Designing better data visualisations
Conversations with Data: #75
Do you want to receive Conversations with Data? Subscribe
Welcome to our latest Conversations with Data newsletter.
We are excited to announce the launch of The Data Journalism Handbook: Towards a Critical Data Practice published by Amsterdam University Press, supported by Google News Initiative and the European Journalism Centre. The handbook provides a rich and panoramic introduction to data journalism, combining both critical reflection and practical insight.
Edited by Liliana Bounegru and Jonathan Gray, it offers a diverse collection of perspectives on how data journalism is done around the world and the broader consequences of datafication in the news, serving as both a textbook and a sourcebook for this emerging field. Read the full book on DataJournalism.com.
Now on to the podcast!
In this week's Conversations with Data episode, we spoke with Maarten Lambrechts, a freelance data visualisation consultant based in Belgium. He talks to us about his journey into the world of data visualisation and how he helps other organisations like the World Bank communicate their numbers by evaluating, designing and developing static and interactive visualisations.
What we asked
Tell us about your career. How did you find your way into data journalism?
Like many people in the field, my career path is filled with twists and turns. I graduated with a degree in Bioengineering, Forestry and Nature Conservation. Then I worked as an engineer for a couple of years before moving to Latin America. I lived in Bolivia for over two years where I worked as an agricultural economist. I also started blogging to keep my friends and family at home informed. That's how I learnt how websites work. I also started writing for a small magazine here in Flanders about my experience abroad. After coming back to Belgium, I became a webmaster. Because of my background in engineering, I had an affinity for numbers. That was when data journalism started to take off. I began experimenting with data journalism and making visualisations. I was later hired as a data journalist by De Tijd, a newspaper based in Flanders. After working a couple of years there, I decided to go freelance.
Most recently the World Bank hired you to help create the World Development Report 2021. Tell us more about that.
The World Development Report is one of the World Bank's flagship publications. It is translated into many languages and is published in print and a PDF. Because this year's topic focused on data, they wanted a digital version with a modern shape. The report's tone of voice is a bit lighter. Readers will find visualisations and animations in the articles. I developed three of the stories. This involved building the storyline, designing the visualisations and programming the stories.
Tell us about those stories you worked on.
The stories focused on competition, the new data economy and how governments could take action to have a healthy ecosystem in the data market. One story I wrote examined how countries regulate the flow of data across borders. Some countries have decided that whenever data is produced, there should always be a local copy within the country. Others have a much more liberal approach that does not control what data is flowing in and out of the country. These different models have different advantages and disadvantages. Another story was about building trust and regulating the data economy. More specifically, it looked at cybersecurity and how countries are governing the protection of privacy and trying to enable the data economy by having good open data laws.
Describe your workflow. What software and tools did you use?
For the visualisations, I used R programming language and ggplot2 package. I made very rough sketches in Google Slides to show how the storyline would unfold. It was a collaborative process. For instance, we had many discussions with the authors of the chapters to develop this. They had a big say in how it should look and how the story would flow. To develop this we had a React-based template to programme the stories.
Who else did you collaborate with on the report?
For the data-heavy visualisations, I worked with Jan Willem Tulp, a freelancer in The Netherlands. We split up the chapters in the book and gave each other some feedback. We also worked with Beyond Words, a data visualisation agency specialising in creating narratives with data. They worked on the more conceptual stories and designed the styling and provided the template to build our stories.
How do you think journalists should handle missing data?
I think one important aspect is the metadata -- the data about the data. Sometimes we must be more explicit about the metadata and maybe even show it. I'm now involved in a new project where we are not just thinking about visualising the data, but also visualising the metadata. For instance, showing how many missing years we have in the data or whether the data is from a government database or a survey. Being more explicit about the metadata can help you as the data journalist and the reader. Data journalists should ask themselves what data is missing and whether they can still draw conclusions from this and tell their story.
What story are you most proud of?
My favourite is a project I worked on a couple of years ago. It's called Rock n' Poll. It's an interactive story where you learn how political polls work and what that means for the uncertainty that is inherent in the polling results. At the time that I built it, I was working at a newspaper and I was really annoyed about how journalists were reporting on political polls. They were focussing on very small differences, well within the margin of error. I wanted to develop something that could explain that uncertainty without using any statistical formulas.
Finally, what advice do you have for journalists who want to begin coding?
One of the best decisions I made in my career as a data journalist and data visualisation expert is to learn ggplot2 as part of the tidyverse. It allows you to build visualisations layer by layer. Once you understand how the system works, you can get really creative and quickly design data visualisations. If you're an Excel user and you're interested in learning to code, start with the tidyverse in R. It's a set of packages that all work well together. I found this easier to work with and understand than base R. As an exercise, I recommend you try to do all of the data operations you'd normally perform manually in Excel and do them in tidyverse. This will allow you to have a script that does all the manual manipulation that you would do in Excel that you can rerun in R.
Our next conversation
Our next Conversations with Data podcast will be a live Q&A with Ben Jones, founder of Data Literacy, an organisation aiming to help you learn the language of data. He will reveal his top tips and tricks for becoming an effective data storyteller. The conversation will be our first live event on our Discord Server! Join the conversation to take part and find out more.
As always, don't forget to let us know who you would like us to feature in our future editions. Share with us your thoughts on our Discord Server. You can also read all of our past editions here or subscribe to the newsletter here.
Tara from the EJC data team,
PS. Are you interested in supporting this newsletter or podcast? Get in touch to discuss sponsorship opportunities.