This is the second of the series of interviews with winners of the Data for a Cause challenge, where volunteers from all over the world design data visualizations for mission-driven organizations, NGOs and charities. In each article of this series, we will focus on one of our amazing volunteers, to share their story or perspective, and to reveal their insights on how they achieved the level of expertise they have today.
Amar is originally from India. He received his Master’s degree in business intelligence and analytics from Saint Joseph's University. Data visualization wasn’t something he initially was interested in pursuing in fact he was interested in data mining. But then Dr. Campbell, his data mining professor, asked him to submit a report with a visual take in it, as part of the coursework. Amarendranath now works as a Data Visualization Analyst at Comcast.
Without further ado, here is Amarendranath’s story.
Where did you find the inspiration for your data visualization for OMSIhelp.org?
When I saw the announcement of the challenge, I told myself: ”don’t mess it up”. Because the topic was sensitive - attitudes towards mental health in the tech industry. I wasn’t sure if any of my previous employers had any mental health programs available either. For two days I kept thinking about what to do and how to do it. I kept going to the website of Open Sourcing Mental Health. I’ve been watching their videos.
For the color scheme: if an organization has a color scheme, then I stick to it. And it did in this case. I’d been brainstorming and seen a couple of ideas on the Internet. For example, the tile map that you see in the visualization on the left comes from the template by Matt Chambers. Also, being active in the Tableau community certainly gave me a couple of ideas. I used all of them to create something good.
I definitely find the Tableau community helpful. Everybody has something to offer – a lot of people, a lot of cool things. You can get lost there. People who encourage me and give me feedback are Chris Love and Pooja Gandhi. Both of them are Tableau Zen Masters. Among others Adam Crahen, Eva Murray and Tamara Gross constantly give me valuable feedback. I hope I’m not forgetting anybody. All these people giving me feedback, telling me what works and what doesn’t, and I make changes. That’s how I work and keep improving. For most of my initial work I give credit to Andy Kriebel (Tableau zen master), because I picked a lot of design tricks from his style. But I think I have to thank the entire community for my inspiration.
How long did it take you to create your data visualization for OMSIhelp.org?
I spent two days just gathering the ideas and then thinking about how I wanted to visualize them, but I think I spent close to two hours on the visualization itself.
If you take a close look at the visualization, it is pretty simple: bar chart, tile map template, and stacked bar chart. I wanted to do something simple and doable.
What I usually do is sit down and sketch it on paper first. I know many people, like Eva Murray, do that. And for the most part, I try to stick to my sketch. It helps me to stay on track and not get overwhelmed with ideas. Other people say that you shouldn’t try to do it. You shouldn’t try to fit your data into a sketch. Nathan Yau in this book 'Data Points' says - let the data do the talking.
So I try to balance these two approaches. I pull my measures and dimensions and see what I can do with them. I try to create a balance and try different techniques. Sometimes I’m successful and sometimes I’m not.
What are some of the challenges you faced when building the visualization, and how did you overcome them?
I was so carried away with brainstorming, I lost track of time. I forgot about the deadline. Also, it was the first time I was visualizing a survey. I hadn’t done it before. Steve Wexler has done a lot of work with surveys. I watched one of his videos from the Tableau conference. That kind of gave me a pat on the back, saying that this could be done.
I definitely wanted to have some emotion in my visual, not just pie charts and bar charts. I wanted it to be dramatic. My biggest challenges were the layout and the color. Because red is a loud color, I kept changing colors, and at one point I asked myself, “do I really want to use red?”. So I went to green, a sub color used on the OSMI website, but it didn’t work, and I had to revert back to red again. Glad, it worked out.
Also, there was a lot of data, so I didn’t know what to narrow it down to. So I started grouping the data and divided my data visualization into three sections.
Finally, I read the 'Storytelling with data' book by Cole Nussbaumer. In her book she says that data visualizations can be pursued like movie narratives, there is an introduction, then you have the drama going on, and then you have a conclusion or climax. So in both of my visualizations for Data for a Cause I did it in order: problem, what is the cause, and conclusion. This is a good format, but I also try to experiment so that I don’t stick with the same thing again and again.
My thanks to Corey Jones for helping me with labels on the viz.
What is your favorite data visualization tool?
Tableau is my favorite data visualization tool. Hands down! Tableau is a great tool. They have a great community and support. They have public sessions all the time. Also, the Tableau public gallery is great. It has a lot of creative and inspirational works.
Recently, my brother showed me D3.js. I don’t see myself using this tool anytime soon, but it is on my radar.
Amarendranath’s data visualizations (click on the images to view the interactive versions).