To hear Rahul Bhargava tell it, too many people are afraid of data.
“The first thing you feel is their trepidation,” says Bhargava, a research scientist at the MIT Center for Civic Media who speaks with organizations and workers about the power of digital information.
The apprehension, he says, is rooted in a common misperception. Data management is seen primarily as a technology challenge—one that requires sophisticated software and technical know-how. Bhargava says, in reality, analyzing data demands little more than the ability to spot patterns and the stories they reveal. As for the math, he says, “Statistics is mostly about counting, and most people are pretty good at that—even my 7-year-old daughter.”
To help people overcome the unease, Bhargava recently helped launch the Data Culture Project, a set of free online tools and activities designed to help organizations build data skills across their teams. The platform aims to take the fear out of data analysis through exercises that are meant to be fun and engaging for employees. One activity asks participants to find commonalities among details about dogs registered in New York City or UFO sightings in Massachusetts.
The project’s underlying premise is that many business leaders mistakenly believe they need to hire scientists to get data right—when all they really need is to build a data culture.
“Building your number-crunching skills doesn’t help you use data across the organization to advance your mission,” says Bhargava, who built the Data Culture Project with Catherine D’Ignazio, an assistant professor at Emerson College.
The project was funded by a Digital Impact Grant, one of 13 awarded since 2016 with the support of the Bill & Melinda Gates Foundation to advance the safe, ethical, and effective use of digital resources for social good around the globe. Digital Impact is an initiative within the Digital Civil Society Lab at the Stanford Center on Philanthropy and Civil Society (Stanford PACS).
Building a Data Playground
From the outset, Bhargava and D’Ignazio concluded that existing data tools, such as DataWrapper, Google Sheets, and Tableau, do a good job of creating visualizations, but fail to provide users with fundamental skills or the right mindset to integrate data into their day-to-day work.
“You want to get people who are otherwise intimidated by data to use their skills, their context, and their knowledge to work with data information to tell stories,” says Bhargava.
Existing tools lack another critical component, according to Bhargava and D’Ignazio: the fun factor.
“We were inspired by Edith Ackermann, the educational psychologist who talked about how people are much more willing to try something risky when they’re in a playful situation,” says Bhargava, who met D’Ignazio when she was pursuing her M.S in media arts at MIT.
The Data Culture Project comprises three, 30- to 45- minute activities centered around videos, specific projects, and discussion topics for organizations to run on their own (Bhargava and D’Ignazio suggest monthly brown bag lunches). One warm-up exercise asks workers to create 3-D sculptures with wires, fuzzy balls and other craft materials. In this way, participants can start to talk, and feel more confident, about ways data connects to their everyday jobs.
In another activity, called “Sketch a Story,” teams compare music lyrics.
“Just knowing that Elvis sang the word ‘love’ 67 times in a song is not all that interesting,” says D’Ignazio. “But once you look across a lot of musicians, you can find patterns of communication. When Elvis is talking about love, maybe he’s talking about one kind of love while Beyoncé talks about another kind of love. That becomes the seed of a story.”
Many tools used by the project are housed under a separate site Bhargava and D’Ignazio developed more than two years ago, Databasic.io. The Data Culture Project isn’t just about the tools; it’s a “lightweight curriculum” for organizations to build data skills across verticals.
“Talking about ‘culture’ conveys the right sense of difficulty,” says Bhargava,” because everyone knows that changing the culture of an organization is hard.”
Bhargava and D’Ignazio are clearly onto something. Some 60 organizations—mostly nonprofits and media companies—asked to be part of the project’s test phase; Bhargava initially thought they’d be lucky to get 15. In the end, participants ranged in size from a two-person nonprofit based in the United States to a 10,000-person international aid agency.
“They were all facing similar problems, which weren’t really about the data but about communication,” says D’Ignazio. “Data was not getting shared across the organization.”
At the Michael Smith Foundation for Health Research, a beta tester, seemingly disparate teams are finding ways to collaborate with data after completing project activities. Now, the Canadian nonprofit plans to host all Data Culture Project content on its intranet and share examples of how employees are using the site’s tools in the jobs, according to Maija Tiesmaki, a senior evaluator at the foundation.
Next, Bhargava and D’Ignazio are looking to add new activities focused on the art of storytelling with data and to get users to join discussion groups where they can share experiences and insights. They are also hoping other creators of data analysis tools will be inspired to design more services for beginners, not just power users.
“In Silicon Valley and among researchers, it can be hard to remember that technology is intimidating to a large swath of people,” says Bhargava.