The newest hockey stick

Yes that blue line is retractions in scientific journals. Dr Roger Pielke Jr. notes on his blog that:

The Wall Street Journal reports that retractions of scientific papers have surged in recent years, with the top 3 journals issuing retractions being PNAS, Science and Nature.  The graph above shows the increase in the rate of retracted papers.

Pharmalot provides a summary:
[T]here were just 22 retraction notices that appeared in journals 10 years ago, but 139 were published in 2006 and by last year, the number reached 339. Through July of this year, there were a total 210 retractions, according to Thomson Reuters Web of Science, which maintains an index of 11,600 peer-reviewed journals.

Meanwhile, retractions related to fraud rose more than sevenfold between 2004 and 2009, exceeding a twofold rise traced to mistakes, according to an analysis published in the Journal of Medical Ethics. After studying 742 papers that were withdrawn from 2000 to 2010, the analysis found that 73.5 percent were retracted simply for error, but 26.6 percent were retracted for fraud. Ominously, 31.8 percent of retracted papers were not noted as retracted (read the abstract).

It should also be noted that there are more journals now than in 1977 and we have rapid publishing tools instead of a drafting table and a typewriter.

Read the rest here http://rogerpielkejr.blogspot.com/2011/08/surging-retractions-in-scientific.html

I wonder if the Monnett polar bear paper will become a new data point on the graph above? It certainly started out as one.

About these ads

73 thoughts on “The newest hockey stick

  1. Actually a good sign. Shows that the power of external auditing via the Web is starting to wear down the old aristocrats. They can’t get away with nearly as much nonsense now.

  2. What would be interesting is someone with access to citations databases to list the percentage of climate papers referencing tainted or withdrawn papers.

    After all, if you rely on something that is wrong, its likely that your conclusions are also wrong

  3. Mann that’s a lot of retractions.

    So, essentially, statistically model speaking, a third of your study will be retracted. I put forth the IPCC AR4 as observed proof.

    Why did they even bother getting a higher education, I wonder? What do they tell their folks? Sorry dad, mom, you ruined yourself getting your offspring an education in academia, and all your offspring could do was to not pay attention so as not to screw up, even with the help of others likeminded screw ups who took the time to review and point out the obvious flaws. How big a failure can these people be with all the money and help they get to try and get it right?

  4. Isn’t that just precious.

    Note the upward blip during the 1998 El Nino stands out like a sore thumb. Coincidence?

  5. 0.035% of publications are retracted. Always watch the actual numbers. It is certainly interesting that the rate is up from pretty much zero thirty years ago. But the actual numbers suggest very few papers actually are retracted. The question is, is this surge due to greater tendency for wrong or fraudulent science, or a greater tendency for there to be actual honesty in admitting that your paper was wrong or fraudulent, or just more people getting caught? I suspect this is simply an improved a posteriori vetting process: the internet probably plays a big role, as there is great public scrutiny of scientific findings and papers. So problems with various papers are probably being spotted more often after the fact than used to be the case. The problem is that we now are getting some idea of how many errors/frauds are NOT caught a priori by the peer review process, only a posteriori by public scrutiny.

  6. It used to be publish-or-perish. Now it is publish-or-starve, publish-or-your-university-starves.

    Of course researchers crank them out.

  7. What!? You mean the scientific process works? For some discipline’s at least. You know, those disciplines where replication and data sharing are the norm instead of the exception.

    I wouldn’t expect a lot of AGW alarmist retractions; as UnFrozenCavemen commented, “They have no shame”, along with no honesty, no adherence to proper science process, they do have plenty of gall and some amazing imperial popinjay attitudes.

  8. polistra says on August 12, 2011 at 11:08 am

    Actually a good sign. …

    Nick Stokes said that?
    Tenney Naubamer maybe?
    Hank Whats-his-face perhaps?

  9. Is it becasue people have got better at spotting rubbish or becasue more rubbish is written?
    Either way remember most of these paper have gone through ‘peer review’ to get published in the first place . The process that is supposed to produce perfect and unchallengeable knowledge or so the climate scientists would claim .

  10. If this trend continues (and we have come to learn that ALL trends in climate science are destined to continue in perpetuity) then, some time around the year 2040, every paper submitted for publication will be retracted.

  11. If the vertical axes are to be believed, retractions, even though increasing, are almost zero (less than 0.005% of the total). Of the 35 retractions, it would be interesting to know what branch of science was the leader and which had the most fraudulent papers. I doubt climate science recorded any of either. A better indicator would be how many fraudulent Nobel prizes were given away in the past 40 years.

  12. Nick says:
    August 12, 2011 at 11:15 am
    What would be interesting is someone with access to citations databases to list the percentage of climate papers referencing tainted or withdrawn papers.

    After all, if you rely on something that is wrong, its likely that your conclusions are also wrong

    ——————————————————————————————————-

    I.e. MBH98, and all the “adjusted” temperature data sets.

  13. Oh come on guys – simply the rise in publications accounts for a fair amount of this.

    Besides – when was the last correction one of y’all made on what are frequently bogus claims? When?

    Just saying…

  14. The graph is potentially deceiving. A raw number of publications is compared to a ratio. This might have the tendency to make one think that perhaps the rise of retractions is simply following the rise of published papers. In reality, QA has really gone down, the ratio of good to bad is what is peaking fast, not the number of bad.

    I do agree with Nick though, I’d like to see a cross-comparison of citations of withdrawn papers.

  15. Ged says:
    August 12, 2011 at 11:10 am

    This is a disturbing trend.

    Not at all. There are always false reports, errors, omissions, and some fraud. As I wrote earlier this week on Wm. Brigg’s site, you can go back to Newton, no scientific slouch, he, to find circular logic and some downright fudging of results. It is probably that we are more able to spot this now, and more willing to call it out. T’is a good thing.

  16. So, clearly this should be added to the Warmlist, the list of those things caused by global warming. As termperatures increase, so do scientific journal retractions.

  17. Hmmm – These curves seem to correlate with the US of A’s accelerating national debt. Could it be that wasteful government spending fosters more sloppy, low quality research, willful ‘scientific’ malfeasance, and out right fraud, requiring increased retractions upon discovery?

  18. As I read the chart, it’s 35/100,000 not 1/3rd.
    I suspect a previous poster is right that this trend has more to do with the growth of the internet and there being more people willing and able to review papers after they are published.

  19. KnR said:

    Is it becasue people have got better at spotting rubbish or becasue more rubbish is written?

    I think you are right on both points. However, several years ago, before the World Wide Web, I was involved in a research project for a small start-up company. Early on, we hired a researcher to dig up as many papers on the subject of our project as she could find, both from the local university library and from several online data bases. In going through the papers, I would say that fully 90% of them contained major errors, were rushed/incomplete, or attempted to state unjustified conclusions. We wrote these off to ‘publish of perish’ where they had to publish something/anything in order to keep their job. We also assumed that the reviewers didn’t want to break the author’s rice bowl, much less be responsible for them losing their position.

  20. Bystander-No, by definition the increase in the number of papers cannot be the explanation for why the portion of papers published being retracted has increased. It is not merely an increase in the number of retractions, but in the fraction of studies being retracted. Learn to read graphs more carefully.

  21. With the finiancial situation I would guess that the numbers of papers will drop. The percentage that are politically motivated will stay high as they still get some funding. Again, I would guess the number of retractions will stay high, along with the number of pointless papers, until the Billions from poorly overseen Gubmint spending funds disappear.

  22. Bystander says:
    August 12, 2011 at 12:19 pm
    Oh come on guys – simply the rise in publications accounts for a fair amount of this.

    Besides – when was the last correction one of y’all made on what are frequently bogus claims? When?

    Just saying…

    So what was the rise in publications, then? Anything to back up your claim?

    And which claims, specifically, are frequently made, bogus and uncorrected?

    Just asking…

  23. “Is it becasue people have got better at spotting rubbish or becasue more rubbish is written?”

    Certainly the latter weighs in heavily – in climate science alone we see garbage daily (count 4 dead polar bears in the sea in a flyover, take the area of the visible strip, divide it into the area of the arctic ocean not under ice and multiply it by 4 to find out how many bears drowned). I suspect a lot of the garbage didn’t get weeded out with the detractions despite the trend. And don’t forget two things. 1) Climate science has benefited from CAGW-faithful publishers so its unlikely that they have culled many offerings. 2) The detractions for the body of literature on this relatively new science haven’t occurred yet. The blue line is going to be all by itself above the bar chart in the next 20 years as models, white lab coats and horn-rimmed glasses get jettisoned.

  24. This is not the equilibrium rate of retraction. Positive feedback promises many retractions in the pipeline. And it’s the 30 year trend matters – short term variation is noise. Recent flattening can be attributed to to China – just wait until Chinese peer review gets to western standards. Scientists have looked at all explanatory variables and only anthropogenic retractions can explain the trend. Anything else is voodoo science.

  25. Next month headline today: M. Mann, of hokey schtick fame, is suing for copyright infringement over statistical hockey stick visualization.

  26. I don’t believe that any scientist familiar with the academic situation these days would NOT conclude that corruption of science is a serious problem these days. A problem that has gotten gradually worse over the last decades.

    The ever growing numbers of papers retracted for cause is simply an symptom of the larger issue of the loss of scientific and personal integrity.

    This is not rocket science [ grin ]

  27. Dave Springer says:
    August 12, 2011 at 11:25 am

    Isn’t that just precious.

    Note the upward blip during the 1998 El Nino stands out like a sore thumb. Coincidence?
    ==================================================================

    LMAO!!!! Global warming causes fraudulent papers to be published!!!

  28. tom t says:
    August 12, 2011 at 12:29 pm
    I would have thought that better technology would lead to more accurate results.

    —-

    It does, that’s 35/100K mistakes that are being caught, that previously wouldn’t have been caught.

  29. Looks to me like the number of retractions is small enough to be considered uninteresting.

    Of course, given that trend lines are the “important thing” in science, this means that in a few years all scientific papers will be retracted. Perhaps before they’re published. It’s very definitely worse than we thought! In a very few decades there will be no science!!!”

    Drawling a line between 2000 and 2009 we have a 5 fold increase in retractions. Draw the line and you’ll see the approaching catastrophe. Clearly there’s no choice but for the UN to take over all scientific research and publishing. There are only a very few years left in which to act to save science!

    /sarc_off

  30. Interesting.

    I wonder if the number of papers retracted is anywhere close to the number of papers that should be retracted.

  31. Bystander says:
    August 12, 2011 at 12:19 pm

    Oh come on guys – simply the rise in publications accounts for a fair amount of this.

    Besides – when was the last correction one of y’all made on what are frequently bogus claims? When?

    Just [babbling]……..
    =====================================================

    I’m wondering, what part of “retractions per 100k” do you not understand.

  32. Dave Springer says:
    August 12, 2011 at 11:25 am

    Isn’t that just precious.

    Note the upward blip during the 1998 El Nino stands out like a sore thumb. Coincidence?
    =================================================================

    lol, proof positive that global warming causes fraudulent publications. OMG!!! It’s worse than we thought!

  33. Several of the influential hurricane papers in Science, Nature, and PNAS by various authors I won’t name should have been “corrected” or retracted.

  34. @KnR
    “Is it becasue people have got better at spotting rubbish or (sic) becasue more rubbish is written?
    Could it be both?

  35. I agree with TTCA, the main difference is the greater public scrutiny… this has forced people to admit to errors that previously they might not have had to (I have seen wrong papers where the authors refused to withdrawal…preferring the “rope-a-dope” method of defense.). One of the cornerstones of science has been the assumption that people will always be as honest as possible (it is as flawed as the assumption that a journalist can be truly objective), and another is that the peer review process with “set things straight”.

    Speaking only for myself (as a physicist):

    1) Papers should always be accompanied by algorithms, data and sample code (it need not be your “production code”) needed to reproduce your results. Compared to the hard work of preparing the manuscript itself, this is not particularly an odious task.

    2) Peer reviews should become part of the corpus of the paper (as supplementary information), upon its acceptance. This includes for the journal accepting the paper as well as reviews from any journals to which this paper may previously have been rejected. The only point of the confidential review process is the protect the authors intellectual knowledge until publication of their work.

    3) Authors should be willing to provide any pertinent emails (private email servers as well as university/government/corporation) if an inquiry is made about the factual accuracy of the manuscript, or the process by which data were selected for publication. (I have done so in the past, it is a completely painless process…if you are being an honest broker. Always assuming that somebody is looking over your shoulders certainly reduces the proclivity for “spinning the results” in a more desirable fashion.)

  36. Bystander says: August 12, 2011 at 12:19 pm
    [Oh come on guys – simply the rise in publications accounts for a fair amount of this.]

    Thus, one confuses volume with percentage, the retractions are a rate (a percentage), if the quality of the work remains constant the rate of retraction does not change. If the quality decreases the rate of retraction increases.
    A rise in publictaion would cause a rise in the Number or retractions, Not the rate (percentage) of retractions.

    Thus, Bystander is once again confused by actual data.

  37. Bystander says:
    August 12, 2011 at 12:19 pm
    “Besides – when was the last correction one of y’all made on what are frequently bogus claims? When?”

    When, Bystander, was the last time you actually made a factual argument instead of just pissing at people?

  38. Given the rise of published papers it’s no wonder that the quality goes down. Of course, of the non-retracted papers, 95% will be junk nevertheless. Sturgeon’s Law applies. Well it was 95% in 1996, maybe it’s 98% now.

  39. This is a naive observation. But,

    COULD IPCC REPORT YEARS – 2001 & 2007 – BE RELATED TO THE INFLECTION POINTS?

    Why those two years for a change in the steady low rise?

  40. Ironically when it comes to the ‘Team’ retractions seem not to be an issue for, example Phil Jones infamous Chinese ‘the dog eat my home work’ UHI study , no sign of that being retracted . So although there would seem to be a lot being retracted that does not mean lots more should not be retracted given their ‘quality ‘

  41. SorryOff topic ! but…You just HAVE to read this article the best summary ever!!

    I really hope your translation from Swedish to English make enough sense. The article i written by Lars Bern who is a climate sceptic with his roots long in the enviromental mowement. One of the initiators to ” Det naturliga steget” (The natural step). His has a backround from top management with enviromental issues as speciality. All the individuals namned in the article are key scaremongers in sweden. His summary of what will be the consecvences of the “green agenda” is spot on. Its worth its one article Anthony !!

    [Note: To make sure Anthony sees it, your link should be posted in Tips & Notes. And it would be helpful if it was in English. ~dbs, mod.]

    http://www.newsmill.se/artikel/2011/08/12/klimatpolitiken-har-f-rv-rrat-skuldkrisen

  42. huishi says:
    August 12, 2011 at 2:37 pm

    I wonder if the number of papers retracted is anywhere close to the number of papers that should be retracted.

    Nowhere close. I see bad mistakes in my field all the time. They’re easy to catch because it’s an operational science…just today one of my coworkers just couldn’t understand the results someone else got. I’m familiar with that area and everything made sense except one result (which was very important in the paper)…I ran the numbers on that result and you find that it’s physically impossible to be true. Yet somehow it got through review just fine…go figure. I’ve seen similar errors in the past. Heck, we once had a guy that had bad/falsified (take your pick, though I prefer falsified) data and tried to support it by pointing to a paper that had an obvious flaw…and the thing he was pointing to was the obvious flaw (he didn’t realize it was a flaw of course). When presented with at least a dozen other papers and even hard experimental evidence from the lab, he refused to back down…still left with a PhD though…go figure. As much as joke about the above, the number of papers from my lab that should probably be retracted ranges on the order of half a dozen or so…and other labs may be worse.

    So the real question is – why don’t these retractions happen? And the answer is that most of them don’t matter! Often, things get superseded so fast that they’re not that important within even a few years, and by that point no one will care about a retraction or not (until someone uses that paper for a starting point, LOL). Also, to call out people for a retraction can be asking for pain in the realm of funding/proposals, and when it’s your own group, it’s basically suicidal. What stinks is when this is in the area of measurement science, when one published wrong result can keep better stuff from getting published (at least in appropriate journals) because the wrong results “appears” better than the actual, good result.

    -Scott

  43. Given Pielke’s long obsession with climate science, I am curious as to why he has ignored it on this occasion.
    How hard would it be to filter out the non-climate science papers ? Other studies have succeeded in doing so, why not this one ?
    Roger, we all know you want to call climate science into question, so why didn’t you ?

  44. I wonder if this number tracks closely to the number of PHDs issued. With more competition comes a higher likelihood of someone publishing anything to try and get ahead.

  45. Sorry if this has already been covered in above comments. But I hope that someone is keeping records and copies of the published propaganda, as with the increasing rate of the collaspe of the ‘green scam’ the evidence will be retracted and may become difficult to access. Kind of like deleting emails. The whole ‘green scheme’ is really nasty and the players seem to show no remorse for their actions. Sad.

  46. A bit of perspective here. The current figure for retractions is 35 per 100,000 papers or 0.035%. Compare that to errors in other professions or product faults.

    Although the link does not specify, it appears that these results are consistent with what has been known for a very long time. That retractions and fraud occur most often in Biomedical Journals. As with the Lancet MMR retraction, and the fraud perpetrated by William McBride on the effects of drugs on birth defects, many of these studies are authored by medical practitioners dabbling in science. There is also the perennial problem of studies carried out or financed by drug companies on their own products.

    The fact is almost all retractions are made at the request of authors discovering their error, or in the case of fraud, by discoveries in house or by other scientists contacting the institutions involved who notify the publishers.

    The link also stetes
    It’s unclear whether the actual amount of misconduct and legitimate error has grown; it may just be that we’re picking up on more of it,” he continues.

  47. Check the right hand scale people.
    These science rag retractions have nearly caught up to CO2 in parts-per-million!
    350 vs 392.
    But the real question then is… has Nature been a net source or net sink? :-)

  48. Taking into account the different types of scales used, this is nothing to worry about. The appearance of the graph is far more alarming than the actual data it discloses. Of far greater concern to me is the zero percent retractions back in the 1980′s.
    Making no particular point of it, I found the following interesting assertions on the net: 1)”Very few journals use more than two reviewers, the only journals I know of that regularly use more than two are Nature and Science (they use four or five) ”
    2)”Science and Nature declined to publish the results of Lauterbur, the inventor of MRI, who later won the Nobel Prize for it.”
    No, on second thought, this graph IS alarming. With statistical significance rated at 95 %, probability alone tells us we ought to be getting 500 redactions per ten thousand papers. Admittedly many will have significance over 95%, but there must also be allowances for error and fraud. The graph is alarming, but not in the way it seems intended to be.

  49. Interesting. In light of WUWT’s recent coverage of the ivory tower and its publications (for example, the recent grad student’s coral reef article, one might be inclined to increase one’s skepticism about “studies”.

    Another chart shows the percent of bad papers out of the total number published each year. That too is growing.

    http://pmretract.heroku.com/byyear

    With all those tens of thousands of journals out there, you’d think at least one of them would publish my “Unprecedented global warming driving calamitous rise in road kill” study.

  50. Philip Shehan says: August 12, 2011 at 9:51 pm

    Thanks for that comment. And other commentators.

    My interest in this is the effects of and distortion resulting from [social] policy.

    From these posters on Jo Novas site and link from Rafe Champion’s comment (#16)

    http://joannenova.com.au/2011/08/we-reclaimed-the-word-skeptic-next-we-reclaim-scientist/#more-16452

    further studies have been posted:

    Academic Fraud & Illegitimacy BS Frey & CREMA (Switzerland)

    http://www2.warwick.ac.uk/fac/soc/economics/news/events/frey.pdf

    (though P31 titled 2002-2005 ….2.212 scientists provided complete responses’
    This should include the description of the sampling method + total sampled to meaningful to reader.

    and this earlier article (April 2006)

    http://economistsview.typepad.com/economistsview/2006/04/scientific_frau.html

  51. RE: Yes that blue line is retractions in scientific journals.

    There you have it, increases in CO2 cause increases in retractions. ;-)

  52. AS has already been pointed out, this is 0.035% of all papers – basically “an insignificant trace”, as some people would call it.

    Also, looking at the original data, it would appear that it is based on an analysis of MEDICAL publications, and so says nothing, either positive of negative, about climate science.

    Also, the phrase “peer reviewed” is much misused and misunderstood. It should not refer just to the process of getting a paper published in the first place – that is actually just the first step in peer review, which begins in its true sense after publication when other scientists in the same field will read and comment on the published paper. This commentary will sometimes reveal errors leading to correction, amendment or sometimes even withdrawal of the original paper. It is possible that what this article shows is that the peer review process is actually working.

  53. 30 to 35 papers retracted for every 100,000 published. Of the retracted papers, some 26% were withdrawn for fraud, meaning 8 or 9 papers out of 100,000 are shown to be fraudulent.

    I think we all know why Mssrs. Watts and Pielke and the WSJ are flogging this dead horse.

  54. Only one climate science related retraction – the “Wegman Report”, and that’s a paper the so-called “skeptics” used as proof that climate scientists colluded in their studies. The report has become a touchstone among climate change naysayers, and now it’s been retracted. Well, well, well.

Comments are closed.