Skip to content
 

“I agree entirely that the way to go is to build some model of attitudes and how they’re affected by recent weather and to fit such a model to “thick” data—rather than to zip in and try to grab statistically significant stylized facts about people’s cognitive illusions in this area.”

Angus Reynolds sent me a long email. I’ll share it in a moment but first here’s my reply:

I don’t have much to say here, except that:

1. It’s nearly a year later but Christmas is coming again so here’s my post.

2. Yes, the effects of local weather on climate change attitudes do seem worth studying in a more systematic way. I agree entirely that the way to go is to build some model of attitudes and how they’re affected by recent weather and to fit such a model to “thick” data—rather than to zip in and try to grab statistically significant stylized facts about people’s cognitive illusions in this area.

There’s some general principle here, I think, which is worth exploring further.

3. Yes, they really do say “Happy holidays” here. Or “Have a good winter break.” Also “Merry Christmas” etc. I’ve never heard anyone say “Season’s greetings”—that just sounds like something you’d see on a Hallmark card.

And here’s what Reynolds sent me:

I figure it’s likely you will not reply given it’s nearly Christmas and you may have already come across it. But I thought it might be worth covering on your blog. I think it fits the theme of probably-correct-social-science-that-could-be-done-better, which you do seem to cover a bit.

Yesterday a study was published in the Proceedings of the National Academy of Sciences linking the proportion of people in an area who believe that climate change is happening with the most recent local weather record.

They conclude that people who have experienced more recent record highs are more likely to accept climate change than people who have experienced record lows (actually it’s slightly more complicated—“the number of days per year for which the year of the record high temperature is more recent than the year of the record low temperature”).

While I find the claim plausible if not likely I don’t really understand the analysis they have done, and find it all a bit weird.
I think they may have fallen into the camp of – got access to lots of noisy data, built a metric, done some statistical tests with significant p-values and then come up with a narrative to fit the results. Something weird is going on with the significance levels- “Levels of significance (*5%; **1%; ***10%).”
Wouldn’t that normally be *** .1% ? Or just (*.05; **.01; ***.001), so you don’t end up making typos with percentages.

Interesting that usually you have to explain to people that weather is not climate (and that weather data is noisy), but in this case it’s how people’s experience of one (which is still going to be noisy) effects their belief about the other.

So I don’t think they have really collected data that truly tests their theory, and then built and tested a model to explain the data. I notice from looking at the maps of America that pretty much everywhere has had “red” levels of TMax and the regions that have the most blue levels are mostly in the mid-west. Wouldn’t it be better to build in a model that also factors in other predictors of belief like region, climate in that region, wealth, levels of education etc. Shouldn’t it also include record highs for each season, rather than just the year? A particularly hot winter or chilly summer might convince someone more/less that climate change is happening than yet another stinking hot Californian summer.
In Australia it feels like we have been living through a handful of “Once in a lifetime weather events” each year.
I guess there’s just so many different maximums that can be broken – hottest day, month, season or year.

The more I think about it the more things I’d want to check, but I’ve already got more than enough ideas for my own research so I should stop.

What I think would be cool if someone could build a model that tracks change in social attitudes about climate change over time (and if local weather events have any effect).

Cheers and Happy Holidays (which apparently is the preferred season greeting in New York)

Leave a Reply