When writing for an audience of your peers or a non- specialised audience, how you write matters as much as what you write. Communicating your message with integrity is a way to enhance trust in your publication and ensure that you are conveying your message effectively and respectfully.
In this series on Communicating with Integrity, we are discussing some of the key factors to incorporate when preparing your manuscript. You can read part one, Laying a Foundation for Trust.
In this article, we consider how what is not said can impact trust in a publication, particularly when communicating with a non-specialised audience, be it patients, policymakers or the general public.
Be specific about the level of certainty of any claims, using numerical rather than verbal expressions.
Readers can interpret verbal indicators of uncertainty differently, e.g. “It is likely that X is a cause of Y” can imply to a reader near certainty or only moderate certainty in a relationship. This is particularly relevant when you are communicating to a non-specialised audience.
Communicating numerically reduces the risk of misinterpretation, e.g. it is better to say that “X is estimated to cause more than 50% of the cases of Y in Switzerland.”.
When expressing risk as fold change, e.g. “7 times more likely”, the expression should be accompanied by the actual risk scores and range of confidence intervals.
Acknowledge sources of uncertainty.
In all research publications, it is important to include a limitations section in which factors that may impact the generalisability or certainty of the findings are described. This should be transferred to any secondary publications, whether this is a systematic review or a blog post about the research. Communicating limitations or uncertainty is an important part of protecting against loss of trust if evidence changes [1].
Important potentially limiting factors to translate to a publication include:
- Distinct population (was the study conducted in a specific environment or with a demographic group that may limit generalisability?)
- Sample size (was the study small or large and did it control for relevant independent variables),
- Effect size (just because a result is statistically significant does not mean that it is meaningful, is the effect size described?)
- Level of evidence (are discussed studies only in animal models or case reports?)
Do not make any unsubstantiated claims or fallacies.
An author may believe that their study conclusively proves their hypothesis. However, it can be easy in these circumstances to commit a logical fallacy because of the certainty of the conclusion. E.g. “Smoking is the main cause of lung cancer in the UK. Jan has lung cancer. Therefore, Jan’s smoking must have caused their lung cancer.”.
Everyone is vulnerable to making such errors and so, as an author and a reader, you should be critical when reading your own or others’ work for such fallacies or overstated conclusions.
References
- Dries, C., McDowell, M., Rebitschek, F. G., & Leuker, C. (2024). When evidence changes: Communicating uncertainty protects against a loss of trust. Public Understanding of Science, 33(6), 777-794. https://doi.org/10.1177/09636625241228449
Further reading
Morgan et al., ‘Communicating with integrity – Supporting researchers with best practice in communication’ League of European Research Universities. Online:
https://www.leru.org/publications/communicating-with-integrity-supporting-researchers-with-best-practice-in-communication (Accessed 8/1/25)
(Featured image declaration: by andryszekk from Pixabay)
