Tiny Little Keyholes

...

I’m particularly irritated this morning by a post at Real Climate addressing the potential plagiarism by the Steig et al team in a correction first mentioned at a Climate Audit post authored by Dr. Hu McCulloch.  Before beginning Hu is a professor of Economics and Finance at Ohio State University who of all the regular commentators at CA is in my opinion the least critical of RC scientists.  In fact for someone who clearly has a hot head like myself, his calmness in the face of idiocy in science can be well… a little frustrating.  Dr. McCulloch’s web page is located here where he demonstrates his very powerful statistical knowledge in relation to climatology in his own publications.

Recently when Steig et al published a Corrigendum on their recent paper widely discussed in blogland, it was noted here Significance and then elsewhere that the corrections were the same values Hu had calculated and with great politeness emailed to the authors back in February this year.

Here is the email from February which is typical of Dr. McCulloch’s tone.

Date: Sat, 28 Feb 2009 15:51:44 -0500

To: steig, dschneid, srutherford,mann, josefino.c.comiso,Drew.T.Shindell

From: Hu McCulloch

Subject: Comment on serial correlation in Steig et al 2009

Dear Dr. Steig and co-authors,

FYI, I have recently posted a comment on your 2009 paper in Nature

on Climate Audit, at http://www.climateaudit.org/?p=5341 .

While I was able to replicate or virtually replicate the 1957-2006 trends you report

on p 460 for the three regions and the continent as a whole, the 95% Confidence

Intervals you report appear to have taken no account of serial correlation

in the regression errors. When this is done, the CI’s are substantially wider

than you report.

Any reactions, by comments there or by e-mail, would be welcome!

— Hu McCulloch

J. Huston McCulloch

Economics Dept. voice (614) 292-0382

Ohio State Univ. FAX (614) 292-3906, attn. J.H. McCulloch

1945 N. High St.

Columbus, OH 43210

URL:

And HERE is the link to Dr. McCulloch’s presentation including code and all details for the reader to verify the correctness of my statements.  After the corregendum was published, Hu was somewhat offended to notice that he had not been credited with the changes and wrote Nature to request corrections.  We all know the habit of RC to not recognize anyone who posts at CA apparently that also extends to highly respected and too polite professors in other fields.

Now this is the Air Vent, and here we don’t like to mince words too much.  Recently RC actually took the time to report this blog to the editor of nature in reply to a request for data, accusing me of making false claims about data availability.  I’m not happy about that as it is to my knowledge dishonest.  In fact we still have not been given access to the data or code for several aspects of this paper but the point is to show the extreme lack of class RC proprietors can demonstrate in any attempt to save a bad paper.

Therefore I’m already ticked and this morning I decided to check RC for the first time in two weeks to see what was going on and I read this:

Resolving technical issues in science

Where RC claims that they didn’t need to mention Dr. McCulloch because the methods are known.

Plagiarism is the appropriation of another person’s ideas, processes, results, or words without giving appropriate credit.

I see it right there “Processes” without giving credit.   I’ve put a trace on the Corregendum (linked here) for changes and have checked daily and it still doesn’t mention Dr. McCulloch.  But now we get to the Air Vent part, be prepared to get pissed off, sorry about that it isn’t my fault.

There is however a different way of criticizing scientific papers that is prevalent in blogs like ClimateAudit. This involves challenging, ‘by all means necessary’, any paper whose conclusions are not liked. This can be based on simple typos, basic misunderstandings of the issues and ‘guilt by association’ though there is sometimes the occasional interesting point. Since these claims are rarely assessed to see if there is any actual impact on the main result, the outcome is a series of misleading critiques, regardless of whether any of these criticisms are in fact even valid or salient, that give the impression that every one of these papers is worthless and that all their authors incompetent at best and dishonest at worst. It is the equivalent of claiming to have found spelling errors in a newspaper article. Fun for a while, but basically irrelevant for understanding any issue or judging the worth of the journalist.

While commentary — even quite negative commentary — of papers on blogs is entirely reasonable (after all, we do it here occasionally), claims that a particular paper has been ‘discredited’ or ‘falsified’ that have not withstood (at minimum) the process of peer-review should be viewed with extreme skepticism. So should accusations of dishonesty or misconduct that have not already been conclusively and unequivocally substantiated.

I have seen many papers discredited and falsified in the past year in blogland and can’t for the life of me understand why alleged scientists would still claim with straight faces the hockey sticks are real.  This point by itself calls into question the genuineness of the authors but we’re all familiar with that.  So we need peer reveiw by others with greater authority now to see when plagiarism happens.  Perhaps we need police to see a car accident, or a fireman to see a fire?

This isn’t even the best part though, check out this beauty.

This brings us to the recent claim by Hu McCulloch that a post on ClimateAudit.org, detailing an error in Steig et al’s paper in Nature on Antarctic temperature change, was not given due credit by Steig et al. when they published a Corrigendum earlier this month. In this case, McCulloch’s comment on the paper were perfectly valid, but he chose to avoid the context of normal scientific exchange — instead posting his comments on ClimateAudit.org — and then playing a game of ‘gotcha’ by claiming plagiarism when he wasn’t cited.

First the characterization of gotcha doesn’t make any damn sense because this is right from Hu’s article where he claims some affect but still significant trend.

When the implied standard errors are appropriately corrected, many of the reported trends in Table 1 do remain significant, but only at a greatly reduced level.

We now know that the trends are highly exaggerated (by over 2 times in total)  and a properly done paper would not have significant trends through many of the regions, however Hu’s comments are hardly the work of a gotcha game.  Papers are public information, despite what our leftist RC proprieters would apparently like (had to say it at least once).  They apparently are objecting to people finding errors in a public science blog like CA or tAV no matter how politely they are worded and how professionally they are done.

Then RC makes the claim that somehow Hu accused them of plagiarism but in fact neither Hu, myself or CA claim plagerism on their blogs or posts.  I originally just mentioned non-credit, SteveM presented the definition which is an implication of potential impropriety, however it is not an accusation and Dr McCulloch put his in his letter to Nature.

According to your Editorial Policies, “Plagiarism is when an author attempts to

pass off someone else’s work as his or her own.” There is no submission

date published with the Corrigendum, but if it this was after Feb. 28, I would

submit that this Corrigendum constitutes plagiarism as you define it.

Where Dr. McCulloch points out that if there was notice prior to Feb 28th, it would not be plagiarism.  Not unsurprisingly this is exactly what the team decided to claim.  This paragraph is the key to their real position inside all the noise.

Had Dr. McCulloch been the first person to make Steig et al. aware of the error in the paper, or had he written directly to Nature at any time prior to the submission of the Corrigendum, it would have been appropriate to acknowledge him and the authors would have been happy to do so.

Oddly, they claim to have noticed the error themselves prior to Feb 28th as nobody else is credited so if it was noticed outside the authors, they would be plagiarizing another individual and then decided to delay five months until the end of July to tell anyone.  Why wouldn’t they give a name here?  Dr. Steig noticed in January or Dr. Comiso noticed prior to …. It’s an awkward and slightly more suspicious acknowledgment of prior knowledge when no names are given.

What’s more, why not reply politely in private to Dr. McCulloch’s email back in Feb that they were aware of the issue and intending a correction?  There was no such email, instead we get a five month delay, a republication of the same numbers Dr. McCulloch presented and the claim that they already knew the problem.

So they have found the little keyhole in which their credit can still squeeze.

It is the Air Vent after all, so whereas it may not be necessary to say, I feel it would be remiss of me to not mention that there is another possibility for the direction events took place.  What if Hu’s post really upset them, and really did seem like a gotcha to them as they have stated here in the somewhat emotional RC post.  After the initial shock, they took time to figure out he was right, then they took their sweet time to post changes to let the paper stand as long as possible before publishing the exact same thing with absolutely no credit given to the professor that stung their otherwise obvious perfection.  Planning instead to stand on their paper’s claim of having done the calculations as had been alleged!

There will be no credit to Dr. McCulloch who probably didn’t really care anyway beyond the idiotic rudeness of the RC team.  There will instead be the claim that an unknown individual who is apparently an author noticed the error prior to Dr. McCulloch and it took 5 months to make the corrections known to Nature.

Also, I have to mention that had Steig ‘actually’ released the code as he falsely cliamed was already done back when the paper was released this issue would have been more obvious as the code section would have been noticed missing or he would have a demonstration that it actually existed but wasn’t implemented.  THIS IS THE CODE WE’VE BEEN ASKING FOR!

Finally, I’ll end this long post with one more point.

Lest there be any confusion about this, we note that, as discussed in the Corrigendum, the error has no impact on the main conclusions in the paper.

It would have been amazingly difficult and far less headline worthy to publish a paper with a claim of total warming of 0.12 C/Decade +/- 0.12 C/Decade.  I hate that they always claim ‘no effect on conclusions’ especially when it turns most of the Antarctic into insignificant trends.  It’s been demonstrated on CA by both Ryan and Roman so I won’t bother finding the pictures.

I will also make the suggestion that the verification stats were the key to this paper.  It’s more than a little odd that the verification stats were messed up in this fashion and that was somehow missed during publication and review.  I wonder if they did know the stats weren’t as they claimed but perhaps they knew quite a bit earlier in the publication process.  Just speculating RC, after all  it’s not like you have been open about this.

Would it really have hurt to say thanks to Dr. McCulloch who also noticed the issue?