Communicating uncertainty in research is hard, so we worked on how best to do it – here’s what you can watch out for

An interdisciplinary group convened by research organization CIFAR — including a cosmologist, meteorologist, genetic counselor, and investigative journalist — spent three days examining how scientists communicate uncertainty, concluding it is too often treated as a failure to hide rather than a normal, even generative, part of research. Examples ranged from weather forecasters producing different maps from identical data to the 2014 BICEP2 episode, where a heralded cosmological discovery turned out to be interstellar dust. A journalist in the group noted that 'beyond a reasonable doubt' gets conflated with certainty, a confusion blamed for wrongful convictions.

Qualitative research from March 2026 found scientists increasingly frame uncertainty communication as translation — conveying complex evidence to audiences expecting simple answers, under pressure from journalists and editors who want a clean, quotable finding. The group noted that how uncertainty lands depends less on the data than on who delivers it and how much the audience already trusts that messenger, whether a scientist, doctor, or podcaster.

The full dispatch is available from the source below.