Friday, February 25, 2011

How to Lie with Statistics

When I was young and impressionable there was a book How to Lie with Statistics that greatly impressed me. It makes you leery of arguments bolster with "data" and "graphs" because it teaches you to ask questions.

There is an infamous hockey stick graph. Here is a version from a screen grab of the video further down this post:

This shows a very scary sudden and big rise in temperatures. This is the kind of graph that Al Gore tramped around the world to "sell" global warming. The graph is impressive. But ti includes the infamous "hide the decline" trick revealed when the East Anglia CRU (Climate Research Unit) was hacked and e-mails revealed that some funny business was going on by major climate scientists to doctor their data. The above is a graph of temperatures when you splice in some surface temperature data.

Here is the same data without the extra surface temperature data:

Wow! Quite a difference. The top one should scare you and make you want to immediately sign up for Al Gore's program to fight global warming. The bottom one should make you say "what the hell is going in?"

If you want the story, watch this video of Richard Muller giving a talk at the CITRIS centre at UC Berkely on climate science. Here is the talk's title and abstract:
Global Warming -- The Current Status: The Science, the Scandal, the Prospects for a Treaty

Speaker/Performer: Richard Muller, Professor, Dept. of Physics, UC Berkeley

Recent events in the field of climate change have confused both the public and many "experts." I will try to elucidate what has been happening. Two out of three climate groups show no global warming for the past 13 years. What does that mean? Why does the third group (led by Jim Hansen) disagree? Why was there no treaty at Copenhagen? (It wasn't political, but technical!) Why do we hear so little about the Copenhagen follow-up meeting, this December in Cancun? What really happened in the Climategate scandal? How serious are the mistakes that embarrassed the IPCC (e.g. their claim that the Himalayas might melt in a few decades, subsequently retracted)? How reliable are the predictions of future global warming? (Pretty reliable, in my opinion.) I will attempt to give a non-partisan analysis.

At 20:55 he announces his own temperature analysis group. While I disagree with Richard Muller about how real global warming is, I applaud his setting up of this group -- Berkeley Earth Surface Temperature -- will use all the data, do it in a transparent way, and use standard statistics, publish this on the web, and allow third parties free and easy access to all data and analysis to check his group's work. Excellent!

At 25:24 he discusses how the IPCC models for Antarctica don't agree with the facts. Muller says that bother him. Yes! It should. I like the fact that Muller is honest. I am much more inclined to accept analysis from Muller than IPCC. Muller is a real scientist with a real honesty about data and theory and its limits.

At 29:48 he talks about "climategate" which is what the discrepancy of the two images at the top of this post are all about.

At 34:39 he reviews all the problems in the IPCC reports on global warming.

At 35:26 he discusses oil reserves. I get a chuckle out of this because he shows how the "amount of reserves" for a country like Canada suddenly showed hundreds of billions of barrels of oil that weren't there before! Magic! Yes... the magic of price changes.

The video is full of good things. I don't fully agree with Richard Muller, but he is the kind of guy that I could have a productive discussion with, and it is possible he could change my mind. I say that because he strikes me as honest about climatology. I'm willing to find new facts and re-interpret theories, but only with real scientists. The IPCC stuff is junk science. Al Gore is a snake oil salesman who doesn't tell you he will profit if you buy into his carbon trading schemes.

No comments: