Update - see below for a comment at WUWT from the paper's author.
Deniers at WUWT are a bit miffed at a study that focused on their confusion about weather and climate (well, not exactly - but its conclusions are consistent with the confusion often seen at WUWT). Anthony Watts posted two press releases about an analysis of how people look for information on climate change (archived here, latest update here). He prefaced the press releases with two short comments in addition to his headline:
The two press releases are:Dueling “weather is not climate” press releases – see if you can spot the politically biased one
- One is issued by the University of Rhode Island, where the researcher is based. Anthony's comment was "Results vary by political ideology, education levels".
- The other is on Springer, the publisher of Climatic Change, the journal in which the paper was published. Anthony's comment was: "Political ideology, education levels affect when people search for climate information."
The paper itself is in the journal Climatic Change. Here is a pre-publication version of the paper. The abstract is (my para breaks):
Learning about the causes and consequences of climate change can be an important avenue for supporting mitigation policy and efficient adaptation. This paper uses internet search activity data, a distinctly revealed preference approach, to examine if local weather fluctuations cause people to seek information about climate change.
The results suggest that weather fluctuations do have an effect on climate change related search behavior, however not always in ways that are consistent with the projected impacts of climate change. While search activity increases with extreme heat in summer and extended periods of no rainfall and declines in extreme cold in winter, search activity also increases with colder winter and spring average temperatures. Some of the surprising results are magnified when heterogeneity by political ideology and educational attainment in responsiveness is modeled, which could suggest that different people have different perceptions about what types of weather define climate change or that climate science deniers seek information through Google.
However, the results also indicate that for all groups in the political and educational spectrum, there exist weather events consistent with the predicted impacts of climate change that elicit increased information seeking.
That's probably interesting information in its own right, particularly if you are a science communicator wanting to help people understand more about weather, climate and climate change.
As for what Anthony Watts is trying to say, maybe it's worth looking at the way the two press releases are constructed, to see the extent of the political bias. Feel free to add your own views.
I picked out the main paragraphs in which voting habits and level of education were mentioned, on the basis that this was clearly of greatest interest to Anthony Watts, going by his two short comments.
From the Springer press release, for which Anthony's comment was: "Political ideology, education levels affect when people search for climate information.":
Republicans search the Net for information about the weather, climate change and global warming during extremely hot or cold spells. Democrats google these terms when they experience changes in the average temperatures. ...
...People from varying political and educational backgrounds reach for their devices at different times to check out information on climate change. Republicans and people from less educated areas do more relevant searches during periods of extreme temperatures, while Democrats and residents of well-educated areas do so when they experience changes in average temperatures.
The above press release discusses searches according to voting preference, education levels and timing of searches - or what prompted the search, which is consistent with Anthony's comment.
From the U Rhode Island press release, for which Anthony's comment was "Results vary by political ideology, education levels":
A University of Rhode Island researcher analyzed Internet search trends and weather patterns and has concluded that people across the United States seek information about climate change when they experience unusual or severe weather events in their area. But findings differed based on political ideology and education levels....
...When Lang compared search data in regions of the country with differing political views and education levels, his results suggest that some groups may see climate change differently. For example, Democratic leaning regions and those with higher education levels were more likely to seek information about climate change when average summer temperatures were above normal, whereas those in Republican and less educated areas sought climate change information when they experienced extreme heat.
This other press release also discusses searches according to voting preference, education levels and timing of searches - or what prompted the search, not all of which Anthony commented on.
From the paper itself (my bold italics):
In addition, I use this framework to examine heterogeneity by political ideology and educational attainment in the types of weather fluctuations that cause online information seeking. The exact query was “climate change” + “global warming”, which counts any search including either of the phrases “climate change” or “global warming” about climate change. The results indicate that all groups in the political and educational spectrum have weather triggers that elicit increased information seeking and are consistent with climate change. Republican and less educated areas increase search activity in response to changes in extreme temperatures consistent with climate change, whereas more Democratic and well educated areas increase search activity in response to changes in average temperatures consistent with climate change. This could indicate that different types of people experience weather differently or have different perceptions about what type of weather defines climate change.
The paper itself also discusses searches according to voting preference, education levels and timing of searches - or what prompted the search.
I admit to not seeing a great deal of difference in the manner in which this particular aspect of the study was reported. One led off with a comparison of Republicans and Democrats. The other didn't use the words "Republican" and "Democrat" in the opening paragraph but still stated there were differences based on political preference. Anthony's comment suggests that he read there were differences in the timing of searches (or what weather prompted the searches). But there weren't. Both press releases reported the same thing.
I then went to the comment section of WUWT to see if anyone else was able to figure out what Anthony Watts was referring to. It was no help at all (archived here, latest update here). One person misread the passages as equating education level with voting preference. But they didn't. They simply reported that results were similar for people living in areas where Republicans dominate as for people living in areas where the education standard is lower. ("Republican and less educated areas"). It didn't suggest that these two areas are the same - although they may well be the same.
So not only is Anthony Watts unable to articulate what he means, none of his readers so far are able to shed any light on the matter either.
I haven't discussed the paper itself. I'll leave that up to you to decide whether the paper is sound from a research design perspective and whether the conclusions are valid. What questions it may raise. For example: what, if any, implications there are for communicating climate science. What implications, if any, there are in regard to addressing inequalities in education across the USA. What further research could flow from the study?
In the WUWT comments, there is one from the author of the paper. C Lang says:
July 18, 2014 at 6:32 am
I’m the author of the paper being discussed here. I’ve enjoyed reading some comments (especially, “I bet Lang did a lot of drinking in college”), though haven’t read them all.
One thing I wanted to clear up is the notion that I’m equating political ideology and education levels. Not at all true. First, a caveat about any conclusions that can be drawn. The data I’m using are aggregate, essentially at the metropolitan area level. So I do not know what how individuals with differing levels of education or differing political ideologies respond to weather changes. What I can infer is how individuals living in metro areas of differing education levels and differing politics respond to weather changes. In the model, I interact metro characteristics of % with college degree and % Democrat separately. That is, the model allows these characteristics to have no effect or opposite effects. However, the results suggest that the effects align for metros with a large percentage of college graduates and a large percentage of Democrats.Added by Sou 11:55 pm AEST 18 July 2014
Lang, C. (2014). "Do Weather Fluctuations Cause People to Seek Information about Climate Change?" Climatic Change. DOI 10.1007/s10584-014-1180-6