Climate change consensus study backfires on researchers

A study of 12,000 scientific papers searching for a consensus on climate change has backfired on its lead author, John Cook of the website SkepticalScience, after it found more papers skeptical of human-caused climate change than papers supportive of it.

Of the papers whose published abstracts definitively take one side or the other in the controversial debate, 78 papers expressly rejected a dominant human role in climate change, compared with 65 papers that expressly blamed humans for climate change.

The unexpected result is being called a spectacular “own goal” for lead author John Cook, whose SkepticalScience website heavily pushes the idea that scientists overwhelmingly believed in human-caused global warming.

Cook’s study concentrated on the “abstracts”, or published summaries, of peer-reviewed scientific papers.


  1. Hmmm I could swear the link wasn’t there the other day. Must be the human-induced heatwave we’ve had in Australia has fried my brain a little 😉

    Unfortunately appears to be a denialist website, so quoting it is one of those self-referencing arguments, like proponents of Intelligent Design quoting the Bible as source. It does however have a link to the Der Speigel article, but the article appears to be in German so it’s impossible to verify Axels findings or notrickzones analysis. Unless you can read German.

  2. “Now, I haven’t crunched those numbers as I was reporting the work of those who claimed to have…”

    Hmmm – I’m sure this article had a link on the weekend to the man who *did* actually crunch the numbers – Andrew Montford, that well known non-scientist who is a rabid climate change denier.

    I assume the link was removed to try and add some credibility to what is just a rehash of someone elses ‘analysis’.

    Because quoting anything by Montford is an own-goal in itself.

  3. Ian, here are the specifics on how your “65 vs 78” reference misrepresents our paper:

    We categorised climate papers according to 7 different categories regarding the level of endorsement of anthropogenic global warming (AGW):

    1 Explicit Endorsement of AGW with quantification
    2 Explicit Endorsement of AGW without quantification
    3 Implicit Endorsement of AGW
    4 Neutral
    5 Implicit Rejection of AGW
    6 Explicit Rejection of AGW without quantification
    7 Explicit Rejection of AGW with quantification

    What your source material is doing is comparing the number of papers in category 1 to the number of papers in categories 5, 6 and 7. Categories 5, 6 and 7 are the total of all papers that reject AGW. A more appropriate comparison would be a comparison of all endorsement papers (categories 1, 2, 3) versus all rejection papers (categories 5, 6, 7). This comparison finds 3896 endorsement papers versus 78 rejection papers.

    In other words, he’s trying to pull a fast one. He ignores 3831 endorsement papers (e.g., 98% of the papers endorsing human-caused global warming) then claims there’s little endorsement of human-caused global warming. It’s quite a transparent ploy – I recommend going back and rereading his blog post with this in mind.

  4. Hi John

    The essence of my criticism in the initial email was that a perusal of abstracts was a flawed idea given that papers sometimes contained data that could be interpreted differently to the conclusion reached in the abstract.

    Your point that 3,896 papers out of 12,000 endorsed human-caused global warming sort of illustrates my point, as it relies on the substance of the papers.

    I think you and I could agree that most climate change papers endorse AGW, but as others whose work I sourced in that news story noted, the explicit abstract endorsement figures were 65 for vs 78 against, which is more than a little ironic.

    Now, I haven’t crunched those numbers as I was reporting the work of those who claimed to have…If you are telling me that the source material is wrong, fine, write me a rebuttal and I’ll include it above….

    If in fact it does work out at the numbers I quoted, based on the parameters laid out, then you will have to accept my gentle teasing.

    I would add, of course, that 3,900 odd out of 12,000 does not add up to a 97% endorsement either.


  5. Hi Ian,

    I noticed you misrepresented our private correspondance on Watts Up With That and continue to misrepresent it here also. Let me add some context to your misinformation. In your email to me, you stated that “A survey of abstracts will be meaningless”, questioning the validity of an analysis of abstracts – which are just the summary paragraph at the start of the paper.

    I responded with “We’ve anticipated this possible limitation in the study design and have used an independent method to measure the level of consensus in the ‘study proper’.” Now that the paper is published, you’ve had ample opportunity to investigate for yourself exactly what I was referring to – inviting the actual scientists who authored the climate papers to rate the level of endorsement of anthropogenic global warming (AGWO) of their very own papers. The result of this independent measure of consensus was a 97.2% consensus among climate papers self-rated as stating a position on AGW. This is consistent with our abstract analysis, which found 97.1% consensus among abstracts stating a position on AGW.

    Science is indeed based on evidence. See our FAQ for more discussion of that issue: and then I heartily recommend you and your readers peruse the many lines of empirical evidence for human-caused global warming which have been readily available for many years at

    Finally, you again misrepresent me with your statement that 65 papers endorse AGW while 78 reject AGW. Our paper found 3896 papers by 10,188 scientists in over 70 countries that endorse human-caused global warming. You mentioned problems with mathematics?

  6. Actually Leslie, I’ve already proven you fibbed on another thread, so calling this a “gob-smacking lie” is a bit rich.

    However, John Cook’s survey was methodologically-flawed, and he admitted there was a risk of this when I emailed him a couple of weeks ago.

    “Consensus” in science is meaningless, because science is not based upon opinion but upon empirical data that can be falsified or verified. Opinions of the committee “consensus” variety actually prove that the science is not settled. We don’t, for example, still have scientists issuing “consensus” findings that gravity probably exists.

    The irony in Cook’s “consensus” own goal is that, of all the papers whose abstracts actually expressed a firm opinion, 65 claimed human caused global warming is real, while 78 said it was not. If you can turn that into a 97% consensus in favour of human caused global warming, I hope you are not in charge of large accounts because it shows you have a problem with mathematics.

  7. Heh heh.
    What complete and utter nonsense.
    You people have no shame at all do you.
    The study is the FIFTH such study to produce a consensus figure of 97%.
    You know the denial industry is desperate when they resort to this level of gob-smacking lie.

  8. Oh, poor deluded John Cook. Time to go back to Kindergarten and start banging square pegs into round holes all over again. With some brute force he might get his imagined AGW/CC 97% CONsensus to fit the data next time around. 🙂

Comments are closed.