Communicating Scientific Uncertainty

Questionnaire Results

In answering the pre-workshop questionnaire, the meeting participants identified many issues that became topics of discussion at the workshop. This section provides a brief overview of the questionnaire results.

First, workshop participants provided input on how to define uncertainty. Many participants noted that at its core, uncertainty is about lack of information or knowledge. Some participants delved deeper and explored types of uncertainty, including epistemic or ontic uncertainty and uncertainty that is a result of forcing, models, and chaos. In addition participants noted that a variety of factors influence uncertainty, including the amount of data available, consistency across data sets and data types, and judgments about plausibility of data. Some participants noted that uncertainty is at times viewed as unreliability. However, as one participant noted, “uncertainty is not ignorance. Tacking back and forth across a stream of data is not the same thing as flip-flopping.”

When asked to discuss how uncertainty is addressed, participants’ responses often varied by discipline. Scientists discussed the use of confidence intervals, probabilistic statements, and statistical tests as a way to evaluate levels of uncertainty. They noted that addressing uncertainty depends on its source. To address forcing uncertainty, multiple scenarios are run. Use of multiple models helps address uncertainty in models. And multiple runs of the same model helps evaluate internal variability or chaos in the system. Finally, some respondents noted that uncertainty should be addressed with transparency when sharing results.

Lawyers had a very different perspective when considering uncertainty. In the legal context, several lawyers noted that uncertainty in facts and data are used as a weapon in the arsenal in adversarial and advocacy settings, while recognizing that rules of evidence and rules of legal ethics constrain how uncertainty is portrayed in such settings. In the policy and management realm, uncertainty is addressed through the precautionary principle, calls for research, monitoring and adaptive management.

Journalists often noted efforts to minimize uncertainty in their work through the use of multiple sources. In addition, some respondents noted that if too much uncertainty exists, a story may not be written at all. Finally, while constrained for space, some journalists indicated that discussing uncertainty may provide context for a news story.

When asked about challenges with scientific uncertainty, the respondents discussed both general challenges and challenges that relate specifically to communicating scientific uncertainty. General challenges include the following:

•           Lack of understanding by the public, journalists, lawyers, judges and even scientists

•           Misinterpretation of uncertainty as lack of credibility

•           Use of uncertainty as an advocacy tool (lawyers and advocates using it to manufacture doubt, scientists using it to obtain recognition)

•           Need for simplicity and clarity means that scientific uncertainty may not be discussed

•           Uncertainty can lead to inaction as people may want firm predictions of the future before taking action.

Meeting participants either mentioned communication challenges that relate specifically to journalism or discussed the issue broadly. For journalists, communicating uncertainty is challenge as it requires journalists to choose words carefully to remain accurate. While some view journalists as “dumbing down” the issues, journalists are writing for an audience with short attention spans who do not want details in many instances and who comes to the story voluntarily, making it difficult to effectively convey scientific uncertainty. In other words, nuance may be lost in the interest of brevity. Lack of consensus can lead to confused communication. “Balanced reporting” can distort public perception of levels of uncertainty. Some journalists may become advocates, especially as social media expands, meaning that uncertainty may be shared in a distorted way.

Questionnaire respondents also provided more general considerations about the challenges of communicating scientific uncertainty. These include the lack of standards, guidelines and best practices for how to communicate uncertainty. Some noted that mainstream scientists may avoid public communication, giving more voice to biased information-sharing by those with an agenda.

Participants noted that lawyers, scientists and journalists are all targeting different audiences or the same audience for different reasons, leading to uncertainty being shared in different ways. When used for special-interest or political messaging, scientific uncertainty may be misrepresented. Finally, some noted that scientific uncertainty can be difficult to explain, so there are challenges for lawyers and journalists to understand and interpret scientific uncertainty.

When asked how to overcome these challenges, meeting participants provided several ideas. They called for clear and simple communication, but done carefully to ensure accuracy. Considering mechanisms and styles of communication, some called for multi-direction (not one-way) communication with audience, aiming for a dialogue. Some discussed the need to communicate in consistent, easily understandable ways, like the IPCC approach to discussing scientific uncertainty.

In trying to overcome miscommunication about uncertainty, participants found no easy solutions but provided some suggestions. These included calling for scientists to avoid taking sides and avoid advocacy, or to clarify when something is a fact versus an interpretation of the facts. Similarly participants indicated the need for scientists and others to indicate knowledge gaps and acknowledge uncertainty, and to avoid red-flag statements that can be misquoted.

Questionnaire respondents also reflected on the ethics and professional norms of communicating scientific uncertainty. For scientists, there is no unified code of ethics, but clarity, accuracy, transparency, neutrality, honesty, and indicating knowledge gaps were all identified by various respondents as key elements of the professional or institutional norms that influence communication of scientific uncertainty. In contrast, many lawyers noted competing ethical obligations—on the one hand advocates for a client and selecting data that best support the client and on the other hand the duty to the court to refrain from using false data. For journalists, ethical codes are in flux as technologies and the profession’s business models evolve. Independence, promoting understanding, accuracy, and fairness were identified as professional norms that provide context for how to address scientific uncertainty in communication.

Finally, when asked what fields might serve as good examples for effective communication of scientific uncertainty, one or more participants noted the following: insurance industry, medical industry (especially doctors), meteorology, and climate science (especially the IPCC).

In summary, the results of the questionnaire indicate how the three different professions, scientists, lawyers, and journalists, treat scientific uncertainty. Journalists often seek to reduce uncertainty with the use of multiple sources and research to identify scientific uncertainty. Lawyers and advocates often seek to maximize communication of scientific uncertainty as a tool to support a position. Scientists often seek to analyze and explain uncertainty, moving the field toward consensus.