When designing an infographic, the key elements to consider are the audience, and the context in which they will receive the information. The initial glance at a visual can either engage an audience or turn them away. Graphics need to be simple enough to be understood yet with enough substance to encourage further action.
The data you are presenting will have a great influence on the way the graphic is styled. If you have emotive data there is more room to connect with the audience using images, colours and text. As mentioned in the previous blog, this is shown wonderfully by the commemoration graphic.
If, however, the data is not emotive there are a number of ways to clearly and effectively present your message. Initially the designer needs to be clear on the reason they are creating visualisations. The highest priority persona needs to be identified. If the graphic tries to meet every stakeholder need, it may miss them all. Secondly, select the statistics that align with your strategic questions. Measuring client input for example may not change significantly day to day. If you design a threshold for concern when client input falls outside the threshold, a designer could create an interesting graphic on these trends. This will ensure your graphic contains meaningful content. Then, you must ask yourself, “What action do we want people to take, following this insight?”. The information can stimulate action, you want to show more than just the monitoring of customer behaviour or performance tracking. You want to align people with your mission, and backed by evidence encourage them to act.
The clearer the story you present, the easier it is for the audience to know where action is required. There are a number of different chart types that can be used to present data. Specifically, there are groups of chart types more suited to certain types of data relationships. The five chart groups span, categorical, hierarchical, relationship, time and spatial data.
Mapping multivariate relationships depend on what the data is presenting. For categorical and hierarchical data, the simpler the better. Here however, you may need a more complex graph to portray the required message. Scatter and bubble plots are the simplest graphs to use, heat maps, force-directed networks and Sankey diagrams may require more effort from the reader but may provide a higher reward.
To plot changes over time line, slope and area graphs are typically used. The more complex Gantt chart is good at showing not only basic time lines, but providing an extra level of detail.
Finally, mapping spatial data can be one of the most engaging chart types if designed correctly. Designers may use choropleth maps, symbol maps, connection maps, location maps and the slightly more complex Dorling Cartogram.
Once the chart type and data are selected, colour needs to be carefully considered. Although it may not seem it, colour is tricky. It is said;
“Colours are perhaps the visual property that people most often misuse in visualization without being aware of it.”
– Robert Kosara
Designers must apply colour, primarily to improve legibility and meaning, then lastly to decorate. Colour can be used to;
- show magnitude and order,
- show differences between variables or quantities,
- draw maximum contrast; and
- establish consistency.
The two choropleth maps below show the difference colour can make to the improve clarity. On the left, the choropleth map has used a variety of colours that have no connection, therefore each block of colour on the map seems unconnected and it is difficult to spot trends or differences. The second choropleth map uses a colour gradient. Without looking at the legend the audience can tell that the darker areas will be of more significant/higher density than the paler areas. These instant deductions can then be confirmed by the clearly labelled legend.
This rule applies to all chart types, not only maps. Colour gradients can be used to show an increase in profit over time in bar graphs, the proportion of the population in sankey diagrams and the difference in school attainment in 100% waffle charts.
If followed, the guidelines discussed above will help minimise the friction between interpreting (effort) and understanding (reward) a graphic based on the audience context. In order to be successful, visualisations need to be well thought out and time spent on planning and design.
“Like data governance standards and policies, data visualization should also
have a set of standards for design, development and storytelling.”
– Cecile Horsky, Automation & Data Assurance at Walgreens
As outlined by Cecile Horsky, visualisations need to meet certain standards to achieve their purpose. An effective visualisation is just clear information, presented in an appealing manner with an interesting message.