Wednesday, September 30, 2009

Peer Review and Yamal

The anonymous "Bishop Hill" writes, in reply to my post:
David also takes aim at Climate Audit's not being peer reviewed. I think this argument is completely overdone. Watson and Crick weren't peer reviewed. Einstein wasn't either. Didn't stop them being right.
Well, Steve McIntyre isn't Einstein. Enough said.

Nor is he Crick or Watson. In any case, their work appeared in the best journals of their day (Annalen der Physik and Nature, respectively), and some rather distinguished editors approved of their publication there. (Peer review didn't become common until after WW2.) They didn't appear on a blog, nor where they accepted immediately. Or without the input from experts in the field.

It's important here to remember what science is and what it isn't. Blog posting aren't science. They don't look the same. They don't read the same. They have far different standards of scholarship. They have far different standards of publishing. It's kind of silly to even have to go into this, because there is no comparison.

And by the way, peer-review isn't an appeal to authority, so let's drop that canard. Nor does it say a paper is right. Plenty of peer-reviewed papers have turned out to be incorrect. As Robert Park says, passing peer-review means that a paper is not obviously wrong. And that it follows accepted methods of scholarship. These are important, and an important distinction. Of course, the only authority any scientific paper really has is its agreement with empirical data.

If Steve McIntyre really thinks he has an important point, then he should write up his findings and submit them to a journal. I'm sure he could collect the page charges from a tip jar. These days you can even ask that it be given express treatment.

11 comments:

boulderSolar said...

Please discuss how peer review of the MB98 series of papers did not include Ian Jolliffe who is the preeminent authority on non centered PCA statistics. This type of statistical analysis was used extensively by Mann. When Jolliffe finally learned about what Mann did (through a webblog no less!) he said it was garbage. Here is his comments at CA: http://www.climateaudit.org/?p=3601

andrewt said...

Bishop Hill does not look worth reading, he says:

Briffa made the startling claim that the coldest year of the millennium was AD 1032, a statement that, if true, would have completely overturned the idea of the Medieval Warm Period

Its a non sequiter, as a volcanic eruption is one obvious explanation for a cold year. More worrying though when I had a quick look Briffa et al's paper I find that Bishop Hill has failed to tell us that they inferred 1033 was the 2nd warmest year (after 1045) in 1000 years. Why wouldn't he tell us that too?

andrewt said...

And Anthony Watt looks even less worth reading, announcing for some tree ring data:

"The trend is flat as road kill for the past 2000 years, though it does show an ever so slight cooling."

but in the linked paper the researchers say:

"the chronology was instead statistically prewhitened and the residuals from a general autoregression model were used for estimating past climate variability. This means that the resulting reconstructions are representative of
interannual to multidecadal timescales only and will not show century- to millennial-scale changes."


Steve McIntyre does point out the error in comments but subsequent comentators still make amazing inferences from the observation that de-trended data has no trend, e.g.:

"A brilliant effort. But I am deeply worried as to why the data was suppressed in the first place and why it took so long to surface. There is more than just a few unscrupulous scientists involved in this matter."

teqzilla said...

The point that Hill makes is that it is obviously the substance of an argument which matters more than the forum it appears in.

It is what is written and not where it is written which determines whether something is science or not.

If all you're going to post on this subject is that a blog post is not the same thing as a scientific paper - a fact which I doubt anyone needed informing of - I don't see the point in you posting about it at all.

Francis Turner said...

The BishopHill post is an attempt at explaning in fairly simple terms what Steve M has found. He is (and I mean this in a good way) merely taking talking points from various of Steve M's posts and therefore when he mentions the temperature in AD1032 he is unlikely to have looked to see what the temp reported was in AD1033

BTW I haven't checked the papers to see if you are getting them mixed up but I should point out the 1032 temp is Briffa 1995 - Polar Urals - not Briffa 2000 - Yamal. Many people (including Briffa himself) seem to think the 1995 paper was somewhat flawed as it had only 3 trees in it at key points. The main thrust of Steve's work (and Bishop Hills rewrite) is Briffa 2000

I am fairly sure Steve M will write up a version of what he has found for publication. So far we've had no more than a week or so of being able to see what Briffa 2000 was based on. For a paper published 9 years ago the fact that no one has been able to see the raw data before now is a pretty poor showing. It is important in this case because other papers H&S2002, Esper 2002, Gurskaya 2007, which cover a very similar area (H&S in fact provided the cores Briffa used), do not show the same 20th century hockey stick. It would be good to figure out why and therefore whether Briffa is actually correct or not.

RobinL said...

It appears the issue here is that 'science' has been, and is being, co-opted to some degree by 'non-scientists' in order to further an agenda - political, social, economic - other than that of establishing the facts.

If so, issues such as peer review seem largely, er, academic.

In the absence of incontrovertible evidence, hypothesis and assertion are all we have to go on. Even if that boils down to the propaganda of two opposing factions.

Off-topic, but on that calculation, I'd favour the 'do nothing' approach every time.

Sean said...

You have serious questions to answer, be those questions from a child, a spastic, or black man, or a poor man, you answer them all the same.

Your arguments disgust me, you are an insult to Karl Popper and the idea of Falsification, what you are saying is that science is a closed shop, you are very much mistaken.

I look forward to you sobering up and answering the questions posed my SmT forensic statistical work.

Unknown said...

It is a truism to say that peer-review isn't an appeal to authority.

Peers/authority - see the dichotomy?

Steve McIntyre is definitely well versed in analysing statistical works. And statistics is most usefully applied to understand subjects where nobody has a high degree of certainty.

Whether Peer Review is applied in that pure peer way consistently across different scientific disciplines is a more interesting question. For example I suspect it works well in high energy physics where there is a certain spread of competing hypothesises which no one in Fleet Street cares about. But in subjects where the application of morals and politics becomes ever more entangled I think there should be caution about the simple appliance of a certainty of Peer Reviews worth.

I am sure there may be a better Meta review of peer review and its application somewhere. However I personally think in most sciences, at best, Peer Review really works as a refining process, whereas in climate science, it seems it is being cynically oversold and used merely as a means to an end. That really could be bad for science if you accept that AGW is nowadays promoted as the poster boy for science above AIDs, Cancer and Energy research, and any more possible abstruse blue sky subjects that may not prove interesting right now - like most of the useful science we take for granted today but never thought of the origins of its birth.

Pogo said...

I think that it's worth remembering that, in reality, "peer review" is little different from getting school kids to mark their friends' exam papers.

Arthur Dent said...

I wonder how much you actually know about the process of peer review? For example were you aware of the fact that some highly prestigious journals now ask the author for names of people who have the expertise to peer review their submitted paper? Wegman in his review of the original MBH papers critically pointed out that most of the papers published in this field came from a close nit group of scientists (who were hardly likely to be hyper critical of each others work)

Anonymous said...

I first visited Steve's site for more than 4 years.

In the early times, he repeatedly tried to publish comments on Climate papers the journals the Climate papers initially appeared in.

These comments seemed to be rejected as being too long for a comment or too short for an article.

He then found them too insubstantial to appear as stand-alone climate papers. At some point, he started asking for underlying data, and that data was refused, thus stymying the effect to publish his questions properly.

In addition, he's established that the CRU temperature thingy has lost the underlying data. He's established that the published records of Antarctic weather stations was being mixed up. The Hockey team is diligent in their efforts to deny him credit for any of this.

Just to set the record straight.