Rod Liddle

Stats and climate change – a response to Jim Ryan

Stats and climate change – a response to Jim Ryan
Text settings
Comments

I find it genuinely difficult to debate with people who deny my right to debate; this is the case with the climate change lobby. The danger, if you don’t watch out, is that the arrogance and certitude of the AGW lobby pushes one towards an ever more antithetical position. This is a flawed, human, response – very similar to the flaws exhibited by those climate change monkeys sending dodgy emails to one another. If you work for, and are paid by, an institution which accepts climate change as a fact, then you will be disinclined to accept scientific evidence to the contrary. You hold climate change as an article of faith, and also a conduit for remuneration. This is how science becomes poisoned; but it happens in almost every scientific endeavour, and always has done. Scientists become trapped by their own paradigms; they are reluctant to let go of ideas. This is why it usually takes a generation before paradigms change. But change they always do. Remember that a generation ago we were worried by global cooling and the coming of the next ice age.

I have no expertise whatsoever in meteorology, but I do have a bit of knowledge about stats, and randomness and chance – and it is this that leads me to a broadly sceptical point of view regarding AGW. Jim Ryan kindly responded to my blog about the UEA debacle with a lengthy and pretty rational argument, to which he appended a list of many organisations which sign up to AGW. What he didn’t say, however, was that these organisations often heavily qualify their belief in man-made climate change, suggesting that it is “probable” or “heavily probable” or “likely”. Fine. And there are many more which will not go even this far.

But it is another part of Jim’s response that interested me, because it involves statistics and displayed the almost universal misunderstanding of statistics and chance. He wrote:

“Rod, you visit a 100 tumour specialists and 97 tell you you require an operation to treat the condition. The other 3 say it is benign and does not require any treatment “

The implication being that of course the 97 are right, and that any rational person would not question this supposed fact. A 97% certainty is pretty much a certainty, full stop, isn’t it?

Well, no. Suppose the tumour which the doctors believe afflicts me is a fairly rare type of tumour, one which affects only, say, one in 5000 people. If that is the case then the likelihood that I do not have that tumour, and that those 97% of surgeons have made a wrong diagnosis, and that I therefore do not need an operation, is far, far higher than the likelihood that I do have a tumour and do need it operated upon. Jim’s analogy utilizes that difficult thing to supposedly prove his point, the false positive.