Presenting scientific evidence for maximum impact

Effective science communication is important. When scientific evidence is applied to policy and decision-making, it mixes with many different types of knowledge, information and opinions.

Scientists are trained to avoid over-interpreting or over-generalising their results, and to be crystal clear about the limitations of their work. As a result, scientific evidence can sound tentative when compared to the certainty of opinions and strongly-held views.

In crowded conversations where many different views and forms of knowledge are present, how can scientists present their evidence in a way that will "stick"?

There's a good reason why science communication is its own discipline - it's not an easy task! General principles of good science communication include knowing your audience, being a good listener, and using plain language. But how about some more specific instructions?

As well as those general principles, we can draw upon research testing the success of different communication techniques and their effect on users' understanding. A particularly active area of research is the communication of climate science, where different sources of uncertainty are common. For example, uncertainty might come from differing model estimates, experts, or values expressed as ranges.

A systematic review of communications about uncertainty in climate-related science (Kause et al. 2021) showed that users' understanding of the information was improved when scientists:

• describe the broader context of data variation and consensus within the discipline

• take care when presenting ranges and probabilities (see detail below)

• use visual formats such as density plots and maps to show the distribution of the data

• present maps to illustrate patterns

• illustrate findings with anecdotes and analogies.

When presenting probabilities and ranges, scientists can aid understanding by:

• giving a central estimate (like a mean) as well as a range

• helping users understand probabilities by presenting a verbal description (such as "likely") together with a numerical interpretation (such as 66%-100%)

• stating the probability range in numbers rather than using < and > (such as 66%-100%, rather than >66%)

• using positive probability expressions (such as "likely") rather than negative expressions (such as "unlikely")

• avoiding double negatives.

I'm also really interested in framing - that is, how does the way an idea is presented affect how people feel about it? I'll unpack that a bit more in a future post!

Disclaimer: not advice, just my thoughts and opinions.

Sources and further reading

• An easy-to-digest introduction to science communication, with links to other resources:

  Natural Hazards Commission Toka Tū Ake. Top tips for effective science communication: Lessons from our 2022/2023 webinars

• The systematic review where I sourced the detailed recommendations above:

  Kause, A., de Bruin, W. B., Domingos, S., Mittal, N., Lowe, J., & Fung, F. (2021). Communications about uncertainty in scientific climate-related findings: a qualitative systematic review. Environmental Research Letters, 16(5), 053005.