From its inception, Quartz has been on the forefront of many types of journalism: data, visual, interactive, sensor, and others. Led by our Things team, which is chiefly responsible for making these “things” that combine reporting with technical skills, our 2020 work continues to reflect that tradition. We’re proud to submit our team’s work for consideration of recognition from the Sigma Awards.
Our work consistently takes the form that is most appropriate for our readers and members to find understanding and insight. That is clear in “The coronavirus epidemic is changing emoji usage on Venmo,” where the data and visualization are the story and the text secondary. It’s also apparent in “Coronavirus antibody tests aren’t as accurate as they seem,” in which a customizable simulator takes center stage. But of course we sometimes let words drive the narrative, as you can see in our “Welcome to Leeside” project.
Our team’s vast array of skills is readily apparent in our work. Compare the abilities expressed with pieces like “emoji usage on Venmo,” “Mexico is illegally destroying protected mangrove trees to build an $8 billion oil refinery,” “antibody tests,” and “The Covid-19 pandemic has made canned foods, shelf-stable milk, and vodka seem better.” They run the gamut from collecting and analyzing massive datasets, to detecting changes in satellite images, to coding probabilistic simulations, to dynamically collecting and displaying reader responses.
We aren’t limited to quantitative methods: We use our numeracy and data skills to augment our qualitative reporting. That’s evident in “This is the year Hong Kong began speaking the Communist Party’s authoritarian language,” where a hunch from a reporter based in the city was borne out through a robust analysis of 165,000 public statements and transcripts we collected. It also showed in “This is the chance your local bank branch will close by July,” where our analysis of the dynamics driving changes in retail banking was localized to the readers in every US zip code.
Our work is beautifully designed and delights. It is built for users on desktop and mobile equally. It takes care to pay attention to accessibility, including making sure visual elements are described appropriately for people using screen readers, and has a dark mode for people who need a low-brightness or different-contrast experience.
Finally, and perhaps most importantly: We don’t take our readers’ knowledge of data or statistics for granted. We regularly produce work that helps them gain a greater appreciation for not just our work but the entire media industry. To help with our members’ understanding of the torrent of pandemic-related data, we published “The data deluge,” a resource of data fundamentals. And when the US government’s initial Paycheck Protection Program loan data was released, we didn’t rush into it to provide commodity coverage about the recipients of the largest loans or the areas with the most financing. We provided context to our readers with “Here’s what we know is wrong with the PPP data.” Without these kinds of resources, many readers would be ill equipped to understand some of the most important stories of the year.
It’s no surprise that on average, work made by or assisted by Things team members in 2020 had more readers than Quartz articles that did not. It’s also no surprise that the team’s work was some of the most successful at converting our readers into paying members who, no doubt, think of our work not as data journalism or visual journalism but simply as journalism—which is just as it should be.
Description of portfolio:
Coronavirus antibody tests aren’t as accurate as they seem
By simulating the sensitivity and specificity of antibody tests on a hypothetical population, we left our readers with a greater understanding of how even tests with accuracy figures near 100% can lead to false results.
Welcome to Leeside
While it looks like a traditional high-production news article, it’s designed—from its content to its layout to its interactivity—as a museum exhibit about a not-so-distant future that the world is headed towards. The lens allowed us to merge real-world data and dynamics with simulations and imagined artifacts to create an immersive story of a community dealing with the realities of climate change.
Mexico is illegally destroying protected mangrove trees to build an $8 billion oil refinery
In a classic example of using contemporary techniques to augment traditional reporting, we used satellite imagery to not only show the destruction of mangrove trees visually, but also quantitatively by using software to measure the amount of destroyed vegetation that can be seen in the images.
The Covid-19 pandemic has made canned foods, shelf-stable milk, and vodka seem better
We began collecting responses from our readers on two simple questions: Is a grocery item generally good, bad, or neither, and now that the pandemic has set in is it good, bad, or neither. We charted the results live and were able to show which items had become more and less desirable.
This is the year Hong Kong began speaking the Communist Party’s authoritarian language
We analyzed the language used in 165,000 public statements to show the steep rise in authoritarian language that was showing up in government statements as Beijing started more forcefully inserting itself into Hong Kong’s governance.
This is the chance your local bank branch will close by July
We analyzed the data on just under 100,000 US bank branches from the FDIC and crafted a multivariate model to determine the probability that a branch location closes in a given year. As we explain in our findings and methodology we automatically insert the data about your area and a nearby bank. Alternatively you can manually select almost any bank branch in the US.
The data deluge
This is a field guide to help our members to understand and navigate data and data analysis during the pandemic and beyond. Of particular note are the pieces “The challenge of percent change,” “How to compare data that change over time,” and “Covid-19 is a reminder to beware of seasonality in data”
The main metric economists use to measure inequality is deeply flawed
This was another service to our readers to help them understand a commonly reported metric. Through clear diagrams, the promise and limitations of the Gini coefficient are made apparent with simple, plain language.
Here’s what we know is wrong with the PPP data
Our soup-to-nuts analysis of the original PPP data release’s flaws involved not just the data itself, but comparisons to databases from the US Postal Service, Social Security Administration, and Census Bureau. That allowed us to discover data about applicants that the government had intended to keep private.
The coronavirus epidemic is changing emoji usage on Venmo
We collected data from Venmo, analyzed it, and created a story that treats the data and visualization as primary content.