LeafByNiggle
Well-known member
To apply this analogy to AGW you would have to assume that the authors of all 12,000 papers were in some sense “asked” for their opinion on AGW. Then the fact that 8,000 of them “declined” to offer an opinion would mean something. It would indicate, as you suggest, that those 8,000 are unlikely to be well-represented by the 4,000 who did “answer” the question.Try this example: Of 100 people asked “Do you believe extraterrestrial life exists?”, one said no, ten said yes, the rest were unwilling to take a position. What is the justification for ignoring the 90% of respondents who were ambivalent?
But that is not the case. As I pointed out earlier, the initial 12,000 were not asked anything. They were just “found” by the first-level keyword search of climate science articles. Contrary to what you might think, there are many research topics in this field that do not have anything to do with the AGW question. The fact that those 8,000 papers were included in the first level search is merely a product of the inadequacy of the simple keyword search to find the relevant set of papers. The fact that the authors of those 8,000 papers took no position on the AGW question does not have any bearing on the accuracy of the statistical projection based on the remaining 4,000.