George Will's latest column about climate change largely consists of gloating over the world's failure to act about it. I guess that's what right-wing hack writers do, and isn't really worthy of comment, other than to note that, his carefully cultivated reputation as an intelligent, reasonable conservative to the contrary, a right-wing hack is all he is.
But he does briefly claim scientific backing for his view that global warming is bogus:
So let's click on that link. Here's what we find:
[BBC] - Do you agree that from 1995 to the present there has been no statistically-significant global warming
[Jones] Yes, but only just. I also calculated the trend for the period 1995 to 2009. This trend (0.12C per decade) is positive, but not significant at the 95% significance level. The positive trend is quite close to the significance level. Achieving statistical significance in scientific terms is much more likely for longer periods, and much less likely for shorter periods.
So the reason " there has been no statistically significant warming" for 15 years is that you wouldn't expect significance over such a short period of time to begin with.
By analogy, suppose you had a weighted coin that landed 'heads' 75% of the time. If you flipped it 15 times and got 11 heads - a perfectly normal result for that coin - your evidence that it was weighted wouldn't be significant at the 95% level, because there's a 5.9% chance that you could get 11 or more heads in 15 flips by sheer dumb luck. You'd need more flips to demonstrate that the coin was indeed weighted.
Suppose further that you had indeed flipped the coin more times, that you'd flipped it 30 times and gotten 11 heads in the first 15 flips, and another 11 heads in the second 15 flips, for 22 heads in 30 flips altogether. Your cumulative result showing the coin's bias would be significant at the 95% level, even though neither the first set of 15 flips nor the second set of 15 flips would be significant at that level by themselves.
At the link, Jones provides a table that shows that the warming trend over the 1975-2009 period is significant. So what he's saying is that we can detect significant warming over the longer period but not over one of the shorter periods that the longer period can be broken down into. It's exactly the same thing.
What Will is saying is that it's somehow a problem for the 'coin is weighted' camp that the last 15 flips, taken by themselves, don't show that the coin is weighted. And that's simply ridiculous. Those last 15 flips indeed don't show anything by themselves. But they don't undermine the claim that the coin is biased. In fact, they support that claim - just not at a level that's statistically significant using a 95% standard.
And, as Jones points out, the same is true of the data for 1995-2009: they support the claim, but not at a level that's significant at the 95% level.
But Will turns this on its head and would have you believe that this is evidence that global warming isn't happening. He's lying.
