Data for Social Good: A Q & A with Visual Artist Vivian Peng

Vivian Peng

Vivian Peng

Q&A

Nearly a million children die of pneumonia each year, and in a new campaign, Doctors Without Borders is calling on Pfizer and GSK to lower their pneumonia vaccine prices so more kids can be protected. Vivian Peng, a visual artist and recent Columbia graduate, is on the creative side of this push, A Fair Shot, aimed at making pneumonia vaccines affordable in developing countries.

Peng’s interest in public health began after volunteering on an HIV-education project in Tanzania one summer. A 2009 graduate of the University of California at Berkeley, she worked in advertising for several years before moving to New York to study at Columbia’s Mailman School of Public Health. With coursework in data visualization and programming in R through our data science certification program, Peng graduated in 2014. In April, she spoke at the New York R Conference on “Storytelling with Data,” sharing the roster with her former professors Mike Malecki and Jared Lander. She recently shared some career tips.

How did you come to work at Doctors Without Borders?

I’ve wanted to work there since I was little and even studied premed at Berkeley to prepare myself. At Columbia, I got the chance to work at Doctors Without Borders as a web intern for two semesters, which I parlayed into freelance design work for two years. In April, I started fulltime on the pneumonia vaccine campaign.

What was your first successful design project?

As a web intern I designed graphics in my spare time and with permission, posted them on social media. In 2013, I was asked to work on an earlier affordable-vaccine advocacy project. I pitched the idea for a white board illustration, which came out like this: Dear GAVI, Please Let Us Access Your Discounted Vaccine Prices. Gavi Alliance, a foundation that provides vaccines in many developing countries, published this response. Doctors Without Borders launched its A Fair Shot campaign to keep the pressure on.

“Peng mapped the changing demographics of San Francisco’s Mission neighborhood in her data visualization class at Columbia.” (Courtesy of Vivian Peng)

At Columbia, you looked at gentrification in San Francisco’s Mission neighborhood for your final data visualization project—why?

I lived in the Bay Area for several years as an undergrad. Since then, the neighborhood has gentrified as Twitter, Google, and other tech companies have moved in. I was curious to see how the displacement of Hispanics and other ethnic groups would affect the long-term health of those groups.

What did you learn?

U.S. Census data confirmed that fewer ethnic groups now live in the Mission. A bar chart could have worked for representing the data but I wanted to dig into D3 and try something more ambitious. I tried Googling the idea I had in mind with no success. Finally, at Mike Malecki’s suggestion, I tried “circle packing,” a technique for fitting circles of different sizes into a small space.

Would you use a circle graph today?

Probably not. They’re not the best way to show population change over time since the human eye has trouble perceiving relative circle areas. Think about spare change. Is a quarter two times, or 10 times, bigger than a dime? Who knows! Today, I’d probably design an animation to describe the Mission’s changing demographics. (I did make a series of animated GIFs that summarize my master’s thesis in eight animations)

Doctors Without Borders

Pneumonia vaccine prices are all over the map, as Peng shows in the above graphic. (Courtesy of Doctors Without Borders)

What’s next at Doctors Without Borders?

I’m hoping to build an interactive map that shows how widely pneumonia vaccine prices vary by country. Prices are all over the map, with lots of data missing. I recently put the project on GitHub to crowd source some of this information.

Why does data visualization fascinate you?

In public health we’re sitting on lots of data we don’t know how to communicate to the public. We think the data will speak for itself but unless you’re an expert in a particular topic, you may not understand why the data matters. Data visualization can turn raw numbers into a clear and powerful message.

What are your favorite data visualization tools?

I like R for data munging, base graphics, and maps. I like D3, the open-source Java Script library, for anything interactive. I live in Adobe Illustrator, but it limits your work to a one-time use. Coding lets you share and reproduce your graphic and also update it when new data becomes available.

Do you have advice for aspiring data visualizers?

Learn to code. The learning curve may be steep, but the payoff is huge. Find a topic you’re obsessed with--in my case that was public health. If you’re really interested in answering a question, you’ll find a way. Also, familiarize yourself with the work of pros like Mike Bostock at the Times (who wrote D3) and Edward Tufte at Yale.

What do you do when you get stuck?

Drink a beer, or post a question on Stack Overflow. There are super-expert people there who will look at your code and guide you. I also go to R Meetups, which are really huge in NYC. 


550 W. 120th St., Northwest Corner 1401, New York, NY 10027    212-854-5660
©2018 Columbia University