When a range becomes a promise, uncertainty vs brevity
For science and government communicators, a confident explanation should preserve what is known, what is assumed and what could change. That work begins in the script and carries through to the final frame.
‘Make it more confident’ is an ambiguous instruction for a science communication team. It might mean replacing a cumbersome sentence with a direct one. It might mean removing the condition that makes the sentence accurate. Before acting on that feedback, the writer needs to establish which change is being requested.
A paper published in Minerva on 29 October gives this everyday editorial problem a wider institutional context. Chae-Won Lee examines how governments make unsettled science usable for decisions, drawing on the UK’s BSE crisis and South Korea’s humidifier-disinfectant disaster. The distinction at the centre of the paper is between organising evidence sufficiently to act and allowing that position to become resistant to revision. The latter can leave unresolved questions looking settled in the public account. [1]
This is a historical and conceptual analysis, rather than a test of particular communication techniques. Its useful challenge for a communication team is to keep the basis of a decision visible. A program can proceed while its expected results remain conditional.
What the edit changes
Consider two versions of a project summary, using invented figures for illustration.
The program will reduce household water use by 20 per cent.
The model estimates a 10–20 per cent reduction in participating households’ water use, assuming the equipment is installed and used as intended.
The second version gives the reader a different proposition to evaluate. It identifies a modelled estimate, retains the range and specifies whose water use is being discussed. It also makes an assumption available for scrutiny. The first version chooses the most favourable figure and presents it as the future result.
The longer sentence may still need editing. Perhaps the installation assumption belongs in the next sentence, where there is room to explain it. Perhaps the audience first needs to understand what the equipment does. Those are decisions about sequence and emphasis. Deleting the assumption entirely would change the claim.
Before approving a revision, compare what each version allows a reader to conclude. Would someone reasonably read the new line as a measured result, a guarantee or a prediction for every household? If the evidence supports none of those readings, the copy needs another pass.
Keep the qualification in the picture
That review should continue beyond the script. In an animation based on this example, a water-use gauge might fall smoothly to a single value and stop. If the narration mentions a range while the gauge displays only the best outcome, the two parts of the explanation disagree.
One option would be to show the estimated band, with a label identifying it as a projection for participating households. The conditions could then appear alongside a simple demonstration of how the equipment is used. The design would give the uncertainty a visible role in the explanation.
Choosing that treatment requires a conversation with the subject expert. A range might describe variation between households, different model assumptions or uncertainty around an estimate. Those meanings are not interchangeable. The designer needs to know what the endpoints represent before choosing how to draw them.
This is an editorial recommendation, not a claim that one visual treatment has been shown to work for every audience. A useful test is to ask viewers what the graphic tells them about the expected outcome and how much it could vary. Their explanation can reveal whether the qualification survived the design.
Be specific about what is uncertain
There is a reasonable concern behind requests for greater confidence: will acknowledging uncertainty make the organisation look unreliable?
Older experimental work offers some reassurance, with an important distinction about presentation. In a 2023 paper, John Kerr and colleagues reported studies involving 10,519 and 2,309 participants, using COVID-19 statistics and projections. Numerical ranges had little effect on perceived source trustworthiness, although trust in the numbers could fall slightly. Vague statements that uncertainty existed, without quantifying it, could reduce perceived trustworthiness of both the numbers and their source. [2]
Those findings concern responses to particular statistical messages. They do not establish that adding uncertainty will build trust in an organisation, or that every kind of uncertainty can sensibly be expressed as a range.
For a writer, the useful question is what the reader needs to understand about the limitation. ‘Results may vary’ leaves much unexplained. Naming the relevant condition gives someone a way to interpret the estimate. Where the evidence cannot support a numerical range, say what remains unknown and why it matters to the decision. Invented precision would defeat the purpose.
The same discipline applies in the other direction. Keep established findings stated firmly. Adding ‘possibly’ to every sentence can obscure which parts of an explanation are well supported and which remain provisional.
Make the next revision possible
A public explanation also needs a life beyond its release date. If it includes a forecast, decide who will review it when observations arrive. If an assumption changes, the team should be able to find the affected claim and update the relevant page, caption or sequence.
For a video, keep important projections and their dates in editable text layers where practical. Retain the source and assumption beside the corresponding line in the working script. These small production choices make it easier to correct the explanation without reconstructing the reasoning months later.
The audience needs a clear account of why the organisation is acting now, what the evidence supports and what would warrant reconsideration. An approval process should leave those distinctions intact.
When the next round of feedback asks for a more confident message, ask the reviewer to identify the sentence or frame that needs changing. Then agree on the stronger wording and check exactly what it promises.
Sources
[1] Lee, C.-W. (2026). Governing Scientific Uncertainty: Public Certainty Regimes and Democratic Decision-Making. Minerva. Published 29 October 2026. Comparative historical and conceptual analysis.
[2] Kerr, J., van der Bles, A.-M., Dryhurst, S., Schneider, C. R., Chopurian, V., Freeman, A. L. J. and van der Linden, S. (2023). The effects of communicating uncertainty around statistics, on public trust. Royal Society Open Science, 10, 230604. Published 22 November 2023. Experimental research using COVID-19 statistics and projections.
