Understanding Uncertainty: How to Improve Communication Around Climate Change Evidence

This piece, first published on July 6, 2016, is being republished as part of the Chicago Policy Review‘s 20th Anniversary Series. Please visit us here to learn more about the series from our Executive Editors.

Although the majority of the scientific community agrees that Earth’s climate is warming, there is still considerable public debate about whether or not climate change is happening and how serious the government’s response should be. Indeed, a recent report from Yale University finds that only 47 percent of conservative Republicans in the United States think global warming is happening. Furthermore, only about 16 percent of American voters understand that at least 90 percent of climate scientists believe that human-caused global warming is occurring.

This disparity suggests that the obstacle is not a lack of evidence, but rather miscommunication. Some organizations have tried to address the problem by developing specific lexicons, or vocabularies, to standardize the statistical meaning of common scientific phrases. However, since vocabulary is not consistent across organizations, it confuses the public when scientists communicate uncertainty.

For example, when a scientist at the Intergovernmental Panel on Climate Change (IPCC) says that an event is “very likely,” they mean that there is at least a 90 percent chance of the event occurring. Recent research found that British and Australian citizens interpreted this same phrase to include probability as low as 65 percent. To address this 25 percent difference, Emily Ho, David Budescu, Mandeep Dhami, and David Mandel, writing in Behavioral Science & Policy, built an evidence-based lexicon based on citizen survey responses instead of scientists’ jargon in order to improve the communication of information around climate change.

In their study, consistency in the interpretation of verbal probabilities increased by 63 percent in the United Kingdom and by 100 percent in Australia after using the new evidence-based lexicon. As the leading international body for the assessment of climate change established by the United Nations, it is concerning that the IPCC’s lexicon produced a mean interpretation consistency of only 27 percent in the UK and 25 percent in Australia.

In the study, participants read eight sentences that included at least one of the following probability phrases: very unlikely, unlikely, likely, and very likely. They were then asked to estimate the numerical range represented by each phrase, both contextually in a sentence and independent of context. Based on the aggregated responses, the researchers constructed two lexicons to categorize how people interpret sentences using probability phrases.

The first lexicon used the peak value method, which represents participants’ best estimates of an equivalent numerical value (between 0 and 1) when the phrases were presented independently. The second lexicon used the membership function method, which shows how well a certain numerical value substitutes for a given phrase in the context of a sentence (0 representing “not at all” and 1 representing “absolutely”). Both methods produced significant differences from the IPCC guidelines, which led to the improvement of accuracy using the study’s alternatives.


According to Ho et al., there is considerable inconsistency in conveying uncertainty in climate science because the general public prefers verbal phrases to numeric values. Words are perceived as more intuitive than numbers, but this results in a range of interpretations of what phrases, such as “very likely,” mean. The study finds that in addition to miscommunication between scientists and the general public, people also incorrectly believe that they share the same interpretations of these terms with their peers. This multi-layered problem of misunderstanding explains why the statistics in the Yale Global Warming report are so striking.

Based on the success of their improved lexicon, the researchers advocate for further application of behavioral science to refining communication as it relates to describing uncertainty. Although uncertainty will always be present in climate science, policy makers can prepare for uncertainty as long as they adequately understand the probability of environmental events.

Article source: Ho, Emily, David Budescu, Mandeep Dhami, and David Mandel. “Improving the Communication of Uncertainty in Climate Science and Intelligence Analysis.” Behavioral Science & Policy, 2015.

Featured photo: cc/(starekase, photo ID: 26831525, from iStock by Getty Images)

Sarah Guminski
Sarah ('17) is a staff writer for Urban Affairs. She is interested in urban social policy.

Comments are closed.