Watts Up With That (WUWT) has new post called Fuzzy math: In a new soon to be published paper, John Cook claims ‘consensus’ on 32.6% of scientific papers that endorse AGW. It’s about a recent paper published today by John Cook of Skeptical Science. The basic aim of the study by John Cook was to get a sense of how many peer-reviewed papers that specifically adressed “global warming” or “global climate change” explicitly addressed whether or not global was “human caused”.
Firstly, the title of the WUWT post doesn’t really make sense. If the goal of the study was to determine what fraction of papers endorsed anthropogenic global warming (AGW), how can the study find that 32.6% of scientific papers that endorse AGW, endorse AGW. Surely 100% of papers that endorse AGW, endorse AGW? Talk about fuzzy maths!
The basic premise of the WUWT post is how can one claim a consensus if only 32.6% of the papers endorse AGW. Well, let’s see what the study actually did.
In March 2012, the researchers used the ISI Web of Science database to search for peer-reviewed academic articles published between 1991 and 2011 using two topic searches: “global warming” and “global climate change”.
After limiting the selection to peer-reviewed climate science, the study considered 11 994 papers written by 29 083 authors in 1980 different scientific journals.
Okay, so the study identified 11994 paper written between 1991 and 2011 that satisfy the topic search of “global warming” or “global climate change”. A little surprised that there are as many as 1980 different scientific journals, but that is apparently what the press release (which I haven’t been able to find) says. So, what did they do next.
The abstracts from these papers were randomly distributed between a team of 24 volunteers recruited through the “myth-busting” website skepticalscience.com, who used set criteria to determine the level to which the abstracts endorsed that humans are the primary cause of global warming. Each abstract was analyzed by two independent, anonymous raters.
From the 11 994 papers, 32.6 per cent endorsed AGW, 66.4 per cent stated no position on AGW, 0.7 per cent rejected AGW and in 0.3 per cent of papers, the authors said the cause of global warming was uncertain.
So, the abstract of each paper was assessed by 2 of 24 volunteers to determine if the paper endorsed AGW, rejected AGW, said it was uncertain, or stated no position with regards to AGW. The results seem quite clear to me. Only 1% rejected AGW or explicitly stated that it was uncertain. Almost one-third explicitly endorsed AGW, and almost two-thirds said nothing with regards to AGW. That seems like a strong consensus to me. The only way this wouldn’t be a strong consensus is if the two-thirds who said nothing with regards to AGW did so because they were uncertain. But, they actually include a category for those papers who claim that it is uncertain (0.3 %). It seems quite reasonable to me that quite a large fraction of papers on “global warming” or “global climate change” might not address anything to do with whether global warming is anthropogenic or not.
Therefore, it seems clear that 97% of papers published in peer-reviewed journals between 1991 and 2011 that address the issue of AGW, endorse AGW. Only 3% either reject it or claim that their results suggest that it is uncertain. Quite a consensus. What I find odd about this WUWT post is that they even attempt to disagree with this. I thought that one of their main themes was that there are indeed a lot of peer-reviewed climate science papers that endorse AGW but that this is because there is some fundamental problem with professional climate scientists. They are biased, suffer from groupthink, or somehow benefit from endorsing AGW. Surely , in order for this to be true they should be accepting the results of this study and then attempting to show that such a strong consensus indicates some kind of issue with climate science. If they are attempting to suggest that there is a problem with this study, does that not then weaken their argument that such a strong consensus implies a problem. Surely, they can’t have it both ways.