Scientific visualisation and data storytelling

Because a rigorous result that its audience doesn't understand informs no decision: evidence only helps you decide when it is understood. The need arises when a sound analysis has to reach a committee, a reviewer, a funder or a non-statistical audience, and the figure says less than the data contain or more than they can support. Many sound analyses end up only half communicated, not for lack of rigour, but because nobody has worked on translating the result into a clear figure and narrative. Translating without betraying the method is what allows the analysis to be understood, defended and used. Scientific visualisation is not a cosmetic layer on top of the analysis: it is the stage at which a rigorous finding becomes a message that someone can use to decide.

No. Every figure is a methodological decision: choosing what to show, on which scale and at what level of aggregation determines what readers can assess for themselves. The chart type should follow from the structure of the data and the question; visual styling comes later. A figure can be beautifully composed and still hide the distribution or exaggerate a difference. In a review of 703 papers in top-quartile physiology journals, 85.6% included a bar chart and only 13.4% a scatter plot, even though very different distributions can produce the same bar chart (Weissgerber et al., 2015). Our job is to reduce noise without removing the evidence needed to judge the result: show the relevant variability, use honest scales and make sure no visual choice suggests more than the data support.

By showing it clearly rather than hiding it in a note. A point estimate without an interval or measure of spread looks more precise than it is. We choose the display according to what the reader needs to interpret: confidence or credible intervals, the data distribution, effect size and, where useful, individual observations. The measure matters: in a review of 703 papers in physiology journals, 77.6% of bar charts showed the mean ± standard error and only 15.3% the mean ± standard deviation, and standard errors can make differences look larger, particularly when sample sizes are unequal (Weissgerber et al., 2015). Communicating uncertainty does not weaken a finding: it shows how far the evidence reaches and prevents overstated certainty from undermining the credibility of the whole result.

Yes. Reviewers, committees and funders approach the same results with different questions, so we adapt the visual hierarchy, level of detail and format to each setting. For a journal, we prepare figures that meet its requirements and reporting guidelines; for a funding application or committee, material whose message can stand on its own; for a presentation, visual support that helps you explain and defend the findings. In publishing, figures count from an early stage: the Senior Deputy Editor of Science Advances describes forming a second impression of a manuscript from its figures and their legends, and notes that many reviewers start there (Warren, 2022). This does not mean changing the conclusion for each audience: we decide what needs to be seen first and what can sit in the background, while keeping every figure within what the analysis supports.

Both. We can design a figure from the results or review one your team already has, as a visual second opinion: a tightly scoped piece of work that improves a figure for a manuscript, viva or committee without rebuilding the full set of graphics. In either case, we check that the display matches the analysis, that the visual hierarchy helps readers find the main result and that it does not imply more than the data support. We return the revised figure and explain the relevant choices about scale, aggregation, colour or annotation. If the concern is not how the figure reads but whether the underlying analysis is correct, a methodological second opinion is the right service.

The deciding criterion is whether the audience will read the material or operate it. Scientific visualisation and data storytelling fit when they read it: a figure or a visual narrative designed to communicate a specific finding, without the user changing assumptions or recalculating results. A Shiny app fits when they operate the artefact: entering parameters, recalculating scenarios, filtering and downloading results. They do not answer the same question. Visualisation delivers a communication piece that conveys a finding or a narrative that is already defined; the app delivers a tool that the team interrogates with its own data. The two can be combined in the same project, because an interactive app includes carefully designed visualisations. For a paper, a committee or a presentation that people consult, the answer is visualisation and narrative: commissioning an operable tool in that scenario adds complexity that nobody will use.

We need the results of the analysis and the communication context: who the material is for, what decision it has to inform, what the main finding is and what limits of interpretation apply. ADI designs the figures, translates the result and builds the narrative. The client decides what to communicate and to whom, within what the analysis can support, and retains scientific authorship of the work. Before a figure is designed, we define which decision it has to inform and which part of the analysis needs to be made visible, because a figure without a clear communication goal ends up showing everything and clarifying little. There is one line the engagement does not cross: interpreting the results remains the research team's responsibility. What is built is the communication of a finding that is already established, with the representation choices documented so that they can be defended against anyone who questions them.

You receive the figures and the narrative, prepared for their audience: publication figures that meet the journal's formatting requirements, material for funding applications and committees in which the message stands on its own, and visual support for presentations and for defending results. The service is for anyone who needs a rigorous finding to reach the people who have to assess it or decide with it. Each figure comes with a strict correspondence between what it shows and what the analysis says, with uncertainty represented in a way people can follow and attention drawn to the main finding without distorting the rest. Beyond individual figures, you receive the narrative sequence: what is seen first, what stays with the reader and how the same result is adapted to different audiences. The analysis itself is not included, because the engagement starts from results already obtained. The material you receive makes the evidence usable without betraying the method that produced it.