For the first year of the Covid-19 pandemic, I wrote down every siren, helicopter, and construction noise I heard. I charted those noises, and included other sounds and events within earshot from my home. Each line in the project represents a day in my personal soundscape from March 29, 2020 to March 29, 2021.

To me, the ambulance siren symbolized the aural manifestation of the invisible disease infecting the city, and an outward, palpable symbol of the fear many were experiencing at home. I live in New York City, by many standards the U.S.’s ground zero of the pandemic. The density of the city, with its hospitals, helipads, and construction sites, obviously comes with noise, but the new uproar made me ask, “What is the shape of racket, or the color of intrusion? What would a composition of sonic disturbance look like?”

My observations tracked with what was happening locally (such as protests) or personally (binge watching to pass time, or meditation to help process it). Weather is noted as well, as it correlates to what acoustic activities could happen during say, a hurricane. During lockdown I also heard some of these sounds in a virtually or in a “meta” way: on podcasts, video and phone calls, and on TV.

By logging the sounds from mundane to noteworthy, I visually illustrated them to understand somehow, to make sense of it all. It revealed itself as a sort of orchestral score, with densely layered notes. I realized I had experienced stages of grief… anger, depression, and finally, acceptance. Since capturing the pandemic sounds, my senses have sharpened and I am keenly focused on the role that quiet plays in my audible environment now.

Q&A with Heather Jones

During lockdown many people found distractions in puzzles, Netflix, baking, etc. But “The Siren Project” leaned right into the pandemic. How did that influence your perception of what was happening?

It made me think about it every day, all day, instead of turning it off completely. I did the Netflix and puzzle thing, too, but kept noting the sounds throughout. I thought it would be a three-month project, but as it went on my perception changed as I saw that the pandemic would be way bigger than anyone originally imagined.

The “musical score” has both personal notes (what you were watching, attempts to meditate, deaths of family) and public ones (the protests, the clapping for workers, the Blue Angels flyover). Did you intend for the data to be more of a personal record or a register of our collective memory?

Probably more of a data collecting project and general noise account, less of a personal journal. But I always like the human part of charts and graphs, little notes and factoids always make them less sterile and richer to me.

What is “The Siren Project” to you now? 

An exercise in endurance?! 

What was the sound of the construction noise? It plays so prominently in the score. How did the sound of building infrastructure feel to you, especially against the backdrop of the other pandemic noises like the sirens and the helicopters? 

I felt I had to cleave to the will of the developers, and wake or work on their schedule. It made sense that the city continued to build (because it never, ever stops) and also that people renovated home spaces while stuck in them. But I couldn’t believe I was the only one suffering from the clamor, so started looking at stats, like from 3-1-1 complaints going up 3,000%, or new bills being introduced to eliminate helicopter traffic.

The sirens responded to the pandemic, the news copters reported it, but construction sites went on like it wasn’t happening at all, like, “Hey this is NY and we need those luxury condos and capitalism doesn’t stop for anything.” Some people I shared the project with thought I should delineate a marking like between a rototiller verus a jackhammer, or a fire engine versus an ambulance. But that just seemed cruel.

Most of the sounds are man-made noises, including the fans and white noises you used. But there are also sounds of nature, like the birds. What were the trends you noticed with the birds? 

That I strained a lot to hear them, or weeks went by without hearing one. Now when I get out of the city my hearing/senses are really sensitive, and chirping seems louder.

Was all the data collection manual? How did you keep it organized?

Yes, on little squares of paper and pens and paperclips; it was very analog. Then I input my raw notes into an Excel spreadsheet and eventually exported to Adobe Illustrator.

What was the biggest challenge with the project?

The daily grind, and wondering if I was training my ear to hear the tiniest of noises at the same time wanting to block it all out. 

From a creative standpoint, how did you cope with re-living every day of the pandemic when you finally put the data into the viz? I imagine this was a very manual process, so was it tough to go through each day again as you transcribed the score? 

Yes, inputting it and trying to wrestle it into something “beautiful” was another challenge altogether, but I tried to focus on what was possibly different about every day, or discover trends at the end of it. It did feel good to “check out” as some feel about that time, like a pause on the rat race, a chance to stop and create something.

The post The Siren Project: A Daily Account of Pandemic Sounds appeared first on Nightingale.

Under her art direction, the final design—which stretched across the entire front and back page of the paper—won gold medals from the Society for News Design in multiple categories including “Story Design,” “Front Page,” and “Combination Print & Digital.” Carrie, herself, also won gold in the “Individual Portfolio” category.

In the following conversation, Carrie talks through the process of capturing one million deaths in print media. This interview was first published on “Newspaper Design.” It has been lightly edited.

How did you conceive the award-winning “One Million: A Nation’s Immeasurable Grief”?

When I was asked to do this, I knew it would have to have news value, be informative, be sophisticated but also carry the weight of the loss somehow. I was looking for a way to do all of that.

I heard the graphics editor Jeremy White, who was working on this, already had ideas. The dots concept was his—I just saw it and saw some of the shapes the dots might take. When I saw the United States formed by the dots and each of them was a person, I knew that was the image. It was informative, emotional, and simple, while also being very complex. It did a lot of work without overwhelming the viewer. Adding the annotation that each dot was a human was something I felt got to the heart of the image, so I knew we’d need to include that somehow.

There was talk of making this front page bigger than the others to make the statement that this was a historic moment. So my managers, Fred Bierman and Andrew Sondern, and I wanted to make it a wraparound page one if we could, though we knew it would be a challenge.

Was it even possible? We needed to see a print version of the graphic. Then we needed to figure out how it would print best on newsprint. We really wanted the dots to print well. This was very complicated and involved trying various dot sizes and ink combinations (grayscale and black & white) and opacities. We also had to figure out whether it was worth adding outlines for the country, states, counties, etc. We actually did an eight-page press test to make sure we were doing the right combination for print so that it would be readable. After that we just had to wait until we were closer to the date to finalize the map with the latest data.

“We really wanted the dots to print well. This was very complicated and involved trying various dot sizes and ink combinations and opacities.”

At the same time we had to convince editors at The New York Times to just run this on the front page and clear the back page of ads for this paper. One thing on the front page of The New York Times?! And it’s not even a story start. But once people saw this, they started to get on board. They found it as moving as we did. Tom Bodkin and Tom Jolly were very helpful in helping us move forward.

Then at last, there was writing of the display type, which took a village. But I believe where we landed was right.

It was a sleepless night both before and after this went to press.

How have readers responded to that unconventional approach?

We never do something like this for page one. Never. Readers took note. Of the responses I heard, people found value in this treatment in remembering loss and recognizing what we, as a nation had been through. That meant a lot to me, because that’s why I do what I do.

“We never do something like this for page one. Never.”

Which software have you used to make it?

I know our graphics team has special software to gather and analyze data, and they had been gathering since the start of the pandemic. For me, to make sure it would print well, I placed the high res .tiff I received from Jeremy into an InDesign document with the single dot annotation and the display type and exported it as a pdf with our press settings to insure everything would stay together and print well. Then I had to place onto the live pages in Newsgate as a single image. But with the normal page elements (like the NYT flag) undisturbed in our system.

This page won so many top awards in different categories of the SND competition. Why is this work striking from your point of view?

First, I just want to say that I am so honored and proud of this work receiving recognition.  Everyone involved worked very hard and with great care. I think it’s striking because to me, this feels like more than a newspaper page. This feels like more than a layout. This feels like a moment of recognition and memoriam that we visualized.

“This feels like more than a newspaper page.”

That is something that I didn’t know was possible but that is how I see this page. This feels like more than a layout. This feels like a moment of recognition and memoriam that we visualized. That is something that I didn’t know was possible but that is how I see this.

As an artist, what is the emotional feeling that passes through you when doing this?

I’ll be honest, I cried a lot working on this. Ans I’m okay with that. This was such a massive loss and I think really letting myself feel that, lets me know that I’m doing work that is honest and has depth.

To whom are you going to dedicate this award?

To the readers and the future readers (two of which are my kids). Journalism makes the world a better place, and I’m glad I am still a part of that effort.

How do you approach your projects?

I always ask myself what’s the core of this piece? What makes it special? What does it need to convey? And I let that inform how I envision it in print. Then I like to talk with any editors involved, make sure I know everything I should know and am thinking correctly about the project. Then I usually start to write ideas, words, concepts, and then start drawing pages. That translates to the computer and then eventually to the page.

How do editors and artists work closely in The New York Times?

It depends on the team and the project. I think my best projects have been because of good communication and collaboration with editors. I think every project needs to be about the story first and if you can’t work closely with an editor, you might be missing out on ways to elevate a piece in print.

Do you think AI is going to create a revolution in the field of data visualisation and information graphics?

I think it will be a great asset but not a revolutionary tool. You still need a human to evaluate certain parts of projects, be sensitive to tone and nuance. I’m also interested to see how news organizations will keep readers trust and maintain editorial excellence as they begin to use these tools.

The post Capturing One Million Deaths on a Page: A Chat with NYT’s Carrie Mifsud appeared first on Nightingale.

I’m told by the authors that this project grew out of a loose collaboration over many years, its origins going back to 1995.  At the time, Paul Kahn was running the Dynamic Diagrams agency and Hugh Dubberly was Design Director at Netscape where they collaborated on a series of celebrated maps of early web-based applications. Twenty years later, they both found themselves teaching in the Information Design and Data Visualization MFA program at Northeastern University where they met Liuhuaying Yang when she was a graduate student in that program.

We asked the authors to talk about what prompted them to write about these visualizations, their collaborative archive called the COVIC database, and what the data visualization community can learn from this kind of reflection.

JF: Why did the team write this essay?

Paul: I was looking regularly at The New York Times and The Washington Post and was impressed by the drama and ingenuity of the visualizations they published. The death milestones were, intentionally, the most dramatic. I started subscribing to the Financial Times at the start of the pandemic. The FT’s view of the world from The City in London was so entirely materialistic, it was a balance to reading the US news. That is how I discovered John Burn-Murdoch’s work and saw how influential his line charts became during the ‘flatten the curve’ phase.  It was Liuhuaying’s contributions that really gave us a sense of the visualizations coming from Singapore and China. And Hugh was analyzing the way The New York Times was devoting blocks of front-page space to pandemic visualizations week after week. We were sending each other examples and commenting on them for many months.

I saw there were great examples coming from many different news organizations that focused on the quantity of loss. The emerging pattern of strategies to visualize death milestones struck me as something worth writing about. Once it started, it was natural for it to become a collaboration project. It was possible because the three of us had views, ideas, and experiences that complimented each other.

JF: What is a death milestone? Is this something unique to the COVID-19 pandemic? 

Paul: In the essay, we say these milestones are visualizations of big numbers and points of reflection. In our initial email discussions, we were comparing the online visualizations to the physical COVID-19 memorials that were popping up in different countries, and comparing those to war memorials. Along with the online visualizations, we collected links to stories and videos about physical memorials. There was A World Remembers in New Zealand’s STUFF that described memorials in the US, England, Germany, Italy, Argentina, South Africa, Indonesia, India, and Russia. We learned from this and decided to limit the essay to comparisons of the online examples.

I do think the visualization of death milestones is unique to the COVID-19 pandemic because they appeared while the event was happening. The Oklahoma City National Memorial or Mamayev Kurgan Memorial to the Battle of Stalingrad or The Nanjing Massacre Memorial Wall is created to remember events that happened in the past. Some of the visualizations we present in the article were certainly reaching for closure, but closure wasn’t there. 

Liuhuaying: I was inspired to search for examples of death memorials. How do we link to what visual representation? What do we think of death? I still remembered Paul’s comment on those physical representations: “a lot of flags, a lot of melting ice figures, a lot of nails, a lot of names, the wall of faces”. It seems that we have a kind of “convention” when we memorize deaths. Then how is it different in a pandemic context, particularly in COVID-19 one? Also, compared to physical representations, we master other kinds of techniques, such as data visualizations with interactive web pages and animations. How would this make things different?

JF:  The introduction to Unimaginable Death mentions the COVID-19 Online Visualization Collection (COVIC). Can you explain that collection and how it relates to this project?

Paul: COVIC began in March 2020. That was a special moment for many people around the world. I participated in a Center for Design conversation with Dr. Isabelle Boutron from the University of Paris and this is when she started what became the Covid-19 – living NMA initiative to map all the COVID-19 clinical trials. Last week at a talk at the Rumsey Map Center, Jessica Martin from Bloomberg CityLab recounted how it was that moment when she and Laura Bliss came up with the idea to ask readers to submit homemade maps of their lives that became The Quarantine Atlas. It was a moment when everyone had to stop and look around for something to do. That gave birth to many projects.

From the beginning, COVIC was a collection of visualizations. Initially, it was a spreadsheet of articles that contained visualizations, with metadata about the publisher, language, country, date, and subject. I sent an email to a network of friends, colleagues, and former students in many countries asking them to send in links, which seeded the international nature of the collection. It was Hugh who looked at this and said: This is about the figures. It was at that point, when we had a few hundred articles, that we went back to record and classify each individual figure. We tried to brainstorm a way to automate this, to extract and clip the visualizations, but we did it ‘by hand’ so to speak. We had many undergraduate and graduate students from Northeastern helping with support from the Center for Design, along with volunteers who contacted us.

Liuhuaying: I gathered examples for COVIC whenever I saw them, especially those from the Chinese community. They are numerous and many are fantastic. It’s also interesting for us to see how different or similar the visual strategies can be by comparing it to the English-speaking community.

The first example I contributed to COVIC was my own project. In early Feb 2020, I created an interactive website titled “Dynamic tracking of COVID-19 in Singapore” for zaobao.sg, the digital platform of the Chinese dailies in Singapore. We visualized the daily situation reports released by the Ministry of Health to inform and communicate with local audiences about the COVID-19 situation in Singapore. As the situation evolved, we kept updating the data every day until June 2nd. The aim of the visualization is to help better picture the relationship between cases within a cluster, how these clusters are interrelated and activities-based, and how cases are disseminated to various hospitals.

I presented this project in the Summer school course and on other occasions to share my visual strategies, intentions, and challenges. A comment I heard most is how unique or uncommon it was that we visualized individuals. I thought it was attributed to data availability until I saw how other examples in COVIC managed to visualize the trees in the forest.

Paul: In May 2020, I taught remotely a summer school course about collecting visualizations during the pandemic. I was in France, my students were in their homes around the US, and my guest speakers were in many parts of the world.  The discussion with students and presentations by Liuhuaying and others helped to solidify the ideas. We wanted to create something that could be used for teaching and research when the pandemic was no longer happening. We could see the pandemic was inspiring and challenging the data visualization community, it was changing data journalism, and it was producing elaborate yet ephemeral results. The data was changing, things were appearing and disappearing, paywalls were opening and closing, and vaccines, mutations, and social and political issues were morphing every few weeks. I thought then, and still think, it is important to gather with an open mind about the role visualization plays, then classify to support later search and filtering.

COVIC is an opportunistic collection. I am not a social scientist, so I had to learn that this is a ‘thing’ or at least a recognized method. We have collected what was presented to us, what people sent us, what we came across, and what was linked to what we read. COVIC is a large and organized sample from an infinite set.

Our version 1.0 was made with Google Sheets, an Amazon S3 server for storing images, and custom code to visualize the figures. That version worked but didn’t scale. We had no idea how big it would get. When we approached 10,000 figures we had to migrate to the current version 2.0, which is managed in Airtable bases and visualized with a custom Javascript SAP. The public COVIC Visualizer is available to everyone and documented on the COVIC website. The metadata can be shared on request in CSV format.

Hugh: Early discourse on design history tends to focus on individual artifacts, this poster, that book, or perhaps a particular subway signage system. But design practice explores ‘spaces of possibilities’ or ‘solution spaces’ with any designed artifact being just one of many possible choices considered. Design history, particularly as it relates to visualizations, is beginning to recognize this fact. Comparative histories (and critiques) require collections of related work; collections, however, require databases for managing and for accessing works.

So, in addition to recording a specific set of visualizations of COVID-19, we also saw the COVIC project as exploring an emerging approach to comparative design history. From that frame, the article on “Unimaginable Death” might be seen as an example of the sort of comparative histories that COVIC supports.

Paul: Exactly. “Unimaginable Death” is simply an example of the kind of analysis and reflection that can be built up from the collection. We came up with a subject – death milestones – and we found a rich set of examples to discuss and learn from. Now we’re thrilled that Nightingale is giving us the chance to present this work to your readers. We hope it will inspire not only discussion but more articles and projects that use this material and this methodology.

JF: One of the challenging aspects of the pandemic was our inability to control the spread of the virus and the resulting waves of infections which lead to inevitable milestones. Did your team have some kind of reaction to the nearing of the milestones? Were you following certain papers and search results or was it more anticipatory?

Paul: I don’t think milestones are inevitable. We mention examples of countries where one-hundred-thousand deaths, half-a-million deaths go by and no data visualization milestone appears online, as far as we can tell. 

I went through my emails to Hugh and Liuhuaying and found that the first set we shared in December 2021 started “from Italy in March 2020 through Germany in November 2021”. Once we had that set I began to anticipate later milestones that would occur in the US and Spain the following year. Liuhuaying found the Chinese examples that were a response to the first surge of deaths. I did search for visualizations in the Spanish media when that country crossed the hundred-thousand mark. And like everyone in the world, we anticipated the one-million event in the US. We all anticipated that we could not complete the essay until after they appeared.

JF: The dataviz world isn’t that big, did your team have a direct dialogue with any of the journalists featured in the collection? 

Paul: I know Liuhuaying and Irene de la Torres Arenas from teaching at the IDV program. They were both kind enough to speak to my class during the first pandemic summer. Irene was working at BBC doing COVID-19 visualizations, though now she is designing for FT. We didn’t talk about death milestones or the specific challenge of representing large numbers at that time. Six months later Irene led the team that produced the 1-year milestone that visualized COVID-19 data as flowers.

Liuhuaying: I had asked Spe Chen, a data visualization designer at The Straits Times who designed the floral icons for the example “Remembering the 5 million lives lost to Covid-19” about her inspirations, especially whether she was aware of the BBC “petal” design. 

She was impressed by Poppy Field – Visualising War Fatalities and searching for visualizations representing the death toll. Inspired by Japanese altars, she decided to make flowers and tried many types of flowers. It was only in the middle of her flower ideation that she saw the BBC “petal” design.

Paul: We heard from several designers indirectly. We learned the backstory of The New York Times examples from the Times Insider series, The Project Behind a Front Page Full of Names and On the Front Page, a Wall of Grief. Later there was Clare Santoro’s Nightingale piece about Alyssa Fowers’ “Cut Short”, and the PolicyViz podcast interview with Aliza Aufrichtig about “Voices of a Grieving Nation.” I also learned about “1 million U.S. COVID-19 deaths” from another PolicyViz interview with Danielle Alberti, and it is also discussed in Behind the Scenes with Axios Data Visualization in Nightingale.

JF: After collecting so many charts about our inability to visualize such huge numbers of human loss, do you each have examples that you find effective?

Liuhuaying: It depends on how we define the effectiveness in this context. I love the three projects in flower examples because of their high emotional values. The flower element linked to funerals and memorials across cultures provides an extra function to the charts: compassion and consolation.

Those examples of dots also provide emotional values by triggering our awe of life: so many lives passed away, sorrow, life is vulnerable, and we need to be more serious about the situation. And aesthetically, both expressed the beauty of lives. However, as always expected in war memorials, we would like to send a bouquet of flowers, and may the deceased rest in peace.

Paul: When Hugh introduced the metaphor of the Forest and the Trees into the essay, it helped me see the fundamental design question that everyone had to face. The designs are not about death and loss, per se. They are about communicating quantities and what quantities are made from. This can be done by carefully labeling the axis of a line chart or area chart. That thought process disengages the visualizations from the emotional experience. And then I recognized that the ones that moved me personally were the designs that re-engaged my emotions by playing with expectations. The designs that ‘violated the grid’ either literally — the death spike on the US map that went through The New York Times masthead — or figuratively by shapeshifting from bars to dot swarms — both the NBC “Seeing the Scale” and The New York Times “How America Reached One Million” do this while engaging the user in parallax scrolling – or animating changing numbers to mimic organic growth – like the unfolding flower petals of the BBC and The Straits Times animations.

Hugh: Building on Paul’s comments, part of what intrigued me about this project — looking at the ways COVID-19 deaths are represented and the ways milestones were marked — was how this type of situation has been addressed before. The AIDS Memorial quilt is an amazing example. Also, the Vietnam War Memorial.  And the Stolpersteine (stumbling stone) memorials in Germany, more than 90,000 small brass plaques place on or near the homes of victims of the Nazis. (See “How Germans Remember the Holocaust” in The Atlantic) These memorials make each individual uniquely visible while also placing the individuals (and the viewers) in the context of unimaginably large numbers. They evoke a mixed set of emotions: awe, wonder, horror, despair, guilt, and more.

JF: How do you want our community to use an essay like this? Did you have any design intentions as you were writing it?

Paul: We want designers to look at these examples and ask themselves hard questions. Should visualizations of death milestones afford mourning? Should they be vectors of social change? Should they impress us to behave in ways that reduce the spread of COVID-19 death? I think the examples in COVIC demonstrate that the data visualization community has collectively said ‘yes’ to all of these questions. 

I watched a talk by Mushon Zer-Avi last night. Mushon gave this keynote, Friction & Flow — a Design Theory of Change, at the Better World x Design conference in Providence RI. He makes interesting use of ‘lessons from the pandemic’ in his talk and builds on the distinction between affordance and signifier.

As Mushon built the distinction between activism that focuses on signifiers (occupy wall street) vs activism that focuses on affordances (canceling debt), I thought we could apply something similar to the death milestones. The visualizations that are clearly assembled from individual stories, whether those stories are represented by a single sentence, or a phrase, identifying a ‘type’ of person, or presenting an audio portrait, are affording our personal confrontation with loss.

The visualizations that describe what ‘the country’ or ‘the world’ has lost as a quantity, compared to a line of buses or the population of a city, reaching this point on the Y-axis, spread across time, distributed in our collective geography, are generating signifiers. They signify the quantities they represent. They do not afford our engagement in the loss of life they are made from. Or maybe, as Liuhuaying says, they do when the signifier invokes an affordance that we understand, such as placing flowers on a grave.

We want people to recognize each other’s work and build on the patterns we observed. 

JF:  How can readers find the examples you discuss?

Printing the essay as a supplement to Nightingale Magazine, Issue 2, has many advantages. It affords everyone the opportunity to see these images side by side on the printed page. But a disadvantage is that we could not print the links to the original stories for each of the 41 examples. This online interview affords us that chance. 

Let me offer everyone this table with the story titles linked. I hope people will use this to explore the examples in context and experience the interactivity, animation, and audio found in many of the pieces.

For more visualizations, please visit the COVID-19 Online Visualization Collection (COVIC). 

Other articles by the authors: 

• The Pandemic That Launched a Thousand Visualizations
• COVIC: Collecting Visualizations of Covid-19 to Outline a Space of Possibilities
• Trails, an interactive visual landscape in collaboration with Dario Rodighiero that combined the image collection into a 2-D space

The post Behind the Scenes of “Unimaginable Death” appeared first on Nightingale.

Over the past two years, our team at NLC built this into a powerful data storytelling tool with almost 5,000 policies tracked across 800 cities. Through the process, we not only gathered invaluable insights into innovative city policies, but we also discovered the power of data storytelling to democratize shared lessons, improve local advocacy, and evaluate policy interventions.

Democratize shared lessons

The early stages of the COVID-19 pandemic saw a proliferation of data graphics highlighting trend lines that showcased the change in case numbers, death rates, and other relevant public health information. These types of data visualizations were incredibly valuable for the general public and for city leaders to assess the situation within their communities. Building off these baseline statistics, our Local Action Tracker combined data-driven trends on popular policy areas with narrative-oriented examples of how cities were addressing their public health, economic, and social challenges. Rather than reinventing the wheel for good policy initiatives, local elected officials were able to identify relevant insights from peer cities and apply them in their own contexts. For example, many cities, ranging from Nashville, Tennessee to Puyallup, Washington, partnered with local businesses to provide gift cards as an incentive for receiving vaccines last summer.

Data storytelling humanizes aggregated, high-level statistics and roots them in the reality of the lived experiences of the stories’ “characters.” Faced with sharply dropping city revenues and increasing expenditures, we saw mayors, council members and other local leaders navigate uncharted waters. City leaders in Cincinnati, Ohio, took a pay cut to their own salaries amidst revenue shortfalls and prioritized city budget spending on supporting those most impacted by the pandemic, such as local small businesses.  By showcasing their stories using data in real time, we were able to help them navigate these challenges together and democratize their shared lessons.

Improve local advocacy

In addition to humanizing high-level statistics, data storytelling also has the power to connect the dots of individual anecdotes and weave them into a compelling thematic narrative. For example, it was immediately clear from the beginning of the pandemic that cities would face added financial pressure due to an overall decrease in economic activity. However, it was not clear how much this would impact cities. There were stories of some cities whose budgets had been hit exceptionally hard while other cities appeared to have experienced only minor setbacks. So, what was driving these differences?

Our data storytelling efforts uncovered a key storyline demonstrating the critical role of municipal tax revenue structures, as cities heavily reliant on more dynamic sources such as income and sales tax saw more immediate financial impact compared to cities dependent on property tax. Overall, our research showed that cities in 2020 experienced $90 billion in revenue loss. This data-driven insight empowered cities to collectively advocate for federal assistance to ensure that essential city services were maintained. The financial challenges that they faced were not one-off examples, but rather part of a collective story that, when amplified, eventually led to the passage of the American Rescue Plan Act (ARPA) State and Local Fiscal Recovery Funds, which provided more than $65 billion to all municipalities across the country.

Evaluate policy interventions

Data storytelling is an effective tool to both find human stories amidst large datasets and identify data-driven themes woven throughout individual anecdotes. It can also provide us with a lens to reflect and evaluate where we have come from in order to inform best practices in the future. Our COVID-19 Local Action Tracker recently celebrated its two-year anniversary. In celebrating the occasion, we collected secondary data indicators that measured the top five key themes of policies that cities had pursued – public health, city operations, infrastructure, housing, and economic/workforce development. In the graphic below, we examined how local government employment numbers had fluctuated during the past two years and were able to demonstrate the effectiveness of ARPA in significantly increasing those job numbers. We can thus see that there is significant value both in leveraging data storytelling in real-time to shine light on current scenarios and in using it as a reflective tool to identify successful policies.

Our team’s experience building out the COVID-19 Local Action Tracker during the past two years has taught us valuable lessons about how to use data to tell city stories. We empathized with our stories’ heroes recognizing them as more than dots on a scatterplot and instead as public servants fighting for their local residents. We energized cities so that they could see themselves as more than singular entities and instead as part of a broader community collectively experiencing a similar storyline. And with the benefit of hindsight, we evaluated the success of both federal and local policies to identify and promote best practices for cities to adopt moving forward in building back better from this pandemic.

The post Telling the Story of Urban Innovation and Pandemic Response with Data appeared first on Nightingale.

That day, I spent ten hours running a piece of cloth through my typewriter. I selected a few lines from the Book of Common Prayer to overtype 549 times, in all caps: “FROM PESTILENCE FAMINE BATTLE MURDER AND FROM SUDDEN DEATH GOOD LORD FROM LIGHTNING AND TEMPEST, FROM PLAGUE DELIVER US.” I’d recently learned that the Book of Common Prayer was compiled in the 16th century during the Black Death. That context may explain why the book’s many pleas for deliverance from plague still radiate urgency. 

Eventually, I began to document daily case counts in my region of Iowa on a larger textile. I recorded cases every day, each with a stamped prayer, “FROM PLAGUE DELIVER US.” Prayers tend to be said on behalf of those closest to us, and a county was the smallest geographic area for which case counts were available. Working with data from Johnson County, IA, I recorded just the new cases each day, excluding recovery data or deaths. My intention for the final product was to situate both the act of prayer and the practice of art making as nuanced methods for approaching uncertain situations. 

What happens when we can both see and touch data? I worked on DEMAND/PRAYER for about 20 hours a week over 18 weeks. I anticipated the project would be a sad one; I didn’t expect it to make me shake with anger while I sewed. It’s one thing to model cases digitally, and another to perform a small physical act of labor that acknowledges each person in a dataset. My anger was fueled by numbers from a relatively small area, compared to the pandemic’s global reach. I wondered if (and how) public policy outcomes might change if legislators, and the public, had a more embodied relationship to case data. As the writer Jenn Shapland claims, “I feel that there is a price we pay for disembodiment.” What might happen if those shaping public policy could touch and see the lives being impacted by this pandemic? 

There are important elements of the pandemic in Johnson County that this textile does not capture, such as which demographic groups bore disproportionately high caseloads. Additionally, the piece represents all new cases in the county over the course of a year with a reference to a Christian text, the Book of Common Prayer. But, perhaps only a third of Johnson County’s residents are religious, and the local religious groups include Jews, Muslims, Buddhists, Hindus, and Christians. Given these demographics, my work is an example of how data art differs from data visualization. As an artist, my work recycles canonical Christian texts to assert that mystical and spiritual experiences happen within, and sometimes despite, established social structures.

Cloth is a practical medium for my data physicalization projects: even durable paper can become worn out from repetitive stamping or typing. A second reason I use cloth is “computing’s historical dependency on textile design, its production methods, and its labor models.” Binary code was invented during the industrial revolution to automate weaving patterned cloth. Core memory, the first viable and widely adopted re-writable computer memory, was a conductive fabric which coded ones and zeros in magnetic beads as positive or negative charges. Core memory was woven by hand, and its fabrication resembled a combination of bead looming and pin loom weaving. A touchscreen technology often found in consumer electronics like tablets and smartphones relies on conductive grids that operate similarly to core memory planes. The anthropologist Stephen Monteiro draws parallels between needlework, touchscreen gaming, and image-based social media interfaces, arguing that “personal touchscreen device use resembles the actions, strategies, and conditions of craft production.”

Some data physicalizations are self-conscious of these connections, demonstrating integration between medium and message. Visualizing blockchain technology with yarn, knitting a blanket of sleep pattern data, and encoding the pace of global warming into winter wear—in these projects, yarn is a considered material choice. However, some textile data physicalizations leave me wondering why textiles were chosen as a medium for that data. To engage a certain demographic as makers or viewers of the data object? Because knitting, crochet, and embroidery are financially accessible techniques for making data physical? There’s nothing wrong with either motivation. However, many makers working at the intersection of textiles and data are still unaware of the degree to which information society depends on textile technology. When we choose textiles as a medium, we have an opportunity to highlight that connection.

Artistic data physicalizations are laborious, and they’re not immediate enough for many of the ways we use data. Still, I am drawn to the ways textiles can make associative leaps and humanize data: as if by giving data a body, we glimpse its soul. 

Learn more about gendered labor, textiles, and the 1969 moon landing:

1. Skilled seamstresses sewed Neil Armstrong’s space suit by hand, using techniques adapted from lingerie construction.
2. A workforce of unnamed women, some of whom were retired or laid-off textile workers, hand wove the space shuttle’s navigation program.
3. A workforce comprised mostly of Diné (Navajo) women fabricated the microchips for Apollo 11.

The post The Soul of Data: Data Physicalizations on Fabric appeared first on Nightingale.

In response to my frustration, I turned to crafting and visualization. Others have written of the power of data visualization as grief and the therapeutic effects of knitting. I’ve personally experienced the benefits of knitting on mental health, as I knit non-stop during the first six months of the pandemic to combat pandemic-induced anxiety and alleviate the tedium of endless Zoom meetings. It seemed logical to me to combine crafting with visualization – merging my professional interest in data with my personal love of knitting and fiber craft – to help process the trauma of living through a pandemic.

There is an established history of mixing yarn and visualization. One of the best known examples from the knitting community is the temperature scarf. The knitter adds a row to the scarf every day in a color that corresponds to that day’s temperature; after 365 days, she has a scarf that shows the weather in that location for a full year. Other fiber-based examples of data visualization include a handknit climate-change sweater, an embroidered visualization of garbage washed up on a beach, the periodic table interpreted in cross stitch, and crocheted hyperbolic planes for use as mathematical models.

I decided to visualize the U.S. COVID daily fatalities in 2020 as a way to process my frustration and grief. While I predominately knit and sew, handweaving ended up being the right method for this project. I had recently purchased a 2-inch hexagonal pin loom that would yield a reasonable-sized visualization of a year’s worth of data, using one hexagon to represent each day. I used data under a CC BY license from The Atlantic’s COVID Tracking Project.

As with every visualization, visualizing with yarn required making design decisions that affected the accuracy and efficacy of the final visualization. The biggest planning challenge was choosing colors. Due to the macabre nature of visualizing COVID fatalities, I decided to use a simple color gradient from white to blood red. This forced me to plot data on a logarithmic scale, as I couldn’t reasonably find more than five yarn colorways on this color spectrum that coordinated. I ended up buying eleven different balls of yarn in white, pinks, and reds to determine which worked together best (though I’m still not 100% satisfied with one of the pinks). The final color key is:

• white = 0 U.S. COVID deaths
• light pink = 1-9 U.S. COVID deaths
• dark pink = 10-99 U.S. COVID deaths
• red = 100-999 U.S. COVID deaths
• dark red = over 1000 U.S. COVID deaths

The process of creating the visualization was both soothing and saddening. The repetitive nature of weaving and joining together 366 hexagons (2020 was a leap year) was comforting. This feeling was punctuated by grief, especially when weaving the 163 dark red hexagons that each represented over 1,000 U.S. COVID deaths per day; a 2-inch hexagon is a small thing to encapsulate such a profound loss. The weaving process also made me think of those lost to the pandemic whose deaths are not included in the official count, especially in the early stages of the pandemic before we had reliable testing. I explicitly recognize how incomplete the final visualization is, both through the presence of data artifacts (e.g., lower reported fatalities on Sundays and Mondays) and difficulty in representing the emotional toll of the national loss.

The final visualization is over a foot wide and almost 8 feet long and used almost 1000 yards of yarn (technical specifics are available on my Ravelry page). I spent almost three months of early 2021 planning, weaving, and joining the hexagons together, with most of the hand work done during endless hours on Zoom. While I could wear this visualization as a scarf or a shawl, I think it’s more impactful when viewed as a whole, so the final visualization became a wall hanging. It now hangs in my university office where I use it to engage others in conversations about data, visualization, and shared anxiety around the ongoing pandemic.

I’m still frustrated by our inadequate collective response to the pandemic to the point where I’m already planning the 2021 version of this visualization. I’m heartbroken that the 2021 U.S. COVID daily fatality visualization will look similar to the 2020 visualization despite the existence of vaccines and known prevention strategies like ventilation and masking. I had originally hoped to visualize U.S. daily vaccination data for 2021, but this year has simply not provided enough relief from COVID. So for me and my own mental health, I’m coming back to crafting and visualization as a way to work through the continued trauma of living in a pandemic.

The post Crafting a COVID Visualization: How I Processed Pandemic Anxiety and Grief with Yarn appeared first on Nightingale.

On July 27th, 2021, the CDC posted a Health Advisory which raised the possibility of COVID-19 hospitalizations overwhelming existing healthcare capacity and advised healthcare professionals of the “urgent need to increase Covid-19 vaccination coverage” as well as the fact that the majority of hospitalizations and deaths are happening among the unvaccinated.  This message was meant to convey that we could reach a situation in which patients needing lifesaving care would be turned away by hospitals that would not have the capacity to care for them.  This is a very significant threat to public health.

With that in mind, take a minute to review the original visualization, which was the lead graphic in a health advisory.  I’m going to spend the rest of this article exploring ways to communicate the same message with greater clarity and urgency.  In the end, I hope you’ll agree that a graphic presenting fewer variables per map with a  simplified color scale and a more explicit message is a clearer form of explanatory communication for the intended audience of non-datavis-professionals. 

What message is the map trying to get across?  

The first step of iterating on a design like this involves establishing a sense of the original designer’s intended message and inventorying the informational elements presented in support of that message. This message was posted by the CDC under its emergency response page and designed to “notify public health practitioners and clinicians about the urgent need to increase COVID-19 vaccination coverage“ as the nation approached 650,000 Covid-19 fatalities. I used the CDC’s Crisis & Emergency Risk Communication (CERC) manual to evaluate where this message falls on the spectrum of emergency risk communication and inform my redesign. 

For the rest of this article, I’ll accept the designer’s thresholds of 40% vaccination and 100 cases per 100,000 population as the thresholds between “low” and “high” in each variable and refer to the groupings that way. There are some challenging elements in deciphering the message of the original graphic which I’ll cover in the next section. For now, with the five color encodings that counties are broken into and the level of hazard described in the health advisory text, my take on the designer’s intended message is:

“Counties with low vaccination rates are experiencing a wave of high Covid-19 infection rates.  That wave is less pronounced in counties with high vaccination rates”

With that core message in mind, breaking down the rest of the information presented becomes an exercise in inventorying the questions that can be answered with the original graphic. The first answerable question that stands out is:

1. Where are the low vaccination rate counties with high case rates?  

Some might argue that this is the only thing the designer wanted to communicate. If that were the designer’s intent, the graphic could be completed with only the darkest purple:

Because the designer also chose to include a color for counties with low vaccination rates and low case counts, I think that information is meant to add context for the audience:

The addition of the low vaccination and low case rate counties lets the reader answer an additional question:

2. What proportion of low vaccination rate counties have high case rates?

Because the original graphic presents counties with high vaccination rates and low case rates in white, one could argue that the designer’s intent did not go further than the above questions. Since the original graphic did include a color for counties with high vaccination rates and high case rates, I think the designer intended for the audience to be able to answer the same questions about high vaccination rate counties. 

These additional questions lead to an important comparative question that can also be answered with the information presented in the original graphic:

3. How do high vaccination rate regions compare to low vaccination rate regions in terms of case rates?

If presented clearly, the information lets the audience draw insights about the relatively high proportion of low vaccination counties experiencing high case rates when compared with high vaccine rate counties. This drives home the health advisory’s message of the “urgent need to increase Covid-19 vaccination coverage.” Ultimately, given the threat of overwhelming hospitals with infected patients, the information presented should be a source of significant concern.

The audience should walk away from the final design thinking:

“There are a LOT of counties with low vaccination rates, and a LOT of those counties are experiencing high case rates!”

The converse is also important:

“Counties with high vaccination rates seem a lot safer.  Look at New England!”

Let’s audit what’s confusing:

After you take a minute to read through the legend and wrap your head around what it’s communicating, can you tell how many counties have high vaccination rates (40 percent and above) and high seven-day case rates (100 or more cases per 100k population)? It’s not easy to understand because the chart combines multiple variables converted into threshold-divided categories encoded by a single color scale. This is a LOT of information in a tight space. The lightest purple is not easily distinguished, particularly in smaller counties.  As an experiment, look at the same visualization after Photoshop was used to swap out the light purple (counties >=40 percent vaccinated and 100 or more cases per 100k) with light green.

With the re-encoding, you can see that high vaccination / high case rate counties are quite rare, they’re often close to low vaccination / high case rate counties, and there are almost none of them in the Northeast. The audience has a cleaner feel for the geography of low vaccination rate counties and a better sense of the large cluster of high case rates in Missouri, Arkansas, Louisiana, Mississippi, Alabama, and Georgia with this re-encoding. It’s interesting to note that Florida is experiencing high case rates in counties both above and below 40 percent vaccination rate. A quick Photoshop color replacement has exposed interesting insights.

Can the legend be easier to digest?  What if we break out the legend along the 40 percent vaccination rate division?

This is slightly better, but it’s still challenging to decipher the two categories of counties. More on that later.

Why is Texas (which isn’t reporting vaccination rate data) so visually prominent? It may be useful to bring this to peoples’ attention, but that doesn’t feel like the central message of this visualization. The unreported data could be added to a footnote or encoded in a lighter and less attention-grabbing color.

Can you tell the date of the data on the map? It’s hard to find, but deep in the fine print underneath, the date is Fri Jul 23, 2021, which was four days prior to the publication of the Health Advisory and three days older than the most recent available data on that day.  When information is changing rapidly, it’s possible to have the methodology for making a graphic pre-approved so that it can be updated immediately prior to publication.

Now it’s time to get the data and start to draft a new visualization.

The day that the CDC published the original graphic, I was able to download the most recent update of their source data which covered up to 26 July. Current data can be downloaded from the CDC here. With the data in hand, a redesign is possible.  This is an area where having many fresh sets of eyes and friends who are willing to give blunt feedback comes in handy. I got a lot of blank stares when I asked people to interpret the original graphic, so I started by prototyping ways to simplify the presentation. I typically sketch out concepts by hand before getting too into the weeds with code. The original graphic lists the thresholds “40” and “100” four different times and presents each category in a way that invites comparison. With that in mind, I wanted to reduce the repetitive text while highlighting comparisons between groups.

To draw attention to the proportion of each vaccination rate group, I could use either bar charts or pie charts. Bar charts offer the ability to compare raw counts across the two vaccination rate groups. Pie charts offer a more intuitive communication of proportions for a lay audience as long as the proportions are relatively large. Separate choropleth maps for each vaccination rate group can partially make up for the loss of total-count comparisons in abandoning bar charts.  Splitting the data into separate visualizations for each group also enables common color encoding and the potential to use column-oriented plots to convey grouping with fewer titles and explanations.

Below is a rough draft of how this visualization might work. Individual elements were made in a Jupyter notebook using the Pandas and Geopandas Python libraries. With this draft, I adhered to the CDC’s original graphic in not highlighting state boundaries, which both keeps the graphic as apolitical as possible and keeps the focus on county vaccination rate as a differentiating factor.

Breaking out the data into two separate maps does highlight the Southeast, but that was where the Health Advisory’s most pressing concern was centered at the time. User testing on friends with fresh eyes indicated that they were more comfortable with fewer variables on each map, taking much less time to decode what they were being shown. Each quickly expressed concern about counties in the southeast with low vaccination rates and high case rates. The column orientation also seemed intuitive to them.

Finally, the title can be tightened up.  

The original title, “Counties by Percentage of Population Fully Vaccinated and 7-Day Case Rate” is perhaps a good science-y title for communication that is not part of an urgent public health advisory. There is, however, room to be more direct here.  The CDC’s Crisis & Emergency Risk Communication (CERC) manual is an excellent reference for crafting this kind of messaging. Even when directed at an audience of clinical professionals, figure titles in a Health Advisory expressing the “urgent need to increase Covid-19 vaccination coverage” should convey urgency. According to the CERC manual, an advisory like this is designed to “explain, persuade, and empower decision-making.”  A headline highlighting the bottom line message can be useful in framing the information presented in a visualization. It can also be useful to physicians working to persuade patients to get vaccinated. With that in mind, something like “Counties with Low Vaccination Rates are Experiencing a Fourth Wave of Covid-19 Infections” might be appropriate. If “fourth wave” is too alarmist, another option might be “Surging Case Rates in Counties Under 40 Percent Vaccinated Pose Significant Threat to Public Health.”  Additionally, some sense of the relative risk between high vaccination rate and low vaccination rate counties might be an appropriate subtitle. In this case, a relative risk calculation reveals that counties with “under 40% fully vaccinated” are 2.3 times as likely to have case rates of 100 per hundred thousand population or higher.  This is a memorable sound bite that could help spread the core message of the visualization.

What’s potentially wrong with this redesign?

Every visualization has limitations; this one is no exception. In selecting pie charts over bar charts, I removed the reader’s ability to judge relative volume of counties in each vaccination group. I’m also relying on judgment of radial area to convey the proportions of each group with high case rates, which can be problematic with smaller proportions but is ok with the proportions in this data set here. I made up for some of these drawbacks by including the choropleths, but it is important to recognize that county sizes and population densities vary significantly. Garfield County in Montana has about 0.25 residents per square mile while New York City has over 69,0000. The choice to map population-based rates still makes some sacrifices. This approach also fails to communicate the proximity of high-vaccination-rate/high-case-rate counties to low-vaccination-rate/high-case-rate counties. This makes it harder to see that most of Florida has surging case rates and there may be some interplay between low-vaccination-rate counties and high-vaccination-rate counties there. In the end, it’s a judgement call. Alaska, Hawaii, and US territories should be incorporated into the final product. They were omitted in rapid prototyping.

What else can you see that might need improvement?

In closing, let’s review the changes.

First, I established a sense of what message readers should take away from the visualization. Then I inventoried the information being presented. Next, I audited the original visualization for points of confusion that might throw the audience off track. I repackaged the key informational elements in different ways, iterating with paper and pencil and gathering feedback along the way. Ultimately, I broke the original multivariable, single-color-scale choropleth into two separate choropleths, one for each vaccination group. I added pie charts to drive home information about the proportions of high infection rates among each group. Then I updated the data and added a prominent date label to the final draft. I also took the focus off of Texas, where county vaccination data isn’t reported, and moved that information to a footnote. Finally, I gave the reader a bottom-line-up-front title that’s explicit about the hazard and more aligned with the “explain, persuade, and empower” purpose outlined in the CDC’s CERC manual for urgent communications like this Health Advisory.

This kind of clear, intuitive visual presentation of data is important when managing a public health crisis. Even when the intended audience is sophisticated, I think driving home key information with simplicity and directness helps spread the message further and faster. When carefully designed, this graphic can convince health professionals of the urgent hazard the health advisory is trying to communicate and empower them in making decisions to improve vaccination rates in their communities. At the end of the day, we’re talking about potentially heading off a situation in which someone who has a heart attack or is in a car accident does not survive because there are no available resources to care for them. That’s a great mission. Taking the extra steps required to iterate and improve visual communication like this can be both rewarding and fun and should become a standard element of public health educational programs around the country. The same concepts apply to data visualization in a wide variety of other applications.

The post How to Iterate on a CDC Health Advisory Graphic appeared first on Nightingale.

Collecting 160 days worth of personal data

I started collecting a variety of data from the day Boris Johnson announced this plan, as I knew this would be an interesting personal analysis of the next phase of COVID-19. I didn’t know exactly what I was going to visualise when I began so I gathered a variety of data points as a starting point. I chose the following datasets as ones I thought the most likely candidates to be impacted when lockdown restrictions eased:

• The date
• Government milestones
• Distance and quantity of steps I walked
• People (outside of my household) with whom I interacted
• Time spent on my phone
• Phone notifications per day
• Location of my destination

I knew I wanted to analyse the number of people I was interacting with and the steps I took each day as this was the most prominent change in the restrictions. The change was quite drastic: from not being able to leave your home or see anyone back in February, to retail, clubs and pubs open as normal without requiring masks. I also thought that it would be interesting to see if the time spent on my phone would change – I expected to spend less time on my phone as the economy opened up. 

I started to notice some trends within the raw data within my Google Sheets. Google Sheets is a useful tool for remote data collection, which I found extremely helpful as I was gathering six months’ worth of data and didn’t have my laptop with me all of the time. The app allowed for quick updates wherever I was. It was fascinating to try to spot patterns and connections by looking at the raw data only. 

Starting with sketches

One of my favourite parts of the process is scribbling in a notebook. I often sketch as it gives me the freedom to imagine what is possible without the restrictions and limitations of my computer. It’s instructive to look back after your dataviz is complete to study your  journey and thought process. 

When the last phase of the roadmap was announced, and ‘Freedom Day’ was approaching, I started to think more about the datasets I collected and the story I wanted to tell. 

Sketching goes hand in hand with cleaning and analysing the data. When you draw out and sketch, you realise which data you need. As I started to draw, I subconsciously and automatically started to populate my graphs with predictions for what I thought would happen at the end of the process: 

• My step count would go down as  my daily walks would get progressively less due to pubs, coffee shops, and restaurants opening  up
• I would gradually see more people and rarely spend time in the house
• My screen time would lower as I’d be outside more
• My notifications would also decrease as I’d be doing things in person rather than online 

Cleaning 160 days of raw data

After sketching out what I thought the dataviz would look like, I sorted the data into ‘must haves’ (data points that are necessary to include) and ‘nice to haves’ (data points that don’t have to be included, but can be if it works).

The final dataset changed only a little from what I was originally gathering. I included every data point other than the location of where I was going each day, and the distance I had walked, as I didn’t feel it was necessary for the narrative.

I realised that I needed to add in some more context around COVID-19 beyond my personal data, so I also added two more data points that I downloaded from the open source government website (refer to the images below). 

• Number of COVID cases
• Number of people vaccinated 

Learning how to use Flourish

I created my previous CoViz-19 data visualisation completely by hand. I didn’t use any tools other than Adobe Illustrator. It was an extremely time-consuming, laborious process. This time I wanted (and needed) to learn how to use a data visualisation tool like Flourish or Tableau as I had six months worth of data,–double the amount on the last data visualisation.

After unsuccessfully browsing dataviz tool options (there are a lot out there!), I reached out to the Data Visualisation Society (DVS) community on Slack where I received the most helpful comments. (The DVS is a fantastic non-profit that supports the growth, refinement, and the expansion of data visualization knowledge regardless of expertise level.) After some fantastic tips and tricks from fellow data lovers, I landed on Flourish as I don’t have much coding experience. This seemed the most appropriate choice. 

Coincidentally, I saw a perfectly-timed advertisement from Federica Fragapane and I soon completed her brilliant online course: “Data Visualization and Information Design: Create a Visual Model” on Domestika where she teaches how to use Flourish combined with Adobe Illustrator. Watching Federica’s workflow and process was so inspiring and supplied the knowledge I needed to further my skills. 

Designing the final data visualisation 

Using Flourish cut my work down by such a considerable amount, I can’t imagine doing any future work without it. It was such a quick process once my data was clean, and refining and designing the dataviz in  illustrator made for such an enjoyable experience. 

Once I’m happy with my designs, I will typically share them with close colleagues and friends to check if they make sense, are readable from a design and narrative perspective, and that the back end of the data is correct. This practice served me well on my CoViz project.

I was struggling to map four different variables onto one line graph. A fellow designer suggested expanding my current graph into two line graphs: one for personal data and one for government data. This adjustment not only reduced the density of the graphic, but it also allowed the reader to distinguish between the sources. Additionally, a helpful data analyst advised that I try  formulating a seven day rolling average and scale my government data so both the COVID vaccinated number (in the thousands) and the COVID cases (in the millions) could be plotted using the same metric. This scaling  technique is a step of data processing that is applied to independent variables or features of data. It helps to normalise the data within a particular range. I’m always grateful to chat with people of different perspectives, even if it is painful to revise what I thought was a finished piece of work–sharing with others affords a fresh outlook. 

The final CoViz-19 data visualisations

Designing this infographic was both rewarding and satisfying creatively, and it enabled me the opportunity to look back on a historic moment to consider our progress as a nation. In February the future was so unknown. It was scary, intimidating, and overwhelming to face such a long road ahead, but on the other side of lockdown, I treasure the journey this visualization reveals. Though it contains a perspective personal to my social and technological habits, I suspect it’s a similar cadence to others. I hope that when others visualise my lockdown homestretch they can find hope and patience in their own live experiences.

The post CoViz-19 Part 2: A Personal Visualisation of the UK’s ‘Roadmap to Freedom’ appeared first on Nightingale.

In May 2020, Joanna Hutchinson felt powerless reading The New York Times piece An Incalculable Loss, the now-famous, full front-page death toll of US lives that had been lost to COVID-19 to that point. Like many of us, Joanna struggled with the sheer scale of 100,000 lives. She was burdened by overwhelming grief. Joanna is a certified public accountant, a data person. But she never thought of herself as a data visualizer — she doesn’t have a datafam. Instead, she found an outlet for her grief in an unusual collective, among the paper-folding community.

What started as a way of grappling with her own feelings became 100,000 Folds, a collaborative sculpture project to commemorate those lost to the coronavirus pandemic. The topics of scale and community keep coming up in my social feeds lately, so when I read about Joanna’s project I set up a conversation with her to learn more.

Mary Aviles: Tell me about the project and your goals for it.

Joanna Hutchinson: The basic idea is that, along with my participants, we’re folding one piece of origami for each of the first 100,000 COVID deaths in the United States. The origami will be assembled into larger sculptures so that we can visualize this number. And, it’s also kind of unifying to bring each individual piece into a larger sculpture. After all, we’re in this together.

How it works is that I mail pre-cut paper to the participants and they fold it in an act of memorializing those we lost. Then, they mail the paper back to me to be assembled into these larger sculptures. Originally I thought of it as a project to remember the people that died. But, in talking to the people who have joined the project, it’s become important for me to think about those who lived. So many of the people who are suffering are still with us, both those struggling with the virus and those who have lost loved ones.

MA: Where did the idea for 100,000 Folds come from?

JH: I am a finance person by trade. I also have a background in art. In my spare time, I enjoy doing different projects, but I’ve never taken on something of this scale before. This started in late May as the COVID deaths were reaching the 100,000 marker. My job switched to remote so I was staying home and I was by myself a lot. I was grappling with all the uncertainty and the virus was so expansive and horrible. I felt so alone in my grief and then I read The New York Times article. I read each of those names and the descriptions. It was so poignant. 100,000 lives. That really hit me like a ton of bricks.

It was a huge number. I’m desensitized to how big numbers are, in a way, and this has really grounded me. I started thinking about one hundred thousand people. I will never know that many people in my lifetime. It’s just too big. I was thinking about the gravity of the situation. It was really hard, emotionally. Then I began to think about what I could do to remember those people. How could I memorialize them? It occurred to me that I could make something that had one component for every one of those deaths in order to visualize the enormity of the situation. So, that’s how it started. I was reading the news and I was struck by how hard it was to understand a number like that. And, since then, it’s only gotten much worse.

MA: Why did you select origami as your medium?

JH: I settled on origami because it’s simple. The forms that I’m using make a triangle unit. They’re relatively easy to make. I have used units like this in my artwork in the past and I like the simplicity of the folds. It’s repetitive. It’s comforting to make the same folds again and again. It’s rhythmic and sort of peaceful to fold paper. I latched on to origami because it kind of made sense in the same vein as Sadako’s One Thousand Paper Cranes. Origami has a wonderful tradition. And, you can create a lot of them.

I was originally going to fold them all myself, but then I did some calculations and figured that if I was consistent and I could fold every single day, it would take me at least four years to finish. Then, I thought, “Please let us not still be in this pandemic in four years. God forbid.” I just couldn’t take this on by myself. It was too big.

I realized that was a challenge that characterized the pandemic, too. It’s a global problem. It’s this huge thing that we’re all fighting together. I decided to make it collective. There are over 300 participants now. So far, I’ve sent out 140,000 pieces of paper.

MA: Who do you consider your community?

JH: First, I want to speak a bit about the reason that I focus on the United States. It’s not because I only care about my country. I think that, as a wealthy and powerful country, we’ve really dropped the ball on our COVID response. We’ve neglected people and they’ve died. We should have done better. It makes me angry to think about it.

It’s like a Venn diagram. There are some people that want to participate because they lost somebody or they feel deeply about the need to memorialize COVID deaths. People have written to me to say, “I’m doing this project to remember my mother” or “I’m doing this project because it’s a time for grief and I need an outlet for my feelings.” There are a lot of local participants here in Philadelphia. They seem to want to participate because it’s a local project and they want to chip in. But, there are some international participants as well. There are some who are artisans from within the paper-making community. Finally, there are also a bunch of people who make things and are more generally creative who are interested in the art.

MA: How have people heard about the project? What do you think drives their desire for affiliation?

JH: Helen Hiebert has a popular blog, all related to paper art, and she featured my project early on. I also work with a couple of local artist communities. The Soap Box is one of them. They are a print-making space and ‘zine library here in Philadelphia. They’ve been co-sponsoring my online workshops and promoting them in conjunction with the Rotunda, which is another space here that not only houses a lot of art-related projects, but also focuses on social justice and other things.

I hosted four workshops in 2020. I’m gearing up for more starting this month. Participants receive a small bundle of paper, I do a demonstration on how to fold, and I talk about the project. I reserve time for community building where we can chit-chat and talk about why someone might be interested in folding paper for a COVID memorial and about origami. The workshops absolutely help bring people to the project. At one point I considered stopping them to focus on getting the sculptures together, but there’s been so much interest. I think people need a place for collective mourning and COVID remembrance.

It’s been word of mouth, mostly. It makes perfect sense that there would be this sort of Venn diagram of different communities. Some people receive the paper, they do the folding, and they send it back. Some people send me updates on how it’s going and what’s happening and what they’re thinking and working on. I have people that have asked for more paper, so I have some repeat participants. A lot of people post on their social media about the project.

MA: What is your vision for the completed sculptures?

JH: I’m thinking that the sculptures will be a large urn shape. The shape is inspired by a design I made a couple of years ago. This sculpture has the same kind of units and I love the way they interlock. I’m playing with them interlocking backwards and how that can change the shape. I’m going to have so much material that I think it’ll be enough to make something person-sized. If my calculations are right, I can make two sculptures that are five- or six-foot-tall vessels.

There’s a place called Cherry Street Pier here in Philadelphia. It’s an open air place for people to gather. I love the idea of having my pieces there. They have studio spaces for artists and I could build the pieces onsite. Then, it would be open to the public. So anybody could go there and there’s no entrance fee or anything like that. I thought someplace like that would be really wonderful. But, because this is a paper sculpture, I’m concerned about having it out. I’m trying to build it in a way that can either be taken apart or moved.

I’m working on designing the infrastructure. There will be steel armature inside the sculptures. I’m working on designing that to be modular in some way so that I can have it in one place for a while and then it can move to another place. I’ve never made something this big so I’m really excited to start putting it together and figuring that out.

MA: How has this project impacted your view of COVID? How has this project changed you?

JH: I was looking for something personal in the beginning. And now I’m looking to give others the same outlet. Originally, I wanted something to work on by myself to work through my own grief. Now there’s this whole community that’s right there with me. It’s comforting to know that there are others, so many others. It also makes me feel a a great sense of honor.

You can find out how to participate in or support 100,000 Folds by visiting Joanna’s site.

Joanna is certainly not alone in her effort to memorialize COVID deaths. Her interview has me reflecting on the meaning of ‘community’ and what it means to be part of one. As suggested in her post below, strategist and early stage investor, Sari Azout, analogizes community with home.

100,000 Folds feels like a different interpretation. It feels like a group of people with disparate interests and motivations working toward a shared purpose. Outside of their collective participation with Joanna’s organization, they may never interact, but they still seem like a community. She has come to realize that lifting up the living, those seeking solace, has added a new dimension to the project. She extended her workshop series in order to help participants process their grief and find comfort among others.

Joanna Hutchinson is a version of a network weaver. Weavers, connectors, and navigators like her share the critical responsibility of establishing networks. They can build bonds among trusted messengers to form human systems. They cultivate weak ties — the folks we meet throughout the course of our day, through friends-of-friends, or through our extracurricular interests.

COVID has done a number on our weak ties. We’re not going to the dry cleaner or waiting in line for coffee or chatting up other spectators at our kids’ soccer games. Last fall, I facilitated roundtables with attendees of the Urban Land Institute’s conference. We talked about the value of weak ties in our personal networks based on Mark Granovetter’s nearly 50-year-old research. In addition to their importance for our career trajectories, weak ties are critical to our sense of wellbeing. The Atlantic recently covered this topic. The article’s author, Amanda Mull, writes:

The Atlantic article goes on to explain that the lack of communal connections can signal larger civic problems. It can indicate deteriorated social capital with dire consequences. In his book, Bowling Alone, Bob Putnam found that:

Think about what happens when you move to a new city. You have to figure out your new system. You have to choose healthcare providers you like, find a grocery store with your favorites, locate people you want to go for walks with, and draft your preferred take-out restaurant shortlist. These things don’t exist on a traditional map. Establishing even weak ties requires some personal investment. Isolation can evolve from not knowing about the communities that exist around you. Word of mouth and mapping can reveal these networks. Asset or ecosystem mapping is instrumental in revealing supportive networks (or critical gaps).

In 2019, the New Economy Initiative, an advocate for inclusive entrepreneurship in Detroit, mapped the entrepreneurial support network in southeast Michigan. Doing so helped them visualize referral behavior among their grantees. In other words, they were able to see how many meetings new business owners were having, with whom, and for what kind of support. These insights helped them connect with more would-be entrepreneurs and guide those folks through the start-up process and on to generating revenue faster.

Last year, I conducted generative and user experience testing for the United Way of Southeast Michigan during the development of their Connect4Care Kids resource. Parents and caregivers were especially enthusiastic about the Location Finder feature. For many, it showed them childcare options, near home, work, and school, that they hadn’t known about previously. Finding safe and accessible child care is, of course, vital to employment and adult education.

Asset mapping documents what a community has, rather than studying its needs. “Assets” can refer to both people AND place. These asset maps bridge the physical to the digital. Building and maintaining asset maps is a critical investment in human systems infrastructure. Data collection and the frameworks for network or asset mapping can be extraordinarily challenging and often require an understanding of local context. This seems like an untapped opportunity for spatial visualization. While 100,000 Folds doesn’t need an asset map to achieve its purpose, consider the work of community organizations and local governments on the 2020 Census. Asset maps, like this example from Google for addiction support, are invaluable to activating civic efforts like census counts or voter registration.

I’m grateful to my digital communities for helping me check in with all my ties, strong and weak, and for providing my kids with some welcome pandemic distraction. One of the many reasons I value my DVS membership is that I think of it as a multiverse of communities, offering a range of experiences and opportunities to develop meaningful bonds with my peers. Now, more than ever, we must rely on digital communities to establish social tethers. The people and the organizations managing our digital communities can learn responsible stewardship by studying place-based best practices. Data practitioners can broaden their view to understand their potential for contribution to the “ecosystem orchestra.” Through her work with 100,000 Folds, Joanna Hutchinson has demonstrated that you don’t have to go it alone, even if you can’t go anywhere.

Which of your communities has been a lifesaver during COVID?

Additional resources:

• danah boyd’s entire talk on the Future of Information (2018) is fascinating even beyond the citation above, for its coverage of “re-networking” and especially considering our experiences with social instability around the 2020 US presidential election.
• Find more information on/examples of asset mapping: City of Boise, HealthCity’s Participatory Asset Mapping toolkit, and Fair Count’s map of internet installations.

The post 100,000 Folds: The Stabilizing Power of Community appeared first on Nightingale.

This prison-like tally chart started as a light-hearted joke, but it became the first of several personal datasets I collected. I knew that this lockdown was a once-in-a-lifetime experience, so I wanted to document as much as possible.

Over 91 days, I recorded the follow data points:

• The month and year
• Whether I showered, got dressed, or meditated
• Type of exercise and distance
• Online and face-to-face human interaction
• Mental and physical health

Documenting my daily routines became particularly interesting as lockdown provided a different day-to-day experience to my normal routine. Many of the short-term variables were removed from my daily routine, making for a cleaner dataset. On one hand, there was no anxiety-inducing commute from Brighton to London or everyday work stress. On the other, there were no spontaneous catch-ups with friends, holidays to book, or dinner dates.

While this unnatural environment could have a huge impact on anyone’s mental health, it also seemed like an opportunity to understand how I might take charge of my own wellbeing by understanding what factors within my control make me happy.

CoViz-19 posters

As a designer, I often use infographics and data visualisations to reveal patterns, connections, and trends that you would not necessarily see otherwise. Because of this passion, it was only natural for me to turn my dataset into a series of dataviz posters which I have named ‘CoViz-19’, playing on the words ‘dataviz’ and ‘Covid-19’. Visit my Behance page to view them in more detail.

So, what did I learn?

1. The first couple of weeks of lockdown were emotional chaos. It took me 12 days to settle into a stable rhythm.
2. I’m gross for showering once in the middle of five days in June (must have been a bad week). Maybe you noticed this, maybe I shouldn’t have point it out.
3. Running was a coping mechanism rather than a product of happiness. When I felt awful, I knew I had to go outside to feel better, rather than running because I was full of energy. All of my runs occurred on days where I had the lowest amount of human interaction.
4. I spent the most time meditating when I was anxious and needed peace, not because I felt content.
5. I started volunteering at a food bank on April 27th, where over 20 people were together at one time. Being around this many people was overwhelming compared to what I had been experiencing and as a result, the following week my mental health fluctuated dramatically.
6. Zoom quizzes with friends didn’t help provide connection, they only emphasised loneliness.
7. My mental well-being declined during public holidays and birthdays that I couldn’t attend, even though I was connected via Zoom. Digital interactions do not and cannot replace physical interactions.

8. My highest peaks were when I started doing something new, like creating isometric illustrations, making fresh pasta with friends over Zoom (my Italian Nonna was proud), and crafting macrame wall hangings. It’s interesting to note that pasta-making Zooms boosted my mood, but quizzes/virtual birthday parties didn’t. The act of using my hands to create something, compared to sitting still watching a screen, affected my mood considerably.
9. On May 28th, the government introduced the rule of seeing up to six people outdoors. After this date, my mental and physical health line was the most consistent it had been during the 91 days.
10. I moved into a houseshare 15 days before I stopped collecting data. Knowing that living with friends again would be great for my mental health, I wasn’t expecting to become overwhelmed by the second week. Being surrounded by people 24/7 was something I had to adjust to after so long in isolation.

Since this project ended I have adopted three things that I know will help me have a better grasp of my mental health and take charge of my own wellbeing. These include:

1. Personal creative projects
2. Regular meditation
3. Little and often time in solitude

The power of visualising this data has provided not only a nice set of posters capturing a significant point in history, but has also revealed patterns and trends about myself that I wouldn’t have noticed otherwise.

The post CoViz-19: 10 Things I Learnt from 75 Days Alone in Lockdown appeared first on Nightingale.

The project is part of a growing body of urgent visualization work in response to the COVID-19 pandemic. Zhou has worked as a data journalist for about six years, and can’t remember a time when data and visual journalism played such an important and consequential role. “For every piece, the stakes are higher,” she said.

For Zhou, the NYT visual story exemplifies the stakes and complexity of this moment. The piece visualizes air traffic data and traces a number of known COVID-19 cases to show that travelers carrying the virus from China got to the U.S. well before President Donald Trump implemented the travel ban.

Yet the piece came out around when Trump started calling the coronavirus the “Chinese virus.” The European travel ban had been erected a week before. Positive cases and deaths were increasing daily. Conspiracists posited that the virus was man-made in a lab in China. Anti-China and anti-Asian sentiments were brewing.

This context led to very different receptions to the piece in different circles of data journalism. Many of Zhou’s Chinese journalist friends considered some of the design choices to be insensitive and felt that the reporting and framing essentially backed up Trump’s claim. But the strong visuals left a different impression on Zhou’s friends in the U.S. In addition to praising the quality of the presentation, they got the message: The travel ban came too late to prevent the spread of the virus. The administration should take the blame.

Zhou has been tracking and visualizing data related to migrant communities, global mobility, and immigration policies around the world for years now, most recently at Quartz. She has produced award-winning visual stories on domestic and international news. In another example of crisis data reporting, Zhou built a database of deaths from Hurricane Maria with reporters from The Associated Press and Centro de Periodismo Investigativo and visualized how government inaction led to continued deaths long after the disaster.

I was fortunate to get to speak with her about her extensive experience, unique background, and her perspective on data and visual journalism during the pandemic as a journalist from China based in the U.S.

What were the main critiques you heard about the “How the Virus Got Out” visual story?

My data journalist friends in China had been reporting on the pandemic for two months when the piece came out in late March. It wasn’t news to them that the virus spread from China abroad. So when they saw the story, they noticed the design choices and questioned the motives behind it rather than focusing on the information.

One criticism is the use of color. The piece used saturated red to represent people carrying the virus, a typical choice under normal times, but can be seen as dehumanizing now. Also, the number of travelers is much bigger than those carrying the virus. Using particles of similar sizes to represent travelers and positive cases exaggerated the latter. They warned that these design choices would lead to biased conclusions.

The visual provided support for terms like “Chinese virus” and “Wuhan virus” and they worried it would lead to discrimination toward a geographical location and groups of people. They wished The Times had exercised more caution producing the story.

Why do you think two groups of people reading the same visual journalism piece could leave with completely different takeaways?

Despite the same data and visual presentation, the piece triggered people differently. The same data can be used to back the argument that China should take the blame as well as that the US did a poor job in preventing it. Data alone is perhaps neutral and objective, but when we decide what to show and how to show it, data can quickly be weaponized to serve a particular narrative. We have seen this happen again and again during the pandemic.

It made me realize that we need to be more careful in our choices of selecting, analyzing, and presenting data. Even though the majority of the target readers are fine with a story, it should not be framed in a way that harms vulnerable populations. I don’t think it’s the intention of the creators to incite racism or hatred. But impactful stories may lead to unintentional consequences.

It’s a delicate balance between presenting information and nudging readers toward one conclusion or another.

I think it’s a journalist’s responsibility to distill massive data into digestible narratives. Leaving information as is wouldn’t do it. As for this piece, I think The Times did an amazing job packing so much information into such a compelling scrolling experience. Unlike other popular means of presenting data on travelers, the smooth animation of particles really gives the audience an intuitive understanding of the initial spread of the virus. It’s an undeniable piece of art with stunning technical details. I feel like the creators focused more on that, didn’t try to nudge readers one way or another, and pretty much left the information as is.

You mentioned this is a recurring problem during the pandemic. Why so?

We are faced with an invisible virus. Everyone wants to have concrete information that they can take a hold of. We want to know why it happened, what we should do. Data — as many problems as they have — calm people down. Data and visual stories become traffic drivers! Stories that back or challenge people’s existing beliefs, especially on controversial topics, get tons of clicks.

One popular type of charts I’m sure everyone has seen is a good example: for a while, charts that tracked the trajectory of COVID-19 cases by country popped all over social media. They were usually in colorful spaghetti lines with a logged y-axis. It left people with the impression that countries were competing with one another to “flatten the curve.” One could easily call out winners and losers. But the reality is much more nuanced: jurisdictions used different methodologies to report cases and had different testing capacities. Comparing case or death numbers across countries in most circumstances was inaccurate and irresponsible. One might’ve wished that these charts would lead to “losing” countries learning from “winning” countries their “winning” strategies. Instead, these visuals inspired policies like travel restrictions. But people loved these charts! Seeing ourselves in a competition is exciting. Data journalism feeds people’s emotional needs. Perhaps to certain extent, we data journalists are complicit in some of the racist and xenophobic sentiments that have arisen from this information.Ten Considerations Before you Create another Chart about COVID-19To sum it up — #vizresponsibly; which may mean not publishing your visualizations in the public domain at all.medium.com

There must be some examples of good uses of data and visuals during this crisis?

Absolutely. As we wanted to understand how it happened, what we could do now, a lot of it has to be explained with scientific knowledge that the public does not have. Visual journalism helps digest complex information into compelling visuals. It feels old now, but — the iconic image of “flattening the curve” effectively conveys the message that social distancing isn’t about reducing cases, but about saving lives by keeping case numbers below the capacity of the healthcare system. The image democratized the knowledge epidemiologists had had for years, and helped build support for social distancing.

How have past projects you’ve worked on prepared you for visualizing the COVID-19 crisis? What have been the greatest challenges for you, or things you felt less prepared for?

I think this is truly a special moment for data journalists because data all of a sudden became so interesting to people who hadn’t felt that way before. From admin data of regular releases to personal data owned by private companies, you pick a dataset and you get an abnormal chart, a news story.

I’d been tracking data related to immigration and global migration, which became a huge area of focus in this pandemic because of the travel restrictions and changes in immigration policies. The leads and datasets I have followed became a source of new stories. When the U.S. unemployment figures came out, I wrote a story on the outsized impact on the immigrant community by analyzing occupations of immigrants using data from the Census American Community Survey. There’s a cool dataset on how many countries one has visa-free access to based on their passports. I visualized the impact of the pandemic on the power of passports (more powerful passports got hurt more). In another story, migrants losing jobs in high-income countries send back home less money, resulting in a drop in remittance income (and GDP from external transfers) in low-income countries.

The rapid news development is both a blessing and a huge challenge for data journalists, though, because it takes time to collect, analyze and produce a visually compelling story. Most of the data in the field of immigration and migration aren’t timely. They are collected by national governments and released at a monthly, quarterly, or even yearly schedule with at least a few months of lag time. Data from private institutions come with their own biases and privacy concerns. It happened a few times that I had an anecdotal story, but there wasn’t enough data yet to back up my observation.

Will the pandemic change the way data journalists work going forward?

One key thing the pandemic taught all of us is that we need to have empathy working with data. This crisis has reminded us that now when we work with numbers, we need to be aware of the humans behind, be it positive cases, death counts, immigrants, healthcare workers, or unemployment figures. They are very likely someone close to us. These humans will be impacted by the decisions we make related to the data. There might even be immediate policy responses. When the Times set a thousand names in small types on the front page to represent the 100,000 COVID-19 related deaths in the U.S., it felt personal and emotional. Rich and dramatic data points have emerged from the pandemic and can help shape public perceptions in important ways. What sets apart the great work from the good ones is the ability to empathize and humanize.

The post With Great Visualization Comes Great Responsibility appeared first on Nightingale.

As a result, this past week has been an education for me. But more importantly, as a student of health dataviz it’s been interesting to see how quickly data visualization — sprinkled with some context — brought me up to speed on this important concept. In the process, I’ve been thinking a lot about what makes the visualization on “flattening the curve” so successful — and what else we as data visualization practitioners can do to educate people on why it’s valuable to spread out and “squish down” that curve describing COVID-19 cases, so that the health care system can effectively treat all those impacted. More than that, though, this visualization also taught me about what makes a health data visualization successful, regardless of topic.

I first encountered “flattening the curve” as I was walking my dog around the neighborhood, catching up on tweets. Vox had just published a piece with the provocative headline, “How canceled events and self-quarantines save lives, in one chart.”

Those “in one chart” headlines always grab me, so I clicked. But this one left me wanting.

You see, I was on my phone — as surely many of us are when we encounter information these days — and it wasn’t worth squinting to try and read those annotations that explained the concept. The text was simply too faint. All that I noticed were two different-colored area charts. I mean, the chart looked very nice — great color palette — but it was hardly practical to my on-the-go, mobile use case. Even that important line about “health care system capacity” washed over me. I simply didn’t see it as I clicked out of the article and back into Twitter land.

Then, as I read more about the concept and encountered charts with clearer typography, the concept of flattening the curve crystallized. For example, Stephanie Evergreen shared this tweet, where the chart has easier-to-read annotations and a headline that’s action-oriented, not generic.

Do I think the chart directly above looks as nice as the one by Vox? Not really, but it sure educated me better on the concept, as did this one from Drew Harris, a professor at Thomas Jefferson University.

Again: clear, cogent, and to the point. Which is just what we need from a visualization in a time of crisis when we need to quickly educate a public hungry for accurate information.

I’m learning, too, how you can use visualizations to expand an initial concept. That dotted line above about “healthcare system capacity” is so crucial, and it hadn’t occurred to me that there is another key reason to flatten the curve and spread out cases: The healthcare community will learn over time how to respond to COVID-19 and, with more time, research will bring treatments and therapies into play. A few squiggles of a line, including by showing that providers getting sick will negatively affect outcomes in the short term, made all this clear through this version of flattening the curve:

Finally, I realize that this concept of flattening the curve cries out for animation, so that people can see “flattening the curve” unfolding, preferably as someone explains the concept in real-time, just as ABC News did toward the beginning of this piece.

Not many of us have easy access to tools to animate visualizations, especially those on the front lines of providing necessary information (such as local public health officers). But there’s no need to go high-tech here. That local public health officer could just as easily have taken a white board and drawn out the concept, whether in front of local news cameras or presenting at a community meeting. Alternatively, here’s an animated gif that’s licensed for public distribution.

So here’s what I’ve learned about health data visualization as I’ve been educating myself, too, on flattening the curve:

• Headlines matter. Don’t go generic when we need to tell people that they need to pay attention. Use a chart headline that helps orient the reader.
• Annotations are good. I mean, with flattening the curve, you don’t even need numbers. Just make sure your explanations have clear messages in easy-to-read type.
• Know your users. For example, many of us encounter visualizations on COVID-19 while we’re on our phones — which can make it difficult to digest what we see. Keep in mind that mobile experience.
• Animations are great, but if you’re in a face-to-face setting, so is drawing out the concept live on a white board or a sheet of paper.

And, of course, when all else fails, realize that cats are key to clarification on the internet, as this redrawn version of the Vox graphic attests to, posted on Twitter by Anne Marie Darling:

My list above is just a start. I’d love to learn from others on what else we need to do to marshal visualization that make this concept of flattening the curve actionable to all.

The post Flattening the Curve and Expanding My Understanding appeared first on Nightingale.