Who doesn’t love to read about statistics and statisticians? That’s a rhetorical question, my friends, so don’t bother answering. But I will allude to an answer, by telling you that I begin the statistics classes that I teach by asking the students whether they’d like to learn a magic trick. They always say yes. It goes like this:
Next time you are at a social gathering and somebody introduces themselves to you and asks what you do, say these magic words, “I am a statistician.” And…Poof! They will vanish before your eyes! It never fails.
So it’s not surprising that some of us feel left out from time to time. Which explains why the American Statistical Association (of which I am a member) has issued this statement “endorsing” the conclusions of the IPCC report while also admonishing climatologists to include more statisticians in their work.
The ASA recently convened a meeting of statisticians to ask them how they can be more involved with climate change. The statement was their answer. These sort of meetings do not always go well. The ASA had another such confab back in ’95 and invited Chicago high school students to listen to the delights that awaited them if they chose a career in statistics. The lecture was by the distinguished ASA president, who was thorough, as all statisticians are. At the end of his talk, he opened the floor for questions. There was a period of silence when, finally, one brave young man shouted out, “Yeah. Why are you so goofy?” So you can see the danger.
Anyway, except for the blanket political* “endorsement”, given only to show that we’re willing to play along, the rest of the statement is pretty good, including this, “Over the course of four [IPCC] assessment reports, a small number of statisticians have served as authors or reviewers. Although this involvement is encouraging, it does not represent the full range of statistical expertise available.”
And this, “Even in the satellite era ? the best observed period in Earth?s climate history ? there are significant uncertainties in key observational datasets. Reduction of these uncertainties will be crucial for evaluating and better constraining climate models.”
Most importantly, this, “The design and analysis of computer experiments is an area of statistics that is appropriate for aiding the development and use of climate models. Statistically based experimental designs, not currently used in this field, could be more powerful. It is also important to understand how to combine the results of experiments performed with different climate models. Despite their sophistication, climate models remain approximations of a very complex system and systematic model errors must be identified and characterized.”
The main thrust is that climate scientists have not done as well as they could quantifying the uncertainty in their models, results, and speculations, and that statisticians should be more frequently consulted, because if we’re good at anything, this is it.
We’re also not too bad at magic.