Toward best practice framing of uncertainty in scientific publications: A review of Water Resources Research abstracts. Issue 8 (9th August 2017)

• Record Type: Journal Article
• Authors: Guillaume, Joseph H. A.; Helgeson, Casey; Elsawah, Sondoss; Jakeman, Anthony J.; Kummu, Matti
• Is Part Of: Water resources research. Volume 53:Issue 8(2017)
• Page Start: 6744
• Page End: 6762
• Publication Date: 2017-08-09
• Subjects: uncertainty -- uncertainty framing -- textual analysis -- rhetoric -- argumentation -- scientific writing; Hydrology -- Periodicals; 333.91
• DOI: 10.1002/2017WR020609
• Languages: English
• ISSNs: 0043-1397

Abstract

Abstract: Uncertainty is recognized as a key issue in water resources research, among other sciences. Discussions of uncertainty typically focus on tools and techniques applied within an analysis, e.g., uncertainty quantification and model validation. But uncertainty is also addressed outside the analysis, in writing scientific publications. The language that authors use conveys their perspective of the role of uncertainty when interpreting a claim—what we call here "framing" the uncertainty. This article promotes awareness of uncertainty framing in four ways. (1) It proposes a typology of eighteen uncertainty frames, addressing five questions about uncertainty. (2) It describes the context in which uncertainty framing occurs. This is an interdisciplinary topic, involving philosophy of science, science studies, linguistics, rhetoric, and argumentation. (3) We analyze the use of uncertainty frames in a sample of 177 abstracts from the Water Resources Research journal in 2015. This helped develop and tentatively verify the typology, and provides a snapshot of current practice. (4) We make provocative recommendations to achieve a more influential, dynamic science. Current practice in uncertainty framing might be described as carefully considered incremental science. In addition to uncertainty quantification and degree of belief (present in ∼5% of abstracts), uncertainty is addressed by a combination of limiting scope, deferring to further work (∼25%) and indicating evidence is sufficient (∼40%)—or uncertainty is completely ignored (∼8%). There is a need for public debate within our discipline to decide in what context different uncertainty frames are appropriate. Uncertainty framing cannot remain a hidden practice evaluated only by lone reviewers. Plain Language Summary: Scientists address uncertainty not only in how they perform analyses, but also in how they write about their results. Not enough attention has been given to the ways in which scientists describe how uncertainty affects interpretation of their conclusions, which we call "uncertainty frames." We aim to raise awareness of this issue by: (1) proposing a typology of eighteen uncertainty frames, addressing five questions about uncertainty, (2) describing the factors that affect how uncertainty is framed, (3) describing current practice by identifying and analyzing what frames are used in a sample of abstracts published in Water Resources Research in 2015, and (4) providing provocative recommendations on how uncertainty communication could improve. We hope to spark debate within our community about how scientists should be communicating the role of uncertainty in their results. Key Points: Uncertainty framing in scientific articles is an important means of addressing uncertainty; indicating role of uncertainty in a claim A typology of frames describes conclusions in terms of maturity, scope, level of belief, depth of analysis, and relatability to the reader Frequency of frames in abstracts is consistent with carefully considered incremental science; potential to be more influential, dynamic