recorded unrecorded

By Fiona Beer, Leon Rauscher and Stephan Jäger

Refugee crises around the world are heart-wrenching and complex. Measuring and understanding the scale of the movement is difficult. In Europe ongoing flows across the Mediterranean Sea have been amplifying for the past decade. This group decided to take on the challenge of telling a story about this refugee crisis to their fellow students by making a large data sculpture that you walk through.

The piece centers around two mirrors facing each other, with lines and text decorating them, and invites the viewer to walk through. As they state:

This installation tries to show the number we know of, but simultaneously to represent the dark figure of deaths, names, age, origins, and causes of death, we have no idea of.

They acknowledge and embrace that reliable and complete data about migration, refugees, and deaths are impossible to come by.

Echoing the aesthetic of topographic map contour lines, they represent the “known” data on the floor and up the glass; one side for the living, the other for the dead. In addition, the dotted line suggests the data we don’t know, and have little way to estimate – the “dark figure of migrating people and death toll.”

This is all fairly readable from outside the sculpture, but stepping in between the mirrors creates an infinitely repeated landscape. The contour lines extend on forever, echoing the endless journey undertaken by so many migrants. The numbers and lines become hard to read, echoing the truly unknowable nature (a creative approach to the problem of visualizing uncertainty the field struggles with).

I think the core metaphor of the contour lines and the reflections that underlie this piece are resonant and memorable. More importantly, for me they pulled off a visual and physical design that embraces the humanity of the data, rather than treating them as abstract numbers that mask the crisis.

Data Sources: 2018 migratory deaths,  UN Refugee Agency Data Portal

Wealth Inequality in Switzerland

By Christina Bocken, David von Rotz and Silvan Häseli

Switzerland is a wealthy (and expensive) country, but there is large income inequality. This group was inspired by a piece from CBS This Morning to explore this inequality in Switzerland. They found, to their surprise, that just 2.5% of the Swiss population holds more than half all of the wealth. They decided to share this story, and what causes it, with a doll-house sculpture and a board game.

The multi-story dollhouse is divided into 3 sections about wealth ownership, each populated according to how much of the population holds that wealth. As the note on the yellow section states – “65% of people share 3.5% of all the wealth in Switzerland”.

To pull readers into why, they created a modified version of the popular snakes-and-ladders board game – called “Who Becomes a Millionaire”. Walk around back of the dollhouse and you’ll find a large playing board, dice, and instructions. Here’s the thing – the game is rigged. The rules change based on your gender, family background, education, and where you live! Playing as a woman nor born in Switzerland? Sorry, you don’t advance very quickly.

Inequality is hard to tell stories about, so I was excited to see this group take this challenge on. I think the house is a lovely invitation to come investigate the piece, and the board game is a creative pull to get the audience to dig in to the data more deeply. Building around the metaphors of the house, and the rules of the game being rigged, were key “anchors” to their design.

Data Sources: Swiss Federal Statistics Office

Medienkonsum 2020 (Media Consumption 2020)

By Eva Bieli, Valéire Huber and Sofia Leurink

Intrigued by our Media Cloud database of news reporting, and the surge of reporting on coronavirus, this group decided to explore the Swiss news landscape in 2020. News sucks us in; we want to read it, feel responsible to read it, but often end up left feeling powerless and upset by it. They found the top 3 themes written about were all depressing things – fears around coronavirus, aftermath from the assassination of Suleiman in Iran , and the tragedy of the Australian fires. This team asked “How consciously do you consume media? How do you digest bad news?”

This piece is built around a visual and explanatory pun – the idea of “consuming media” like one does food. Mapping each topic onto a pie-chart style plate, they laser-cut words into biscuits based on their frequency of use when discussing each of those top three topics. This is a literal representation of the news we consume, and I can tell you that picking up a cookie etched with “coronavirus” certainly does give you pause!

I think this piece is a delightful provocation to think about the news we consume each day. The representation in food is an inviting way into a thoughtful topic, hopefully breaking down barriers the audience might have to reflecting about their own news consumption. The 3 plates are playful, pulling you over to understand why they are different sizes.

Data Sources: Media Cloud (Swiss national and local media source collections)

The True Cost

By Nathan Blain, Rino Schläfli and Daniel Mankel

The true environmental cost of the goods we produce and purchase is almost impossible to comprehend – there are just too many variables. This group rose to this explanatory challenge, creating an immersive multi-sensory piece that assaults your senses when you choose an environmentally unfriendly option.

The piece present you with a choice between two identical-looking shirts, sit down and press the button associated with the one you want. Once your choice is made, three things happen that represent the environmental cost of the object:

1. The animation and sound of water on a screen increase based on the amount of water used to produce the shirt.
2. Smoke pours out at you (from a fog machine), obscuring your vision and making you pull back, based on the how many greenhouse gases are emitted in the production and transport of the shirt.
3. Your other hand, inserted (trustingly) into a dark box, is sprayed with water based on the amount of water used to produce the shirt. Well, it doesn’t actually spray your hand yet… they almost had it working by the opening so hopefully they’ll get it working in another day.

I’m delighted by their idea of creating a multi-channel assault on your senses. We don’t have good approaches to communicating the un-captured costs of these goods. This piece is a provocation to consider the physical impacts of those choices on your own body, rather than in some far-off rain forest or ocean.

Data Sources: TBD (I’m not sure what they used)

The Mask Collection

By Jasmin Schnellmann, Jasmin Vavrecka and Timo Lanz

This group was struck by the public reaction and news reporting about coronavirus. While certainly it is cause for concern, they realized that global air pollution is a far more serious problem right now, in terms of deaths and scope. Trying to find a way to communicate this risk, they struck on the symbolic power of the medical face mask.

They created two objects for the mythical fashion show about “the mask collection”. The first is a small wallet made out of masks and sized based on the deaths due to Corona virus. The second is a large purse made of sewn together masks, sized by deaths due to air pollution. The viewer reading more closely is likely to be surprised by the “price tags” on each, with the small one having a higher one while the bigger one has a lower cost. These price tags show more detailed information about deaths and media coverage – which are inverses of each other. Air pollution has far more deaths but far less coverage (in 2020).

The objects they created speak very strongly to me. They make you look twice to understand how they are made, with strong red thread echoing a sense of importance and dread. It is a simple visual mapping, but presented as a collection of daily use objects they speak to the choices we make, and the risks we face, each day.

Data Sources: Media Cloud, Guardian UK and NYTimes articles, World Health Organization data portal

My Takeaways

I was impressed by the students creativity, fortitude, and craftsmanship. They created these in just one week(!) with no previous experience with data storytelling. I continuously challenged them to think beyond simple tricks like physical bar charts and such, and I think they rose to the challenge.

From this one week I can picture expanding my approach into a semester-long course that explores the aesthetic principles of physical variables of data sculptures. Externalizing these models and stories creates an opportunity to communicate around and about them, and affords me the opportunity to revisit back on my training in the pedagogy of constructionism. 

This builds on earlier thoughts I captured in a paper with Catherine D’Ignazio at the Pedagogy and Physicalization workshop at DIS 2017. It also has links to projects from many other academics working on these questions of data storytelling and 3d physical materials:

• Data Physicalization workshop materials
• Data Physicalization wiki
• Findings from Dagstuhl seminar on Data Physicalization

I look forward to exploring these techniques with students and building my own data sculpture practice more. I hope they inspire, resonate, and challenge you as well! If you happen to be in Basel this Feb or March, these are on display for a few weeks.

What do data sculptures look like? The wonderful team at dataphys.org has been cataloging, thinking through, and writing about this for years. I could do no better background than they do in their paper about Opportunities and Challenges for Data Physicalization, so just start by reading that.

Ok, welcome back. So what is a data sculpture to me? It is a representation of data created using physical objects in the real world. While charts and graphs in 2D map data onto classical visual variables (size, color, shape, position, etc), data sculptures map data on to an additional set of things — smell, texture, 3D shape, taste, scale, etc. This media gives you a new toolbox with which to create data representations, and requires a new set of skills for creating with.

What I want to do is share some lessons and ideas from my ten years of helping people design and make data sculptures, in a variety of educational settings. Warning: I’m going to put on my Professor hat and share some of my strong opinions about what I think works and what I think doesn’t. I look forward to your constructive disagreements.

This posts teases out three themes I’ve seen with concrete examples from my teaching and beyond. These themes are:

1. Making charts in 3D just scratches the surface;
2. Choosing your materials wisely is critical to your physical data mappings;
3. Moving beyond gimmicks lets you flesh out how to support multiple levels of reading.

Making Charts in 3D

Most folks approach the idea of data sculptures with their existing vocabulary of data visualizations; they simply render an existing chart type in 3D using some physical material. This is all well and good, but I think making 3D charts misses the potential for data sculptures to attract and interact with audiences in memorable and provocative ways. Here are a few examples that went, or could have gone, a little further along that path.

Parallel Coordinates

One of the sparks for this post was a wonderful piece from two folks at Gramener, who wrote about their experience creating a physical data visualization in support of hackathons they run (thanks to Allen Downey for pointing me at this). They created a parallel coordinates chart with bamboo sticks and string to show metadata about the participants at these events.

Parallel coordinates are hard to read, but are powerful because they can show both trends and individual data points (see the great writeup on datavizcatalog.com for more details). Rendering them as a physical spectacle is a wonderful idea to both attract attention and do get to know the data. However, I’m left wondering about missed opportunities in the creation based on the physicality of the sculpture itself.

This project immediately brought to mind earlier work by the Domestic Data Streamers, who prompted attendees of a 2014 arts event to create a similar chart (they called it “Data Strings”). They key difference here is that they asked the participants themselves to create their data points. I think that addresses the main criticism I’d have of the Gramener example — it helped the authors understand and represent the data, while the Data Strings example took advantage of the idea to engage participants more fully.

If you’re going to make a chart in 3D, make sure it is in 3D for a reason. The Domestic Data Streamers participatory invitation is a strong reason.

Fireworks: Fun & Dangerous

Lets tackle another example of a very chart-like data sculpture. A team of students in my 2016 Data Storytelling Studio course decided to analyze fireworks-related injuries in the US. As they thought about how to best represent this in a quick data sculpture prototype, they landed on the idea of painting a mannequin to show where injuries occurred most.

This repurposing of a heat map in 3d form was a clever idea, especially since it used the human body itself in a very relatable way; you couldn’t help but feel the impact of the dark red hands (good color choice). It was a simple comparison story rendered in an emotionally evocative way, clearly intended to caution the audience about being reckless with fireworks! Yes, it is an old technique rendered in 3D, but the physical scale of the body standing in front of you fundamentally changed how you read it — you related to it.

It’s a Mysterbee

Let’s move on to the classic bar chart. A team of students from my 2018 class were digging into data about bee colony collapse across the US. Thinking of how to get people interested in a topic they might not otherwise be engaged with, the students decided to tell a high-level comparison story in honey itself by filling two cups with honey. Each represented a different year, and the amount of honey was based on the total production in each year. They invited the hypothetical audience to dip a cracker in each cup and compare — essentially creating drippy and delicious bar charts.

When was the last time you ate your bar chart? This invitation used a familiar method of reading that would make sense to people, but playfully used the subject of the data itself (the honey) to represent the data in a simple way. They had follow-up material that could support a longer conversation for folks that did stop and try the experiment, so it wasn’t just a one-trick show that ended with questions. The bar chart in 3D supported a comparison, and the cracker lent itself to being a barchart. Their use of the bar chart had a reasoning to it.

Closing Thoughts on Charts in 3D

I’ve been thinking about this theme for a while, because I see it so often. In fact in class, and in our Data Sculptures activity, I explicitly caution against doing this. I don’t mean to say it is never appropriate. In fact a 2013 study from Jansen dug into 3D bar charts to explore if they supported investigation and inquiry more. I took away the lesson that when people physically touched the 3d objects representing the data they did a better job understanding the data. A more recent paper they wrote, from 2016, investigated different approaches to mapping size as a physical variable and how people perceived it. It has a range of interesting findings, such as how spherical surface area was more accurately perceived than volume, but mostly point out that we don’t understand yet how physical variables are perceived.

Take Advantage of Your Material

The second theme I want to flesh out is how much the material matters. Cardboard is light and folds easily; balloons grow and shrink, float and pop; water flows, drips, quenches your thirst and gets you wet. If you’re using any of those materials, take advantage of how people use the material and what it can do. The creators of the Gramener graph acknowledge this for themselves, noting the power of how “feeling every data point was an experience in itself”. Choose your material intentionally to design the look, feel, smell, and taste of your data sculpture. Here are a few examples that flesh out what I mean.

Where is Your Water From

Water is used in a variety of ways across the globe — agriculture, industrial operations, human consumption. A group in my 2019 course used data about this breakdown to create an interactive piece about the future water available to us. They called it Where is Your Water From?. Their interaction was built around people’s perceptions of what water is used for and comparing that to the data.

They key point I want to emphasize with this example is how the invitations they crafted for participants centered around the “wateriness” of water. When asking people to guess how much water is used for different categories, they ask participants to pour the water from one container to another. When looking at data about how much water might be available in the future for us to consume directly, participants were invited to drink the tiny amounts. These two actions are strong examples of using the properties of water to help support the narrative they are trying to tell. They used the physical affordances and human behaviors around water to represent the data story.

Tasting Air Pollution
Taste is a wonderful sense to explore, and playful thing too map data to! Data Cuisine has been leading workshops doing this since around 2014 — their gallery has some wonderful examples of data rendered in food. Many are visual representations, but others alter flavor based on the data. A group of students in my 2017 course were inspired by this idea to sketch out an idea that mapped air quality data onto flavor in a project called Tasting Air Pollution.

Air quality is hard to experience; we don’t see subtle changes and don’t have a good sense of what an abstract air quality number means in terms of our daily experience. Stephanie Posavec and Mariam Quick’s “Air Transformed” piece gets at this in a concreate way. They literally created a set of glasses that obscured what you could see more based on how much pollution was in the air.

This group of students in my class decided to experiment with flavor as a way to represent air quality data. They were particularly curious about how we perceive intensity of flavor, and how gagging or couching on surprising or bad flavors feels like the response you have to polluted air.

To surprise the audience they invite participants to taste different brownies but didn’t telling them that the amount of salt had been increased based on how much pollution in the air there is in different cities. The “goal” brownie has the right amount of salt to make it delicious, while the Beijing brownie tastes horrible. Trust me, I was the test subject when they presented it in class!

Closing Thoughts on Materials

Data sculptures are more than just ink and paper or pixels on a screen. Data sculptures are made of something, and you have to think about what that smoething should be. Be conscious and intentional in your choice of the material you make your data sculpture out of. Consider the affordances, limitations, common uses, and interaction patterns of your material. Choose your material wisely.

Support Deeper Investigation

In my workshops and classes I encourage participants to support many “layers of reading” in their pieces. What dose that mean? Viewers should be able to quickly scan the piece and understand the main story, but should also be able to dig deeper to see more nuance and detail associated with the narrative.

Here’s the thing — most data sculptures I see don’t have many layers of reading. They use some kind of clever gimmick or tongue in cheek pun to make their point. I want to encourage you to move beyond these simple tricks and flesh out a multi-layered story that can be told with multiple uses of your physical mappings. There is a richness in your material and form that you should take advantage of.

Monopoly and Elections
One of the few examples in print that I showcase often is an article published in the New York Times in 2016, entitled The Families Funding the 2016 Election. The narrative focuses on the small number of obscenely wealthy families that were responsible for most of the campaign donations. To tell this story they use the visual metaphor of houses and hotels from from the board game Monopoly; a symbol instantly recognizable to any American kid.

The article opens with a visual pun. They show a mocked up photo of a huge pile of green Monopoly houses blocking the White House, then quickly zoom in to a tiny pile of red Monopoly hotels on top (as the reader scrolls). The whole pile literally obscures the White House and the contrast between the number of red and green pieces instantly reveals the story arc (along with the text superimposed on top). This is playful, effective, and a good example of a data sculpture presented in 2D.

Keep scrolling down the piece the reader discovers why this is even more powerfully used. First off, they continue to represent data with these same physical symbols to compare things like party affiliation of the donors.

Continuing event further down the piece one fnids that they bridge from house satellite imagery to maps showing their locations, and real photos of the houses themselves. This progression of representations is a wonderful example of really pulling all the power out of the physical symbol that you can. They support digging deeper and deeper into the data and the narrative, utilizing this physical representation in different ways throughout.

The Hidden Weight of Food

The water used in food production is becoming a larger topic of discussion as droughts become longer and more frequent. Another group of students in my 2019 course used the data about water cost of foods to create a series of sculptures — the hidden weight of food. They describe the interaction like this:

The hook is a long table with plates of food. Each plate has a fork with a bite-sized piece of food on it, such as a slice of apple. When you lift the fork, you realize it’s much heavier than a slice of apple should be. Upon being surprised and interested to learn more from the exhibit, you read the sign and realize that the weight you are lifting is the weight of the water used to produce that bite of food. For a slice of apple, that’s a full 27 pounds.

I can tell you from experience that it is a very surprising and effective trick, even in the rough prototype form that they build. They took advantage of the fact that we eat food, and that water is heavy. This comparison, between the expected weight of a bite of food vs. the far larger weight of the water used to produce that bite, is a super compelling and surprising story. It tries to capture that surprise and turn it into interest. They considered the subjects of the data (water and food), and used their affordances to design a delightful and evocative data sculpture.

They expanded on this simple and surprising interaction by adding another sculpture that provides more detail. After lifting the forks and becoming engaged with the topic, viewers can walk to the next sculpture, which breaks down the types of water used in the production of an orange, to complement the total volume of water presented in the first sculpture. They use the familiar shape of 2-liter bottles to make a pyramid with colored water representing different types of water. This constructs another physical invitation, digging into the story of water along a different dimension.

Closing Thoughts on Layers of

The lesson? Don’t stop with your initial idea; tease out how you can support your longer narrative using spark that you’ve got. Thee power in these examples is that they used the data sculptures approach to present multiple dimensions of the data story.

They expanded on this simple and surprising interaction by adding another sculpture that provides more detail. After lifting the forks and becoming engaged with the topic, viewers can walk to the next sculpture, which breaks down the types of water used in the production of an orange, to complement the total volume of water presented in the first sculpture. They use the familiar shape of 2-liter bottles to make a pyramid with colored water representing different types of water. This constructs another physical invitation, digging into the story of water along a different dimension.
They power here is that they used the data sculptures approach to present multiple dominions of the data story.

Conclusion

Curious to hear more about my approach to data scultpures? Check out the lecture slides, with notes, from the data sculptures session in my Data Storytelling Studio class on MIT’s Open Courseware site. My thoughts have evolved more since then, but it is a good set of sparks, prompts, and reflections.

Curious to hear more about my approach to data scultpures? Check out the lecture slides, with notes, from the data sculptures session in my Data Storytelling Studio class on MIT’s Open Courseware site. My thoughts have evolved more since then, but it is a good set of sparks, prompts, and reflections.

So not surprisingly I was excited to see Data Cuisine coming to Boston for a workshop by Suzanne Jaschko and Moritz Stefaner! Sadly I’m out of town and can’t make the workshop, but it sparked me thinking about food and data, and creative data representation a bit.

When doing data presentation in a creative medium, you have to choose your mappings and datasets carefully.  I’m often introducing people to more creative techniques for data presentation for the first time, and argue the strongest stories come when the message matches the medium well.  For example, one of their participant projects maps tomato and basil in a dish to the amount of Italian speakers.  This is a fairly culturally loaded mapping, that many would understand.  However, others are more abstract.  One mapped people to noodles to discuss sexual habits. A stronger mapping is the project that makes a joke about “death by chocolate” by creating small caskets to tell the story of common causes of death in Belgium.

Another intriguing example is Dan Barber’s red pepper egg (featured in an episode of Netflix’s Chef’s Table show).

He worked with a farmer to breed super colorful red peppers, then fed a mash of them to chickens to create the red yolk you see above!  Why?  All to start a person about to eat the egg wondering how it got that red.  What did the chicken eat to make that happen?  Why have I never thought about the supply chain going into this egg before?

To me, Barber’s red pepper egg is a wonderful example of data representation as food.  The food chain data in beautifully captured in the red yolk, and it prompts you to ask questions directly aligned with his goals in presenting it.  Wonderful!

More abstract representations of data in food are like a missed opportunity to me.  The artistic merit can be there, but leaves the viewer hungry for more.  A strong mapping between the medium of your data presentation, and the data and story itself, is key to creating a lasting impression.

Recently at the Responsible Visualization event put on the by the Responsible Data Forum I had a wonderful chance to sit down with the amazing Patrick Ball from the Human Rights Data Group and talk through how we help groups learn about working with incomplete data.

With my focus on capacity building, I’m trying to find fun ways for NGOs to learn about accuracy and data at a very basic level. Patrick agues that in fact you need rigorous statistical analysis to do this well, from his background in human rights data. I pushed a bit, asking him is there was a 80/20 shortcut. His response was to paint a great distinction between homogenous and heterogenous observability of data. For instance, there are many examples of questions that don’t require quantitative rigor – case existence, case history, etc.  This sparked a fun conversation about visual techniques for conveying uncertainty.

Watch the video to see the short conversation, or just catch the audio below.

Data can be used for a variety of things.  In thinking about setting up architectures for data use within your organization, you need to focus on two main questions:

• Does the data we have align with our goals?
• How can we use data to further our mission?

Alignment with Your Goals

People see data everywhere now, and get overly excited about it. When you think about using data within your organization, you have to return to the roots of what your organization is all about and make sure the data is in alignment with that.

There are a few common patterns organizations fall into when using data. First, many collect data simply because it is easy to collect, without considering whether and how it can be used.  Second, many tend to focus on quantitative over qualitative data, when in fact the strongest arguments are often made using both.  You have to understand what kind of data you have before you can use it effectively:

All these types of data need to align with your goals.  You can use data in a wide variety of your efforts, from inspiring more activism to changing behavior.  The key piece is your use of data must support those activities.

Using Data to Further Your Mission

Your data is not an end in itself.  It is an asset you can use to do your work more effectively.

You can use data in lots of ways to further your mission.  Three quick examples:

• improve operations: you can monitor engagement on social media campaigns
• spread the message: you can use data in your communications materials to advocate for change in new ways
• bring people together: you can gather around the data to find stories (and paint murals)

Of course there are loads of other things you can do as well. The key here is that This framing encourages you to be goal-centric, rather than technology-centric (which is a big danger when working with data). You don’t want to get lost in the hype around the latest and greatest tools. That approach does help you advance your mission. A beautiful external-facing infographic that doesn’t fit into your ladder of engagement, or includes no call to action, is useless.  A dashboard showing key indicators doesn’t mean much if they aren’t the right key indicators.

I hope this quick intro helps ground some of the hype out there around data use, and help you figure out what architectures to support for data use within your organization.

• if there more food waste in rich areas?
• do more expensive restaurants waste more food?
• do restaurants with more waste go out of business at a higher rate?
• are certain towns more wasteful than others?

This process of asking questions help you move beyond the data you have, to getting the data you need to answer the questions you have.  This question-centric approach is critical to make sure you don’t fall victim to having your dataset in hand be a constraint that stops you from finding an interesting story.

An Example of Getting More Data

So how do you go from these questions to more data?  I encourage folks to go “data shopping” (a term I enjoy stealing from my colleagues at the Tactical Technology Collective).  This involve taking each of your questions and thinking about what other data you need to answer it, and where you might get that data.  Returning to the food waste example above, to answer the question of whether more expensive restaurants waste more food, you need to categorize restaurants as expensive or not.  My students remembered that most restaurant review sites, like Yelp, have a dollar-bill scale that tells you how expensive a restaurant is.

How could you get that data? You could do it by hand, but that would take a while for all the restaurants in the food waste spreadsheet.  Instead, they pointed out that Yelp has an API, and you could write some software to query that and ask Yelp for the dollar-rating of each restaurant on the list.

Types of Data Sources

This examples uses one source of data – a private company.  There are, of course, others. Here’s the list I tend to introduce:

• Private Companies – There is tons of data collected and stored by private companies, and sometimes they will give or sell it to you.
• Governments – There is loads of official data collected by government agencies, and you have a right to the vast majority of it (depending on where you live).
• Non-Profits or Advocacy Groups – Interest groups typically collect datasets to back up and inform the advocacy they are doing.
• Crowdsourcing / Do-It-Yourself – Sometimes the data isn’t there, so you need to make it yourself!

That’s the list I use.  Am I missing a category?

Ways to Get Data

Fine, so there is data in a lot of places… how do we get it?  Here’s my list of techniques:

• Download Open Data – Yes, sometimes the data is just out there waiting for you to find and download it.  This doesn’t mean it is usable, but it is often there.  Usually large non-profits and governments have big data repositories you can poke around.  Sometimes it will be stuck in a PDF or HTML table, but you can still get it out.
• Ask For It – I mean it. Sometimes you just need to make a phone call and ask. A little social engineering goes a long way!
• Scrape It – Far too often the data is out there, but not in a nicely usable form… you need to scrape it from a website.  Scraping involves taking taking data is scattered around a website and using a process to get it all in one place in the same format. Nowadays there are lots of tools to help you scrape websites.
• Manually Collect It – If the data isn’t there, you gotta make it yourself.  This might involve crowd-sourced data collection, a focus group, or asking of social media.

Answering Your Questions

I introduce these two lists, of data sources and ways to get data, in order to support the data shopping process.  With a richer set of data in hand, you’re better positioned to find the most interested and meaningful stories in your data.

In public settings, it can be quite hard to get folks walking by interested in a data-driven argument about your cause.  We often argue that a creative data sculpture can grab their attention… like maybe a vegetable laser cut with some data about food security!

We’ve worked with the Somerville Food Security Coalition a few times, including for our first data mural pilot project!  Recently, we had a chance to come together again around their local data about food security at the Somerville Arts Council’s 2014 Ignite Festival.  The festival celebrates fire and food, which inspired us to laser cut some data onto food and see how people reacted!

Here’s all the veggies we cut – eggplant, cucumber, zucchini, bread, and watermelon:

In addition, we prompted folks to interact with two questions – both of which they could answer with M&Ms and raisins.  Asking folks to take an M&M survey is a highly effective way to get them to interact with their data!

https://twitter.com/rahulbot/status/498880294226001920

https://twitter.com/rahulbot/status/498883181714882560

Here’s a behind-the-scenes video showing the laser cutter in action:

This is cross-posted to the Civic Media blog.

A Corporate Example

The first connection I saw recently was a video ad for Prudential while I listened to Pandora Radio.  They are trying to tell a data story about how long people live after retirement, with the goal of getting them to set up a retirement plan with Prudential. The campaign is very appealing from a data-presentation point of view.  In one ad they asked people how much money they thought they needed for retirement, then gave each a length of ribbon, and had them walk from the center of a circle to the length of the ribbon:

Another let people put a sticker on a big chart to build a histogram of the oldest person they knew:

These are cool, and look fun.  Letting people be the data connects them with the information in a real, body-syntonic way.  I’m sure this makes the people more likely to be interested in Prudential’s product offerings and planning services.

An Academic Example

In the academic realm – my colleague Nathan recently went to the Computer Support Collaborative Work conference, where he learned about the MyPosition project from Nina Valkanova, Robert Walker and others.  Her recent work revolves around concepts of presenting information in public spaces.  Here’s an academic paper describing the MyPosition project.  It allows people stand in front of a projected poll and add their vote by holding up their hand:

Their findings in the paper around social pressure are interesting, as is the fact that people got around the fancy tech to actually engage in the question they were polling.  Also the idea that people used it more when it showed real people’s faces is interesting.  All in all, it presents a fascinating example and some usable insights into how to design these types of public interactive data presentations.

A Community Example

My colleague Sasha Costanza-Chock recently pointed me at the Crossing Boundaries project from the local Urbano Project.  Artists Alison Kotin and Risa Horn worked with 10 local high school students to gather data about local transit and create art pieces that told the data stories they found.

Their pieces are embodied data sculptures – wearable objects that represented the data story they want to tell.  This example is fantastic empowerment, data literacy, and art work.  I enjoy it in so many ways and look forward to talking with the creators sometime in the future.

Be the Turtle

So what’s the takeaway?  As a young participant in a robotics workshop I ran years ago said – “Be the turtle”.  Think about ways you can engage people to actively be the data in the story you’re trying to tell.

(1) Papert built on Freud’s notion of “ego syntonicity”, which concerned the mind.  This presentation I found online digs into this more in relation to computer programming.

One other particularly useful list comes from a classic academic paper called “Narrative Visualization:  Telling Stories with Data” by Edward Segel, Jeffrey Heer (download it here).  The paper meticulously reviews about 60 online visualization, mostly from newspapers, to define some recurring genres.  If you can stomach the academic prose, the paper is worth a read.

Their “genres” focus on 2-d visual presentations of data stories, to be expected based on the title of the paper and the examples they pull from.  However, within that space it is a particularly wonderful list:

• magazine style: “an image embedded in a page of text”
• annotated chart: a traditional chart of graph with textual callouts highlighting specific data points
• partitioned poster: a “multiple view visualization”
• flow chart: a directed series of pieces of information
• comic strip: multiple frames in a linear path
• slide show: a series of visuals presented one at a time to assemble a narrative
• film/video/animation: fairly self-descriptive

These vary based on the number of “frames” (visuals) presented, and how they are shown over time.  This list breaks down the set of techniques differently than I usually do, and that’s a nice thing so I thought I’d share it!

From there they move to a discussion of author- vs. reader-driven approaches.  That’s a wonderful reminder to decide early on whether you are building an exploratory or explanatory presentation.  Are you trying to tell a strong narrative, or showing information and letting the viewer take away a story?

Factoid Stories

There’s only one Eastern Redbud tree in all of Somerville! What’s the story of that tree?  Turns out the leaves change to bright pink in fall, but everything else it yellow and orange.

Sometimes in large sets of data you find the most interesting thing is the story of one particular point.  This could be an “outlier” (a data point not like the others) like the Redbush example above, or it could be the data point that is most common (can we tap more of the Maple trees that dominate Somerville?).  Going in depth on one particular piece of your data can be a type of data story that fascinates and surprises people.

Connection Stories

How come Somerville Ave has some many trees in the best condition? Oh, it was recently renovated… that is why those are all new trees.  There’s a story about more aesthetic outcomes of big street resurfacing projects.

When two aspects of your data seem related, you can tell a story about their connection.  The fancy name for this is “correlation“, and you of course need to be careful attributing causes for the connection.  That said, finding a connection between two aspects of your data can lead to a good story that connects things people otherwise don’t think about together.

Comparison Stories

Walking down Somerville Ave. gives you a good sense of the most populous trees across the city.  That street is a good representative of the tree population in the city as a whole.  Is your street different?

Comparing between sections of your data can a good way to find an illustrative story to tell.  Often one part of your data tells one story, but another part tells a totally different story. Or, as in this example above, maybe there is a more human slice of your data that serves as an exemplar of an overall pattern.

Stories of Change

Turns out there was a big die-off of trees in 2008.  Was the climate weird that year? (I made this up since I don’t have any time-based data)

People like thinking about things changing over time.  We experience and think about the world based on how we interact with it over time.  Telling a story a story about change over time appeals to people’s interest in understanding what caused the change.

“You” Stories

You live on Highland Rd? Did you know that ALL 9 Spruce trees in Somerville are on Highland Rd? Maybe we should rename it “Spruce Rd”?

Another way to find a story in data is to think about how it relates to your life.  People with map literacy like maps because they can place themselves on it.  This personalization of the story creates a connection to the real world meaning of the data and can be a powerful  type of story for small audiences.  Stories about your personal experiences can be grounding and real.

In Conclusion…

This is just one take on the type of data stories that can be told.  Please let me know how you think about this! Telling that story effectively is a whole different topic, but I find the story finding exercise much easier when I introduce a bunch of categories like this.  Most of these benefit from multiple sets of data, so remember to go data “shopping” during your story finding process.

Footnotes:

(1) For instance, I’m a huge fan of Seger and Heer’s Narrative Visualization paper, where they give a catalog of visual storytelling techniques.  Also good is Marije Rooze’s thesis work (particularly the tagged gallery of visualizations from the Guardian and New York Times).