Thursday, August 16, 2007

Reporting and promoting science

Science sells these days. In the push for advertising dollars and readership, the conventional and electronic media strive to link new scientific results to the pressing issues of the day.

And, to be fair, in the push for tenure and grant dollars, we academics can be guilty of the same. The holy grail used to be getting a paper in Nature or Science. Now it is getting a paper in Nature or Science so that the paper will be reported on CNN or the BBC.

The intent, in either case, can be benign. A lot of the published science today, on subjects like climate change, is important news. But in the effort to "frame" - to use the terminology in the Matt Nisbet and Chris Mooney article that created buzz earlier this year - science for public and political consumption, a lot of mistakes are being made. It ranges from the inaccurate reporting and questionable publicity of the mythical lake that would resolve the very real crisis in Darfur to outright abuse of statistics to throw stones at solid science.

Take these three recent headlines... please.

1) Red faces at NASA over climate-change blunder

I call this is a "false positive". There's been a huge and unnecessary uproar over the discovery of a minor mathematical error in the NASA GISS historical temperature dataset. The error means that 1998 was no in fact the warmest year in US history, but is tied with 1934. As the NASA scientists themselves report (pdf), the error has a negligible impact on the global temperature, no impact on global rankings of the warmest years, and absolutely no impact on the evidence for human influence on the climate (see Tamino or Realclimate for details).

2) Warming will pause then full steam ahead, scientists contend

This is the "we didn't read the whole paper". These reports of a "global warming" forecast for the next decade come from a innovative short-term climate modeling study published in Science. The goal of the study was to test the ability to predict climate on decadal or shorter time-scales, a specially developed climate model that explicitly considers the frequency of large-scale atmosphere-ocean oscillations like El Nino (see Tamino). At the end of the study, after a lengthy model validation against observed data from the recent past, the authors discuss the model's predictions for the next ten years, stressing they are contingent on stochastic variables like the occurrence of El Ninos. The headlines made it seem as though the scientific community had confidently concluded that 'warming will pause' for a couple years.

3) Trees won't fix global warming

And finally, "we just didn't understand the paper". The headlines are based on research, presented at last week's ESA meeting, from Duke University's Free Air Carbon Enrichment (FACE) site, where scientists have been testing the effect of higher CO2 levels on tree growth for the past ten years. In the past, scientists had thought that higher atmospheric CO2 would effectively 'fertilize' plants. The Duke experiments showed that this fertilization effect was limited by the availability of water and nutrients (press release). In an effort to link the result to a public issue - carbon offsets - the media stories reported that new research shows planting trees won't work to combat global warming. In other words, planting trees won't take up ANY carbon. Of course it would; all that wood is made of carbon, where else could it come from? The Duke research only showed that there won’t be an extra growth bump because there’s more CO2 levels in the air, not that there won't be any growth at all.

This is the danger of popularizing science. As we saw with the coverage of the Darfur Lake, more attention was given to the initial headline than to the later reports that the 'lake' did not in fact hold any water. Unfortunately, with such quick turnaround in reporting, it is vital that all of us, the one's doing the research, the one's writing the press releases and the one's writing the news story, get it right the first time. It's not a trivial task. But otherwise, these misrepresentations (about global temperature) or misintepretations (about short-term climate prediction) or mistakes (about trees and carbon) make it into the public consciousness.


John Fleck said...

Your proposal that we all try to get it right the first time 'round is spot on. But I don't think it's the solution, simply because the pervasiveness of this sort of problem suggests that the sort of failures you describe are inevitable. They are inevitable for a number of reasons.

First, because there are lots of us involved in this chain (scientists, press officers, journalists) who aren't that good at this science communication thing. It's a really hard problem.

Second, because the idea of "popularization" is murkier that many of us along that chain think (see Hilgartner, Social Studies of Science, Vol. 20, No. 3 (Aug., 1990) for what was for me an eye-opening discussion of this).

And third, because the nature of the science-politics interface is such that someone will always be ready to seize on and exploit the mistakes - not out of intellectual dishonesty, but simply because of the reality of confirmation bias and the inevitable scientization of these political debates.

That's not to say that we shouldn't try to prevent the sort of problem you cite, but merely that we need to recognize that they'll happen no matter what. That means we need to find ways to move forward at this science-media-politics boundary that are robust to the reality of these problems. At this point, I sheepishly admit that I have no earthly idea how to do that. Suggestions welcome. :-)

Simon Donner said...

Oh, definitely. I'm being very simplistic here. I just want to get the discussion started, I don't have the answers either.

If we scientists are rarely able to accurately explain our research on the first try, usually it takes some rehearsal to come up with the most accessible language and analogies, we can hardly expect the university press office, the reporter and the news editor to do so.

My concern is that the research community is accidentally stirring the pot by pushing to publicize our research. That's why I was so hard on the Darfur lake study. I don't doubt that the authors had the best intentions... but in the effort to publicize the research, someone failed to explicitly state a few key facts, like "ancient lake" = "lake with no water", in the press release. Maybe the road to hell is paved with a well-intentioned press release.