COMMUNICATION GUIDE / 06
Scientific communication helps readers see the question, method, evidence, uncertainty, and implication in the right order. Clear communication does not remove complexity; it organises complexity so an audience can understand what the research supports.
Follow the reasoning
Start in order, but do not treat research as a straight line. Later evidence often sends you back to fix an earlier decision—and that is part of doing the work properly.
Decide whether the audience needs technical detail, a decision, an overview, or teaching. Write one sentence stating the main evidence-backed message before designing slides or charts.
Quality check: Different audiences may need different explanations, but the underlying result must remain consistent.Move from problem and objective to methods, results, limitations, interpretation, and next action. Give each section a job and remove details that do not serve it.
Quality check: A presentation is not a report pasted onto slides.Use position and length for precise comparisons, show denominators and uncertainty, label units, order categories meaningfully, and avoid decorative effects that distort values.
Quality check: A table is often better for exact values; a chart is better for patterns and comparisons.Report what was observed with effect estimates and uncertainty, then interpret those results in context. Keep methods reproducible and connect every conclusion to presented evidence.
Quality check: Do not introduce a new result for the first time in the discussion or conclusion.Use one idea per slide, informative titles, readable type, restrained colour, direct chart labels, and speaker notes for details. Rehearse timing and prepare answers for likely methodological questions.
Quality check: If the audience is reading paragraphs, they are not listening to the explanation.Explain missingness, bias, confounding, imprecision, generalisability, and design limits without dismissing the work. State what the findings support and what further evidence is needed.
Quality check: Honest limitations make a conclusion more credible, not weaker.See it in context
Lead with the adjusted effect and uncertainty, show a simple comparison, state the sample and method, then explain that an observational cross-sectional result supports association rather than temporal or causal proof.
Late-night phone use was associated with poorer sleep in this sample
Adjusted effect estimate with its confidence interval and clearly stated reference group
A labelled dot-and-interval plot or other chart matching the result
Sample, setting, outcome definition, adjustment variables, and missing-data handling
Cross-sectional timing and self-report may restrict causal interpretation
Recommend a prospective or intervention study if causal evidence is required
Before you move on
The audience and one central message are defined.
01Every visual answers a question and labels units, groups, and uncertainty.
02Methods and results are accurate enough to evaluate.
03Interpretation distinguishes association, causation, and practical importance.
04Limitations and next steps are specific rather than ceremonial.
05Good communication makes evidence easier to understand, not easier to exaggerate.
Follow the next decision
A good question should shape the literature search; the evidence should sharpen the design; and the methods should make the result possible to understand and challenge.