Can design help make data more accessible, comprehensible and equitable?

Most discussions of data and design these days revolve around ethics and the murky world of data collection by tech businesses. And with good reason. Over the past decade or so, scandal after scandal has revealed how Silicon Valley companies, in particular, have not designed systems well enough to mitigate the unethical (and, in some cases, unlawful) collection and use of data on a massive scale.

There are, however, many designers and artists out there today creating meaningful and valuable work by using data as raw material, while also challenging a narrow definition of data that sees it simply as something that only large tech companies have the scale and ability to collect.

Last week in New York, we gathered four creatives who belong to this group at the Wix Playground space in the city’s Meatpacking District, for an evening of talks and discussion around the topic Data, Narrative and Design. While none of them underestimated the issues relating to data collection and use in modern society, they all talked through projects that use data for building community, for informing people, and for the public interest.

Mona Chalabi, a data journalist and the data editor of The Guardian US, describes her job as making data more digestible to a broad audience. Her work takes often-impenetrable data that reveals socio-economic inequality, endemic racism or government ineptitude and makes it strikingly clear and visible, stripping away any obfuscation. “Often data is saying something really important, but the visualisations make it hard to understand or confusing,” she explains. “I’m trying to make the data memorable.”

She has a number of different methods for making her illustrations and their data memorable. Anyone who is familiar with Mona’s work will know that she mostly creates illustrations that represent the topic visually in some way. “I try to link the subject and the visualisation,” she says, “because I know I’ve only got a second to grab your attention and I want you to know immediately what this illustration is about.”

Another way she makes data approachable and visible is by using it to tell stories, to build narratives. Attention spans, as she notes, are perhaps not what they once were – but, she says, “we can build those attention times back up by giving people information in stages, rather than all at once.” Lots of the best examples of her work employ this technique, constructing a story over a series of images and leading the viewer to a surprising (and often troubling) conclusion, all through using data.

This was also a method employed by Laura Kurgan, Professor of Architecture at the Graduate School of Architecture Planning and Preservation at Columbia University, where she is also the director of the Center for Spatial Research. At our event, she spoke about her project In Plain Sight, which was exhibited at the 16th International Architecture Exhibition in 2018. In this project, she and her colleagues used satellite images, in sequence, to show how quickly electricity returned to Houston after Hurricane Harvey and Puerto Rico after Hurricane Maria (both Category 4 storms).

After 11 days, Houston was back to 100 per cent electricity coverage, whereas after 120 days, still just under 30 per cent of households in Puerto Rico were without power. “The difference between how quickly power came back to Houston and how slowly it returned to Puerto Rico,” Laura says, “shows us that our world is fractured by inequality.” Again, for Laura, building a story, a sequential picture, through data is the best way to land that shocking point.

The evening concluded with a talk from Barron Webster, a designer currently working at Google Creative Lab, who focused on how, as an individual, data can often be alienating and feel like it’s beyond our control – but he also argued that it doesn’t have to be. Instead, it can be two important things: reclaimed for personal knowledge and fulfilment; and secondly, it can be made from scratch for your community.

As an example of the first of these, he showed a project he did back in 2015, when he collected and downloaded data on himself, his actions and movements across a month and then published them in a physical book, titled Buy My Privacy. “Computers analyse this for autocomplete or for serving better ads,” he says. “But there are poetic moments when we use these footprints for ourselves. The project made me more aware of both the footprints I left but also how personal those footprints are.”

Baron also showed two projects that explored how data can be made from scratch for your community. The first was payinterns.nyc, which saw him gather information on which design studios and companies in New York paid their interns a living wage, and then publish that information on an open platform. “I learned that data doesn’t have to be big or hyper-complicated to have an impact,” he says.

A not dissimilar project was shown by Tal Midyan, a creative director and designer currently working at Spotify. He talked the audience through The Viisualizer, a microsite made for the release of Bon Iver’s latest album, i,i. The project collected almost-live data from around the world of who was streaming the album and turned that data into a fun and interactive platform.

It was all about building a sense of community. “If you’re a fan, your experience of music is really personal, especially if you predominantly listen with headphones on,” says Tal. “So if you see people all over the world who are also listening to the same thing, it’s opening your eyes to how communal the album is, it’s making that connection.” He describes the project with a really nice analogy: “It’s almost like you’re watching a live stream of Coachella or a concert, but the camera is turned to the audience, not the artist.”

Barron also showed the audience at the Wix Playground a project called Teachable Machine, a Google Experiments initiative that allows people to train a machine-learning tool based on a dataset provided by them directly. “Typically,” he says, “models get made for the biggest possible audience, like the speech models in Amazon’s Alexa and Google Home. And they do work for most people – but not all people.” Teachable Machine allows people to create their own models based on their own data. “It’s letting these people use their own data to choose how computers see them, instead of having to fit into the ways programmers in Silicon Valley or New York optimise it to see.”

Visibility is also something Mona feels is incredibly important about what data can achieve. “In theory at least, using data means I can show a multitude of experiences rather than just my own,” she explains. “When data is at its best, it shows all the nuance and variation in those experiences – it allows you to see how age, the colour of our skin or where we live can affect everything that happens to us. And if you can do that effectively (and by effective, I mean accurately and sensitively), then people who are so often smudged out in the average can feel seen.”

Barron is not unaware of the slight irony many people will perceive in him speaking on data ethics as a member of Google’s Creative Lab. But he believes his work stays true to a central Google tenet. “There’s this ethos: Focus on the user and all else will follow,” he explains. “Basically, the ethos that I’m talking about is that same ethos. The problem is that it’s very easy for people at these companies, Google included, to forget that or to assume they know better than the user.”

He calls this a “design problem writ large,” because, “it’s very easy to see a problem and to say, ‘I know how to fix that.’ It’s much harder to see a problem that users are facing, and to ask them, ‘How would you solve this? What would you do?’” This requires far more translation, because these users might not have the technical literacy or the design literacy. But when you do make the effort to translate, he argues, “you see more personal results, the users feel more invested in those results, and those products or features often bloom in a way they don’t when technologists or designers are deciding for the consumer.”

In the future, when it comes to data, his hope is that “more companies start to see the value in opening their toolboxes and datasets in a way that is accessible to people who are not technologists and data scientists.” For this to happen, companies need to stop trying to be gatekeepers: “There’s this assumption that people can’t handle technology, that we need to abstract it and simplify it until it doesn’t resemble the technology,” Barron says. “But for a lot of users, understanding the complexity helps them to use it. Machine learning, for instance – the complexity is what makes those technologies work and sing. We’ve got to take a leap of faith and open the box, because people will also figure out the complexity if it provides something of real value for them.”