18 February 2021

Research pick: Apocalypse when? - "Apocalypse now? Communicating extreme forecasts"

We seem to face apocalyptic forecasts on a more and more frequent basis and yet often the predictions do not manifest themselves in the anticipated doom and gloom. Of course, some predictions have long-term consequences such as those surrounding climate change. However, as with all areas of science, the error bars that scientists know only too well can simply look like uncertainty and dithering to some non-scientists.

Research published in the International Journal of Global Warming suggests that the framing of uncertainty that is an essential part of the scientific endeavour leads to confusion among some non-scientists. The railing against this uncertainty is often perceived as “anti-science” but for the lay public it may be more a matter of being anti-uncertainty. People prefer to know for sure what they might expect to happen in their future, especially when it comes to apocalyptic forecasts, rather than to be faced with doubt.

David Rode and Paul Fischbeck of Carnegie Mellon University in Pittsburgh, Pennsylvania, USA, have found that the mere mention in an apocalyptic climate forecast reduces the amount of media attention a given forecast receives. Given that there will be uncertainty, error bars, confidence intervals, and other such matters mentioned in every scientific source, this can lead to a credibility gap. When a report fails to mention the uncertainty, it gains more media traction than a report that does not.

The team has suggested various strategies that might allow the scientific message complete with its uncertainties to reach an appropriate audience without instilling over confidence nor without looking like it is hesitant about the data it presents. The team concludes by alluding to Carl Sagan who warned us that extraordinary predictions require extraordinary caution in communication.

Rode, D.C. and Fischbeck, P.S. (2021) ‘Apocalypse now? Communicating extreme forecasts’, Int. J. Global Warming, Vol. 23, No. 2, pp.191–211.

No comments: