Research produces data, and as researchers, we develop the skills to interpret it and draw out the details, the trends, and the correlations, that are of interest. For a wider audience, we need to present the data in a way that makes these important points as clear as possible. Publishing in a journal, readers don’t want to have to wade through all the data to see whether the arguments stand up. When writing for a general audience, it’s even more important to present data in an easily digested manner.
What is data visualisation?
Data visualisation is simply presenting data in a graphical form. It includes the myriad different forms of graphs and plots you find in almost every academic paper, and also maps, word clouds, timelines, and even animations.
You’ve almost certainly put together some sort of data visualisation before, but doing it well is a skill which is worth developing. Good data visualisations can increase interest in your work, especially among journalists and the general public. It’s a transferrable skill, also very useful in putting together lectures, or commissioned reports, and if you can produce something really compelling, there’s a good chance that your institution’s publicity department will want to help it get seen.
How do I choose the most appropriate data visualisation?
There are so many different ways to visualise data, I find it useful to consider three questions:
- What exactly do you want to communicate?
The visualisation is only going to work if you’re clear what point you’re trying to communicate. Once you’ve decided, that may immediately point to a preferred approach. If your data is geographical, a map is likely to work well. If it shows a change over time, then a timeline is going to be clearest. To show the composition of something, a pie chart can work, but if you want to compare changes in composition under different circumstances, then a stacked column or area chart works best.
- Who is the intended audience?
Just like your writing or presentations, you will need to tailor what you’re doing to your audience. There may be particular charts associated with your field which will be easy for a colleague to interpret, but if they are unfamiliar to the public, it’s best to use something more familiar, like a map, timeline, line graph or pie chart.
- In what format will they be viewing the visualisation?
If you’re putting something up on screen briefly in a conference presentation or lecture, it had best be simple. It’s best to use animation sparingly, but where the format permits, you might use it to show a transition between different states. Online visualisations have the most flexibility, and it’s possible to add options for users to change what is displayed and even explore the data themselves, but such online visualisations can be time-consuming to produce.
If you’re not sure what type of visualisation might work, there are guides online which can help you decide, such as this flowchart, from Infogram.
How much information should I include within one data visualisation?
Data visualisations can summarise huge amounts of data, and that’s a great thing. But there can be a temptation to crowd them with data which is irrelevant to the point you’re trying to make, or difficult to interpret.
The best data visualisations communicate their key message quickly. Ideally, someone looking at it should be able to see the point you are making within a few seconds – the value, the trend, or the correlation of interest should be highlighted and the focus of the visualisation. Including a title also helps reinforce the point you’re making.
Other information on the visualisation should support the point that you are making. For example, if you are showing that a trend in the UK differs from countries in the EU, you might put together a line graph with trend lines for each country, and highlight the UK trend. The point is made immediately, and users can examine the graph more closely to see how the other countries compare. There’s no need to include a trend line for the US; if that comparison is also of interest, it can be saved for another graph.
In some cases, there might be a lot of data on the visualisation – a scatter plot might include hundreds of relevant data points – but it should be immediately clear what point you are making, and any labelling of individual data points can be limited to what helps you make that point.
How can I best use colour in my data visualisations?
Colour is a powerful tool to grab people’s attention. If your visualisation is intended to highlight one key point, you can use a strong colour there, to immediately catch the reader’s eye.
Colours can also help people associate your data with a particular idea. For example, if your data relates to blood, you might choose to feature the colour red. If your timeline needs to differentiate between periods of Conservative and Labour government, you can use blue and red, respectively, to immediately make it clear that this is significant.
Colour gradients can also be a useful tool. For example, on a map, you might use a dark green to show high levels of tree cover in one district, and a light green to show low levels in another. This will only work if the differences are clear – two very similar shades of green may be impossible to distinguish.
Because colours attract people’s attention, you may want to use a standard colour palette through your piece of work, to reinforce the idea that it is a coherent entity. Using very different colours or styles for each visualisation can suggest that you have pulled them from a variety of sources without taking time to adapt them to the needs of the people using them.
How can I make my data visualisation accessible?
Data visualisations are fantastic ways to present information, but some approaches can make them unusable for people with particular disabilities. The whole piece of work should be accessible to everyone, and you can take simple measures which will improve their universal usability.
Don’t rely solely on colour to communicate information. Two colours which look very different to you may appear identical to someone with colour blindness, or even someone who prints your visualisation out in black and white. Try viewing your visualisation in black and white, and seeing if it is still clear. If not, you can use darker and lighter tones, or distinct patterns and shapes, to communicate the information.
Do add labels, legends, and explanations in accompanying text, to help people with visual impairments to understand your work – they may use a screen reader, or struggle to spot the exact values shown on a detailed visualisation. And make sure that any text is large enough to read, and is in a legible font.
Our infographic service can help you present your data in a way that is easily understood by your target audience. Find out more about our infographic service.
How can I produce a video abstract?
A video abstract can accompany research articles. Some journals strongly encourage authors to put together a video abstract. They are a great way to interest potential readers in the highlights of your research, so it’s likely that they will be more in demand in future. Depending on the copyright terms, you may able to re-use them on your institutional webpage or even put them on YouTube to encourage journalists and the public to engage with your work.
Video abstracts work best when they’re short, two to threeminutes, and when they add something to your work. Don’t simply read the written abstract, but imagine that you’re explaining the importance of your findings to a member of the public.
Remember that this is aimed at a general audience who have limited time to view your video, and can read your paper if they are sufficiently interested, so the visualisations should be simple and focus on one key point in a clear and visually interesting way.
Our video abstract service will help you deliver the core message of your research to a wider audience. Find out more about our video abstract service