Let’s practice! Look at these charts and think how you might make them better convey the message.
In the 4 ways to make charts that tell your story, we briefly reviewed what data visualization is and how to make charts that better convey your message. In this post, let’s look at some sample charts and see how we might improve them.
This post was inspired by Cole Nussbaumer Knaflic’s article “5 easy ways to make your data visualization more accessible” (Storytelling with Data) and the positive feedback I got on a 40-minute mini-training I held on this topic with UN colleagues.
A quick recap: What are 4 ways to make clearer and more accessible charts?
To recap: Data visualization is “the visual presentation of numbers” (UK Government Analysis Function), like charts, tables, maps, and infographics. Good data visualizations let you present complex info simply and highlight features in your data.
To make charts that better convey your message, start by choosing the right type for your purpose. (Source of list below: Government Analysis Function, GOV.UK)
- Tables to look up and compare values
- Bar graph to compare the size of different categories
- Line charts to show data over time
- Pie charts to show how parts make up a whole
The 4 ways to make your charts clearer and more accessible are:
- Make the title the takeaway message
- Directly label the data points
- Consider contrast and colors and use white space
- Add alt text
And make sure each chart can stand on its own.
Practice! What do you think of these charts?
So, let’s practice! Here are 3 charts that were inspired by real-world examples. What do you think the authors wanted to convey? Is it the right chart type? What might you suggest to improve them?
2 pie charts to show before/after change?
The pie charts below show the self-reported skills levels “before” and “after” a training. What does this pair tell you? (Does it tell you anything?)
I imagine the authors want to show that their training improved people’s ability to tie shoelaces. But here are some problems with using two pie charts to show the before-to-after change:
- It’s unclear what the change is—did things get better from “before” to “after”?
- Labels are unclear (what’s “average” versus “moderate”?).
- The labels text don’t match the title. What does it mean “Outcome: tie shoelaces effectively” is “high”? One can guess, but it’s doesn’t immediately make sense.
- I’m not sure if the colors mean anything.
A better option would be to use a bar graph.
Now, you can see that the peak shifted from around “average” or “moderate” to “high.” Also, the title now tells the reader the takeaway message.
It still takes me some moments to understand everything, so I’m sure there are better ways to get the message across. But this simple option is still better than using two pie charts.
show lessDoes this bar chart look right?
The next one is a bar chart, again trying to show the change in skills from before to after a training. Does it make sense? Can you see whether the training did indeed improve people’s skills?
The biggest question for this one is: Why are the before and after percentages stacked?? Stacking only makes sense if the pieces are parts of the same whole. Adding them like this is nonsense.
Also, we don’t know what S1, S2, S3 etc. are. So the chart doesn’t tell us anything.
A better bar graph might look like this ↓
Now, the before/after bars are side by side and the y-axis goes up just to 100%.
The Ss (presumably the S stands for “skill”) are also spelled out. Even if we can’t fit the full name in the small space, we could put in some keywords for each S and put the full list in the figure notes/caption.
The title is now the takeaway message so the reader knows what they are looking for in this chart.
show lessIs this the right chart type for the message?
Here’s a pie chart of military expenditures by country. What message do you get from looking at this chart? Does the pie chart do its job?
I hope that was fun and you have a better idea of how to review your own charts.
Feel free to share your thoughts and other recommendations in the comments. I also run online trainings covering material like this. Contact me if you’d like one for your organization.
A few resources to get you started…
I leave you with a few resources.
- Data visualization e-learning (GOV.UK): Short online lessons about data visualization
- Accessible data viz is better data viz: 5 easy ways to make your data visualization more accessible (Cole Nussbaumer Knaflic): Great starting point
Picking the right chart
- Visual vocabulary: Designing with data: Interactive chart to help you pick the right type of visualization
Designing good data visualizations
- Data visualization checklist (Stephanie Evergreen): An excellent checklist (you must register to get the download)
- How to build data visualizations in Excel (Evergreen Data): Instructions!
For dataviz ideas and inspiration
- Beyond the bar (Ama Nyame-Mensah) and When a bar is boring (Stephanie Evergreen): Bar charts are great but if you want to try something else…
- Visual Capitalist
Want writing tips and updates on new content like this?
Plus: Explore my free downloads and connect with me on LinkedIn or Facebook.
Cover image created by Ema@The Clarity Editor.
