Oh good grief. OK. I’m not a climate scientist. I’m not a statistician. I don’t really want to enter into a wider debate about climate change. I haven’t verified who drew this graph or when, or whether the information on it is correct. I just want to point out that this graph, which appears in the Mail on Sunday today, which is purported to ‘finally show’ that scientists are wrong about global warming, taken on its own terms, makes the following prediction: ‘5% of the time, the black line will fall outside the pink area’. The black line is consistently in the pink area. 100% of the time. The graph says *the exact opposite* of what the story says it does.

A quick google of the ‘Global Warming Policy Foundation’ quoted in this story suggests – if the forum, headline and general tone of the article didn’t – that this isn’t simply a case of innocent idiocy.

Here’s the point, though. You can generally prove anything you want with statistics – just add some data or take away some context, or spin the result with some emotional language. Did you know that over 99% of immigrants to the UK have a higher than average number of feet? That’s true. Yesterday, the Nike Air Jordan 13 Retro sold out after just a few hours of release. The headline ‘Are Immigrants Leading to a Shoe Shortage Crisis?’ practically writes itself, doesn’t it?

That graph is meant to be ‘irrefutable’. That’s the best shot. That’s the proton torpedo they think will zoom down the reactor shaft and blow up the whole climate change edifice. It’s an argument presented entirely in their own terms, using only data they presented, framed in language of their choosing. It’s been spun and distorted and shaped as much as they possibly can to get the result they want to get and it **still** says that the scientists who have consistently and accurately predicted that the world is warming were right. That’s their best shot? It’s *rubbish*.

### Like this:

Like Loading...

*Related*

Hi,

before I start I just say want to say that in my opinion the Daily Mail is an utter rag, I really, really despise the paper. Also I definitely think that the climate is increasing in temperature. Nonetheless what the Mail is saying is correct.

Let me quickly explain the 95% degree of confidence and 75% degree of confidence — skip this if you know what I mean.

*****************************************************************

If I have 40 students in my class but they don’t all always come in and let’s say I recorded how many students came in for 2 weeks and I got the following: 34, 34, 35, 36, 32, 34, 33, 36, 35, 35. I could guess that the next class there would be either 34 or 35 students, and I’d be reasonably confident (let’s say 75% confident). I could be much more confident in predicting the next class would have between 32 to 36 students (let’s say 95% confident), so the more confident I want to be the bigger the range of values I have to state, that’s why the 95% light red band is bigger than and includes the 75% dark red band. If in a regular class only 28 students turned up that would fall outside my 95% model, so the chances of that happening is only 5% given that there was no special external events, like Rag Week, and if I had exactly 28 students for several weeks, I would have to say my model was very unlikely to be correct, since the data is out of the range of the predictions so frequently.

*****************************************************************

In this case the dark red and light red areas represents guesses as to how the climate is changing. The light red area representing 95% degree of confidence, and the dark red area representing 75% degree of confidence. The centre of the dark red area is is what climate scientists predict the change in temperature will actually be, so if you drew a line through the centre of the red area, it would be going in an upward direction, like the blue line here:

https://picasaweb.google.com/100245622318001910653/March172013#5856352362282158738

The article says “The estimates – given with 75 per cent and 95 per cent certainty – suggest only a five per cent chance of the real temperature falling outside both bands.” or in other words if the light red band is a 95% estimate of the change in temperature, the chances of the real temperature falling outside that band is 5%, which makes sense.

And the next sentence says “But when the latest official global temperature figures from the Met Office are placed over the predictions, they show how wrong the estimates have been, to the point of falling out of the ‘95 per cent’ band completely.” or in other words, the end of the black line is very near the boundary of the pink area, which is a bad sign for the model that the climate scientists are predicting. So I think it’s fairly OK what they are saying, but I still despise them.

Damian

P.S. I teach statistics so I’ll have to say this, you can’t really _prove_ anything with statistics – there is a confusion about statistics, and I think it’s because it’s taught as part of the Maths curriculum, but statistics isn’t at all like other Maths, in fact, it should be taught as part of modelling theory. In Maths mostly you are presented with a problem (there is a known answer) and you use a pre-determined technique to get the correct answer. Statistics is not at all like that, it really is a branch of modelling theory, where you have a bunch of data, and you suggest a statistical model that might closely fit the data, and it might or might not be a good model based on how successful it is at future predictions of data, but there is definitely no way you can use statistics to prove anything.

omf that is absolutely hilarious.

The average person has one testicle

I don’t know if there is a season for this but as well as the Mail piece I read a blog post by someone promoting some weird cosmic ray theory, completely ignoring the fact that while scientists are subject to confirmation bias, they are all raring to prove each other wrong.

Well, the graph says to me that the predictions from the model have been correct. I think you have a case of fractal graph-fail.