Choosing and using visualisations

Tobias Sturt , the former head of creative at The Guardian 's digital agency and creative director of AddTwo , delves into how to leverage the magic of visualisation, helping your audience see, understand and act on your data.?

Whether we include figures in a company report, ground-breaking insights in a presentation or standout numbers for a press release, data is a vital part of our communications. But it's not always easy.

Data visualisations (charts, graphs, etc.) are crucial tools that help people engage with and understand data. Using them effectively, however, means knowing how they work best. Visuals aren’t just there to make our data look eye-catching. By representing data with simple shapes like squares and circles, we hijack the human visual system to help readers ‘see’ the data without having to think about it. Visualisations magically help people comprehend complex data.

However, the magic only works when we don’t get in the way of the visualisations doing their thing: we must be careful about decorating our charts. Another consideration is that because charts unconsciously communicate, they can alter how a reader thinks about the data. Charts tell stories. And different charts tell different stories.

As a result, when choosing charts for our data, we must think, what story does our data tell? Which chart will make that story evident?

Our data is usually telling one of five types of stories:

1. Comparison.
2. Change.
3. Composition.
4. Correlation.
5. Geographical distribution (which, irritatingly, does not begin with ‘c’).

1. Comparison

We often compare or contrast data points: one brand, country, trend compared to another or how one data point stands out from the rest.?

Without a doubt, bar charts are the best option for comparison. Separate rectangles represent data points, all aligned to the same base, allowing us to compare the comparative sizes of the values.

Of course, bar charts are the default for pretty much every use: everyone recognises them, everyone understands them, and everyone can read them. Only when, however, they are easy to understand. It can be tempting to suspect that your bar chart is boring and liven it up. This is almost always a mistake. Their usefulness lies entirely in their simplicity.?

There’s too much going on in the first version above: too many fonts, colours and visual elements. Calming it all down makes it a lot easier to read.

2. Change

It can be hard to see change over time when you look at a column of numbers, but line charts make it immediately evident.

By joining together the data points with lines, these charts show us how each data point leads to the next. The steepness of each line also gives the reader a clear idea of just how different one point is from another. However, the data must always be related.?

The first chart is perfect data for a line chart, but the second one about changes in age groups makes more sense as a column chart than as a line.

3. Composition

Often with a survey or demographic data, we think about composition. For example, 52% of the population believes one thing, and 48% another.

Pie charts are easy to read and are perfect for composition. Just as we cut a pie into slices, we divide the circle to present each data point. However, they must be simple and not too detailed ― they work best with fewer slices and visible differences in value.

There's too much going on in the first chart, and you should never use 3D in flat data visualisation. Fewer segments and simpler visuals are crucial for pie charts.

4. Correlation

Correlation ― the relation between trends or patterns ― is a tricky data story to express through charts. Traditionally, we convey correlation with scatter plots because you can simultaneously show multiple datasets. However, these are complex charts, and we rarely use them beyond academic papers.

For general audiences, you'll likely chart each dataset separately (with a bar or line, for example) and present two charts side by side. Ensure they're simple, and audiences can compare them easily.

You can pack a lot into a scatter plot, but it takes time to understand them. The two charts side-by-side tell the same story much more quickly.

5. Geographic distribution

Geographical data is simple because there’s a universal way to present it: through maps. People recognise and know how to interpret them.

However, maps are also tricky. Countries are often unusual shapes and not always recognisable out of context. More importantly, massive countries like Canada or Russia can be sparsely populated, meaning whole continents might take up space in our datasets. Not to mention how most of the Earth’s surface is water.

So, as convenient as maps may seem, bar charts or bubble charts may be the correct option for geographical datasets as they allow you to compare countries or a table of the regions.

There isn't any data here for North America, but it still takes up a chunk of the map. This data might make more sense as a bubble table.?

No matter what story you have to tell, knowing how to select the best chart type is key to leveraging the magic of visualisation, helping your audience see, understand and act on your data.

Join Tobias Sturt and Adam Frost for their online masterclass, Data visualisation: A one-day bootcamp, on Thursday 9 February 2022.