogs about climate change, the first thing I look for is who is the author and why should I trust what they have to say? I do trust AT – partly because I’ve been following his blog for quite a while now and he’s seems to know what he’s talking about and acknowledges his own mistakes when he makes them. But a big part of this trust is also that I know who is he. And he is a working physicist with published papers.
This stems back to a question Pekka asked in a recent thread which I wanted to respond to but haven’t had the time lately. Here’s what he said:
One answer to this is to ask an expert in the field. I know people will say I’m appealing to authority here but this is why we have specialists. We can’t all know everything about everything.
There’s so much rubbish on the web. How do we know what to believe and what not to believe? I once took a class of high school students and explored this very topic with them. There are lots of ways to establish credibility of information on the web and one of them is to see who the author of the informatino is. Because of this, I’m much more likely to trust a practising physicist’s opinion on say climate sensitivity than the opinion of a solicitor or a TV weatherman.
> Here is a good test. Go back over all your posts and comments where you merely implied
Moshpit pulls a Joshua on AT just after telling the former:
But he did not really ask a question, of course.
Mosher,
If a scientist discovers something bad will happen if humans continue to do X, then I think they are failing in their duty to society at large by *not* advocating.
Academics have a duty to act as critic and conscience of society. I’ve linked to this before but it’s worth repeating:
> Carrick my favorite moment involving him was some months back, IIRC on Eli’s blog, when he sneeringly attacked Harold Brooks after the latter dared to dispute his wisdom on some aspects of tornado damage. Oops. And whaddayaknow, Carrick just vanished upon being informed (not by HB) as to HB’s credentials, not even pausng to apologize.
Citation needed. All I can find is at Dr. William’s:
http://scienceblogs.com/stoat/2014/03/20/3601/#comment-39848
If Eli would so kind as to name the name, that would be nice.
Judith Curry tweeted:
(My emphasis)
Why the qualification? Why not all scientists, or all academics? Are privately employed scientists free to ignore, even participate in academic malpractice.
Further, seeing this seems to be job specific, was Curry told this by her employer? Can she provide the evidence in the form of clauses in her contract of employment? She cannot plead that it is a general presumption of government service in that we now have several cases of governments restricting the ability of scientists in their employ from speaking out about the misstatement of scientific facts, and of deliberately not retaining research data (included in Curry’s vague “lack of academic research” according to Mosher) as a budget saving measure. If governments can instruct scientists to be quiet on scientific matters, it is not a general obligation on scientists in government employ to speak out when they have evidence of scientific malpractise – still less when they have vague and hand waving suggestions of a vaguely defined “lack of integrity”.
Rachel, you consider the question:
My answer is not the opinion of an expert, but the consensus of expert opinion. So, what I look for are articles or blogs that clearly state the consensus opinion, and the reasons why that is the consensus opinion. I do not necessarily look for agreement with the consensus view. I will happily trust somebody who states the consensus, and the reasons it is accepted, and then points out their problem with the consensus and why they disagree. I may not agree with them, but I will trust them as a source of information. (Unfortunately such forthright sites are very rare.)
What is not required is that I know the academic background of the source. It is true that the best sources on science tend to be other scientists, but an interested non-specialist can pick up the nature of the consensus, the reasons why it is the consensus, and the reasons why they disagree (if they do). Sometimes such non-specialists can be better sources of information on the consensus than the scientists, due to an ability to explain matters clearly and without mathematical (or scientific) jargon. Consequently, while I like Anders as a source of interesting discussion on climate science, his status as a physicist is not relevant to that fact; and his anonymity does not impact on my estimate of the value of his blog.
Further, there are often very good reasons to remain anonymous on the web. Not least of these would be a desire to avoid vindictive complaints to employers or personal abuse directed to yourself or family to your discussion of controversial topics. These reasons can even extend to situations in which you are making private accusations of misconduct. Anonymous whistleblowers are not ipso facto unethical. The only times anonymity should not be retained (in this context) is if the evidence you supply is used as part of a formal complaint (other than initiating the process); or if you are making accusations of misconduct in public forums.
Steve Mosher writes: “Go back over all your posts and comments where you merely implied
that others were engaged in nefarious behavior. Ask yourself if you had air tight evidence.
I’ll wait. It doesn’t take long to find examples.”
Steve, If you have something specific in mind why not point to it? I’m sure ATTP has occasionally slipped from his own principles, but the idea he does so often doesn’t mesh with my experience reading this blog. I did a quick check on a post I thought might be a likely candidate, Bravo, Richard Tol, Bravo!. Instead you’ll find ATTP actually defending Tol in several comments and nowhere does he accuse Tol of a breach of scientific or research integrity.
I don’t know about anyone else around here, but I have a huge problem with this statement in Curry’s post — it is an assertion made w/o any supporting evidence:
Political psychology shares some common challenges with climate science in that much of the research occurs in the political arena (e.g. racism, wars, etc).
Where the hell did this come from?
He’s got to phone Heartland to ask first…
To continue, she seems to conflate the moral responsibility of someone who as the proverbial saying goes says “Stop!” as the car they are in is about to be driven over a cliff. Her aversity to anyone in the field pointing out the implications of the science is where this stems from, but I think that makes her a moral eunuch.
Hold on just one minute there, willard –
=> “Moshpit pulls a Joshua on AT just after telling the former:”
1) Judith mentioned a stat about participation rates at her blog, and I asked her for some of the details related to her stat (which she later supplied). Mosher, because he’s obsessive in his desire to make it about me, tried to make it about me.
2) Mosher showed up here and gave others an assignment to do his research.
Mosher didn’t pull a Joshua. Mosher pulled a mosher.
Here’s the Joshua I had in mind, Joshua:
> I wonder if you hold yourself to the same standard of air tight evidence.
The Moshpit was “but Climategate”.
I can’t believe INTEGRITY ™ was that good a slogan. Neither could Moshpit:
That’s the one, Willard. Not to put words in Eli’s mouth, but I expect the term is Dunning-Kruger.
Thanks for the link as the re-read was fun. The best bit was where Carrick, in full pontification, tells a world-leading expert (possibly even *the* world-leading expert, although I’m not familiar enough with the field to know that for sure) “it is clear you actually know so little about it.” Sweet.
In addition to Carrick’s Marshall McLuhan moment (applause to HB for the subtle reference), that thread was also good for the demolition (in part by moi) of William’s attempted defense of RP Jr.
Rachel
“If a scientist discovers something bad will happen if humans continue to do X, then I think they are failing in their duty to society at large by *not* advocating.”
You may think that but its not relevant what you think.
Look at the rules of the road here you are suppose to stick to the argument. the argument is about what Judith thinks. not what you think.
you are probably right. but the issue is what does judith MEAN by integrity.
THAT you disagree with her is beside the point.
“2) Mosher showed up here and gave others an assignment to do his research.”
No I asked the owner to do the work precisely BECAUSE i dont want the thread to be about him.
he can do the work and reflect. or we can make it about him or me or you or climategate.
I suggest he do the work for himself.
What Judy means by integrity is a facially absurd point to try to address. [Mod: Potentially defamatory] So yes, what anyone may guess about Judy’s view of integrity or indeed her own statements on the subject are quite beside the point.
“Steve, If you have something specific in mind why not point to it? ”
because then the thread would be about that rather than about what Currry means by integrity
Now I could pull a Joshua and hi jack the thread, but instead I suggest that the owner just take some time to reflect.
air tight evidence.. is that a standard he always holds himself to in science and blogging?
Perhaps.
but, look at the way he intrpreted Judiths words and argues that she implies some nefarious intent. is that a statement made with air tight evidence? hardly.
to repeat. he should ask Judith what she means by integrity. Then you get a good discussion.
As it stands he’s defined what she means and intends in such a way that no discussion can take place.
its pretty simple.
have any of you talked to her? written her a mail? asked her what she meant? tried to understand before you criticize? those are real questions.
I will wait for answers. I read the rules of the road here.
“What Judy means by integrity is a facially absurd point to try to address. [Mod: Sentence removed from original comment]”
this appears to violate the rules of the road here.
lets see what happens
> No I asked the owner to do the work precisely BECAUSE i dont want the thread to be about him.
An alternative theory is that it was an apophasis to make it about AT.
Joshua should try this line at Judy’s and see how it goes.
==> “this appears to violate the rules of the road here.”
Perhaps more importantly (moderation policy is a distraction, IMO), it violates the rules of logical reasoning.
That said, Judith’s definition of what is and isn’t integrity has no grounding in empirical analysis. It seems to me that she does not make a consistent attempt to control for subjectivity. She seems complacent in her subjectivity – mostly out of defensiveness (IMO).
> the thread would be about that rather than about what Currry means by integrity
AT editorializes on pontificating about research integrity, which he finds annoying, perhaps because he finds the claim that we should behave with integrity quite trivial.
The main problem to me springs from the very possibility to advocate for INTEGRITY ™. We all stand for more or less the same principles; we more or less all advocate for policies that conflict with one another. To say that one “advocates” for scientific integrity is just a verbal trick. It makes for perfect political slogans, as useful as they are empty
***
Here’s an instance where Judy’s advocating for something:
Judy’s clearly advocating against the IPCC. Advocating “against” is an auditing trick: since Judy does not suggest anything specific (she did hand wave to red teams, though), she can reply to someone who’d say (e.g. Joseph) that she’s advocating something: “what exactly am I advocating for?” It also discharges from the need to provide an alternative. That we “put down” the IPCC is certainly not a consensual policy. This kind of claim can be interpreted as advocacy. The same does not apply to standing for scientific integrity.
No scientist can afford to stand against INTEGRITY ™. So the question, as always, is how we should do this.
Without the possibility to disagree about implementation details, there’s very little advocacy going on.
***
That Judy tries to trademark that principle is suboptimal, to say the least. More so considering that she rubber stamped Carrick’s magnificent fall:
The discussion that follows may be noteworthy, considering that we’re talking about INTEGRITY ™.
I have no idea what Judith’s idea or notion or definition of integrity is, but I know that it is sorely lacking.
We only have to look at how she acted when it was pointed out that she had misquoted and completely mischaracterized an essay by Professor Robert Torcello.
This was done in a post titled: Towards a pragmatic ethics of climate change.
Ethics, integrity? Judith Curry’s only relevance to the subjects are as an example of a climate scientist lacking both.
Perhaps it might be useful to see what Universities have to say about research integrity. The University of Oxford has a page called Academic integrity in research. Here’s the code of practice:
The next paragraph gives a definition of misconduct in research:
Ethics and politics 101 with Judy:
Having to repeat college stuff over and over again is very tiring.
Mosher:
What complete and transparent bullshit. Not even Mosher could have believed it while he wrote it.
And here a transparent shift of the goal posts. Anders specified “air tight evidence” as his standard for accusations of lack of integrity, not for blogging in general.
No! If Curry means something different from standard usage by “integrity”, she should have so defined it, or better, used a different word. As she did not, Anders was perfectly justified in taking the word in its common meaning.
The point that Mosher is trying to elide is that Curry did not talk about “academic integrity” or “integrity” in general, but specifically “research integrity”. Therefore she can only have been talking about breaches of research ethics, or trying to paint something which is not such a breach as such a breach to taint it without justification. Advocacy by scientists is not a breach of research ethics by any definition. If that is what Curry meant, she needs to correct her error in mislabelling her claim.
The approach Mosher takes instead is to allow Curry an absurdly broad definition of terms that are well understood, and then impeach Anders for not noticing Curry’s persuasive definition.
==> “have any of you talked to her? written her a mail? asked her what she meant? tried to understand before you criticize? those are real questions.”
I think it was yesterday, mosher attacked me (he loves to make it about me), because I asked Judith what she meant.
I will say that sometimes mosher is right that I criticize Judith’s reasoning without a sufficient effort to (or an open-mind approach to) understanding what she means. But on the other hand, he also attacks me simply because I question her, or simply because I express disagreement with her reasoning (after having made a sufficient effort to understand). There’s a fine line. But mosher’s effort to appreciate the subtleties finding that line is often lacking. I chalk it up to tribalism.
Steven,
If it’s easy, find some. I try very hard not to excuse others of nefarious behaviour, so am always happy to be corrected or retract something if it’s pointed out. I’m not, however, planning to do that work myself.
Steven,
Judith uses the term research integrity. As others have pointed out, this is normally taken to mean honesty in your research, not how honest/decent you are as a general person. If Judith means common decency rather than research integrity (or, at least, how I understand the term) then I would have more sympathy with her views (not complete agreement, mind you, but more understanding). So, if Judith is using research integrity in a way different to how others would understand it, then maybe she could clarify.
This guy FoxGoose, who I follow on Twitter for the entertainment of reading his unrelenting rage, said this:
One of Curry’s favorite commenters is one Tomas Milanovic, who sounds like a real name but Curry admits only she knows the real identity of. Tomas is seen by the Curry clan as the man with the true scientific authority.
Curry likes to fight her battles by proxy. Anything that will feed her Uncertainty Monster is condoned.
As I’ve been quite critical of Curry above, out of fairness I should say that I’ve never had the impression that her actions are driven by her commercial interests and I think its unfair to suggest that this is the case.
WebHubTelescope says:
August 5, 2014 at 7:14 am
This guy FoxGoose, who I follow on Twitter for the entertainment of reading his unrelenting rage, said ….
One of Curry’s favorite commenters is one Tomas Milanovic, who sounds like a real name but Curry admits only she knows the real identity…….
Curry likes to fight her battles by proxy. Anything that will feed her Uncertainty Monster is condoned.
I’m not sure I’m following you here Webby.
Is the allegation that I’m an alter ego of this Milanovic chap – or a secret proxy agent of Judith Curry?
Neither is remotely true – so my unrelenting rage compels me to point out that you are a [Mod: Be nice, please] victim of “conspiracy ideation” (c. S. Lewandowsky ).
but the issue is what does judith MEAN by integrity.
THAT you disagree with her is beside the point.
How is it beside the point? Surely part of having a debate about ethics is defining what does or does not constitute (un)ethical behaviour. So yes, it’s fair for you to point out what Judith means when she talks about ethical behaviour but it’s also reasonable for others to take issue with her particular definition.
I had a brief Twitter exchange with Philip Moriarity that makes me think I haven’t explained myself as clearly as I should have, or am being a little misunderstood. I’m certainly not arguing that people shouldn’t criticise others. If you want to have a ding-dong scientific battle with someone else, go ahead. I’m involved in something like this myself at the moment. These are, however, typically resolved in the literature and typically don’t include accusations of misconduct (unless there’s clear evidence). My issue is with insinuations of a lack of research integrity in others for which there isn’t much evidence and which isn’t resolved in the scientific literature (and, by research integrity I mean the integrity of someone’s research, not their own personal integrity).
I’m also not suggesting that everything is fine. It was this kind of thing that motivated an earlier post of mine about Elements of truth. There is much, in my opinion, that we could do to improve research standards. Peer-review isn’t perfect. How we fund science isn’t perfect. What we value and what we reward is not perfect. These problems, however, are – in my view at least – more general than specific. They apply across academia (or at least across the physical sciences). There may be some fields that are more prone to problems than others (Ben Goldacre’s books would suggest pharmaceutical research has some big issues), but I’ve seen nothing to suggest that any issues in climate science are particularly different to issues in physics, for example.
So, I’d be all for attempts to improve peer-review, to encourage universities to value quality over quantity, and other changes that could improve the quality of research. This post is really just a criticism of those who seem to feel comfortable insinuating a lack of research integrity in others, but which appears to be based more on a disagreement about science (and maybe politics) than on any actual evidence.
FG,
I thought Joshua’s point was obvious. If you’re going to argue that those who claim to have, or are perceived to have, some credibility should not be anonymous, then you should be consistent.
FoxGoose, Sad to say, but the discussion is not about you.
Tomas Milanovic is one of just a few Curry skeptic commenters (i.e. one of the “deniszens”) hat takes the angle of argument by invoking physics complexity. To Tomas, nothing can be calculated because chaotic complexity and the nature of ergodic processes precludes our ability to predict anything. He is more of a drive-by commenter who skillfully makes the occasional comment to make himself look very serious.
Curry plays off him to the hilt, and all the gullible denizens follow in line. Maddening.
And Then There’s Physics says:
August 5, 2014 at 9:30 am
FG,
I thought Joshua’s point was obvious……..
I’m a bit confused here Anders – are you sure you’re joining in the right conversation?
WebHubTelescope says
August 5, 2014 at 7:14 am
This guy FoxGoose, who I follow on Twitter for the entertainment of reading his unrelenting rage, said this………..
Closely followed by:-
WebHubTelescope says:
August 5, 2014 at 10:21 am
FoxGoose, Sad to say, but the discussion is not about you….
I’m getting even more confused here – it’s all a bit “through the looking glass” now.
Anyway, since key words are now being chopped out of my posts – this is the point where I retire to more open pastures.
Enjoy
FG,
But surely the key point in WHUT’s comment was about anonymity and credibility (real or imagined)?
Endless insinuations of nefarious intent by the usual suspects. Further redundant and arse-numbingly tedious evidence that the ‘sceptics’ have nothing more substantial than an endless smear campaign (aka never-ending audit).
Well, sans any kind of actual scientific case, they must work with what they’ve got.
I think you may have confused WHUT with Joshua.
Anyway, the comment is still completely opaque to me.
I’m anonymous – I claim no special credibility.
What exactly has the fact that some guy I’ve never heard of may do so elsewhere got to do with me?
FG,
This is the comment I’m referring to.
I don’t think it was about you specifically. The argument – I think – is that if you’re suggesting that if I engage in serious discussion with/about Judith on my blog, that I should remove my anonymity because it is now becoming academically serious, then presumably that should apply universally. Yet, Judith seems happy to allow people to post supposedly academically credible arguments on her site despite people being anonymous (to everyone else, at least).
Personally, I don’t really care if someone is anonymous/pseudonymous or not and while this remains on the blogosphere I would not regard it as academically serious.
With an H/T to OilMan:
Hmmmm…. unfounded conclusions based on models that have 100% error. I don’t see the problem with Dr. Curry bringing up the issue of advocacy and integrity.
nickels,
Hmm, 100% wrong, that’s a bad start. Judith is more than welcome to bring up the issue of advocacy and integrity, that’s her right. It would be nice if she could provide some actual evidence that advocacy does influence research integrity as I have yet to see any actual evidence.
Tom Curtis quoted this tweet by Curry:
“Its my job, and the job of all academics paid by the government, to protect research integrity”
I think it fits nicely with the way that paid tobacco advocacy and associated research, being by private companies, isn’t required to have any integrity.
Those of us who remember the previous century will also recall a great deal of research carried out by corporations to prove that their product wasn’t doing any harm to anyone or anything at all and if anything was doing good…
The pattern I am seeing, whether it’s Curry’s ‘research integrity’, insinuations of misconduct everywhere, Mosher’s ‘lack of integrity’ that isn’t quite misconduct, etc. – all of this boils down to a basic Argument Ad Hominem appeal, an avoidance of discussing the science.
In Curry’s most recent post she conflates advocacy regarding policy (advocacy informed by scientific viewpoint) with a lack of integrity, seemingly implying that the science is being driven by the ideology. This is just wrong – if you disagree with the science, discuss the science. If you disagree with the policies derived from the evidence (as Curry does with her frequent ‘do nothing’ advocacy), discuss those policies. But strong feelings about public policies have absolutely nothing to do with the strength of the science that may be informing those feelings.
Foxgoose – If you are having issues with Anders and his(her?) stated background, you are falling prey to a _self-inflicted_ Argument from Authority. And that is your problem, not Anders.
Judge statements and analysis by the content, not by the speaker.
KR
Yes, that is exactly as it appears to me, (eg. comment above).
And as we all know, argumentum ad hominem carries no weight whatsoever. But that’s all they’ve got.
Andrew Adams said “As I’ve been quite critical of Curry above, out of fairness I should say that I’ve never had the impression that her actions are driven by her commercial interests and I think its unfair to suggest that this is the case.”
Briefly:
1) CFAN, a for-profit weather and climate forecasting service, is launched in 2006 by Judy and husband Peter Webster. Judy and CFAN operate out of Atlanta/Georgia Tech, which can be safely called a politically reactionary milieu. CFAN’s focus is on providing services to extractive industries; oil is named, but likely transportation and insurance are included given their concern about weather vulnerabilities. I don’t think it’s radical to suggest that taking a contrarian stance on climate would be helpful in marketing CFAN to executives of such industries based in the U.S. south (many of whom are conveniently GTech alums)..
2) At about the same time Judy, with no prior history of such involvement, becomes active in the blogosphere and immediately focuses on participation at Climate Audit. A little later (IIRC 2008) she takes the odd step of inviting the amateur statistician Steve McIntyre, prominent for his attacks on other climate scientists, to GT for a presentation of his ideas(!).
3) Note that 2006 is also when the Republican Party (which controls Georgia) began its sharp tilt toward climate denial, a stance that became more extreme through 2010, and as noted will have made it increasingly convenient for CFAN marketing for Judy to be seen as a climate contrarian.
4) In 2010 Judy goes clearly rogue and launches Climate etc.
5) As a climate contrarian, Judy gets access to venues that her record of scholarship would not have gotten her, e.g. the U.S. Congress and a profile In Nature. This can only have been a big help to marketing CFAN.
*Maybe* this is all a coincidence, but I doubt it. It’s certainly not “unfair” to suggest it isn’t. You can choose to look the other way, Andrew, but I don’t.
Steve Bloom, while interesting, your evidence is circumstantial at best. It certainly does not rise to the standard of “watertight evidence” mentioned by Anders. Further, it is just one of several theories as to why Curry “jumped the shark”, not of all which support a presumption of a lack of integrity, and I would not make that claim on so circumstantial a case.
In addition to what Tom says, a point I was trying to get across in the post was about accusations of research integrity. That – as I interpret it – implies something about the honesty of someone’s research which, in the physical sciences at least, one can determine by looking at the research itself. That’s why I don’t think one can make such accusations simply on the basis of someone’s public behaviour, and why someone’s public behaviour is largely irrelevant.
Tom, I thought I was clear that people could reasonably disagree on the strength of the case (by no means fully fleshed out as I stated it, note). My point was simply that it’s not “unfair” to conclude as I have.
FWIW, I think it’s helpful to have had the advantages of watching all of this unfold in real time (I don’t know how far back your and Andrew’s close involvement goes) and an inside understanding of U.S. politics, especially the environmental and climate aspects. Yep, with the latter I’m pulling credentials on you, but were the situation reversed I think I’d be hesitant to be quite as firm in my opinions regarding a parallel circumstance in Oz or the UK. Of course, “firm” might not be the right word to describe bending over backward to give someone the benefit of the doubt.
Anders, my point was very much not personal integrity rather than the research variety. I’m unimpressed by Judy’s papers of the last eight years or so, especially the absurd “stadium wave” business, but I’d say that stuff is more a matter of being blindered by poor physical intuition. The latter is perhaps necessary to her contrarianness, but by no means necessitates it.
s/b …very much *about* personal integrity…
Playing the ball is the best way to score.
INTEGRITY ™ – Play the Ball.
It’s fine to have a conflict of interest provided it is fully disclosed. A conflict of interest does != lack of integrity. But there should be full disclosure of the conflict of interest in research papers and other publications.
No democracy without transparency.
I stopped paying any attention after her comments #114, #125 and #231 04/25/10-04/26/10 on Wegman, as part of discussion at Collide-a-Scape
Briefly: #114 & #125 contained at least 6 clear factual errors, all favoring Wegman, and called Deep Climate “reprehensible” several times, including:
“Let me say that this is one of the most reprehensible attacks on a reputable scientist that I have seen, and the so-called tsunami of accusations made in regards to climategate are nothing in compared to the attack on Wegman”
#231 JC asked Keith Kloor to declare the topic off limits
#292 andrewt suggested an apology to Deep Climate might be in order
#19 in later post..
People might read the full comments in context and see if they discern any patterns.
Of course, although some plagiarism in the WR and CSDA articles were already obvious, but there was much more to come.
I have been looking further at our hosts rampant campaign of unwarranted accusations of a “lack of integrity” suggested, by apophasis, to exist by Mosher and was coming up empty. Fortunately Brandon Shollenberger came to the rescue at Climate Etc with this comment:
The accusation has been endorsed by Carrick and Mosher. Carrick in fact admits that the two accusations made by Shollenberger are the only examples he knows of that have lead him to claim that Anders has never “…demonstrated that [he] have a moral compass”, and that he is an example of “Scientific sociopathic behavior at its finest”.
The accusation, it turns out, is related to this comment by Anders at Shollenberger’s blogsite, where he writes:
The actual facts of the case are that after running his program to generate 10,000 pseudo reconstructions from red noise, McIntyre and McKitrick, 2005 selected the 100 pseudo reconstructions with the highest Hockey Stick Index (HSI, =(1902-80 mean minus 1400-1980 mean)/1400-1980 standard deviation). That group of 100 pseudo reconstructions were mentioned in the paper, and included in the supplementary information. All choices of pseudo reconstructions to graph by M&M05 were chosen from that cherry picked selection. That includes the single graph shown together with the MBH98 reconstruction as Fig 1 in M&M05, and the panel of twelve pseudo reconstructions generated by the code included in supplementary material of M&M05, and published in the Wegman Report. At no point in the paper or in supplementary material do M&M05 draw attention to the fact that the selection is a biased selection.
Now, it is clear that Anders made a couple of mistakes. To start with, although the pseudo reconstruction that was published in M&M05 “… were not randomly selected from their sample of 10000”, being in fact selected from their sample of the 1% with the highest HSI; and though the 10 pseudo proxies graphed by the code provided with the supplementary information of M&M05 “… were not randomly selected from their sample of 10000”, the later 10 were not shown in the paper. Further, the eleven pseudo reconstructions graphed either in figure 1 or by the code provided with the paper were not the eleven (or 10) with the highest HSI, they were merely selected from the 1% of pseudo reconstructions with the highest HSI.
It is clear, however, that while these are mistakes, they are trivial mistakes. They do not effect the substance of the issue. A cherry pick of the top 1% of pseudo reconstructions in terms of the HSI is still a cherry pick, and not informing readers either in the paper or notes on the supplementary information that the graphs generated by the program in the SI are a biased selection is a serious breach of normal standards of publication. This is particularly the case as the only tests conducted by M&M05 to show the ability of red noise to generate MBH98 like pseudo reconstructions is by visual comparisons with the cherry picked selection.
That last may seem an odd claim, in that surely a comparison was made using the HSI itself, but in fact M&M05 never publish the HSI of either the MBH98 reconstruction (1.13), the MBH98 580 year PC (0.94), or the MBH 99 reconstruction (1.13). These are 22.7 (MBH98 and 99) and 27.9 (MBH98 PC1) standard deviations less than the mean HSI of the selected 1% of pseudo reconstructions. The are also in the bottom one percentile (MBH98 PC1) or, probably, 2.5%ile (MBH98 & 99) of results for all 10,000 pseudo reconstructions. I say “probably” for the later because M&M05 do not give full statistics. The do state that less than 1% have a HSI less than 1, and that only 27% have a HSI less than 1.5. From the histogram (figure 2), it is possible to determine that about a fifth of that 25% have a HSI less than 1.25.
You can see, therefore, why McIntyre and McKitrick were loath to do more than visual comparisons. Had they done a statistical comparison using their chosen measure of similarity (the HSI), the paper would have (apparently) reported that the MBH98 method applied to red noise generates hockey stick like shapes but that statistical tests show at the 90% (certainly) or 95% (probably) confidence levels, the MBH 98 was not a random outcome from red noise.
This is even clearer using more conventional tests. The peak r-squared comparing the 100 selected hockey sticks to the instrumental record over the period 1902-1980 was 0.35, compared to 0.76 for MBH98. That is seven standard deviations from the mean of the r-squared statistics for the selected pseudo reconstructions. Again, actual numerical statistical tests, as opposed to eyeballing cherry picked graphs, show the MBH98 reconstruction not to have been a chance outcome from red noise. This is something Michael Mann has also demonstrated by other means.
Further, certainly McKitrick (McKitrick 05) and probably McIntyre were using the selected set of pseudo reconstructions in public commentary in 2005.
Going back to Ander’s “perfidy”, it is clear that in this episode his claims were essentially correct. Further, he did not make an accusation of cherry picking as such. He did claimed a selection bias in the graph, but made no claim as to how that would effect the result or as to the motives of McIntyre and McKitrick in imposing the selection bias. His most strenuous criticism is to say some of M&M05 methods “seem odd”, and then to qualify that by indicating that that may merely illustrate his lack of specialist knowledge on the topic. A damnation of M&M05 on the basis of lack of research integrity his comment is not.
The second basis of Anders purported “lack of moral compass” is that he banned Shollenberger from this site because Shollenberger repeatedly claimed that he (Anders) had lied (Shollenberger uses the word “fabricating” which definitely implies intent, and in context implies the claims are false). After a discussion with Shollenberger on his site, I heartily endorse Anders sentiment of never wanting to discuss anything with Shollenberger again. It is not a moral or a personality flaw to dislike discussing things with people who do not discuss in good faith, show a lack of personal integrity, and behave like complete pricks.
Did someone say oil man?
I’m just asking questions, but how much money per year is Curry getting paid in consulting fees to the Oil and Gas Industry? An industry which has decided to suddenly use Curry as well as many weather and climate trending services used by all the other industries in the region.
Do I need to point out how much money the oil and gas industry has dumped into killing Climate Science? There is plenty of data on Exxon’s anti science expenditures.
Why would anyone care what a high schooler like Anthony Watts has to say? Maybe I’m being a knob. What exactly is the uninformed high schooler opinion on scientific integrity? Anyone? Anthony? Bueller?
TC: it wasn’t just McKitrick(2005), the really key presentation was the 05/11/05 talk for CEI + GMI, which Joe Barton’s staffer Peter Spencer later gave to Ed Wegman, and which was effectively the blueprint for the Wegman Report. That’s this:
p.10 claims that the Daly=McIntyre image variant of Lamb(1965): IPCC (1990) Fig. 7.1(c) came from IPCC 1995.
p.11 Uses Huang et al, which Huang’s group had long before declared not usable for that
p.12 Has the infamous, unsupported Deming quote, from JSE, my favorite dog astrology journal, magically transformed to a journal of repute, i.e., Science.
To be clear, it would be published in JSE, but at the time, it was found prepublished at Fred Singer’s website, a curious place for it.
p.18 4 of the carefully-selected simulations
Tom,
Thanks, I did wonder if that would come up again. Your summary seems about right. I think I also acknowledged that error when Brandon was sockpuppeting here.
John, the presentation you link to is dated September 2012. I can see no evidence in it of material post dating 2006, and so it may indeed be a re-presentation of the CEI+GMI talk, but do you have direct evidence of that? Further, I am unfamiliar with the acronyms “CEI” and “GMI”, do you have a translation? 🙂
With regard to McIntyre’s presentation of the cherry picked pseudo reconstructions, I have definitely found him using three of them in a blog post dated Jan 27th, 2005. McIntyre introduces the graph, saying:
The images give every appearance of being drawn from the 100 selected pseudo reconstructions, but there is no acknowledgement that they were drawn from a biased selection by McIntyre.
The image is archived here.
KR –
It’s a fairly standard practice in politics – always accuse your opponent of doing whatever it is you are doing. So a badly-informed bystander just sees people throwing the same accusation at each other and loses interest.
Tom
1) Sorry: CEI = Competive Enterprise Institute, GMI = George Marshall Institute
For the context, see Strange Scholarship in the Wegman Report (SSWR), pp.27-32.
Activity 08 on p.28 is the particular meeting.
(Note:following is more detial than almost anyone would want, but I just had time to grab some bits from a PDF before going off to sleep.)
Bottom line: the provenance of the 05/11/05 PPT is rock-solid.
2) SSWR p.185 describes this, starting with McIntyre’s comment about upcoming meeting, for Cooler Heads Coalition (managed by CEI) and GMI. Most of that was right, except there were more versions of the presentation around, people changed files, some disappeared, etc. Shortly after SSWR was published (late 2010), Dan Vergano of USA Today made FOIA requests to George Mason U and Wegman responded in November 2010 by sending files he had gotten from Peter Spencer (in August/September 2005) which included the PPT file. this was README file supplied by Wegman for Vergano, and the directory had the following, selected for Wegman via Joe Barton’s staffer Peter Spencer. The *’d items were among the “Important Papers”.
#12* Burger, Cubasch (2005)
#23* Esper, et al (2002)
#33 Jones, Mann (2004) the only paper omitted
#41* Mann, Gille, Bradley, et al (2000)
#42* Mann, Jones (2003)
#43* Mann, et al (2005)
#47* Moberg, et al (2005)
#48* Osborn, Briffa (2006)
#49* Rutherford, et al (2005)
#54* von Storch, Zorita, et al (2004)
#59* Wunsch (2006) (Not relevant (N))
#64 IPCC (2001) TAR, Ch 2, Observed Climate Variability and Change
#78 McIntyre, McKitrick (2005) MM05x, 66p raw PowerPoint
Of Spencer’s 14 files, 10 were included in the 17 labeled Important,W.8.2.
The next 5 Important papers were all cited in MM05x (and obvious):
151 Spencer later attended 3 Heartland Institute conferences at their expense, 2008-:
http://www.legistorm.com/trip/list/by/traveler/id/8305/name/Peter_L_Spencer_Jr_.html
37* Mann, Bradley, Hughes (1998)
38* Mann, Bradley, Hughes (1999)
44* McIntyre, McKitrick (2003)
45* McIntyre, McKitrick (2005a)
46* McIntyre, McKitrick (2005b)
36* Mann dissertation (not relevant (n))
59* Wunsch (2002) (not relevant (n))
The WP used all papers from Spencer’s list except #33, then added the 5
obvious papers from MM05x, plus 2 irrelevant items.
The choice of Important papers was almost entirely determined by the set Spencer gave to Wegman, which he said was not “coaching.”
As mentioned in FOIA Facts 1…
“Dan Vergano had filed FOIAs with GMU in 2010, which somewhat oddly were handled by letting Wegman choose the documents to be provided. He supplied a mass of (seemingly-irrelevant) material, but without the usual redactions of personal data and claimed (falsely) that FOIA replies could not be supplied to others, showing he did not understand the rules. FOIAs from 2011-2013 sometimes led to reexamination of the 2010 information, of which some turned out to be quite relevant.”
“Finally, Dan Vergano has posted about 30 relevant files from FOIAs at DocumentCloud. Some will be referenced in posts 2-4.” The PPT file does not have a hardwired date, so the PDF that Vergano posted showed the date he did the conversion for DocumentCloud.
Anyway, although complicated, the provenance is solid: there are reasonably-consistent PDFs that appeared in 2006.
I have a partially updated iteration of SSWR, which fixes some confusion, but I’ve jsut ghrabbed text without much reformatting:
McK05a PDF, 04/04/05, original McKitrick paper for APEC 159
PPT, unknown, but many slides in common with:
MM05x PPT, 05/11/05, but file date varies dynamically 160
McK05 PDF, 07/22/05 (04/05/05), version Wegman Panel likely referenced 161
MM05y PDF, 08/30/05, original version MM05x PPT PDF, 66p 162
MM05z PDF, 06/17/09-current, annotated talk, extra comments, 42p 163
159 An early version was found, with no errata, presumably a copy of the original.
http://web.archive.org/web/20060919154910/www.uoguelph.ca/~rmckitri/research/APE
C-hockey.pdf
160 Raw PowerPoint (PPT), M_M.May11.ppt, by FOIA, W.8.9.2
http://www.documentcloud.org/documents/422182-m-m-may11.html
161 http://www.uoguelph.ca/~rmckitri/research/McKitrick-hockeystick.pdf
http://www.webcitation.org/6AU3dkaql Archived, with following note:
‘Correction of some typos. 7/22/05: Corrected IPCC 1995 ref. on p.5 to IPCC
1990. 10/18/05: Correction of minor grammar and punctuation.’
Since the WR mentioned IPCC 1990 as a source, presumably the 07/22/05
correction, at least was available, although MM05x did not reflect that.
[2010] Multiple versions caused confusion, as McK05 fixed errors after MM05x.
162 MM05y could have been posted on website any day from 08/30/05 to 02/22/06.
Internet Archive only copies file after they have existed a while. It gives the
history, from 02/22/06 – 11/27/08, during which MM05y was unchanged:
Click to access 316.pdf
163 By next snapshot, 08/07/10, the file had changed to MM05z:
Click to access 316.pdf
http://www.webcitation.org/5n3coYFsz archived 01/25/10
Off to sleep.
A significant issue is that Curry’s appreciation of ethics is confused but adamant. One can also say the same about her appreciation of science.
Readers may be interested to read our own take on the issue of scientific integrity, more from the point of life sciences (http://blog.pubpeer.com/?p=164 ). In a nutshell, we feel that too many researchers do cheat and succeed, to the point where it has become a real problem. Regulation has demonstrably failed.
Note however that PubPeer accuses nobody of fraud or misconduct. Commenters simply point out anomalies in papers, many of which are however incredibly difficult to explain other than by misconduct. We argue (read the comments below the blog post) that discussion of published data is always be justified, even if the data imply that misconduct has occurred. If there are signs of misconduct in published work, that’s important information to share publicly, rather than to hide immediately under a cloak of confidentiality during several years of investigation for fraud.
PubPeer,
Thanks. My post certainly isn’t an argument that there aren’t important issues related to research integrity, or misconduct. There clearly are and after a quick read of your post, I have some sympathy with what you say. There clearly is an incentive to publish high impact work and people are rewarded for doing so, but rarely punished if they are wrong in ways that could have been avoided.
I’ll make a couple of comments to try and clarify what I was trying to get at in the post. I’m certainly not against people highlighting flawed work and even making claims of misconduct if warranted. Your post seems to highlight some examples. Philip Moriarty highlighted another particular example. The issue I have is with those who throw around accusations of a lack of research integrity in others, without providing any evidence, and give the impression that they are somehow beyond reproach and are doing everyone a favour by highlight this possible issue.
The other issue may illustrate my physical science bias. Possibly I have too much confidence in the scientific method, but the existence of fundamental laws of physics makes fabricating data a little more difficult than in some other fields. Of course, this doesn’t mean that physical scientists have more integrity than others, it simply means that it’s harder for them to be dishonest because it’s easier to get caught. This doesn’t mean that it doesn’t happen, just that it would appear to be more difficult than in some other fields. Maybe I’m wrong about this, so feel free to convince me that I am, but one reason I see less reason for concern in the physical sciences is that the existence of these fundamental laws provides checks and balances that may not exist elsewhere.
I’ll add that I broadly agree with what you suggest at the end of your post. Trying to audit research is a non-starter. Who could do it? If you have enough knowledge and understanding to do it, you’d be doing the research, not doing the auditing. I’d certainly support stiffer penalties for those who get caught and support the idea that there should be easier access to data. I’ll add, though, that the latter is not necessarily as simple as some might think. My modification would be that work should be reproducible on a reasonable timescale. If your data is really not accessible in any other way, it should be freely available. On the other hand, if someone could reproduce your data in an afternoon by writing a short computer code, then maybe not.
I do think, however, that there are some more fundamental issues related to what is valued by universities and funding bodies and how they incentivise such behaviour, but I’ve probably said enough 🙂
==> ” I think I also acknowledged that error when Brandon was sockpuppeting here.”
Expect a post now about how you’re either immoral or stupid or an idiot because it “makes no sense” to think that Brandon posting here under more than one screenname is sockpuppeting.
That is a fairly typical example of Brandon’s playbook. He assume authority over something which is a matter of opinion (ex: what is and isn’t sockpuppetry), confuses his opinion with fact, and then focuses on that conflation rather than look at the larger issue at hand – in this case the fact that he posted at this site under two screennames.
Brandon’s an interesting person. He thinks that other people are responsible for decisions he makes about what to do with his time. He thinks that if he insults people, it’s their fault if they don’t think it’s funny. And he thinks that if he doesn’t understand something, it can only mean that it “doesn’t make sense.”
A really interesting person.
Not sure why previous comment went into moderation, hopefully this one will get through – although I’ll understand if you think it’s better off to let the whole Brandon handbag fight thing die…it’s perhaps even more juvenile than typical blogospheric banter…
I just found this comment from Brandon over at Judy’s, and I had forgotten how this is another move from his playbook.,
==> :”Basically, Curtis is saying it’s good to ban people from sites because you don’t like their character.”
Italics mine.
I’ve seen this before from him where he assumes as fact that someone has said X,when it is clearly only an opinion about what X said, and inserts a “basically” to adjust for his conflation between fact and opinion. Perhaps not coincidentally, I’ve seen mosher similarly leverage “basically” to conform something someone else has said to fit his purposes or confirm his biases.
I think that move is a very clear example of what I interpret to be Tom’s explanation for why he supports banning Brandon – because it is an example of not exchanging views in good faith.
Tom,
This is probably more for other people’s benefit, than for yours, but I’ve done a bit of investigating regarding this figure.
In the McIntrye & McKitrick (2005) Supplementary Information, you can download a file containing what McIntyre & McKitrick call “a sample of 100 hockey sticks”.
You can also download their code. The relevant parts are here (I think, happy to be corrected if wrong)
which – I think – calculates the hockey-stickness for each of the 10000 samples.
And here
which writes out a file of 100 samples, but ordered in decreasing hockey-stickness (I think).
For what it is worth, the example shown in Figure 1 of McIntyre & McKitrick (2005) is number 70 in their list of 100. In the figure you’ve included, the top left is 100 in the list, the bottom left is number 66 and the bottom right is number 27 (hence MBH98 is top right). I also found 4 of the 10 Wegman figures in the sample of 100 (33, 35, 49, 54). I gave up after finding 4.
So, as I think Tom and others have pointed out, the sample of 100 hockey sticks provided by McIntrye & McKitrick (2005) appears not to be a random sample from their 10000, but a sample selected on the basis of hockey-stickness (i.e., they are the 100 most hockey-stick like out of the sample of 10000). So far I’ve identified the example included in Figure 1 of McIntyre & McKitrick as coming from this sample, 4 of the 10 Wegman examples, and all 3 of the examples that Steve McIntyre (according to Tom’s quote) claims is representative of 99% of the 10000 samples, rather than being an example out of a sample of the 100 most hockey-stick like.
I don’t have easy access to the 9900 that are not included in the sample of 100 so can’t state with certainty that they don’t also have hockey-stick-like shapes, but I guess that Deep Climate and Nick Stokes have probably already looked at this.
Also, FWIW, the random red noise used by McIntrye & McKitrick was also generated using the original Mann, Bradley & Hughes data. I’ve always found this rather confusing since the Mann, Bradley & Hughes data is simply 72 timeseries from around 1400 to about 1970. Quite why McIntrye & McKitrick used these timeseries rather than simply randomly generating 72 trendless timeseries has always been somewhat beyond me.
Joshua,
You might be able to guess.
Anders –
BTW – Don’t know if you’ve seen this:
There’s certainly a flavor of damning with faint praise, but I think that aspect is secondary (and it’s unfortunate that she framed her acknowledgement in such a context. It’s that kind of move on her part that seems to me to be counterproductive to her interest in “building bridges.”
Perhaps if you went out and bought a moral compass, Carrick would value your contributions also: 🙂
Joshua,
Interesting, from Judith I’ll take that as a complement 🙂
Much of the rest on that thread are things I shall largely choose to ignore.
==> “Much of the rest on that thread are things I shall largely choose to ignore.”
Exactly what I’d expect from a scientific sociopath. 🙂
I love how Carrick climbs on his moral and scientific high horses to impugn the science and morality of others, completely oblivious to the completely unscientific nature of his reasoning and the questionable moral implications of his Jell-0 flinging.
I like to call the blogosphere irony-apalooza.
Joshua,
The reason all those versions of your comment were moderated, was because you mistyped your email address and you were regarded as a first time commenter. 🙂
==> “and you were regarded as a first time commenter. :-)”
If only it were true, eh?
Thanks, Tom.
I tried to reconcile all positions to commit some basic truths at Judy’s, but to do that, I felt had to scratch “cherry pick”:
Carrick, after all, scratched his “sociopath” comment to replace it with his “moral compass”. I am thankful for it, even if I had to work harder than expected to get that concession. I don’t think there’s any need to claim that the Auditor cherry picked anything for the criticism to stand.
When that plays out, I may have the time to try to go at Brandon’s to commit some “basic truths”.
No, I won’t go there under a false ID.
Pingback: Anche le moscerine, nel loro piccolo – Ocasapiens - Blog - Repubblica.it
Anders, I have the PC1 from figure 1 of M&M05 as being number 71 on their list published in the SI, counting from 1-100. On the same convention, Deep Climate has already identified all twelve PC1s printed out by the M&M05 script, and used as Wegman Fig 4.4, as being numbers 4, 14, 33, 35, 46, 49, 54, 61, 72, 81, 91, and 100. Nick Stokes shows a selection of 12 essentially randomly chosen PC1s, or which I would say only 1 or 2 have a hockey stick shape (at most). Shollenberger seems to find hockey stick shapes in all of them, however, so go figure. Almost certainly all 12 do have a HSI greater than 1, and hence are hockey sticks by M&M’s definition, which just goes to show how worthless their HSI is as a statistical measure, IMO.
The 100 selected PC1s are certainly not a random sample. That has been noted by Deep Climate from the R code, and I can confirm it by a comparison of the histogram of the 100 selected PC1s to the histogram of all 10,000 presented in Fig 2 of M&M05. I do not think there is any question that M&M cherry picked the PC1s with the most extreme HSI, and have almost exclusively drawn their graphs of red noise generated hockey sticks from that cherry picked selection. I write “almost exclusively” because I know of no counter-example, but have not read everything either have published.
That to me is extraordinary, because despite their purportedly great statistical chops, the only statistical test they use in comparing MBH98 to their red noise generated PC1s is visual comparison of graphs. Indeed, that visual comparison is with the PC1 of MBH98 (ie, the first principle component of the decentered analysis of all proxies that extend back to 1400) which as I note above has a HSI of 0.935, ie, it is not a hockey stick according to M&M05’s definition (see the introduction). It is astonishing that a paper could be published, purportedly being a statistical analysis, but whose most strenuous statistical test of a key feature of the paper is, “well, it looks a bit like it, dunnit”. Perhaps I am not recognizing M&M’s true statistical genius because I am not squinting right.
Tomas Milanovic’ usage of ‘what’ instead of ‘which’ is much like the usage by the poster named Tom Vonk both of who seem to have special interests in chaos theory.
Willard, I think “cherry pick” is entirely justified. I would go along that a softer term might be appropriate if MBH98 was compared to the red noise generated PC1s by any statistical test other than similarity of appearance on the graph. But M&M05 did not even calculate the HSI of MBH98, and certainly do not compare the HSI of MBH98 to their whole sample. Indeed they do not use any statistical test to compare their sample to MBH98. Given then, how essential the hockey stick like appearance of their sample is to their argument, choosing the visually displayed members of their sample exclusively from the 1% with the highest HSI has to count as a cherry pick.
Tom,
Thanks. You’re right, it is 71. I’m using IDL which starts all its array indices from 0. I corrected for that in most of the cases, but forgot in that case (in fact, I didn’t forget – I had the right number on the piece of paper I was using, but decided to double check and then wrote down the wrong number).
What you say about Figure 2 seems obvious now that you mention it. All 100 in the sample must be on the right hand side of the distribution that considers all 10000, so it can’t be a random sample. I’m also aware of all the other sites that have shown this, so was just trying to check for myself.
Joshua, Shollenberger of course misrepresents me. I do not say it is good to ban people, I say it is a matter of moral indifference whether or not you ban somebody in the vast majority of cases. People blog for their own enjoyment. If it is not enjoyable to discuss things with a particular person, they are under no obligation to host the comments of that person. And if you find out how unpleasant it is to discuss something with a particular person on their website rather than your own, your ahead in the game.
Now, clearly there are exceptions to this rule. Some websites are in fact services to the public, and they may want to place more restrictive rules on their moderation practises. But that is not the case here except to the extent that we personally find Anders posts and discussion informative and thought provoking. And that is not a service provided by Anders but rather a consequence of his being a thoughtful and thought provoking person with a pleasant personality. His being a nice person does not place a moral obligation on him to suffer the likes of Shollenberger.
ATTP,
As far as I have understood, what McI&McK have done looking at the 100 most hockey-stickian cases is part of the study. It has been declared by them and there’s nothing wrong with that. Some people may have read more in their results than they have ever had in mind, but that’s not their fault.
It’s not fully obvious, what one should conclude from their work, but picking 100 out of 10000 is a well justified part of their methodology.
Just a quick reminder. The problem arose because you claimed in a post about “new rules” that
“McIntyre & McKitrick 2005 has numerous easily explained issues”.
But you did not say what these easily explained issues were.
If you don’t say what they are or withdraw the comment, people will continue to question your integrity.
Tom,
You might be right as far as criticism of MM05b is concerned, but my objective is rather to pay due diligence to Brandon’s “basic truths”, among which we have “but Wegman’s fig 4.4 is not in MM05b” (paraphrasing), which is supposed to refute AT’s position, and the underlying argument that all suboptimal choices were Wegman’s. To that end, it suffices to show that
(a) Wegman was helped by the Auditor in running the code.
(b) Wegman’s fig 4.1 was in MM05b;
(c) Wegman’s fig 4.1 rests on the selection;
(d) The selection was in the SI;
(e) The selection was not advertised properly in MM05b;
(f) The selection has a substantial effect on the figure.
The first four “basic truths” suffice to show that AT’s position has merit, i.e. that the “selection” can certainly be attributed to MM05b. It also shows that Brandon’s wedge between the Wegman report and MM05b does not work, but this is an independent point from the fact that first point. I am consigning (f) because it is basically the HS argument in reverse; I am not sure AT needs to maintain it for his position to hold. If he does, then we have to define “substantial” and this will never end, which I guess is OK for audits.
I think putting (e) on the table would suffice to refute the claim that MM05b is beyond criticism or what not, which is Brandon’s latest squirrel. That is, if we reject, for argument’s sake, that (a)-(d) does not already undermine Brandon’s wedge.
***
So the latest game state is that Brandon needs to show where in MM05b that selection was advertised. For Brandon now claims that:
I can’t find that label. Can you?
> If you don’t say what they are or withdraw the comment, people will continue to question your integrity.
And if you do, they might still continue to question your integrity.
INTEGRITY ™ – We Question Yours
Pekka,
I’m not sure that it has been declared by them. What they say in the paper is
I can’t find anything in the paper that explicitly says that these 100 were the ones with the highest positive hockey-stick-index. As Tom points out, and as I’ve largely discovered for myself, all comparisons – that we know of – between MBH98 and the results in MM05 are taken from that sample of 100. How that is interpreted by others is that the MBH analysis naturally produces a hockey stick that matches what they presented in 1998 even if the data is trendless random red noise. That – in my view – is not the correct interpretation since the comparison between MM05 and MBH98 is always done on the sample of 100 that were chosen because of their hockey-stickness.
Also, to be clear, even in my comment to Brandon I did not accuse McIntyre & McKitrick of anything. I said
which seems consistent with what you yourself have said. The issue is not simply about MM05, it’s also about the Wegman report that appears to perpetuated this idea that you get MBH98 hockey sticks out of trendless random red noise.
I’ll also add what I mentioned above. From their code, it’s clear that they generated their red noise using the original MBH98 data. I haven’t quite worked out what this means, but it seems odd to me that if you’re trying to show that a signal in some data is spurious and depends only on the method, why would you use the original data to produce the data on which you do your test?
Paul,
Actually, that’s the first time someone’s mentioned what I said there as being the problem. I believe I’ve just explained one above. As I’ve also just mentioned, I still am unclear as to why they used the original MBH98 data to generate their trendless random red noise. Why not simply generate an equal number of timeseries using a random number generator? Tom’s also pointed out that much of their statistical comparison seems to rely on “see, it looks the same”. You can also read RealClimate, DeepClimate, and Nick Stokes blogs if you want to learn more.
Also if you think saying
somehow indicates some lack of moral turpitude, you are a sensitive fellow. I get the impression that there’s little I can do to convince some (yourself included) that I have, or do not have, integrity. Nor do I have any real interest in trying. It just seems a bit like a school playground spat than anything else – “I won’t trust you until you say sorry” is something I think I stopped encountering 30 years ago.
However, as far as I can tell, both MM05 and MBH98 are largely irrelevant. I can go to Google Scholar, search for “temperature reconstructions” and find papers published in the last year that appear to have Hockey Sticks. Sure there are details that have changed since the 1990s, but the broad result is the same. The proxies also vary, as do the methods. So, what does it actually mean? It means that the evidence still suggests that the rise in temperature since the mid-1800s has probably been faster than at any time in the last millenium and has increased temperatures to a value higher than at any time in the last millenium (you can of course draw similar conclusions for the Holocene if you consider the work of Marcott et al. 2013).
One more time: MM05 was part of a carefully-orchestrrated campaign, and what they *said* about the paper was quite consistent, that MBH process generated hockey sticks from red noise.
Go back and look at the May 11, 2005 presentation.
John,
I haven’t managed to find that. There seems to be a presentation from 2012 which says that, but not one from 2005.
Pekka:
This is the crucial part of M&M’s description of their methods as it relates to generating PC1s from red noise:
They then go on to point out that inverted hockey sticks are essentially the same as non-inverted hockey sticks for the purposes of MBH98 and M&M05.
You will notice there is no statement of the selection of 100 PC1s by HSI ranking. Nor is such a selection relevant to the method, as they specify the distribution in terms of HSI of the full 10,000 PC1s, the histogram for which is shown in Fig 2. The only mention of the selection of 100 PC1s is where they indicate, “Computer scripts used to generate simulations, figures and statistics, together with a sample of 100 simulated ‘‘hockey sticks’’ and other supplementary information, are provided in the auxiliary material.” Note that no mention is made of the most obvious property of that sample, ie, that they represent the top 1% of PC1s by HSI ranking. Nor is mention made that the PC1 shown in figure 1 was drawn from that top 100 selection.
In the Readme for the SI, the sample is described as follows:
Again, no mention of the most salient feature of the sample, nor any mention of how it is used in the methods of M&M05. Nor is there any description of the samples in the actual file containing the 100 sample PC1s.
That means the only way you can find out about the biased selection of the sample from M&M05 is either by reading the code by which it is collected, or by calculating the HSIs of the sample, and comparing a histogram of the sample to that of the full population. Nor, anywhere do you find a description of the sole methodological use of the selection, ie, as a biased pool of very high HSI PC1s from which to draw examples for graphs.
So, Pekka, your statements seem to be completely uninformed and, what is worse, show a complete absence of knowledge of the actual methodology of M&M05. It gives the appearance of a statement made not because it was informed comment, but because something had to be said to defend an ideologically important document, and it had a vague plausibility to the uninformed.
I had thought better of you.
I have never considered the results of multiproxy analyses essential. One reason for that is a presentation by one of the first Finnish scientists to discuss global warming in public. He spent some effort in explaining, how little bearing these analyses have on the general understanding of climate change. I don’t really know, what was the reason for his emphasis of this issue, but perhaps it’s related to the attitudes of Finnish scientists who have produced proxy series from lake sediments (including Korttajärvi) and dendrology from pines grown in Lapland. Both of these groups have been highly critical of the way their data has been interpreted. The Korttajärvi data was used erroneously on several occasions, and the data from pines of Lapland shows a particularly warm medieval period.
Extracting reliable information from the past centuries by multiproxy analysis is clearly a difficult task. It has been accepted by all (as far as I know) that the methods used cannot reproduce correctly the strength of variability of the past centuries but tend to make it weaker. The question is not, whether that’s the case but, how strong this effect is.
Anders, see Fig 7 of this May 11th, 2005 document linked by John above. I have also come across graphs showing members of the selected sample in two blog posts by McIntyre in 2005/6, and in McKitrick 2005 (to which I linked above). There are a number of links to columns in mainstream media, or presentations in the same period that are currently dead. Clearly, however, McIntyre and McKitrick were very active pushing their “results” at the time and Mashey’s description seems warranted.
As an aside, the decentered principle component analysis with red noise having properties determined by the MBH98 time series produces “hockey sticks”, being defined as having a HSI >1, 99% of the time. The produces a HSI>2 just less than 1% of the time. This does not mean they look like MBH98 hockey sticks, or any sort of hockey stick, although those with HSI near to or greater than 2 always have a vague resemblance to hockey sticks, if not MBH98, at least.
In fact, when M&M05 claim:
it is simply false. The inflection point of red noise generated HSIs are very rarely at the start of the 1902-1980 calibration period. Rather, they tend to be 50 to 200 years before that, even among high HSI PC1s. Among lower HSI PCIs the inflection can be very shallow, or non existent, as shown by Nick Stoke’s sample Further, while the MBH98 reconstruction correlates well with the temperature series from 1902-1980 (r=0.87 for the full reconstruction, the average correlation for the top 100 is only 0.21, with a best performance of 0.59 and 19% being negatively correlated with temperature.
Of course, by only showing PC1s with HSIs near to or greater than 2, M&M at least are able always to show a sharp inflexion, even though that is not a general property of the PC1s (contrary to their claim).
I wrote my comment about the 100 in 10000 based on recollection, where I mixed different issues. Thus it was totally based on wrong recollection. I apologize for that.
I looked only after writing that comment more closely, what the particular paper of 2005 was about. In that paper the show only a single example referring to it by the sentence
That tells clearly enough that what they show is not an average case but an example from some of which bore a quite remarkable similarity. Picking more examples out of the 100 without telling that the whole set is preselected is misleading.
Pekka, multi proxy reconstruction have factors that damp variation (regression to mean) and factors that exaggerate variation (limited number of regional proxies showing greater influence of regional variation). It is generally accepted at Climate Audit that the net effect is to damp variation but that has not been shown, and countervailing factors tend not to be discussed. I do not think it is the view of the paleoclimate community. Rather the view of the paleoclimate community seems to be that the extent of variation or damping shown differs with time scale by method, and by mix of proxy type.
Tom,
The same claim that the methods tend to damp variability can be found in some of the later papers of the main authors. It was presented quite strongly. Unfortunately i don’t remember exactly, what the paper is.
It’s true that the method can certainly also create spurious variability, but I think that such a case would show up in variability that appears rather random.
The main problem is that every proxy series has variability for other reasons than changing temperatures. They have also varying timing issues and/or delays and smoothing. The coefficient between the observed quantity and temperature is fixed by a direct or indirect comparison with instrumental data. While some of the series get a too high coefficient in that the overall effect of determining the scales of many series and combining them to form a final temperature time series is very likely to lose a significant part of the real signal.
You know this comment from Paul Matthews wins this thread – just likely not the way he intended it:
“If you don’t say what they are or withdraw the comment, people will continue to question your integrity.”
Next time Judith Curry comes with yet another vague accusations, please, Paul Matthews, ask her to come with specific examples. Good luck in getting a clear answer (and no, don’t fill out the blanks yourself, *she* must explain what the basis of her vague accusations are and who she means). I’ve tried once when she made claims inconvenient papers were kept out of the literature – and got no response whatsoever. I guess you have just given me permission to continue to question her integrity based on that non-response.
Pekka – focusing on multi-proxy reconstructions immediately after people have shown your original comment about MacIntyre et al to be wrong is a very climate ball tactic, leading things off at a tangent insntead of the main point, which is the issue of lack of integrity [Mod: potentially defamatory].
> That tells clearly enough that what they show is not an average case but an example from some of which bore a quite remarkable similarity.
The blurb refers to “simulations,” that is Monte Carlo simulations. I thought Monte Carlo simulations were randomized:
http://en.m.wikipedia.org/wiki/Monte_Carlo_method
Selection in Monte Carlo simulations should not go without saying.
***
Also, the blurb refers to the simulations. Not a subset of them. It says that they NEARLY ALWAYS give hockey stick shapes. That’s a bit strong, considering Nick’s simulations.
There is “some of which” later, which indicates two things to me. First, that the NEARLY ALWAYS may refer to ALL the simulations. Second, that the text does not make its selection explicit.
Brandon’s label argument lacks substantiation in the text.
Tom,
A quick question that you may no the answer to. In Fig 2 of MM05, they show the HSI for the centered calculation (top) and for the decentered, or MBH98, calculation. For the centered it suggests that the HSI should have some kind of Gaussian-like distribution between -2 and 2. For the MBH98 method it suggests that it should either lie between +1 and +2 and -1 and -2. Nick Stokes’s post shows a random sample of MBH98 (decentered) PCs (4th figure from the top). There’s a tendency for HS shapes, but not that strong. Do you know if Nick Stokes’s calculation is consistent with the HSI for the MBH98 method always lying having an absolute value greater than 1 (as suggested in MM05 Figure 2) or not?
“Its my job, and the job of all academics paid by the government, to protect research integrity.” Perhaps I’m wrong but this reads to me as though Curry is suggesting that government paid scientists act as the watchdogs of “research integrity”! She may be unaware that such organizations already exist at least for the Public Health Service in the form of the Office of Research Integrity. More to the point, if she does have any evidence for misconduct then the appropriate procedure is to contact the appropriate individuals at the relevant funding agency or institution and make her accusation to them in confidence. That is the usual way that whistleblowers can protect themselves from recrimination or the accused from being publicly harmed if the accusation proves groundless. Of course, because climate change is a total hoax orchestrated by scientists, research institutions and universities, funding agencies, journals and even governments, that is never going to be the avenue that the pseudoskeptics will travel. Its much better to make unfounded pronouncements to a naive public to build an argument. That’s where pseudoskeptic integrity lies.
guthrie,
I spent some effort in looking at multiproxy analyses a couple of years ago, mainly in 2010, when I had the most detailed look on the paper of Kaufman et al on Arctic cooling. That was a nice paper to look at, because the small number of proxy series allowed for looking in detail, how they are related to each other and the ultimate results. I thought – erroneously – that my recollections from that time were good enough, but that was not the case.
==> “which is the issue of lack of integrity of denialists such as MacIntyre.”
Oy. Anders – rules of the road?
What I think that people miss is that they really don’t know the reasons why people do some of the things that they do, and that to judge someone’s “integrity” there needs to be a high bar of proof.
This applies to Judith and “skeptics” and it applies to “realists.”
I think that in general Pekka engages with integrity, even though at times he evades directly confronting errors that he might have made. That, my friends, is human nature. But I don’t know Pekka and so I am not in a position to assess his integrity generally, or really even in a specific incident.
To call such a human failing as a sign of lack of integrity is meaningless, IMO, and exploits what might be meaningful discussions about how to evaluate integrity – for the sake of partisan point scoring in the climate wars. Invariably, we see double-standards w/r/t the application of that method of evaluating “integrity.”
—-
I would like a judgement from the commisioner of Climatball.
Does playing Climateball necessarily imply something about intent? In other words, if I avoid confronting a mistake I made (and I’m not agreeing that is what Pekka did here. He may have just moved on to another issue in a sequence that made it look like an evasion when it wasn’t one), and the reason is because I’m human and not because I’m trying to score a goal for a team – is that Climateball?
And when you get done with that, can you describe for me the sound of one hand clapping?
Joshua,
Yes, I agree.
As I understand it, there is a subclass of CLIMATEBALLTM, called INTEGRITYTM. My understanding of the is that INTEGRITYTM falls within the rules of CLIMATEBALLTM.
Joshua,
I don’t think so and “clap”. Everyone can reproduce these results by playing the ball and clapping with one hand.
One can try to appeal to someone else’s humanity, but that gets us all of ClimateBall. For instance, I did not try to play ClimateBall when I asked Carrick to recant his “psychopath”.
You’re a psychopath.
You’re in denial.
You have no moral compass.
You have no integrity.
Are just bad ClimateBall moves anyway for the established viewpoint’s proponents. They can easily backfire and distract you from playing the ball. If AGW is the only game in town, there’s no reason to seek out complications. Simplification leads to a winning endgame.
INTEGRITY ™ – Play like you’re winning
MM05~
We measured the heights of 10,000 men in a town, and almost all are 6’6″” or taller : here’s a sample of 1-12 from the population.
Unsaid: the sample was from a 100-man subpopulation chosen from those found on the courts at college basketball tournaments.
But the graphs were visible and understandable for a general audience, very few if whom would have the background and motivation to look deeper. The graphs are like the smile of Alice’s Cheshire Cat.
Willard, did you ever get around around to explaining what you mean by ‘peddling’?
No, Vinny. Basically, it’s using someone else’s claim to peddle your own pet topic, e.g. alarmism. It’s like running with someone else’s squirrel.
Peddling ain’t that bad, as long as it does not turn into a food fight. And when it does, a pox in all the houses.
***
The most fantastic one is Tiljander:
This could happen here if some ClimateBall player used Pekka’s comment to peddle it here.
Willard and the fantastic Tiljander: ‘Then it gets even more interesting.’ Alas, no, it didn’t.
But I like the idea of running with someone else’s squirrel.
Words of wisdom, Vinny. Words of wisdom.
See how it will play out at Joe’s. I might be compelled to peddle my bitching hypothesis at Joe’s:
https://twitter.com/TLITB1/status/496999893698097152
I have no idea why Leo provokes that round of ClimateBall.
Annoying misquote cropping up everywhere: It’s “a plague on”.
/pedantic git
I stand corrected:
http://memegenerator.net/instance/53355658
I know it’s not exactly the same, but X ALL THE THINGS requires a verb.
Willard
[emoticon self-redacted]
Anders @4:53 pm, as a quick test on your question, I modelled a linear trend of 0.001 C per annum over the interval 1400-1980. It had a hockey stick index of 1.5. Therefore, by M&M05’s definition, a straight line with a positive linear trend is a hockey stick. It might reasonably be claimed that the straight line is only a hockey stick by their definition because of its artificially low variance. It has, afterall, a standard deviation of just 0.168. That, however, is more than that for MBH98 (0.131), and 521 standard deviations greater than the mean of the top 100 HSI PC1s generated by M&M from white noise (mean 0.019: St dev: 0.0003). It also has a better correlation to the instrumental temperature record (r^2 = 0.33) than does the does the average top 100 PC1 from red noise (mean: 0.094, St Dev: 0.094).
I think it is important to remember, in judging the quality of M&M05 as a contribution to the scientific debate, that it entails that a straight line (indeed, any straight line with a finite trend =/=0) is more of a hockey stick than MBH98, and than 27% of “hockey sticks” generated using the decentered PCA method they use to emulate MBH98 on red noise.
Tom,
Thanks. I was thinking about this last night and if I understand this centering properly, then centering on the instrumental time period – rather than the entire time interval – will have a tendency to shift timeseries that are not flat up or down, relative to where they’d be if centered. I can see that having a tendency to amplify hockey-stick like shapes or even – as you say – linear trends.
Something I’m also trying to understand better is how MM05 generated their trendless red noise. It’s clear (from their code) that they used the original MBH98 data. It seems that the intention was to use this to generate correlated noise. However, as mentioned in this Deep Climate post, they may have used a noise model that retained a low-frequency climate signal
If so, it’s not that surprising that their analysis produced hockey sticks, as it would seem to suggest that they were still there.
ATTP,
As I understand that statement of Ammann and Wahl, it means that the climate signal adds to the long tail of autocorrelations, but not directly to the hockey stick shape.
Pekka,
This is what Martin Vermeer said and he was (is) probably right that I don’t quite get this. Whether I quite get it or not, it does seem that if you don’t remove the hockey stick signal, then you contaminate the noise in a way that makes it more likely that you will find spurious hockey sticks. According to Martin that doesn’t mean that the data contains hockey sticks, but I’m not quite sure that I get the distinction.
Okay, maybe I’m slowly getting this. If you correlate your noise using AR(1), then – I think – if it’s annual data, each data point depends on the previous one. Over long enough time periods, however, this shouldn’t produce any trend or signal. What it appears was done in MM05 was to generate the noise using a model that could have a much longer auto-correlation length (decades instead of a year). If so, you can then produce much longer periods where the data is above or below the mean. Hence, it becomes more likely that you might have some anomalously warm (or cold) period at the end of the data (the period representing that last century) than if the correlation period was much shorter. At least, I think this is the issue, but maybe someone who is more familiar with this than me can clarify. This is explained by caerbannog666 quite well, assuming it’s correct 🙂
ATTP,
What MM05 argues is that the method of MBH98 produces spurious results. That’s their basic point, they use the red noise model to prove that point. They don’t claim that the red noise model is correct description of the reality, it’s just a tool for making the point. They use the actual variability of the time series to fix the statistical properties of the noise to add some quantitativity to their argument. Essentially they say that spurious results are comparable to the observed signal, when the properties of noise are fixed in that way. What others like Nick Stokes have found is that the spurious effects are there, but that they are not quite that strong, when less cherry picking is applied to the analysis. Steve McI has responded that they didn’t really claim that the average strength is that strong, but that they have shown that so strong spurious signals are possible with a non-negligible probability.
The real question is not, whether the blade of the hockey stick can be reproduced, it’s whether the method of MBH98 produces correctly the shaft as the argument is not about recent warming but about MWP. In my view the criticism of M&M is basically correct, but is likely exaggerated. I think it’s also pretty certain that MBH98 didn’t study the method used in the paper carefully enough to tell, how significant its findings were. The way they applied PCA is not standard, and non-standard methods tend to be badly understood. Even now, 16 years later, the question of the actual temperature history and the accuracy of the estimates based on multi-proxy analyses seems to be somewhat open. For this reason most climate scientists concentrate on other evidence, and for this reason the issue remains a favorite topic of the skeptics.
The blade of the hockey stick is the instrumental signal. Since the proxy’s are calibrated to the instrumental signal, the fit in that time period is forced. The instrumental period is also the period of most rapid change so it is a good choice for the calibration.
As to the current state of play see PAGES2K, which surprisingly is closer to MBH98.99 than many of the other subsequent resonstructions.
Eli,
PAGES2K would seem to be the crucial point. Arguing about papers published 10 – 20 years ago seems rather pointless given more recent analyses.
Anders, the issue about straight lines relates to the Hockey Stick Index, as defined in M&M05. No matter the slope, so long as it is neither zero, nor infinite (ie a vertical line), the HSI will tell you that a straight line is a hockey stick. That is a problem with the HSI as a purported measure of “hockey stickness” or of similarity in shape to the MBH98 reconstruction.
With regard to the use of the statistical characteristics of MBH98 data to set the autocorrelation function, that would not preserve the climate signal, ie, the actual dips and rises in the MBH98 data which is presumably due to past climate. Rather, it will preserve the statistical properties, in particular the persistence of autocorrelation. That does make a difference in that the more persistent the autocorrelation, the greater the tendency of decentered PCA to generate shapes with a HSI.
Whether or not that is a problem depends on the argument being made. MBH98 argue that the rapid temperature growth in the 20th century relative to other periods is evidence that that temperature growth is substantially caused by anthropogenic factors. M&M05 argue tacitly (at least) that that implication is not justified because red noise of suitable autocorrelation will generate equivalent profiles of flat reconstructed temperature trends prior to the twentieth century and purportedly high trends in the twentieth century. It is not an adequate counter to this counter-argument to point out that the M&M05 pseudo-reconstructions preserve significant statistical features of the climate signal, and to argue that they should have used a reduced persistence, or excluded the 20th century in determining the statistical properties involved. Such an argument assumes that the persistence in the reconstructed temperature signal is in fact a climate signal, and that sharp rise in the twentieth century is anthropogenic. If the MBH98 argument only stands on this assumption, it becomes circular. In contrast, if M&M05 were to argue that because decentered PCA applied to persistent red noise generates “hockey sticks”, it follows that the hockey stick contains no climate signal, they would be making a circular argument by assuming that none of the persistence in the MBH98 reconstruction was due to the climate signal.
Thus, the relevance of the persistence of autocorrelation used in M&M05 depends entirely on whether or not they are arguing that MBH98 have not proven their case (McIntyre’s line as I read it), or arguing that M&M05 show that there is no climate signal in the MBH98 proxies (the popular interpretation of the paper among “skeptics”). I should note that, IMO, MBH98 have a devestating response to even the weaker claim from M&M05. Specifically, they are not assuming the proxies they use have a consistent climate signal. The proxies they used were chosen because they are known independently to have a persistent climate signal. Further, it is also known independently that the persistence of temperature signals that are independent of climate signals is quite short.
This point alone would be enough to render M&M05 largely irrelevant IMO, even leaving aside the massive technical problems of the paper, and the ability to reproduce effectively equivalent temperature reconstructions to that in MBH98 without decentered PCA or even, or even without PCA at all.
I have seen the abstract page of that PAGES2K article, but Nature Geoscience is for some strange reason one of the very few journals I cannot access through my university account. Can the full paper be found for free somewhere?
Tom,
Thanks. What you say here is what I thought,
Yes, that makes. However, doesn’t that imply that there was already some kind of Hockey Stick feature in that particular timeseries? So, decentered has a tendence to generate shapes with a HSI, but can’t do so unless there is some kind of trend in the data being analysed. So, it might amplify a Hockey Stick-like signal but can’t generate one if there isn’t some kind of signal present in the first place?
I’m going to have to think a bit more about the rest of your comment. What I’m understanding at the moment is that MM05 are suggesting that you can get Hockey Stick profiles from random noise if the autocorrelation length is sufficiently long and therefore that the presence of a Hockey Stick doesn’t imply an anthropogenic signal. Basic physics, however, would argue that correlation over long-timescales doesn’t simply happen by chance and hence that the presence of some kind of Hockey Stick-like signal is indicative of some kind of climate forcing. I’m may not quite have got your point though.
Pekka,
I can’t seem to access it either.
It may be that the PAGES2K work has produced powerful enough and well enough understood methods to put a point on the discussion of the validity of the results. That information was, however, not available, when M&M05 was written.
The main results of MBH98 were in the PC1 due to the decentring that forced all those time series to PC1, which had a significant difference between the two averages. That’s true for all those time series that have that difference due to recent warming, but also for all those that have the difference for some other reason. What M&M05 claims is that getting classified to PC1 is a real possibility even for proxy series that are not significantly dependent on the temperature or have a non-representative period over the latest decades (as the Korttajärvi data has). If the number of such “erroneously” classified time series is large that cancels part of the temperature signal of earlier centuries.
Thus the question is: Was the recent warming the dominant factor in determining, which time series ended up in PC1, and with which coefficients? If the answer is “yes” then the MBH98 results have a good change of being essentially correct, if other factors had an equally important role, then the results are not quantitatively correct. (There are other issues that might affect the correctness as well, but M&M05 was about this effect.)
I add that by being classified in PC1 I mean that the related coefficient is large.
Pekka Pirilä says: August 7, 2014 at 9:48 am
“Can the full paper be found for free somewhere?”
The extensive SI is here.
“Was the recent warming the dominant factor in determining, which time series ended up in PC1, and with which coefficients?”
A key question is, how many PC’s to retain. MM05 kept 2. Mann, I think, said it should be 5. I think Ammann and Wahl did that and found off-centering had no effect.
But the E&E MM2005 paper did an emulation of MBH (Fig 1). They were only able to get a difference by discarding 50 years of the Gaspe cedars, using the excuse that they didn’t go back to 1400. (I posted on that here). They start at 1404, and MBH interpolated the missing 4 years. Otherwise it is very similar. After that, their focus shifted to bristlecones etc.
Nick,
Thanks for the comment. I spent quite a bit of time yesterday reading your posts on this topic. Very useful.
From the MBH98 we can read:
Thus PC1 is essential for the global and hemispheric average, because decentring forces it to take this role, while the next four PCs describe mainly spatial patterns of temperature variability.
Much of the discussion skips totally the spatial structure, because taking it into account makes it much more difficult to understand intuitively, what’s going on in the analysis.
Sorry this doesn’t follow on from recent comments. I’ve not done any physics beyond A-level, so I should probably keep quiet, but to this amateur it’s always hilarious to see Dr Curry attacking other climate scientists for their “advocacy.” To my admittedly untrained eye it appears that Curry misrepresents the science before House of Representative sub-comittees, and she implies that taking no action to reduce carbon dioxide emissions is a sensible policy option. Bizarrely, this behaviour doesn’t seem to count as advocacy. (I don’t see any reason to doubt Dr Curry’s sincerity, by the way, although I haven’t followed her blog or her research v carefully).
With the help of Google I found the full PAGES2K article from this site.
Rob,
I think that Willard and Joshua have both been trying to point out to Judith that her claims that she doesn’t advocate don’t appear entirely consistent with what she says and does. She claims that she doesn’t advocate. To me this is an illustration of why the “advocacy promotes distrust” is a rather silly idea. If I say Judith advocates and she says she doesn’t, what doesn’t that mean? It probably means two basic things. One is that defining advocacy is non-trivial and hence until we can do so (and maybe we can’t) arguing that we shouldn’t advocate doesn’t make sense. The other is that whatever people choose to do, they will probably still say things that others will choose to interpret as implying something about their integrity, simply because they disagree with them in general. So, my view is, just be as honest as you can be (i.e., be honest). Avoiding saying something just because others may use it to imply that you lack research integrity is probably silly because they’ll probably find something else to criticise. On the other hand, avoiding saying things that are patently silly is probably a good idea.
> the issue about straight lines relates to the Hockey Stick Index, as defined in M&M05
Here’s how Huybers words it:
http://onlinelibrary.wiley.com/doi/10.1029/2005GL023395/full
What do you think of this definition of hockey-stickness, Pekka?
Willard,
M&M define hockey-stickness as a simple measure of one property of the time series. It compares a 79 year period to the full 580 year period. Looking at this value makes sense because the same two periods are part of the MBH98 methodology.
Concerning the issue of normalization that leads to those three different values of hockey-stickness the summary of Huybers seems appropriate:
At least I do share the view that there were methodological issues with MBH99, but I’m not surprised at all, when it’s found that M&M give an exaggerated picture of those issues.
To the extent I have read, what Steve McI writes, my impression is that he is technically quite competent, but that he gives too much weight for issues that he finds in main stream research. Most of those issues should be answered by the original authors, they are real enough to warrant that, but in general they do not change anything substantial.
Some other scientist than Mann might have taken a more constructive approach in answering what M&M have written and through that reached a much better outcome. I think that this is true independently on the level of good/bad faith in the activities of M&M.
Thanks, Pekka. I agree with what you say.
I asked my question about hockey-stickiness because I believe it’s the basis of such claim:
Click to access APEC-hockey.pdf
It seems to me that this kind of claim rests on their (arguably idiosyncratic) definition of hockey-stickness.
Would you agree?
Anders –
Not to speak for willard…
==> “One is that defining advocacy is non-trivial and hence until we can do so (and maybe we can’t) arguing that we shouldn’t advocate doesn’t make sense. ”
Well said, and that’s the point I’ve been trying to stress. Judith treats the definition of advocacy as trivial, IMO. Her treatment of the issue is largely subjective, IMO. She doesn’t seem to approach the question in a scientific or empirical manner, IMO. And, IMO, she uses the existing related evidence-based/empirical analyses in a selective manner so as to fit with her “motivations” (as in motivated reasoning).
I think that the right and practice of advocacy is crucially important for the betterment of society. I don’t like to see the notion of advocacy exploited as a weapon with which to fight partisan agendas in the climate wars. The way to prevent such exploitation is with careful attention to applying scientific techniques of analysis. It won’t prevent subjectivity, but it will help.
Judith focuses on the importance of recognizing uncertainty and I support that focus. Where I disagree with her is related to what I feel is her inconsistency in how she prioritizes treatment of uncertainty. The inconsistency I see in how she approaches uncertainty related to “advocacy” and “activism” are part of my larger critique of her inconsistent approach to uncertainty more generally.
Because there is a thread over at Judith’s where many “skeptics” line up to conflate questioning Judith’s approach to “integrity” with saying that “integrity” should never be examined – I should make it clear that I’m perfectly fine with examining the pros and cons of advocacy and the elements of advocacy that might be beneficial or detrimental. I think that is an interesting area for exploration. It is because I think it is an interesting area, that I think that it merits a careful analytical approach .
The talk of Gavin Schmidt What should a climate scientist advocate for linked to on his homepage is a much better discussion of advocacy that what I remember from Judith Curry. I didn’t listen to the talk again. Thus I don’t remember in detail, what I liked and what not, but that I do remember that my basic impression was positive.
Again, I disagree with Pekka. Whilst we can all agree MBH98 isn’t perfect, this from Pekka is just wrong:
“Some other scientist than Mann might have taken a more constructive approach in answering what M&M have written and through that reached a much better outcome.”
Nothing short of an “Oh dear the paper is totally flawed and I retract it and all my work showing a hockey stick shape”, would have been acceptable to M&M. So expecting some sort of ‘better’ outcome suggests a rather different idea of what M&M and their behaviours than I think the evidence warrants.
M&M are not the supreme judges. Thus the requirement is not satisfying them, but answering more specifically the points they made.
It’s surely quite difficult to figure out the actual situation at the time M&M presented their various claims. Thus I may err in my judgment on what would the outcome of alternative approaches have been, but that’s how I see it in hindsight.
Pekka, when you say “Some other scientist than Mann might have taken a more constructive approach in answering what M&M have written ” what planet have you spent the last decade on. McIntyre and McKitrick were never going to accept any answer from Mann.
That is not to say that Mann is not stubborn, but the roots of that are quite clear. MBH was Mann, Bradley and Hughes. Bradley and Hughes were the heavyweights. Everyone went after Mann who was a post doc at the time of publication and then an Asst Prof at UVa. That he survived was a miracle. That the gauntlet brought out his combatativeness not a surprise.
WRT the interesting thing here. There are much better methods of defining and characterizing hockey stick like data. Moreover mm05’s definition is a Tjlander in the making. The sign of the blade’s slope is not a matter of indiference.
Eli,
Yes, I agree. That amazed me. There’s no question that I wouldn’t have survived. FWIW, I also agree with Guthrie. I also think that all this discussion of behaviour is a distraction. It would be wonderful if everyone behaved impeccably. That this isn’t always true doesn’t suddenly make their research questionable, or make research in their field questionable.
Even this post has generated various accusations that I lack integrity. This appears to be based on a comment I made more than 6 months ago that was factually incorrect (acknowledged) and because – from Paul Matthews – I wrote (parenthetically) that there were numerous easily explained issues with MM05. Firstly, I fail to see what is so atrocious about saying that, as it doesn’t immediately say anything negative. It may well be self-evidently true of almost any paper. Secondly, what seems clear is that it’s not that I should explain this, it’s that I should explain it to the satisfaction of my critics (which may well be impossible). Much of these type of criticisms in the climate debate are essentially “you didn’t behave in a way that I regard as acceptable, therefore you lack integrity”.
Since I’m on a roll, I’ve just remember that I had a Twitter exchange with someone who was complaining that some of the Cook et al. data was unavailable. I eventually convinced them (by providing a link) that the data they wanted was indeed available. The criticism then became “but it’s not in the order that I want”. I then pointed out that it could be re-ordered using excel (I may also have pointed out that one of my children could probably help if they didn’t know how). The criticism then became “but that’s going to take me time to do”. So essentially it went from “data doesn’t exist”, to “data exists but isn’t in the right form”, to “I don’t have the time to do the work that’s needed”. It just seemed like a mini-illustrating of much of the criticisms in the climate debate.
Joshua,
Your comment made me think a little more about my issues with this type of thing. I agree that these types of things are interesting and worthy of study. However, my general view is that it;’s impossible to research these things (and I would include any study of how biases or behaviour might influence research) in a truly independent way. Those doing the research are not formally independent and much of what they’re studying may (and probably does) influence them as well as those they’re studying (as you yourself indicated).
So, what I think I find myself responding to negatively, are those examples where someone appears to suggest that they are somehow able to study these things in an objective way. I would regard that as impossible and hence that anyone who doesn’t recognise this as either lacking in self-awareness or lacking in understanding of what they’re actually studying.
Eli,
I know that sometimes those opposing conclusions of main stream scientists don’t follow “the rules”. Responding in kind to them may, however, be a counterproductive choice for a scientist. The only real strength of science is its truthfulness, and that should be given even more weight in the choice of the way of responding. It’s better to respond to claims by people like M&M in the same way as a substantially similar argument were answered when presented by a established scientist. In this I assume that there’s real substance behind the argument, while it may be formulated differently from the way the scientist would have formulated it.
My impression after all the discussion is that there really was substance behind the arguments of M&M05. The arguments may have been exaggerated and partly erroneous, but they were not obvious nonsense, but worth responding in more detail and differently than they were.
omfg i am a layperson freaked by physical events happening in this world, with some hope science will fix things in time…and after reading ya’ll..shit. no damn way. if scientists are this bogged down in crap that doesnt matter, nada is gonna b done in time. kill.me.(and all of us) now.
Here’s another instance where hockey-stickness matters:
I don’t think that “introduces a small bias” would carry the same weight as “nearly always produces a hockey stick”.
TC:this was more or less covered in earlier post but WebCitation seems down right now, but Wayback machine has the right items anyway. So, to cover the provenance:
MM05x was the original PPT, which Peter Spencer gave to Wegman ~Sept 2005, and Wegman sent to Vergano in November 2010.
MM05y (via Wayback) was posted on the GMI website by Feb 22, 2006, and it is a PDF that completely matches the PPT above, but converting variable date into August 30. 2005. PDF Properties tells us Mark Herlong (GMI) created it 08:30PM on August 30, 2005.
This date is actually important, because from Yasmin Said’s silly 2007 talk, slide 3:
““Dr. Edward Wegman was approached by Dr. Jerry Coffey on 1 September 2005… After the initial contact, Dr. Wegman received materials and a visit from Congressional Staffer Peter Spencer.”
While it cannot be proved, a plausible inference is that Spencer, Coffey and Herlong cooperated to get a PDF produced for Coffey to show Wegman, and then Herlong gave the PPT to Spencer, who supplied it to Wegman … Wegman and Said showed zero competence in understanding the papers (as documented in excruciating detail), and the easiest thing to read was MM05x … and its memes propagated into the Wegman Report with no challenge.
MM05z is an annotated version of MM05x, with transcript. By PDF Properties, this was created by GMI’s Mark Herlong on 06/17/09, and at some point replacing MM05y.
(When I was writing Strange Scholarship in the Wegman Report, this caused some confusion.)
p.5 gives the date of talk: May 11, 2005
It is well worth reading the transcript around the slides, to see what they were actually saying, but especially p.13:
“To show how strong this effect was, we wrote a paper published not long ago in Geophysical Research Letters where we fed a type of random numbers called “red noise” into the program and we showed that it reliably produces a hockey stick-shaped first principal component, even when there is no trend in the underlying process. It is just a curious thing. …
Figure 7 shows four graphs. Three of them are examples of “red noise” yielding hockey sticks from the Mann algorithm and one of them is the actual proxy data portion of their temperature reconstruction. If you have trouble telling which is which, that is the trouble. “
(But, read the whole thing. How many of the audience in Washington DC would get the idea that the upward blade of the hockey stick was an artifact of the MBH shorted-centered PCA, and thus non-existent in the real world? How many would have had the paleoclimate knowledge and stats to delve deeper? Sadly, the 2005 transcript does not identify the questioners, unlike one of the earlier GMI meetings that show Singer and Michaels effectively coaching McIntyre and McKitrick in 2003. Do read Michaels on p.23 and Singer’s comment on p.25. The first is the first I’ve found of the Pal Review meme, and the second the complaint about combining reconstruction (shaft) with blade (instrumental).
Well, this is a thread on integrity…
@- ATTP
“So, what I think I find myself responding to negatively, are those examples where someone appears to suggest that they are somehow able to study these things in an objective way. I would regard that as impossible and hence that anyone who doesn’t recognise this as either lacking in self-awareness or lacking in understanding of what they’re actually studying.”
For some reason this makes me think of the ‘Mike Hulme’ type of view of science that Dr Curry seems to favour, that it is shaped by social, political and personal contexts. From that POV it is a short step to view any advocacy {as expediently defined} is a clear sign that the science promulgated by those scientists is not objective.
According to this analysis, Hansen, to take a real example, who strongly advocates for a policy response to the implications of his research MUST be subject to strong doubts about the integrity of his research BECAUSE of his advocacy. The two are not separable.
Advocates of this view, that social values are always embedded in the science, seem to believe that they arrive at this conclusion by objective means. It forces them to doubt the integrity of any research published by someone who is also an activist for one side or the other at the policy level.
Dr Curry may seem to get of more lightly than Hansen for presenting science to politicians.
Gavin Schmidt’s remix of Stephen Schnieder’s take on science and personal advocacy seems to have covered the subject best.
Here are links to the sources and the responses when this was raised at the end of last year.
http://www.realclimate.org/index.php/archives/2013/12/agu-talk-on-science-and-advocacy/
Or just a transcript.
http://www.yaleclimateconnections.org/2013/12/agu13-gavin-schmidt-speaking-up-and-speaking-out/
Rather like the dissection of MBH98 this does seem like déjà vue all over again….
Michelle,
Well, that may be a bit extreme 🙂 There are many scientists who are not bogged down by this nonsense and who do so by simply ignoring it. I’d also quite happily be not writing this blog and maybe by doing so am contributing to the nonsense but would like to think I’m helping to refute some of it. Doesn’t always work, but maybe it sometimes does.
Izen,
Indeed, that was exactly the kind of thing I was thinking of.
So we can assume Dr Curry has a hate on for M&M?
Has she commented the fact that they didn’t detrend, and for that matter, what detrending is?
Detrending is removing the effects of water from tree ring growth. Apparently folks who study this actually think trees will grow more if you feed them water. M&M are the only humans on earth who disagree and have published that belief. (I don’t know about you, but I think they are wrong. I think plants need water.)
Pekka, agreed, the value was pointing out that the proper PCA centering was superior and that there are some issues with some of the proxys, however, contrast Nick’s treatment of the Gaspe issue with McIntyre’s. A rational conversation would go, hey, start your analysis in 1404 not 1400, oh yeah is the response, that makes things clearer. . .
> A rational conversation would go,
A conversation simpliciter would go:
[Nick] where is the M&M equivalent of Amman’s fig? And if there is one, how come Wegman didn’t know about it?
[The Auditor] […]
Almost anything but crickets, squirrels or flames would do.
Source: http://wattsupwiththat.com/2011/06/02/manns-hockey-stick-climategate-and-foi-in-a-nutshell/#comment-674924
Krugman:
My emphasis. Don’t treat any of this sort of thing as new, or limited to climate science, or not motivated by right wing politics.
I haven’t read them, but Perlstein’s books sound like a good entry point for those who need to learn about the development of this tendency in U.S. politics.
As Eli surely knows, there might have been more value in pointing out value of centered PCA if it had been done in 1999, rather than in 2004/2005, given, from Wikipedia:
“The original MBH98 and MBH99 papers avoided undue representation of large numbers of tree ring proxies by using a principal component analysis step to summarise these proxy networks, but from 2001 Mann stopped using this method and introduced a multivariate Climate Field Reconstruction (CFR) technique based on the regularized expectation–maximization (RegEM) method which did not require this PCA step. A paper he published jointly with Scott Rutherford examined the accuracy of this method, and discussed the issue that regression methods of reconstruction tended to underestimate the amplitude of variation.”
John – you mean that Mann and co-workers knew/ found out that the original MBH98 wasn’t so good, and actually improved things?
But why come along years later and slag off the original paper instead of tackling the new stuff? They can’t have been playing politics can they?
Which of course takes us back to integrity, and [Mod: potentially defamatory].
guthrie: statistical methods evolve, unsurprisingly. John Tukey had some good words as well as some influence where I used to work. 🙂
The original PCA, even decentered, was “good enough” given the particular datasets and the fact that it didn’t apply to *all* the data.
MM05 was just one more round in the endless attempts to discredit the hockey-stick’s blade (instrumental) by all this statistical stuff about the reconstruction (shaft). Some people seem to think that if there were any error, even minor in the shaft, that the blade was a hoax.
I’d always wondered how an obscure Canadian economist (and later McIntyre) got totally plugged into the Washington thinktank scene by 2001, introduced to James Inhofe, brought to Washington multiple times, hooked into CEI + GMI, coached by the best, gotten front-page and multiple exposures in Wall Street Journal, etc, etc … but fortunately, I added the 2nd edition of “Taken by Storm” to my collection, and lo and behold, p.66 told more than the 1st edition.
I’d prefer people didn’t accuse others of lacking in integrity without being clear what they mean by the word integrity. Many people take the word integrity, especially when it’s preceded by “scientific” or “research”, to mean misconduct. Or things like fabricating data and plagiarism. These are not accusations to be made lightly on someone’s blog. Firstly, the blog owner can be sued for defamation if the accusations are false, and also, there are proper channels to use for such things and this is not one of them.
AnOilMan @7:25 PM, tree ring widths are in fact detrended when you want to extract a climate signal. The reason, however, is not related to water. If it were, that would indicate that rainfall was greater in the past in all regions, which is not true. Rather, as trees grow, they their tree rings become successively smaller. It is this biological noise from the aging of the trees that is removed by detrending in order to extract the climate signal.
That aside, this is the first time I remember seeing McIntyre and McKitrick of not using detrended tree ring data in their criticisms of MBH98, and particularly those criticisms in M&M05. Do you have a source for the claim?
Above I said of M&M05’s claims that “A sharp inflection was regularly observed at the start of the 1902 – 1980 ‘‘calibration period’’” was “simply false”. I can now put some numbers to that statement. Specifically, I tested for the inflexion point by finding the start year of the strongest 50 year trend in the 1850-1949 interval (ie, the start years 1850-1900).
This is not a perfect test. For instance, (based on visual inspection and pixel counts) the inflection point of the PC1 shown on panel ten of the Wegman report graph (x5631) has its inflection point around 1717. That on panel seven (x5989) has its inflection point around 1831. Because these inflection points occur earlier than the test range, they were not picked up, and a later date assigned to the inflection point of these series (1877 and 1870 respectively). What is more, with such early infection points the overall trend from the inflection point to 1902 is gradual, and small fluctuations from the trend may have caused the test to pick a later point in any event. The test will not indicate too early a date for inflection points, however, both because late inflection points imply a higher trend more readily distinguished, and because by visual inspection it is clear that the inflection points of MBH 98 and MBH99 (1899) follow that of all top 100 HSI PC1s from M&M05.
Given these limitations, however, I can report based on my test that the inflection points of the top 100 PC1s have a mean of 1874.8 and a standard deviation 9.65 years. The distribution is left skewed (ie skewed towards earlier dates). The latest inflection points in the top 100 are x7413 and x8943, both in 1895. The MBH98 & 99 inflection points lie 2.5 standard deviations above the mean of the top 100, and the start of the calibration period lies 2.8 standard deviations above the mean. Looked at another way, just 25% of top 100 PC1s have an inflection point within two standard deviations, and just 4% have a calibration period within 1 standard deviation of the start of the calibration period.
This is just one of a number of statistical tests on the performance of red noise in generating “hockey sticks” that M&M05 simply did not report (and may well not have done). It is different in that they make a claim contradicted by the test when performed, based I presume on the one “statistical test” they did perform – the “does it kind of look like it” test.
A challenge.
If you guys al think MBH is robust, reproduce them.
Yeah I know some claim to have done so, but I mean here, online. Using only the data and code available publically and from the very start – i.e. first selecting the proxies.
Publish each step, the inputs and the results for comment.
If they stand up then it should be reasonably straightforward as a crowd type excersize ?
Morph – that’s unnecessary. We’ve got better, newer scientific observations and work which show the hockey stick shape. You might as well ask us to reproduce the Millikan oil drop experiment. In science, there’s no need to endlessly repeat older experiments when newer methods and capabilities have been invented and give much greater accuracy and precision.
Morph, why should we do something that has already been done and reported in the scientific literature?
Also, what is the use? There has been a significant increase in available proxies, better understanding of methods to be used to extract information, etc. etc.
Morph, why do we need to do what has already been done before?
Further, you are missing the point. I (at least) and I believe few others here would use MBH98 as an estimate of past NH temperatures. It is an obsolete paper some of whose methodological choices could be improved upon (and have in more recent studies). It was also working on a very limited proxy set. That does not mean it was not very good for what it was, ie, a first effort at quantitative reconstructions of past temperatures. Modern reconstructions still show temperature ranges that fall within the MBH98 uncertainty range always, or almost always. However, the model-T Ford was very good for what it was as well (ie the first mass produced car). Doesn’t mean I want one for my familly car.
The point of this discussion is that the “skeptical” discussion of MBH98 is premised on very poor analysis, and even when correct has massively overstated conclusions. The interesting and valid criticisms of MBH98 had already taken place before M&M put pen to paper, including by Mann himself. Further, the method of MBH98 had already been abandonded by Mann and others based on those criticisms in the literature.
So, having made the point clear, do you agree with it? Of do you follow McIntyre and McKitrick in thinking “it kind of looks like it” is a valid statistical test?
Eli, innocent bunny that he is, inquires if anyone has ever asked McKitrick for the complete set of 10K red noise samples?? Would that not be a FOIA of fun?
The Auditor has a category for W & A:
http://climateaudit.org/category/mbh98/wahl-and-amman
Around 50 blog posts.
***
The Auditor has no category for H05, only a tag for all stuff Huybers:
http://climateaudit.org/tag/huybers/
Three posts for H05.
Hmmm. Big hmmm.
Since I’m on the road to hell, I might as well recall the last ClimateBall episode where I mentioned Huybers:
This was my conclusion from a discussion at Judy’s about mediation. The Auditor appeared in the comments to recall his W&A war story. I asked him why he has not told his other one about Huybers, whom he considers “serious”. He told me he had forgotten about it. There was also a big difference between the mediation promoted by Judy and the unilateral collaboration offered to Huybers.
The road to hell showcases lots of non-serious distractions.
Anyway.
Willard,
I actually don’t know much about Steve McIntyre’s responses to Ammann & Wahl and to Huyber et al. or what happened as a consequence. Is it interesting enough to delve into, or best avoided?
Dear ATTP,
Just a tip, no need to constantly state that something is – your opinion -.
It’s your blog, we know – and this is just my opinion – it’s your opionion.
victorpetri,
I’m not convinced you’re trying to be helpful. I’m not doing it to make it clear that it’s my blog, I’m doing it to make it clear that it is my opinion and that it is not something that I’m stating as fact. Although you may find it hard to believe, my opinion can change, or be changed.
@ATTP
It’s a blog, not a scientific article, it is about your opinions, by the very definition of a blog.
> Is it interesting enough to delve into, or best avoided?
I guess that depends upon what you find interesting, AT.
I don’t think this is mandatory reading. Neither seem to think CA readers if you look at the timestamp of the comments of the second blog post about H05:
There’s a strange bump in the comments between 2005-09-19 and 2006-06-10.
***
I never read them [W&A05 and H05] thoroughly, except to get an idea of what was discussed at the Auditor’s or in the beloved Bishop’s political hit job. There’s no need to get into any of this, unless you want to judge MM05b, like PaulM friendlily reminds you. Most Climate Ball ™ players seem not to care much.
I returned to Huybers because his name got mentioned in the Wegman thread at Nick’s, e.g..:
http://moyhu.blogspot.com/2011/06/effect-of-selection-in-wegman-report.html?showComment=1308139675520
TCO has more colourful comments on honour and INTEGRITY ™ that I won’t repeat here. But if “reading the blog” teached me anything, it’s to trust his instinct. For all his faults, TCO has a good intuition when the Auditor throws a smokescreen. If you want more formal criticisms, you should check for pete, who made the Auditor delete a thread he seldom if ever mentions nowadays.
***
If you want to say something about MM05b, you might need to read the relevant discussion. What interest me, for now, is to pay due diligence to Brandon’s posturing that all criticisms of MM05b are to be put on Wegman’s reading. It made me discover that the “basic facts” he tried to impose on you rest on a very peculiar reading of MM05b.
But now I’m onto something that Huybers himself dismissed:
That choice makes sense to a climate scientist that would prefer to improve climate science methods. I’m not a climate scientist. As a ninja, I’d rather improve ways to argue for the significance or not of the bias that MM05b showed.
I thought this was quite relevant: such parsing efforts are the most expedient way to hell. If you want me to stop bringing this here, I will. Only now do I realize what my tumblog should have been about.
Auditors are quite good at parsing. Editors are better.
Tom Curtis: I do hate getting called out by you guys. 🙂
Its a well known fact that I make a lot of mistakes, but I never get called on it by ‘them’. Like, never ever, ever.
> It’s a blog, not a scientific article, it is about your opinions, by the very definition of a blog.
Well, VictorP, here’s your chance. Go see Brandon:
Tell him that they are all opinions, including his own impressions, since both AT’s and Brandon’s claims come from blogs.
If you can tell me how the last sentence is connected to the previous ones, that would be nice too.
While this is now !4 years old, for context, see a few pages from Strange Scholarship in the Wegman Report:
p.5 brief background
p.13-17 Background
p.20 Graph from Wahl and Ammann (2006), Wegman doing his best to discredit it
pp.25-32 From 1998 GCSCT meeting (American Petroleum Institute, and then to McIntyre & McKitrick’s getting plugged in with Washington thinktanks, Inhofe, Barton’s folks, etc.*
pp.134-142 fanciful graphs
*In 2010, I couldn’t figure out how McKitrick got connected with Myron Ebell (CEI, Cooler Heads Coalition) to talk for Congress 10/11/01. But years later, I got the 2nd Edition of Essex & McKitrick “Taken by Storm” and on p.66 we find:
“Around this time, in March 2001, Chris was helping organize the Nerenberg Lecture at the University of Western Ontario. That year’s speaker was S. Fred Singer. ”
The UWO Nerenberg Lecture is run by the Applied Mathematics dept (where Essex is), described as:
“The Nerenberg Lecture Series recognizes accomplished people having extraordinary and authentic things to say to a broad audience on the great ideas of our age relating to science and mathematics.”
I found most of the lectures, which looked interesting, by strong people … but also:
2001: Fred Singer’s talk, Wayback 06/05/01 from McKitrick’s website at U of Guelph.
2009: Sallie Baliunas
2012: Monckton delivers Nerenberg lecture, which was at
http://www.uwo.ca/sci/news/nerenberg_2012_monckton.html, but seems to have since disappeared. Searches for monckton at UWO website seem to find nothing relevant any more.
You can watch the lecture at YouTube It did get mentioned elsewhere.
Although, John, based on that evidence I would think some pre-existing connection was just as likely. McK.’s Cornwall affiliation implies deep-dyed wingnuttery.
Steve,
His apparent signing of the Evangelical Declaration on Global Warming would also indicate that he’s already decided on his beliefs with regards to global warming. I’m always amazed that those who claim to be interested in studying biases that might influence someone’s research never seem to mention those that seem to be explicitly stated. Just because someone has stated a bias, doesn’t mean that it’s not a bias.
See McKitrick section of pp. 138-139 of PDF at Crescendo to Climategate Cacophony.
Note “A call to Truth: Prudence, and Protection of the Poor: an Evangelical Response to Global Warming” but that was (2006).
It was obvious in 2010 that McKitrick had been well-connected for a while, and I think it highly unlikely that he came to his views only *after* Singer’s visit … but the issue was how an obscure Canadian economist got totally plugged into the Washington DC thinktanks, got paid trips there, etc, etc … something rarely afforded to the Cornwall signers or OISM signers or many others.
I’d be happy to be able to push the evidenced date earlier, but so far, have not been able to.
Yeah, it would be hard to find direct evidence going so far back, but his 1996 thesis (the earliest Google Scholar hit) on carbon tax modeling seems like the sort of thing that would have gotten attention from the DC libertarian ‘tanks (which IIRC at the time were pushing cap-and-trade as an alternative to a tax).
Then he was co-author on this attack on endangered species protections in 1999 and started in on his economic activity/temp trends stuff in with this in 2000, given which it would have been amazing if he hadn’t come to their attention well before Singer’s appearance. IOW such a range of activity by a young economist at that time is anything but obscure in the context we’re discussing.
His more overtly political stuff (e.g. multiple attacks on Kyoto) does seem to be post Singer’s appearance, but 2001 is also the year he got tenure.
McIntyre having sought him in particular out in 2001/2 would have made perfect sense given this background, although I’m sure there’s more to that story too than we’ll ever know.
Couldn’t resist looking a bit more:
Actually his first attack on Kyoto seems to have been this in March 2000, although this was the academic rather than the overtly political arena. But get a load of the abstract:
All the stages of denial in one neat package!
No topic is listed, but I notice that two months after giving the above talk he appeared at the University of Chicago Department of Economics to give another. I’m going to guess the topic was a repeat. In any case, for McK that appearance was truly hitting the big time.
Finally, I notice a 1998 comment on SO2 trading.
All of this may not have been a careful plan on his part to get the sort of attention he did in the early 2000s, but it’s exactly what one would look like.
John,
I’m not sure if you indicated somewhere that Wegman also signed this Open Letter to the Secretary-General of the United Nations:
http://scienceandpublicpolicy.org/reprint/open_letter_to_un.html
The usual line has been “but this was in 2007”.
***
There’s also the whole starting point of the Bishop’s story:
http://rabett.blogspot.com/2011/04/exposing-herself-to-art.html
That our beloved Bishop, the genial Lucia, and the Auditor use different words to refer to the private mailing list where the Auditor got in touch with Hans Erren, McKitrick, and alii, sounds quite strange. At the very least, it tends to show that the “I’m not a sceptic” line depends on what “sceptic” means.
Oh, did McI say the mailing list was the point of contact with McKitrick specifically? Did he say when?
Steve: yes, the history is there … but I wouldn’t be surprised if there were plenty of such around.
Go back to GCSCT 1998 for which a key piece was:
“Identify, recruit and train a team of five independent scientists to participate in media outreach. These will be individuals who do not have a long history of visibility and/or participation in the climate change debate. Rather, this team will consist of new faces who will add their voices to those recognized scientists who already are vocal. ”
So far, there is no evidence of contact between McKitrick and Washington thinktanks earlier, whereas the sequence and timing are very suggestive:
a) *Applied mathematician* Essex invites Singer, always alert recruiter
b) Singer meets McKitrick
c) ~6 months later, McKitrick in Washington, and later Essex+McKtriick (for Taken by Storm)
Again, I’d love to see an earlier connection than whenever Essex/Singer happened.
Willard:
Wegman: yes, SSWR p.66, part of A.3 Time to Move On?
Since science is about examining the evidence, and “Andthenthere’sphysics” stated that making accusations about someone is “something for which – in my opinion – you should have almost watertight evidence.”, I’d like to examine the evidence and offer another point of view.
Fact-Judith Curry didn’t use the term “personal integrity” or “professional integrity” and yet the arguments seem to be directed at her as if she did.
Fact-“Integrity” means more than just moral uprightness, or honesty and virtue.
It ALSO means-#2 “the state of being whole and undivided.”
“synonyms: unity, unification, coherence, cohesion, togetherness, solidarity
antonyms: division
#3″ the condition of being unified, unimpaired, or sound in construction.”
“the structural integrity of the novel”
When the structural integrity of a building or ship fails, for example, we don’t assume that the building or the ship were dishonest or lacking in moral uprightness, because that’s not the definition that applies in that situation. Likewise, why wouldn’t every scientist feel it a personal duty to protect the unity, coherence and soundness of scientific research? Especially in the case of “government scientists” whose research and discoveries are paid for by the public and intended to benefit society as a whole? Why would any scientist, no matter who employs them, want to divide or cause discord in scientific research?
Andthentheresphysics” said-“There’s a reason we often use the term research community.” A community is defined as “a feeling of fellowship with others, as a result of sharing common attitudes, interests, and goals.” I cannot fathom that the scientific research community wouldn’t share the same attitude, interest or goal of protecting the “integrity”, soundness, unity, coherence of the research it does….
It is only my opinion, but based only on the tweets presented and the response by Andthentheresphysics, there seems to be a lot of assumptions about Judith Curry’s motivations, personal beliefs and thoughts presented, but I don’t see “almost watertight evidence” at all.
I’d say that Chicago appearance is even more suggestive, John, but opinions will vary.
After a quick search I see that CEI had him at a May 2000 DC briefing, just after Chicago. There may have been prior contacts, but that March talk seems to have made him a star.
SourceWatch has him back in DC for CEI in October 2000.
A quote from an interview with Malcom Hughes- In Conversation-2007-
http://www.abc.net.au/radionational/programs/inconversation/malcolm-hughes/3399750#transcript
Regarding MBH99
Robyn Williams: “And it’s pretty clear that as the end of the hockey stick was going up you’re looking at the affect presumably of the industrial revolution and the human input really?”
Malcolm Hughes: “Well that’s what folks have then gone on to try to test in two ways. First just to see you know again ours was the first attempt and on the paper that attracted most attention, which was published a year later in 1999, the title contained the words ‘inferences, uncertainties and limitations’. But nobody seems to want to know about that; particularly those who wish to try to make political capital one way or the other out of our results. But they’re there and they’re in bold and they’re about 16 points, you know.”
Hughes willingly admits and points to something that rarely, if ever, comes up in discussions about the original hockey stick paper. There were 16 points in the original detailing the “inferences, uncertainties, and limitations” of their analysis…which makes it truly odd that anyone would view it as “settled science” of any kind.
Thanks, that earlier CEI connection is clear evidence, and pushes it back a year, good stuff. I’ve been accumulating updates for SSWR and that will go in.
Now I wonder how the Essex-MCKitrick connection happened and when.
Aphan,
Do you mean by me, or do you mean by others in the comments?
Aphan,
I don’t think anyone has, at least not here.
What it did settle is that there wasn’t large-scale unforced climate variability for the period studied. It didn’t do so with precision, which is the sense in which it’s not settled science, but it did do it, as has been confirmed by subsequent work.
Assumptions, Aphan? I think not. She has provided loads of evidence. Read back through all the relevant posts and comments here for the details.
If you see a specific claim you can prove is wrong, fire away.
John, Essex has contrarian pubs dating back to 1991, plus Guelph and London are university towns not far apart, so it would have been hard for McK. to not know about Essex.
ATTP:
McKitrick is claimed as a signatory by The Cornwall Alliance itself. That page is missing from their recently revised website, but the wayback machine comes through for us: Prominent Signers of An Evangelical Declaration on Global Warming.
Further ATTP:
Assuming they’re sincere, signers of that document aren’t just biased toward AGW denial, it’s literally an article of faith with them. The Evangelical Declaration on Global Warming is unambiguous (my emphasis):
…
If “Science is a way of trying not to fool yourself” (Feynman, who else?), then scientists who sign the Declaration are announcing their willingness to fool themselves rather than believe that their deity would allow AGW. To accept the scientific consensus, they’d have to abandon a bedrock tenet of their religion!
OTOH, there are principled scientists like Katharine Hayhoe, who professes Evangelical Christianity but is willing to accept what her science is telling her, and is reaching out to her coreligionists on the urgency of AGW. I think that illustrates how personal religious belief can be, denomination notwitstanding.
Steve: yes, I’ve visited that area a dozen times (ie, which included London, Waterloo, Guelph and Toronto (where McIntyre is ), and driven around, so it’s obvious they are handy. The question is: how far back can one *prove* connections.
You’ve helpfully pushed the MCKitrick:thinktanks one back at least a year or two, so if you come up with any more or Essex/McKitrick it roll be interesting.
(we’ve got houseguests on top of 3 days of Hot Chips Conference, so I’m mostly out of it, limited to iPhone :-))
Mal,
Precisely!
Tom Curtis: I don’t think I’m out of my mind. I did read up on tree rings, and a few papers papers on the subject. Several dealt with with tree rings that needed moisture detrending.
This isn’t the same, but;
http://en.wikipedia.org/wiki/Dendrochronology
“Adequate moisture and a long growing season result in a wide ring.”
Here’s one paper trying to discern the impacts of moisture;
Click to access tree_rings.pdf
I don’t normally worship but when I do worship, I worship at the Church of Batman the Redeemer.
Mal,
Thanks. Useful. I had linked to that a few times in the past, but only just noticed that it was no longer available.
MBH99 Figure 3 on p.761 says “two standard error limits (shaded)”, and the shading persisted into IPCC TAR.
Analogy:
a) the average is 100, +/- 1
b ) Aha, the average is .99, so a) is WRONG, WRONG ,WRONG. throw it out
McKitrick is also a very smooth presenter. I wonder how much that’s a factor in his usefulness to bodies like CEI.
> Judith Curry didn’t use the term “personal integrity” or “professional integrity” and yet the arguments seem to be directed at her as if she did.
Here’s the first sentence with the word “integrity” in the relevant post at Judy’s:
We should survey scientists to know if that part of Judy’s job can refer to something we might call “professional integrity”.
INTEGRITY ™ – Parsing Our Way to Hell
> they’re in bold and they’re about 16 points, you know
That’s 14 point too small:
http://blog.guykawasaki.com/2005/12/the_102030_rule.html
Considering our guesstimated average age for Judy’s Denizens, I’d say 30 points is a bit too small.
His Cape covers us all.
40+, easy.
John, McKitrick’s record of public activity really does make it seem like he was waiting for tenure before piping up about AGW (although there was also the imminence of the TAR as an impetus), so I wouldn’t expect to find a public trace of the Essex connection. (I did look a little.) Note that for the March 2000 talk, McK credited three people with help (and who are they?), *not* including Essex, so I think that’s some evidence that the substantive contact was after that. That McK, as a junior economist, needed someone like Essex to give his output scientific weight seems obvious enough. The (not very) subsequent Donner book connection seems more interesting, although I haven’t read your material on that.
willard writes: “That’s 14 point too small:”
The “16 points” refers to the number of caveats in the paper, not the font size.
Kevin O’Neill, I think the 16 points refers to the font size of the title, for which I don’t think Willard’s rule of thumb is applicable 😉 The paper does use the text string “uncertain” eight times outside of the title, the text string “infer” once outside the title, and the text string “limit” three times outside of the title, for a total of 12 uses. It also has six sections including the abstract, and twelve paragraphs. I don’t think any of these statistics can be stretched to mean “16 points”. I think the reference to the font size is an ironical attempt to draw attention to the fact that M&M effectively ignore a quarter of the discussion in the paper, ie, that part dealing with the uncertainties and factors that effect them.
Willard/Kevin/Tom-
Why Hughes would be talking about 16 point font size is beyond me, but hilarious.
ATTP and Steve Bloom- On assumptions, what I said specifically was-
“It is only my opinion, but based only on the tweets presented and the response by Andthentheresphysics, there seems to be a lot of assumptions about Judith Curry’s motivations, personal beliefs and thoughts presented, but I don’t see “almost watertight evidence” at all.”
Since ATTP didn’t provide any evidence other than Curry’s tweet for the basis of the arguments in his post, and since Judith Curry has not presented a definition of what she meant by the term “research integrity”, then all anyone can possibly use as arguments are their own personal assumptions. Assumptions are not evidence until they become verified facts.
Steve Bloom- “What it did settle is that there wasn’t large-scale unforced climate variability for the period studied. It didn’t do so with precision, which is the sense in which it’s not settled science, but it did do it, as has been confirmed by subsequent work.”
It didn’t even address the climate of the globe as a whole, and more recent studies show a great deal more variability over the same period, especially when taking the Southern Hemisphere into consideration.
Aphan,
By your argument, no-one can ever comment on anything anyone says unless they explain all their terms. Which is ridiculous. If someone uses a phrase, then the reader is entitled to assume it has its usual meaning. If the author of the phrase wants it to have some other meaning, then the onus is on them to explain that.
To the the man on the street, research integrity implies the avoidance of plagiarism and fabrication of data, not suppressing or concealing data, and the ethical treatment of human and animal subjects. I personally think it is very unprofessional to accuse someone of one or more of these things on a blog. For two reasons: if the accusation is false then it’s defamatory; and secondly, there are official channels to report this kind of thing and a blog is not one of them.
Aphan:
1) It is very kind of you to play the humpty-dumpty card on behalf of Curry, but the fact is that Curry must be presumed to be talking in our common language. Ergo the meanings of the words she used are the common meanings shared by all of us. If you can only defend her opinions by arguing that she, and only she, is to be master of her words in the humpty-dumpty sense, then you, and she, have already lost the argument.
2) Hughes talked about font size as an ironical way to draw attention to the fact that, having totally ignored a large section of MBH99, attention to which was specifically drawn by the title, much of M&M’s criticism amounts to arguing that MBH did not pay attention to the issued discussed in the section M&M ignore.
“It didn’t even address the climate of the globe as a whole, and more recent studies show a great deal more variability over the same period, especially when taking the Southern Hemisphere into consideration.”
Perhaps Aphan would
1) favor us with citations to the more recent reconstructions .. and explain his information sources from which he got that comment, especially given the well-known differences in variability.
2) explain why a reconstruction of NH+SH disproves a reconstruction explici5tly says NH
(That’s like saying that a study that explicitly studied male heights was wrong because it did not also study female heights, especially since very few female heights were available at the time..)
Aphan –
Actually, I kind of agree with you about the problems with assuming meanings of fairly ambiguous and inflammatory terminology. In such cases, I think that the first response should be to ask for clarification.
But I do also think that an author has a responsibility, in such cases, to be careful about defining their terms. I criticize Judith quite a bit about a tendency to use such terms w/o providing clear definitions. I think that to do so is unscientific at a fairly fundamental level.
When people use somewhat ambiguous and inflammatory terms in such a manner, it is easily predictable that it will elicit strongly critical responses. Certainly, Judith could anticipate such responses. Why, then, would she use such terms in such a manner when her goal is to diminish the impact of tribalism and politicization on the science of climate change?
Let me ask you a question – what do you think Judith’s use of “research integrity” might have meant?
Joshua, “research integrity” is neither ambiguous nor inflammatory. It is probably defined by every university, academic journal and government research funding body in the globe. That it was used in an inflammatory comment by Curry does not make the term itself either ambiguous or inflammatory. If, however, she used it incorrectly in making an inflammatory comment, that does speak to her personal integrity.
Rachel-
“By your argument, no-one can ever comment on anything anyone says unless they explain all their terms. ”
Really? You find that to be a logic response? My argument is that taking offense without any proof that offense was intended makes one a thin skinned fool.
“If someone uses a phrase, then the reader is entitled to assume it has its usual meaning.”
Exactly. And the word “integrity” has TWO meanings. One that applies to human beings who are capable of making moral decisions, acting in ways that are virtuous or not etc. And a 2nd that is just as common to me as the 1st, that applies to the structure, unity, completeness of inanimate things. (Your knowledge, usage of both terms is clearly different than my own experience)
Since “research” is incapable of making moral choices such as whether or not it wants to be honest, or virtuous, the LOGICAL conclusion is that the 2nd definition, which applies “things” like research, rather than people, is even more properly used than the first.
“I personally think it is very unprofessional to accuse someone of one or more of these things on a blog. For two reasons: if the accusation is false then it’s defamatory; and secondly, there are official channels to report this kind of thing and a blog is not one of them.”
The trail of evidence shows Michael Mann calling Judith Curry “anti-science” after her Senate testimony and making various defamatory statements about her PRIOR to Curry’s tweet about research integrity. He also wrote an Op Ed about “If you see something, say something”, yet seems to delight in saying something without providing any evidence to support his claims.
Maybe he was called out on this blog prior for his unprofessional, defamatory statements. I have yet to read the entire site. If he wasn’t, then there would be a double standard, correct?
Tom –
The meaning is fairly ambiguous to me in the sense that it could mean fraudulent use of, or fabricating data – standards that are objectively measurable and easily attained, or it could mean intellectual honesty and complete objectivity – which are standards that are more subjective, difficult to measure, and virtually impossible to achieve. Part of the reason why research integrity has to be defined by so many entities all over the globe is because there is a subjective element. For example, in my experience in some cultural contexts the “integrity” of plagiarism plays out differently (I have had students who clearly felt that it wasn’t necessary to identify the source of information as long as the provided me with the “correct” information).
As for it being inflammatory – good point. It is not inherently inflammatory, although it can certainly be used in an obviously inflammatory way – which is what I was going for with the “used in such a manner” (meaning in an inflammatory manner).
Nice tries, Aphan, but no cigar.
“In such cases, I think that the first response should be to ask for clarification.”
From Judy? Bwahahahahahaha!
[Mod: This comment has been removed from the moderator – we’re not going to get anywhere with this so I’m bringing this discussion to an end]
==> “From Judy? Bwahahahahahaha!”
Yes, well, with her you aren’t very likely to get a response. I will agree there. I still think it should be the first response, however.
[Mod: This comment has been removed by the moderator]
[Mod: This comment has been removed by the moderator]
[Mod: This comment has been removed by the moderator because it’s a reply to a deleted comment]
http://lmgtfy.com/?q=site%3Ahttp%3A%2F%2Fjudithcurry.com+integrity
See also:
Notice the title that contains the word “integrity”.
Also note that it was in response to Pekka’s comment that won the thread.
Earlier in the thread, I pointed to Dr. Curry’s commetns, which people can read for themselves. I’d summarize:
1) Mention Wegman, giving Wegman Report strong credibility.
2)Make strong mis-statements of simple facts about it and him.
3) Label Deep Climate “reprehensible” twice as attacking Wegman … i.e., exposing plagiarism in the Wegman Report, which was obvious once found, and later strongly confirmed by academic plagiarism experts.
4) Call this far worse than Climategate.
5 ) Then halt discussion:
“Mauri, I am not making any further comments on Wegman, I have remarked on my relative ignorance on the other thread, and for my regrets at being baited into commenting (it pushed one of my “buttons”, usually I don’t let this happened) and my comments were uncharacteristically unmeasured and unproportionate (in other words “bad move”).”
In my searching of the archived thread, *she* was the one who first mentioned Wegman, 3 times in comment @114, to attach Mann, IPCC, Wahl&Ammann, Climactic Change.
I wouldn’t call this last comment an apology, which was certainly owed Deep Climate, but again, I’d urge people to read the actual comment linked earlier. Basically, she made intense opinions on a topic she displayed ignorance of, and when that was shown, shifted blame elsewhere for being “baited” into a topic that she’d actually started, and ruled further discussion out.
This do it for you Aphan
They don’t check the research integrity of the computers. IT does that and the animal IRB handles whether the sheep and goats have been behaving.
Pingback: Some kudos | And Then There's Physics
BBD has picke the wrong theme music for Judith’s political road show
Absent a brass band rendition of A Rambling Wreck From Georgia Tech ,
the proper anthem is this
BBD has picked the wrong theme for Judith’s political road show
Absent a brass band rendition of A Rambling Wreck From Georgia Tech ,
the proper anthem may be heard here ”
Russell
I stand corrected.
I am bound upon a wheel of fire.
🙂
Pingback: Political activism | …and Then There's Physics
Pingback: Basic Truths, Part Two | Izuru