Scientists losing war of words over climate change

by Catherine Brahic, New Scientist, February 10, 2009

Who understands the probabilities of climate change? Certainly not the general public, if psychological tests on volunteers in the US are to be believed.

The public, it seems, thinks climate scientists are less certain about their conclusions than they actually are. The results could explain why the minority views of "climate sceptics" get proportionally more attention from the general public than those of climate scientists.

Scientists are by their nature reluctant to express results as absolutely certain, and climatologists are no exception. Future projections based on climate models always come with error bars -- an indication of how likely the data is to be accurate.

Spelling it out

In an attempt to make this tool clearer to policymakers and the general public, the Intergovernmental Panel on Climate Change (IPCC) adopted in its last report, published in 2007, seven verbal expressions of certainty:

• "Virtually certain" (considered more than 99% likely to be correct)

• "Very likely" (more than 90%)

• "Likely" (more than 66%)

• "More likely than not" (more than 50%)

• "Unlikely" (less than 33%)

• "Very unlikely" (less than 10%)

• "Exceptionally unlikely" (less than 5%)

They also used the expressions "very high confidence" and "high confidence" to modify statements that had at least a 9 out of 10 (very high) or an 8 out of 10 (high) chance of being correct. The numerical translations were included in a footnote at the beginning of the summary for policymakers. The degrees of confidence then trickle down to the public through media coverage.

Low estimates

David Budescu of the psychology department at Fordham University in New York and colleagues asked 223 volunteers to read sentences from the IPCC reports that used these expressions. For example: "It is very likely that hot extremes, heat waves and heavy precipitation events will continue to become more frequent."

They then asked participants to estimate on a scale of 0 to 1 the probability conveyed by each sentence.

Participants tended to underestimate the certainty of the sentences. Three quarters of respondents thought "very likely" meant less than 90% certain, and nearly half thought "very likely" meant less than 66% certain. Public understanding of climate changewas slightly better if the readers were given a legend to refer to.

The researchers recommend that documents relating to climate science use both words and numbers to express uncertainty. For example: "It is very likely (more than 90% likely) that in future New Scientist will include numbers to describe the certainty of climate change forecasts."

Journal reference: Psychological Science (DOI: 10.1111/j.1467-9280.2009.02284.x)

Link to article: http://www.newscientist.com/article/dn16576-scientists-losing-war-of-words-over-climate-change.html