Good graph design
Data should be presented in a way that is easy to understand. Good graph design can help your audience understand your data quickly and help you engage others with data
The rule of thumb is that you should be able to understand the data presented in under a minute.
There are 6 elements to consider in good graph display:
- Run charts/SPC charts are a good way to show data over time
- Bar charts show data clearer than pie charts
Other types of charts that might be helpful include Pareto charts, Histogram, Scatter plots and Funnel plots
- Don’t put too much data on one chart; use small multiples (same format, same scale)
- Use one measure per chart
- Remove unnecessary borders, backgrounds and grid lines
- Default orientation should be landscape
- Avoid distracting graphics or animations
Title: This should clearly describe the chart—including what, where and when in large clear font
Axis:
- Use clear labels
- Suppress unnecessary decimal places
- Don’t allow percentage charts with a vertical axis that goes past 100%
- Simplify the scale—include (e.g. thousands) in title
Legend:
- Consider omitting the legend and labelling directly
- Don’t let legend steal space from the data
- Scale axes appropriately so they don’t deceive—charts show patterns and people may not read the values
- Avoid large areas of white space unless the intent is to draw attention to stability/lack of variation.
- For bar charts start the scale at 0 and continue just past highest bar.
- For run and SPC charts:
- the data should occupy about half of the graph’s vertical scale with the remaining half of the scale as white space;
- if data can’t go past an absolute value (e.g. percentage values of 0–100%) then the vertical axis shouldn’t either;
- always use whole numbers when possible
- Be consistent & clear
- Minimise number of colours used
- Consider that there may be colour blind individuals looking at your chart, so use colours with care (avoid using red and green to differentiate)
- Think about what happens if someone prints the chart
Add annotations to the graph to show any testing
Here is an example of how using these principles can improve the display of data
Before: It is difficult to see what is happening on this graph. It is cluttered and having more than one ward on the same graph makes it hard to be able to understand quickly
After: Using the principles above , we have split the graphs into single wards ( This is the graph for ward 3 ). We have removed things that aren't needed from the graph e.g. grid lines, data labels and we have changed the scale to make it clearer. We have clearly labelled that axes, removed the legend and updated the colours. It is now much easier to understand the chart in under a minute. The next step would be to ask questions to get insight into what is happening.
*A median line had been added to allow run chart rules to be applied. For more detail on run charts and how they can be used to tell your if your changes are resulting in improvement see here
Content updated September 2021