It’s the size of the smile on your face when you see someone. It’s the decision to stay in touch with a niece long after family events end. It’s the fragile silence between you and a friend who couldn’t support a recent life choice.

As a data designer, I’ve always been obsessed with how we categorise the intangible. Recently, I set out to map the people who have shaped me. I didn’t want a balance sheet, but I did want to see the patterns. A relationship always evolves; this would only represent a snapshot in time.

The Taxonomy of Connection

I began by listing every person I care about. First from memory, then verified by my friends list on Facebook. But as I opened my spreadsheet, the questions started to flood in. Can family members count as friends? For example, my nieces and I have been chatting nonstop for years now. We grew fond of each other through the circumstance of birth, but we stayed in touch by choice. Does that make them friends? And what about friends who aren’t supportive of my life choices? We were very close 7-8 months ago, but we are not now. Are we still friends? If I exclude her from this, does that mean I have given up on our friendship? Also, I use the term “friend” very loosely. I am naturally familiar with strangers. Is my new neighbour — with whom I have shared a few cups of tea — my friend?

To make sense of the friend list, I distilled friendship into three core metrics, scored on a scale of one to three, three being the highest rank possible: 

• Reliability: Loyalty, faithfulness, and the feeling of being safe.
• Empathy: Supportiveness, kindness, and open communication.
• Joy: Playfulness, liveliness, and shared common ground (though one might question whether friendship is required for common ground; for the sake of this visualisation, I decided it was).

I also added two judgment values: Duration (how long we have been friends), and Contact (how recently we spoke). To keep the data honest, I limited the scope to friends I had contact with in the last 24 months. I chose 24 months as a mark because it’s the period since my daughter was born. Spoiler alert: In a time when I often felt lonely as a new mother, the data showed me I was actually deeply loved.

From Sketching to Scripting

In my notebook, the design evolved rather quickly into a series of “tiles.” I remember having the visual in my head for a while, and I felt as if I were a vessel letting it out onto the paper. I wanted something that would represent the scale’s levels easily. Level one was a simple base; level three added complex detail. 

Initially, I used background colors to denote duration, but the palette was too loud. It made the story about “how good I am at making friends” rather than “how these friendships built me.”

Then came the pivot. Usually, I build these visualizations by clicking the mouse. A thorough process of copying, pasting, and double-checking layers in Illustrator and Figma would easily take three hours. But, inspired by the “vision to execution with a click” movement, I turned to Claude and Gemini.

I asked Gemini to help me write the prompt for Claude. It generated a Python script that processed my Excel file and generated stacked layers as PNG files. Claude taught me how to install Python on my Mac. (Honestly, I felt like I was back in the 90s, typing into a terminal to launch a game.) Then, “Boom. Your tiles are ready.” With a single click, the assets were generated. A few back-and-forths with Claude, and the grid was aligned. The work was done.

The Cost of Efficiency

As I looked at the finished folder, a strange feeling washed over me: I didn’t recognize the data. By automating the execution, I had accidentally bypassed the data familiarization stage — that meditative hour where you handle each data point with care and remember the person behind it. The tiles were beautiful, but they felt distant.

It raised a fundamental question for our field:
If the AI builds the layers, are we co-creators? Or are we just curators of our own memories?

The Tokens of Gratitude

Despite the digital distance, the final grid is a testament to my life. These tiles are me. They represent the people who stayed through puberty, the ones who signed my wedding book, and the new friendship that started when I collected my son from preschool, which grew close.

This project is more than a visualization; it’s a token of gratitude. It captures a snapshot of my soul as it exists in 2026. Shaped by humans, rendered by machines, and held together by the voluntary, personal relationships that make life worth mapping.

The post The Tiles That Made Me: Mapping Friendship through the Lens of AI appeared first on Nightingale.

I lived in London for a decade, the final three and a half years spent in Brixton. There, I would frequent two distinct green spaces. Brockwell Park, a large public park where the neighbourhoods of Brixton and Herne Hill meet; and Brixton Orchard, a small community garden situated directly opposite Lambeth Town Hall. It was during this period that I developed a keen interest in community-focused urban green spaces. In the winter of 2023, I left London for the densely populated city of Taipei, the Taiwanese capital. I moved into an apartment within walking distance of Da’an Park, the city’s central green expanse, but the true joy was the city itself: greenery along every street and tucked into every corner with plants of every species tumbling down off balconies. From my kitchen window, I could see the man who lived in the top floor apartment opposite had cultivated a veritable jungle atop his roof, in which he would emerge daily at sunset to water his plants and sit quietly on a bench he had nestled under his palms. This set-up is quite common in Taipei. As Clarissa Wei described the city for The New York Times, it is “a literal urban jungle—ferns and large elephant ear plants sprout through the crevices of roofs and sidewalks with wild abandon”. Across the street from my apartment, a pocket-sized neighbourhood park was a constant theatre of intergenerational life, teeming with both children and the over-seventies. I was surrounded by nature in the heart of a capital city, and I loved it. 

Urban greening, or green infrastructure, is the deliberate integration of vegetation—street trees, parks, green roofs, and living walls—with urban development to provide ecological, environmental, and cultural benefits. The protection of nature in urban spaces is essential for sustaining natural ecological cycles, and provides crucial cultural ecosystem services: it softens the harshness of urban infrastructure, ensures a critical connection to nature, improves general wellbeing, and fosters social interaction. Taipei relies heavily on green spaces of all types for public life. The city serves as a compelling case study for urban greening in compact cities, defined by its “top-down” planning approach born of its limited land area. A study by Peilei Fan, professor of urban and regional planning, and colleagues, found most neighbourhood centres and one subcenter in the city “exhibit both high compactness and good green accessibility”. The central city rests on the ancient Taipei basin, bounded by major rivers (like the Tamsui) with steep, mountainous terrain rising abruptly on most sides of the basin. Other examples of compact cities in East Asia include Hong Kong and Singapore. London, by contrast, is a city that grew organically over centuries, resulting in its pattern of “urban villages”.

Since 2010, Taipei’s urban planning policy has shifted its focus from a broad, visible green strategy toward practical green landscaping schemes. This strategy gained momentum around 2014 with the emergence of an urban regeneration programme and the prioritisation of the development of small green spaces, such as river corridor greens and pocket parks. Through government funding and the lease of state land, the Taipei Beautiful Programme and the Open Green Project have delivered the development of small green spaces across the city. A complementary effort, the Garden City initiative further promoted urban agriculture, including citizen farms (allotments), community plots, and rooftop gardens, particularly on school buildings. I used open source data from the Taipei Government Open Data archive to illustrate the development of green spaces in Taipei City over the last fifty years. I first translated the raw data from Mandarin, followed by exploratory analysis using Python. I plotted the time-series data as a raster-style graphic, aiming for an aesthetic that evokes an interconnected urban ecosystem. I was interested in mapping the locations of developed green spaces, and I relied on data from the Taipei City Water Green Space Atlas. I manually classified all the green spaces listed in the atlas according to the definitions of the Garden City initiative.

Effective data visualisations and well-planned urban spaces seamlessly blend aesthetics and utility, aimed at enriching public experience and understanding. In data visualisation, this means selecting graphs, layouts, and interactive elements that communicate the underlying data with clarity to engage the target audience. Similarly, in urban greening, this translates to designing spaces and infrastructure that provide ecological benefit and prioritise the needs of the local community. Form follows function. Munari viewed the designer as a mediator, bridging the gap between expert knowledge and public life: “The designer of today re-establishes the long-lost contact between art and the public, between living people and art as a living thing.” Without effective design in data visualisation, the data remains inaccessible. With it, visualisation can empower the public to understand, question, and engage with critical issues. Urban greening, by its very nature, is public-facing. The design of urban green spaces is about making a city a usable and sustaining environment.

Of course, data visualisation can also act as a bridge between academic research, expert knowledge, and public understanding within urban greening contexts. As someone whose academic background is in Neuroscience, and not Urban Greening, I write this from the perspective of the public that these policies need to engage; the public must be integral to the decision-making process for the planning and design of their local green spaces. The ‘Citizen Dialog Kit’, an open-source toolkit, was developed to leverage situated visualisation within public spaces through a set of interactive, wirelessly networked displays. By making local environmental data visible and understandable, it invites discussion and feedback from diverse community members who might not typically participate in traditional planning meetings. Residents can see the tangible results of past efforts and then contribute to ongoing dialogues about future greening priorities. In a recent study, socio-environmental scientist Thomas Mattijssen and colleagues presented a participatory application of GIS that bridges the gap between data-driven and citizen-centred urban greening. The authors used spatial modelling in community workshops where residents contributed local knowledge to enable researchers and local citizens to jointly identify greening criteria, translate them into indicators, and pinpoint potential greening locations. This accessibility fosters democratic participation in environmental decision-making.

Visualisations serve as a powerful tool to engage community members and translate complex datasets into compelling narratives, increasing both understanding and acceptance of environmental initiatives. For example, RisingEMOTIONS, a data physicalisation and public art installation situated outside the East Boston Public Library in 2020, aimed to engage communities directly affected by sea-level rise, encouraging their participation in planning adaptation strategies. Looking ahead, emerging technologies offer significant opportunities to enhance the role of data visualisation in urban greening, and more broadly, climate policies. Innovations such as AI-assisted personalisation, mobile technologies for context-aware experiences, and real-time environmental sensing can amplify its impact, creating tools—imagine an app that allows a Taipei resident to visualise the projected, real-time impact of a new pocket part on local air quality—that make planning tangible. For a comprehensive look at how data visualisation can be leveraged to address sustainability goals, a recent article by social computing professor Narges Mahyar provides an excellent review.

In adhering to Munari’s principles, where design prioritises functionality, clarity, and intrinsic beauty, the design of data visualisations and urban green spaces finds its highest purpose in its utility to the public: “Art shall not be separated from life.” At their best, both disciplines demonstrate that “beauty arises from functionality”—whether in the efficient form of a bar chart communicating a climate trend or the purposeful design of a pocket park managing water runoff—both are functional in their communion with the public. Moreover, data visualisation can become an indispensable tool for urban greening. By effectively communicating complex planning concepts and policies, visualisation fosters community involvement and informs policy decisions that directly benefit local communities.

All images designed and provided by the author.

The post From Pixels to Parks: The Intersection of Data Visualisation and Urban Greening appeared first on Nightingale.

“LA on the Move”, our exhibition organized by Metro Art at Union Station in Los Angeles, California, opened in October and will remain on view through next year. Through larger-than-life graphics, a massive 3D map, playful character designs, and even animal sounds, we’ve created an immersive experience that asks Angelenos to see themselves reflected in the lives of coyotes, mountain lions, monarch butterflies, red-tailed hawks, and california kingsnakes.

From City Animals to Union Station

The seeds of “LA on the Move” were planted at ArtCenter College of Design, where we first encountered the City Animals class taught by Santiago Lombeyda and Ivan Cruz. “It was a topic I hadn’t really thought about before,” Jen recalls. “The interaction of humans and animals in LA County—it was super intriguing. The more we got to know the projects and the students, the more excited we became.” Then there was a chance to have an exhibition that pulled together a lot of these concepts and also showcased the student work, created in association with Metro Arts at Union Station. From there it just started rolling”.

The final projects from the City Animals class focused on speculative projects that explored how Angelenos could redesign their homes, backyards, and neighborhoods to better integrate with the natural world. Jason explains, “The projects that the students did were really about how people in LA could think about the intersection of the built environment—their homes, their yards, their backyards—with the natural world”. From there, we led two intensive workshop sessions with the students, working side by side to visualize ecological data in bold, accessible ways that were displayed in the ArtCenter student center for the following month.

From there, we were connected with Arroyos & Foothills Conservancy, a non-profit organization focused on preserving and restoring natural open spaces and wildlife habitats. They became an essential partner, sharing datasets on animal sightings, migration patterns, and habitat corridors across LA County as well as expert advice and access to Southern California’s environmental researchers.

The research process: Data meets daily life

“I think the first thing that we did, and what we always do, is begin with research,” Jen explains. “but in time, we leaned on the expertise of our friends at Arroyos & Foothills Conservancy—they were incredibly helpful. The other part, I think that’s very important, is collecting anecdotal information when you’re talking to people that live in Los Angeles about their experiences”.

For us, stepping away from the data is essential. “It’s important to step away from the facts and the figures, and start talking to people because the experience that Californians have with wildlife is completely different than a New Yorker’s,” Jen says. “You can’t just go about your business like a city dweller and ignore nature in California. It’s integrated into your day-to-day experience”.

Los Angeles, we discovered, is extraordinary in its biodiversity. Jason notes, “Los Angeles has such a unique environment. And what we found is that it’s actually one of the three areas in the world that is considered a biodiversity hotspot“. This became a cornerstone of the exhibition—LA isn’t just a city with some nature on the edges; it’s where wildness lives alongside urbanity in remarkable, sometimes precarious, ways.

Five animals, five stories

We chose to focus the exhibition on five species: coyotes, pumas (mountain lions), red-tailed hawks, california kingsnakes, and monarch butterflies. Each animal became a character in the larger narrative of LA residents navigating neighborhoods, dating scenes, commutes, and survival just like the humans around them.

“One of the first things that you drew was the coyote that says: ‘I love LA.’ That’s one of the featured images in the show,” Jason recalls. For Jen, this illustration became a statement of intent: “A human says, I love LA—and we all know this phrase—but animals live there too. What’s their role in this? So, we wanted to make sure that the animals and humans get equal time in this show”.

The personification of the animals was deliberate and humorous. Jen explains, “The more you learn about animals, how they’re mating with other animals, for instance, you think about the LA dating scene, and then you think about animals, which have some funny crossovers. As we have these neighborhoods in a city, they also have their neighborhoods.” Jason chimes in, “For example, a monarch butterfly says, ‘Hey babe, let’s overwinter in Mexico’—a line that could just as easily come from an Angeleno planning a winter getaway…” Jen adds, “And the monarch is saying like, I’ve got a really busy schedule.” Jason elaborates: “They have this multi-generational migration habit where up to five generations of butterflies will go from Central Mexico all the way up to Nova Scotia and Southern Canada and then back again. And they do this over five different generations. Even more remarkable—five generations later they’ll return to the same tree”.

The California kingsnake became another favorite. “Well, it’s not an LA Dodgers hat. Thank you very much,” Jen jokes, describing the snake’s illustrated headwear. “It’s a Los Angeles hat”. The kingsnake’s ability to live almost anywhere—from woodland to wetlands to suburban basements—made it a perfect symbol of LA’s adaptability. As we say, “you live in my backyard.”

Navigating the hard truths

While humor runs through the exhibition, we didn’t shy away from difficult realities. Rattlesnakes, for instance, posed a design challenge. “I made this drawing. When you might be on a hike, you may encounter a rattlesnake. And this is frightening, right?” Jen recalls. “There was like a discussion about making the rattlesnake so it wasn’t so intimidating, which was funny because I was like, well, a rattlesnake is intimidating and very scary, and you can’t really take animals and smooth out all the rough edges, right? Because that’s not what they are”!

The story of P-22, the famous mountain lion, underscored the fragility of human-wildlife interactions. Jason reflects, “Take the story of P-22—a famous mountain lion that was known around the Mount Wilson Observatory. And eventually, through a series of interactions with humans (and despite best intentions) he dies”. The exhibition addresses this directly, including data on rat poison’s devastating impact on mountain lions and the importance of hazing techniques—like carrying a can filled with coins—to maintain healthy boundaries.

“Even though we anthropomorphized the animals, we shouldn’t forget the fact that there are negative results of some of our interactions with the animals. We should be mindful of that”.

Making data visible and inviting

One of our core practices is taking complex datasets and transforming them into visuals that invite exploration rather than intimidation. “Part of what we do is find information and basically make it much more understandable to the general public and to ourselves,” Jen explains. “Like rat poison killing pumas, right? We made this diagram so that we have the data there, but you can just see it more clearly”.

A standout piece in the exhibition is the massive chart “Animal Species at Risk in California”, which visualizes 930 species by class and phylum, showing which are extinct, endangered, or imperiled. Working with data visualization collaborator Paul Buffa, we transformed this overwhelming dataset into the shape of a California poppy—the state’s native flower.

“If I saw this information in spreadsheets, I would be very intimidated because it’s just a lot of information,” Jen admits. “But since we put it into this California poppy, which is a native plant, it invites you over to explore it. You don’t have to look at every single detail, but it is fascinating”.

The wall also includes a Sankey diagram comparing California’s at-risk species to global standards—revealing that California has considerably more species in danger. And the bar chart showing imperiled species? “It literally towers over your head. It’s about seven and a half feet tall, so we wanted it to have a physical relation to how you encounter the data”.

The iconic title wall: Observing Union Station

The exhibition’s title wall features three illustrated characters walking across a vibrant gradient backdrop—each carrying something that subtly references animal behavior. Jen describes how these characters emerged: “We were standing in Union Station, and I could see people walking through, going from the trains to the entrance, and it gave me this idea about what kind of people would be walking through LA and walking particularly in Union Station”.

The older gentleman carries a bag of groceries, echoing how animals travel to forage and transport food. The young woman holds a bundle of flowers, referencing seed distribution—how seeds attach to animal coats or are eaten and deposited elsewhere. “All said and done, the more time you spend with the exhibition, you know every element is intentional and thought out and has a relationship to the information that we learn as we go along,” Jen explains.

The massive 3D map: Placing yourself in the data

Perhaps the most captivating element of LA on the Move is the enormous 3D map, created in collaboration with Julian Hoffmann Anton. This wasn’t just a cartographic exercise—it became a months-long process of negotiation, expansion, and refinement.

“Every project we do, we discuss a map component,” Jen says. “And sometimes we have time to do it, and sometimes we don’t because what starts as a simple map becomes very complex. It’s because a map is political. You can’t leave anyone off because they’ll notice”.

Initially, the map focused narrowly on downtown LA and Union Station. But through conversations with Metro Arts staff and community input, it expanded dramatically—eventually encompassing all of LA County and parts of Orange and San Bernardino Counties. “We were pushed and pushed on the map, but that’s not a bad thing. It’s a much more inclusive map, so when visitors come to Union Station, they can find themselves”.

In addition to showing every detail of the city, the map tracks sightings of all five featured species across the region, revealing fascinating patterns. Mountain lion sightings appear surprisingly far south of downtown; California kingsnakes cluster in parks and mountains but occasionally show up near Marina Del Rey; while coyote sightings may reflect research centers as much as actual populations.

“I’ve never seen a map of this scale, physically, of this detail,” Jason marvels. “It’s an extremely detailed 3D rendering of the entire metro area”. And because it wraps around a corner, visitors can find neighborhoods that might have been cropped out of a conventional map. Jen describes a photograph of a man pointing to the side panel: “He’s finding himself, which we wouldn’t have had in our original idea”.

Adding Sound: Activating the Space

For the first time in a Data Vandals project, we incorporated audio. “I pushed for this because we wanted to activate the space as much as possible,” Jen says. “We’re dealing with walls, and we wanted ways to expand these rectangles out”.

Visitors can hear the sounds of pumas, coyotes, and hawks. “I thought, okay, if I’m walking through Union Station, what is it like to hear some of these animals?” Jen explains. The sounds are surprising—sometimes beautiful, sometimes unsettling. Jason describes, “The mountain lion has lots of really low growls, more aggressive than a purr, and I found those to be unsettling”. Coyote calls also sound strange and a bit frightening, but these sound elements ground the exhibition in sensory reality, reminding visitors that these are not cartoons but living, breathing neighbors.

Iconic cutouts and LA signage culture

Atop each wall, we placed large cutouts of the animals lifted high on Sintra board to add height and visual drama. Jason says, “We wanted them to refer to the history of the Hollywood back lot, even the Hollywood sign itself.”

Jen reflects on LA’s distinctive signage culture: “I think the signage is very different from anything you ever really see on the East Coast; in New York we don’t have that kind of sign culture and I find it fascinating and really attractive”.

The billboard aesthetic also responds to Union Station’s architecture—a stunning 1930s Art Deco space with soaring ceilings and intricate tilework. “Union Station is so gorgeous, you want to try to do it justice. Something that iconic, you worry that whatever you do is going to be overwhelmed”. To honor the building, we photographed the tile floors and extracted colors to integrate into our palette, creating a dialogue between the historic architecture and our contemporary street-style graphics.

As the exhibition settles into its year-long run, we hope it becomes a recurring destination; a place where commuters pause for five extra minutes, where families return to discover new details, where Angelenos see their neighborhoods reflected in a 3D landscape populated by shared species.

“I just want people to enjoy it and have fun with it and see themselves in the data,” Jen says. “It’s so fun to see the different types of people, and I feel like I could draw those people and put them into the exhibition. It reflects a lot of our intentions”.

Jason hopes for depth and revisitation: “I’d love that the exhibition is very detailed; you can return to it over and over and learn something new each time that you revisit it”. And Jen adds with a laugh, “I hope it brings us back to California again and again –  we love LA “!

“LA on the Move” is on view at Union Station through 2026.

For more information: https://datavandals.com/la-on-the-move.

The post LA on the Move: Data Vandals Bring Wildlife and Humans Together at Union Station appeared first on Nightingale.

We reflect on our experiences arising from a recent computer science graduate class about data feminism, during which we explored the idea of being data detectives. In this report, we explain what we mean by Data Detective as an active approach where we, as individuals, could approach the underlying questions, as suggested by D’Ignazio and Klein “Data science by whom? Data science for whom? Data science with whose interests in mind?”. By connecting individually through personal reflection, data literacy, and critical engagement, our goal is to inform and inspire those who are interested in integrating similar methods into their classes.

As our society continues to evolve, more and more of the information we need is stored as data, and many of these repositories are growing and becoming what we refer to as Big Data. In this process, data becomes more challenging and less accessible to us as individuals. We, as visualization researchers, work on the creation of visualizations as at least part of the solution to this problem. However, much of our data is still not visualized, and even when it is, individuals still often find it challenging to understand. How do we cope with this? How do we teach our students to cope with this continually expanding problem? 

In our data feminism class, we introduced concepts such as visual variables, physicalizations, assumptions about knowledge development (e.g., Positivism and Interpretivism), along with reflection and discussion on reading the book Data feminism. We then explored developing an active practice through which we would document our investigations of both qualitative and quantitative data under various themes. We now term this active practice as being a Data Detective. 

Our concept of Data Detective is modeled on detective work in a more general sense, where a person uses coherent, time-based, record-keeping of their activities to gain a better understanding of that which they initially do not know but want to understand. Thus, to act as a Data Detective is to discover and conduct purposeful, documented, and reflective actions needed to gain access to the desired data. This detective work ideally results in access to the desired data, an understanding of the data, and the detective process involved. 

The term Data Detective appears in various contexts, making it important to clarify our specific approach. Unlike children’s books that suggest counting objects (like red cars versus white cars), or Harford’s statistical literacy guide with its ten rules for making sense of statistics, or visualization workshops for children by providing a gamified sense of accomplishment. Our approach also differs from Inselberg’s multidimensional data detective work, which focuses on analyzing existing visualizations, and from data activism approaches that emphasize community engagement.

We visualized our investigative approaches as journeys: beginning by defining a critical problem or question, identifying gaps (“What’s missing in the picture?”), searching for data, and confronting barriers—missing, partial, or deliberately obscured datasets. Each step provided unique insights into the relationships between data, society, and power dynamics.

Examples

Throughout the semester, students undertook diverse projects with strong societal relevance, including topics such as gender bias in politics, barriers faced by women in entrepreneurship, the functions and ideology of pockets constrained by historical gender roles, and gender representation within STEM academia. 

One student examining women’s representation in political institutions vividly illustrated the practical challenges of data detective work. Initial exploration quickly highlighted systemic data gaps as key datasets were fragmented or unavailable. The student navigated through a frustrating landscape marked by opaque official sources, partial records, and silences. Despite challenges, this data detective journey offered significant emotional and intellectual rewards. The student discovered patterns of marginalization, for instance, women are frequently relegated to peripheral roles rather than core decision-making positions. Each painstakingly gathered dataset provided clarity about structural inequalities. Ultimately, the effort became a tangible act of resistance against invisibility and marginalization.

Another student explored the challenges women encounter in achieving tenure in Canadian academia. Initially optimistic, the student encountered considerable barriers, including incomplete or outdated datasets and inconsistent categorization across institutions. Interviews became essential to fill these gaps, highlighting how data detective work can require alternative methods beyond computational data collection. The journey revealed systemic biases: women disproportionately assigned tasks correlated with lower job satisfaction and hindered career progression.

Working with both big data & personal data

As we moved through the process, we found ourselves blending two approaches to data visualization that are often kept separate: working with big data and working with personal data. Big data showed up in the external datasets we chose to investigate, such as government records, institutional statistics, or public health databases. These are the kinds of large-scale, structured data commonly associated with the term big data.

On the other hand, personal data and visualization came into play as we reflected on our own experiences navigating these data landscapes. By documenting our paths through note-taking, diagramming, and visualizing our steps, we deepened our understanding of the datasets themselves and uncovered what was missing, what was hard to access, and where our questions should lead next.  

Central to our pedagogy was encouraging students to critically reflect on their data practices. We structured reflective exercises to surface the implicit power dynamics in data collection and usage. Students were prompted regularly to question: Whose data are we using? Who collected it, and for whose benefit? Who controls access, and how does that affect analysis?

This reflexivity deepened our critical engagement, enabling us to overcome technical challenges and interpret the implications of their findings from the data.

We suggest one possible pathway to actively take on the role of being a Data Detective:

• Initially clarify what one is looking for—this is before one has the data.
• Develop a timeline starting from the current moment, which will track the process by which one gains or loses access to the data.
• Choose a currently promising direction to find more information (could be: ask a person, search on the web, go to an institution, etc.)
• Collect and reflect on the information collected, filling in one’s timeline, with data, facts, responses, including emotional and frustration level responses.

Actively conducting Data Detective projects in our class, where we used personal visualization of our detective process to teach us about both institutional and personal data, whilst revealing many factors about our society. Each data point gathered and each visualization created represents a small act of making the invisible visible, contributing to more equitable and inclusive understandings of our complex social world.

Acknowledgement

We thank our colleagues and reviewers for their thoughtful comments. This research was funded in part by NFRFR-2022-00570 (A Co-Design Exploration), NSERC Discovery Grant: Interactive Visualization RGPIN-2019-07192, and Canada Research Chair in Data Visualization CRC-2019-00368.

The post Exploring Data Detective Practices as a Class Activity appeared first on Nightingale.

I was 14 then, and football was more than a mere sport. I vividly remember my mornings began with practice before school, lacing up my shoes. Yet, in those training sessions, I often felt the weight of being different. Few girls around me cared about sports, but even fewer about football. I envied the boys, who seemed to belong to a world I was desperate to be part of. Hoping one day I’d witness women players in the limelight.

Jump to the year 2025, women’s football is gaining its momentum. The current UEFA  Women’s Euro is  the largest event as the attendance has been record-breaking. Over 600,000 tickets were sold, setting the stage for the most attended Women’s Euro ever.

Before it was Messi, Cristiano, Maradona, now young football enthusiasts can look up to Aitana Bonmatí, Alessia Russo, Sam Kerr, Alexia Putellas, Patri and a whole bunch of amazing women footballers.

Decoding Aitana Bonmatí

Football is a sport that requires immense strategic, physical, and mental endurance. One of the most prominent names in women’s football is Aitana Bonmati, who began her career at the age of 17. Aitana Bonmatí won the Ballon d’Or, which is the most prestigious individual prize in both men’s and women’s football, not one but three times in a row. Aitana Bonmatí grew up in a home where standing up for what she believed in was simply part of life. From the start, her parents made a bold choice to challenge tradition, placing her mother’s surname before her father’s. In doing so, Aitana became one of the first in Spain to carry her mother’s name first—a quiet but powerful statement of equality that would shape the person she was destined to become. She was the lone girl in the boys’ team.

Bonmatí exhibits extraordinary technical finesse, seamlessly maneuvering the ball through tight spaces while maintaining constant awareness of her surroundings. Her vision and spatial intelligence empower her to dictate the tempo of matches. Moreover, her relentless off-the-ball movement and impressive stamina consistently create opportunities for her teammates and drive Spain’s dynamic, high-intensity style of play.

“We need to have women in more powerful positions that are making decisions, so when that 10-year-old girl is looking up and wondering, ‘What can I do and what do I want to be when I get older?’ She has the opportunity to do and be whatever she wants.”

 — Abby Wambach

The debate over gender in sport is nothing new, and football is no exception. For decades, women have been underrepresented, underpaid, and held back by systemic barriers. While the popularity of women’s football has surged in recent years, it still doesn’t rival the global dominance of the men’s game. But on a positive note, female footballers are becoming global superstars and commercialisation is growing in all areas of the game. The 14-year-old girl in me would be overjoyed to see her favorite sport shattering barriers and achieving unprecedented growth.

The post Breaking Stereotypes in Women’s Football appeared first on Nightingale.

As a former organiser of UX Camp Hamburg—an annual UX bar camp—and now of the DataViz Meetup Hamburg, one of the many things I am curious about is: can I use data visualisation as a tool to foster community and exchange? Thus, I took the opportunity to give a workshop at the above mentioned bar camp. To create a together-datavis about the interests of the participants over time. In a slot of 45 minutes only. To see what happens. To make people talk and think.

Understanding the challenge

I sketched it for myself first. Defined colours for interests, made a timeline on an x-axis, thought about specific moments that helped me remember what I was into at that time and drew that onto the y-axis. An easy exercise.

But how should this work out with more people? How many colours would we need? How can we draw it into one chart? All at once? How many participants would there even be? How many different interests would they have? How old would they be? I was swaying between “forget it, that is going to be too complex” and “let’s find out”.

So, I tried. I organised my process and wrote down a prompt to identify what data types there would be: “Which topics were you interested in at a certain moment of your life?” Hence, I needed to find a visualisation that presents data by different people, development over time and interests grouped into categories. How many manifestations these data types would have, I would know for sure only then, in the workshop.

Finding a visual idea

I skimmed through datavizproject.com and sketched ideas. I was searching for a flexible type of visualisation that everyone could draw at the same time. Everything showing data by area size dropped out. We would have needed to calculate it beforehand. The data should be collected while drawing. Besides, I decided to show time spans instead of a whole timeline for each participant, to simplify the process and address the problem that there would be people of all sorts of ages.

I checked out charts with distinct symbols as Column Sparklines. They turned out to be not flexible enough. Bee Swarm Plots. Too much chaos. Just bar charts out of different coloured lines stacked upon each other. How long would that take until everyone got hold of a pen in a particular colour? A heatmap would be cool with squares of transparent paper we could layer! “But who would buy so much paper and cut all the squares?” I was asking myself.

In search of different inspiration, I took Visualising Complexity (by Darjan Hil and Nicole Lachenmeier) with me to the playground. Instead of looking into it, I bathed it in the contents of my water bottle. This book being all wet, Dear Data (by Stefanie Posavec and Giorgia Lupi) became my bedtime reading. And there it was. Giorgia Lupi’s “Phone addiction” visualisation: different circles spread over the postcard, surrounded by the data points.

Each participant could work on their own “circle”. The “circles” would hold different symbols embodying different interests. And the circles wouldn’t be circles, but forms depending on the decades the participant has already spent on earth. They will be spread over a canvas. This way, everyone would be able to work on their own datavis at the same time. We wouldn’t need different colours.

Planning the details

I sketched the idea. Planned the structure of the workshop. Squeezing everything into 45 minutes. Drew the legend beforehand. Made an analogue presentation to be independent from technical restrictions at the venue. I collected the material and bought paper on a huge roll. One thing I learned from previous workshops: better bring your own material and be prepared for any local conditions. Since I found pens in grey and blue, I came up with another piece of information to add to the drawing: symbols for interests that are still relevant today would be blue, while all others would be grey.

My plan: Give an introduction and explain the data visualisation we are going to build. Collect interests together and cluster them into categories. Assign a symbol to each category. Give the participants some time to sketch their own part. At the same time, pre-draw the bigger forms according to the different ages. Put everything together.

What happened?

We ended up in the canteen. There we had a big table to draw on, and no space to stick post-its to. So, the collection of interests and clustering took quite some time. We had to use the windows. The final drawing worked out quite well. At least with the participants not heading off for lunch. The benefit of being so present in public: everyone else could take a look at our emerging drawing.

All in all, I gained the experience that producing a data visualisation with more people at once can work out—if one defines everything beforehand. It is still usable for self-reflection and talking then. Probably more than any dot voting. However, I don’t think participants can learn much about data visualisation, and they sadly cannot become creative on their own. Still, I hope that with a bit more time and in a calmer setting, we will achieve even better results. What’s your take on this? Have you ever tried anything similar? I am still curious.

The post Datavis as a Community Experience: How to (Not) Create a Datavis in a Group appeared first on Nightingale.

So when I taught a new Human-Computer Interaction (HCI) course where many students would be engaging with the field’s literature for the first time, I knew I needed to get creative. I teach Computer Science at Mount Holyoke College, a small, private, predominantly undergraduate liberal-arts college in Western Massachusetts. When I set about the task of designing the new course (an intermediate-level elective for undergraduate Computer Science majors), I set an objective to expose students to the broad assortment of areas in HCI through engaging with exciting new literature — the final weeks of the course would be at the same time as the largest HCI conference in the world (ACM CHI), after all! However, I knew that, while students would be familiar with reading academic texts generally, this might be their first time engaging with Computer Science literature (broadly) and almost certainly their first time reading HCI literature (more specifically). Therefore, I wanted to design an activity which would help students get more comfortable navigating new texts in a way that felt fun and approachable, but would build strong skills they could apply to future readings.

I ultimately designed an activity in which students create data comics as a means to better understand the structure and content of research papers containing human-subject studies. Inspired by past work about how creating data comics (data-driven stories in a comic strip-like form) might benefit researchers, I designed this activity to use the process of creating data comics to benefit readers’ skills.  The big idea is this: in order to create a data comic, a student must both find the pertinent information they need to tell the story of that paper and understand enough of what they’ve found in order to summarize it. Further, because creating a data comic may feel more fun, creative, and low-stakes than other deliverable formats that students are familiar with (e.g., reports or presentations), students may be able to engage with this difficult work with less fear and stress. 

In this short report, I will provide an introduction to data comics generally, explain the activity I designed involving them, and reflect on the opportunities and challenges of conducting this kind of activity.

What are data comics?

Data comics are a type of narrative data visualization which present a data-driven story in a comic strip-like form. While data comics may look like any other comic strip at first glance, they incorporate visualizations into their data-based narratives, using different combinations of visualizations, (narrative) flow, narration, and words and pictures (see above Figure for examples). 

Data comics can be a particularly powerful tool in educational contexts because they leverage, break down, and communicate potentially complex information in an approachable format. Authors have written about the potentially helpful role of data comics in a variety of contexts including helping people make sense of their personal data and better understanding how to approach unfamiliar visualizations through reading and creating explanatory comics (for a comprehensive survey of data comics in education, see Boucher et al.’s 2023 survey). Further, creating data comics provides students an opportunity to practice both high- and low-order cognitive tasks (e.g., finding and summarizing) in a creative, low-stress context — ingredients which Psychology and Education research tell us contribute to long-term learning.

The application of data comics which most inspired the activity I designed was Wang et al.’s work on data comics as a means to report controlled user studies. In their paper, Wang et al. describe how authors of scientific papers could use data comics as a means to report information about their user studies in a format that might be more accessible to both experts and non-expert readers. While my students aren’t (often) authors of scientific papers, they were readers of them, so I wondered: could making data comics help student readers understand the structure and content of papers? To try this out, I modified Wang et al.’s workshop procedure (described in their publication and the workshop website) into the following activity for students which would be possible to accomplish within a limited class time.

The activity

In this activity, students created a data comic for an existing research paper containing a human-subject study during one 75-minute class session plus a 10-minute pre-class preparatory discussion. While I hoped that students would come out with a good understanding of the paper that they’d read, the primary objective of this activity was to build students’ confidence in finding and summarizing key pieces of information in an academic paper so that they could apply those skills to future reading tasks. To make the initial paper navigation smoother, we spent 10 minutes in the prior class session discussing the sections of a “typical”’ research paper and their high-level purposes. For instance, we talked about how abstracts are a summary of the work as a whole (and thus serve as great overviews), but often do not contain critical details about methods, results, and impacts which can be found elsewhere. The purpose of this discussion was to provide a general road-map for students to recall, apply, and expand in the following class, while they were actually creating their comics.

At the start of the main class session, I introduced the concept of data comics and students explored examples of existing data comics. The goal of this introduction is to help students get an idea of what data comics can feel and look like. We used Bach et al.’s data comic gallery as a starting place, combined with other examples students found elsewhere online. 

Then, I divided students into small groups of two or three and asked each group to pick a paper to convert into data comic form. In the inaugural version of this activity, I selected three short papers from ACM CHI for groups to choose from. Each paper incorporated a human-subject user-study of some kind to connect the exercise to topics students had seen earlier in the course related to human-centered design methods. Although I selected the papers in this iteration, the search process could alternatively have been student-driven with students either independently proposing papers or consulting proceedings together, depending on the goals and time constraints.

Each group was then given several sheets of plain paper and a set of colored markers to create their data comics. While there are lots of great tools for creating digital data comics, I intentionally chose to have students create comics with paper and markers to reduce the friction that comes with learning a new tool and the perceived pressure to try to make something that “looks nice.” This philosophy is consistent with substantial existing work on the benefits of creating paper-based, lo-fi prototypes of visualizations to facilitate idea generation and divergent thinking. Further, I wanted to focus students’ attention on the “fun” of being creative — I’ve learned that college students often don’t get to play with markers in class as much as they might like!

Then, it was time to dive into comic making, focusing on creating comics with a simplified 3-part structure. Given the amount of time students had to create their comics, I asked students to focus on finding the information required to tell a story with the following 3-part structure:

• Motivation & Question: Explain the researchers’ central research question(s) and why they matter
• Methodology: Explain what the researchers did to try to answer their research question(s)
• Results: Explain what the researchers learned from their experiment(s), focusing on the most important outcomes

For each part, students were asked to find the information in the paper in the relevant section(s), summarize the most important pieces of information together as a group, and decide the best way to communicate that summarized information in their comic through a combination of images, text, and visualizations.
At the end of class, each group shared their creations with their classmates. Of the comics created in that inaugural class, several focused on Gui et al.’s paper “A Field Study on Pedestrians’ Thoughts toward a Car with Gazing Eyes”  (perhaps because of its cute “self-driving car with eyes” concept!). You can see components of three different groups’ comics for this paper in the Figure below.

Possibilities and challenges

Overall, I found the first version of this activity to be quite successful, in terms of both positive student reception and accomplishing learning goals. While most groups did not produce complete, polished comics in the 75 minute session, they all engaged with their chosen paper deeply over the session and wrestled both with the format and content in productive ways. 

Additionally, students reported that they loved this activity: in their end-of-week reflections, they repeatedly described the activity as the highlight of their week. Students’ comments indicated that they found it both fun and extremely helpful for furthering their understanding of how to approach and read papers in the future, emphasizing that the act of creating something new based on the reading was particularly impactful. While these initial impressions were volunteered as a part of a broader weekly reflection assignment for the course (and thus may not reflect all student reactions or reflections), they indicate that this activity was a positive experience overall for many. I plan to collect more systematic feedback from students regarding what precisely worked (and didn’t work) when I repeat this activity again.

Despite my students’ generally positive reaction, there are certainly challenges to conducting this kind of activity which I’d suggest readers think about if they are considering doing something similar in their own context.

Allocating the Right Amount of Time

First, selecting the right amount of time for this activity can be a challenge. In the initial version of this activity, my students made their data comics over the course of one 75-minute class session, supplemented with a short 10-minute introductory lesson in the prior class. Though I do think students accomplished enough deep work in this time to ultimately improve their reading skills, few of them came away with a fully complete comic. Additionally, while students shared their comics with their classmates, we did not have time for students to give each other feedback on their comics or for students to refine their comics based on that feedback. As discussed by Boucher et al., engaging in these kinds of feedback loops is critical to both developing more polished, effective comics as well as cementing learning.

One approach to picking the “right’” amount of time for this activity may be to think about how complex the main learning objective is for the session and allot an amount of time to match it.For instance, an implementation of this activity which mainly aims to build students’ skills for finding information may require less time than versions that focus on the summarization and presentation aspects, because finding information is a less complex task than summarizing it (according to Bloom’s “Taxonomy of educational objectives”). In situations where the activity time is fixed, it may also be possible to incorporate pre- or post-activity work to support in-class activity time. For example, Boucher et al. had workshop participants identify a visualization to explain in a data comics before beginning their workshop session and Wang et al. asked participants to identify a dataset to use between the first and second session in their 3-session sequence.

Considering Existing Familiarity With & Orientation Toward Key Ideas

Second, while comics are enjoyed by a diverse group of people throughout the world, they are not universally understood. Instead, readers must learn how to decode the visual and linguistic conventions in comics, like any other form of narrative. One impact of this reality is that students who are less familiar with comics may face an extra barrier to their learning. Therefore, educators who are considering this activity should consider students’ existing familiarity with comics and allot additional time and practice to account for familiarization (e.g., by allotting additional time to analyze the format or work with existing comics prior to asking students to make their own).

In addition to comics, it is important to consider students’ familiarity and comfort with visualizations. As previously observed by Wang et al., it can sometimes be a challenge to get students to integrate visualizations into their comics, depending on their existing experiences with the topic. During this iteration of the activity, I observed that some, but not all, of the groups incorporated visualizations into their comics, though it is unclear whether this was because they were uncomfortable with using visualizations or just ran out of time (see Figure below for an example of one group’s use of a timeline and pair of pie charts to summarize the methods and results). Educators whose students are less familiar or comfortable with making and using visualizations may find tools like Boucher et al.’s “Comic Construction Kit” or Bach et al.’s data comic design patterns cards helpful to scaffold this challenge and re-direct students’ energy toward learning objectives.

Further, convincing students that creating comics is a worthwhile learning activity may be difficult depending on their existing orientations toward this kind of activity. While work in Educational Psychology has shown that creative activities like drawing can be beneficial to learning in STEM fields, students may not view it this way, depending on their existing beliefs about these activities. For instance, while my students were enthusiastic about creating comics as a component of a Computer Science course, my institution is a liberal-arts college which highly emphasizes interdisciplinarity and takes a pretty broad view of what Computer Science is and how it can be taught. However, educators at institutions which take a more traditional view of what the “work” of their field is, the acceptable pedagogies used to teach it, or which abide by a stronger science/art divide may need to do additional work in order to get student buy-in.

Selecting the Right Comic Format for Your Paper Type

Third, while it may be possible to create a data comic for any academic paper, the 3-part format described in this article may need to be modified for papers without experimental studies. When I initially designed this activity, I knew that students would be working with papers containing human-subject studies because we had covered related methods earlier in the course. Therefore, the 3-part Question/Methodology/Result narrative structure used by my students was picked with these kinds of papers in mind. However, these three sections may not meaningfully encapsulate other types of papers which do not use experiments as the basis of their claims (e.g., theoretical or position papers). Educators who want students to create comics for non-study papers should consider their main components and select a structure with that in mind. For instance, data comics for theoretical or position papers could instead map out the steps or pillars of the argument being made and how they relate to each other.

Conclusion

In conclusion, creating a data comic based on an existing research paper may be an effective learning activity because it forces students to practice both finding pieces of information of interest within the paper’s unfamiliar structure and digest the information they find in order to transform it into a new form — two stumbling blocks for those new to reading academic papers. I am planning to bring similar activities to my other courses and I hope that this article inspires other educators to bring data comic creation activities into their work as well.

The post Learning to Read Academic Papers by Making Data Comics appeared first on Nightingale.

The experiment

The goal of the experiment was to find out if age impacts engagement in interactive data stories. We recruited 2,400 participants from the UK and split them into four age cohorts. We did not use names typically connotated with generations to avoid stereotypes. Instead, we called each age cohort by their age range: 18 – 27, 28 – 43, 44 – 59, and 60+.

Three interactive designs

The researchers tested three interactive data story examples, each inspired by award-winning publishers:

1. “Make a Guess” (Design One): this design asked users to guess an answer to a question and then surprised the user when revealing the correct response. Inspired by the New York Times ‘You Draw It: How Family Income Predicts Children’s College Chances’ 
2. “Breaking the Fourth Wall” (Design Two): this visualization personalized the experience by asking users to input their name and then integrating their name into the narrative. Inspired by ABC Story Lab ‘See how global warming has changed the world since your childhood’ 
3. “Exploration” (Design Three): this reader-driven design lets users click through interactive cards, encouraging free exploration to create their own narrative. Inspired by The Pudding ‘A Visual Guide to the Aztec Pantheon’

Participants interacted with one randomly-allocated design, answered a 22-question engagement survey, and provided optional qualitative feedback.

Key findings

The study’s quantitative analysis confirmed a statistically significant difference in engagement scores (p = 0.03), with the 18 – 27 cohort (mean score: 112) more engaged than the 60+ cohort (mean: 109). 

• Younger Audiences (18 – 27): As digital natives, they noticed interactivity, mentioning it multiple times in their feedback. They comprehended the intent of engagement inside the data story where, for example, they observed that Design Three allowed for data discovery through exploration. Aesthetics also mattered; criticism of “dated” colors or “cluttered” layouts was common, with 12 instances for Design One alone. They preferred concise, gradually revealed text, making information digestible and intuitive.
• Older Audiences (60+): This group faced usability challenges, reporting confusion or distraction from interactive elements like scrolling. They desired more contextual text to fill knowledge gaps, with multiple instances requesting additional information. Their lower engagement scores reflect these barriers. Furthermore, this age cohort mostly did not observe the intent of the data story author to use interactivity to engage. There was comparatively little mention of interactivity with this age group.

Design recommendations

For authors, these findings translate into practical strategies to create inclusive interactive data stories that resonate across age groups:

For Younger Audiences

• Embrace Bold Interactivity: Leverage narrative design patterns like “Make a Guess” or “Exploration” to engage tech-savvy users. Explicit interactive elements align with their expectations, boosting engagement by making them active participants.
• Prioritize Aesthetics: Clean, modern interfaces with vibrant color palettes and minimal clutter are critical. Avoid outdated or busy designs, as younger users are quick to disengage when aesthetics fall short.

For Older Audiences

• Simplify Usability: Ensure interactive elements, like scrolling or buttons, are intuitive and clearly marked. Complex interactions can lead to missed data, undermining the narrative’s message.

Provide Context: Offer optional detailed text to provide background information, addressing potential knowledge gaps without overwhelming the primary narrative.

Why this is important for authors of interactive data stories

This study empowers authors to move beyond one-size-fits-all designs. By understanding age-related preferences, you can craft data stories that resonate with your target audience. For younger users, it’s about creating visually appealing experiences that leverage their digital fluency. For older users, it’s about reducing barriers and providing context to ensure accessibility. The study’s emphasis on inclusivity aligns with the broader push for universal design in visualization, ensuring no audience is left behind.

Moreover, the findings challenge assumptions. Younger users don’t just “crave” multimedia—they engage deeply when interactivity is purposeful. Older users aren’t inherently tech-averse; they simply need designs that prioritize usability. By applying these insights, authors can create visualizations that not only inform but also inspire, fostering engagement across generations.

N. Errey et al., “An Age-based Study into Interactive Narrative Visualization Engagement,” in IEEE Computer Graphics and Applications, doi: 10.1109/MCG.2025.3591817. Read a preprint available here: https://arxiv.org/html/2507.12734v2

The post Why audience age matters: The Influence of Audience Age on Engagement with Interactive Narrative Visualization appeared first on Nightingale.

The internet “reads” your interests, and gradually your network expands. At some point, I became part of the Flowers and Figures community, joining others who are interested in data art and passionate about creative projects.

My path

At first, I was just studying the works of others, hesitating to try it myself, though ideas for data encoding had been circling in my head. As is often the case, my first attempt happened by accident. A few members met offline at a café just before Christmas to get to know each other and to attempt creating a data art project right there, on paper, using pens or markers, based on each person’s own data.

I loved it! Both the process and the result, even if it looked quite simple.

After that, I was eager to create something of my own—something real, not on paper.

As with any data visualization project, the key steps are:

• choose a topic
• find data
• and, in this case, invent a way to encode the data into visuals—the most creative part, in my view

An important decision is the path you take in encoding and presenting data art:

• One way is symbolic—any geometric figure, flower, or petal can mean anything, depending on what data value or category you assign to it
• The other way is to keep it as close as possible to the real object in the data, which isn’t always feasible.

The first step for every new community member is entering their data into a shared Google Sheet. This forms a dataset that can be used to create a full-fledged piece of data art. A community portrait—sometimes called data badges—is a popular format in data viz spaces, especially within communities or at conferences and events.

The range of themes for data encoding in the community is wide: geometric shapes, flowers (matching the community’s name), even bugs, and coats of arms. I wanted to create a community portrait that stood out from the gallery. And this time, I really wanted the project to exist physically; to photograph it and then compile a digital version.

According to the format rules, the portrait had to show:

• skills—drawing, crafting, data, data viz
• proficiency level in each skill
• name, gender (optional)

Dataset: Google Sheet where each new participant fills in a row about themselves.

The idea of physical data art

For representing the skills, I chose colorful plastic cable ties—commonly used to bundle wires and cables. They’re a simple and effective way to connect elements, widely used in construction and daily life. My daughter, a student, used longer ones to secure her rolled-up architectural drawings. These ties can withstand quite a bit of stress.

I wasn’t interested in technical specs though—the important factors were color and minimal size. There wasn’t much variety in colors—most sets offered standard combinations: red, blue, green, yellow, and orange. So I worked with what I had.

The encoding took shape

• Colorful ties = skills: drawing, crafting, data, data viz
• Proficiency level = length of the tie: trimmed literally to 1, 2, 3 cm or fully cut off if the level is 0

The hardest part was figuring out what to attach the ties to. I tried wooden coffee stirrers—too long. Chopsticks—needed cutting or grouping data for 2–3 people, which felt clunky and undermined the clarity of the concept.

I browsed various craft supplies, school kits, and eventually found counting rods—those plastic sticks used in early math learning kits for first-graders. Perfect for the project: 6 mm in diameter, 6 cm long—exactly what I needed.

One stick = one participant.

Everything else fell into place: gender represented by the color of the stick—black or white, painted with gouache. Labels repeated this info. At first, I tried writing names with a gel pen, but eventually moved to printed labels.

The result

I tried different layouts for the finished sticks. You can’t twist them too much—names become unreadable, lighting matters, shadows too. The final shot of the stick layout became the data art piece. The legend was made in Figma, and the whole composition was assembled there too.

The data art includes information from just a portion of the community—it’s grown a lot, and photographing the full dataset in one frame was technically impossible at home. I really didn’t want to use photo compositing. I added numbers to the printed name labels so participants could find themselves quickly, since names repeat. The whole process took about two weeks.

The sticks themselves turned out charming, and during our offline meetups I can hand them out to participants. They’re nice to touch and sort through—each with its own texture. Honestly, I didn’t want to put them down. But all things end—and the data art now fits neatly into a small box from a gadget.

I’ve seen breathtakingly beautiful projects shared in our channel—complex constructions from paper and thread, beads, even 3D-printed pieces, and what amazed me most—made of clay.

I couldn’t wait to share my result. I didn’t expect the post in the community to get so many comments and positive feedback on my modest effort. A short moment of fame—delightful and inspiring.

Now I’m thinking of making something material on a socially meaningful topic. To do that, I’ll need to: find the data, come up with an encoding in a specific material, and bring it to life. And in my most ambitious plans—participate in a data viz competition!

The post Plastic Portrait: Visualizing Technical Skills Through Cable Ties appeared first on Nightingale.

Figure 1: (a) left: a solitary tree standing erect on a sunny day, (b) middle: a solitary tree swayed on a windy day, (c) right: a collection of trees on the same windy day

On our way to the university, we go past trees shown in Figures 1b and 1c. The other—featured in Figure 1a—was one one of us found while hiking in the high Himalayas. What differences do you spot? It is these differences—the way they stand (some ramrod straight, some swayed; some on their own, some along with friends) or the backdrops against which they are set (one was a bright, sunny day, the other gloomier in contrast) that could, if we are willing to notice, point to memorable peculiarities in more serious settings, such as creating a time plot. Tasked with depicting the evolution of variables, modern analysts’ default gravitation is towards line diagrams shown in Figure 2. The (fractal) trees we propose, by winning you out of that culture-based addiction which becomes tedious far too quickly, help you focus on the contrasts between two phases—adjoining or not—of the evolution, and retain that observation far more solidly.

Winds that whisper degrees of change

What are we offering? A platform on which the interaction between two phases in a time series may be staged through a tree. This tree is made using properties of both phases. The extent to which one phase is different from the other is shown by the way the tree sways. And the force that is doing the swaying—possibly a gust of wind—is suggested through the amount of leaves that get torn away. The backdrop expresses the storm that triggers this wind. And all this, collectively, showcasing action, lets readers feel the amount of difference between the two phases examined and the “winds of change” that are responsible, against their faces.

Figure 2: Time series showing usage of common idioms over the years. Structural breaks are marked by vertical separators. A tree is planted (in Fig. 3 below) at a “present”: a place where one phase ends and another begins. In general, these phases need not be separated by change-points.

To underscore a comic and impish possibility, we choose to demonstrate all this through n-grams made out of idioms involving “winds” or “change”. Please see Figure 2, but the methods apply just as well to any other time series. These are (scaled) numbers of times these phrases were deployed in books and media over the years. The task at hand could be comparing how the usage has changed over two sections of history. Some structural breaks are expected, that is, some upticks can be explained: the rise in “the winds of change” around 1990 correspond to Scorpions’ release of the song commemorating the fall of the Berlin Wall, the one in “blowing in the wind” to Bob Dylan’s one around 1963, the one in “candle the wind” to Elton John’s 1997 rendition, in loving memory of Princess Diana. There are weird people—and you are in the clutches of two of them—who nitpick endlessly on where these break points should be. We will spare you the technical details for now, but there is mathematical backing behind the vertical separators you see on Figure 2. Although the phases to compare need not be defined by change points, we choose two such phases to bring out the differences most glaringly. So, for us, the pre-change phase is 1975 – 1988 and the post-change one 1989 – 2000.

Figure 3: trees showing how the pre-change phase (1975-1988, on the left) clashes with the post-change phase (1989-2000, on the right) for (a) left: “winds of change” and (b) right: “blowing in the wind”.

The intent is to construct a tree that, through its very presence, would narrate both the extent and the type of shift a time series went through as it transitioned from the pre-change to the post-change phase. The placement of the tree would point out an imaginary present that would create a past-future split. Every shift in the ongoing intensity of idiom-usage may be thought to be caused by a breeze flowing from the region of higher intensity to the region of lower intensity, making the tree sway in the direction of its movement. How much the tree would sway depends on how strong the wind.

A look at the first natural tree (Figure 1a), for instance, would suggest the absence of a wind, making the past and the future this tree was trying to separate, more or less equal, in properties. In contrast, the second natural tree (Figure 1b) would suggest a strong wind blowing from its left, or the past, implying a higher intensity on the pre-change phase as opposed to the post-change one. In our n-gram examples, the average usage of an idiom over a period may serve as its intensity. These are marked for “winds of change” and “blowing in the wind” on Figure 2. For both those cases, if we focus on 1988, the post-change intensity was more dominant. This is why both the resulting trees in Figure 3 sway to the left under the influence of winds blowing from the right. The post/pre-change intensity ratio guides the way the trees grow. If this ratio is more than one, the twigs twist to the left, giving off the impression of a breeze blowing from the right. If the ratio is extremely more than one, the twigs twist quite a lot. Similarly, they twist to the right in case the ratio is less than one. For instance, for “winds of change,” this ratio is 2782/1764 = 1.577 while for “blowing in the wind”, it is 999/618 = 1.616. “Blowing in the wind”, therefore, endured the bigger change, as shown by tree 3b, through its stronger tilt to the left, in comparison to 3a. Regardless of the type of wind blowing, the trees, through their structures, can reveal fluctuations among the numbers in each phase. In case the fluctuations are similar—such as for “winds of change”: 361/249 = 1.449, Figure 3a—a twig, at each point of split, will not deviate too much from its sibling, suggesting an overall solidity, condensing the similarity of variation among the pre- and post-change values. In case the fluctuations are different—such as for “blowing in the wind”: 237/109 = 2.174, Figure 3b, the deviations will be more, suggesting a structural instability, or a flexibility, representing the unsureness the tree has, in deciding how to sway.

These trees, and their positioning, therefore, arrest drifts over non-participatory or uncontrollable time domains, freezing temporal flows (Figure 2) to geographical invariables: the past fixed to their left, the future, to their right (Figure 3), making obvious, through repetitions (next section) small shifts that may go unnoticed in ordinary line diagrams.

Repetitions, refinements and absurdities

A tree could, therefore, be an apt instrument to witness first-hand both the extent and nature of turmoil phases of a time series tolerate. Once a simple tree is erected, principles of good visualisation may be deployed to better its potency. One such is the principle of repetition. The breaking up of twigs may be done many times—each time following the same rule of tilt and spread (outlined in the previous section)—to stress the large-scale or eventual impact (through the leaves at the top that these subdivisions trigger) of even small differences between two phases. Here, they lead to fractal-type self-similar structures, but in our earlier essays, we have used the same principle of repetition to amplify or make blatant minor dissimilarities unearthing non-fractal settings. Robert Fathauer, a renowned sculptor based in Mesa, Arizona, is alleged to have quipped “Humans have a huge ability to get better at complex tasks with repetition”. We have paid heed. Those interested may bring in further tweaks or other modifications. The number of repetitions may be employed to suggest the difference in the lengths of the pre- and post-change phases.

Figure 4: Tighter (i.e., slender or narrower) trees result when the pre- and post-change volatilities are similar. Even against the same background, it is possible to think of trees of varying spreads.

We can exploit the background to incorporate another piece of detail. It could be something endogenous or internal such as the extent of change (the absolute or relative differences in the pre- and post-change intensities, instead of the ratios shown in the previous section, in case the extent of tilt is not revealing enough) or something exogenous or external such as the average number of times similar phrases were used in non-English texts, providing a context or reference, in a way. The darker the background, that is, the more ominous the sky, the stronger could be the breeze blowing. Much like the gloomier day shown in Figures 1b and 1c. We have sampled, in Figure 4, situations that show how different trees can sway in different ways on days that are equally sunny or gloomy. The difference between the background sky and the foreground tree-wind combination can be highlighted through the number of leaves being torn away. The higher the friction, the more the number.

Figure 5: On absurdity. Adjacent trees are expected to experience the same environment and consequently, sway similarly, in unison. This, however, may change when one tree condenses a specific change environment for one time series, the other, for another, with the two series sharing some underlying commonality (such as a grand average magnitude of change) shown through the common background or sky.

Absurdity is valued in many ways. Edward Tufte, in his work Visual Explanation, makes reference to  Mark Tansey’s 1984 painting The Myth of Depth, which shows many characters on a boat adrift amidst an ocean, while one, Jackson Pollock, walking, as if miraculously, on water, and, as if, stressing the title or the purpose behind the painting. In our context, once a sufficient definition of a background is worked out, similar phases from many different time series can be placed next to each other (much like the collection of trees in Figure 1c) against a common background triggering visual absurdities shown in Figure 5: neighbouring trees, though of potentially differing spreads or canopies, are expected to sway in the same direction much like they did on iFgure 3. This is contradicted when, even against a common backdrop (representing a pooled average change), the trees are employed to show not just differences between phases of the same series, but between pairs of corresponding phases of many.

When it comes to graphing time series, we believe it is not a failure of execution but a collapse of taste that brings an art to eclipse. The dashboards we offer showcase many alternatives, combinations and many ways of creating a tree. Each, through expressing motion, engineers a sense of reality. The backdrop changes. The sky brightens or darkens. A weak or a strong wind billows. The tree, swayed at times, however, stands resolute overall—a silent sentinel observing the switching of moods. And, faithfully, the foreground-background combination chronicles the unending skirmish between the fixed and the fugitive.

Acknowledgements

This work was supported, in part, by the Research Enhancement Grant awarded to the second author by the American Mathematical Society and the Simons Foundation for the 2024 – 2027 cycle. Bentley University’s summer research grant is also gratefully acknowledged.

The post Winds of Change and How They Sway Fatalist Fractalists appeared first on Nightingale.

As the saying goes, a picture plus a thousand words is better than two pictures or two thousand words. Here we consider two ways that words and pictures are combined on the page: statistical graphics (also called data graphics or information visualization) and comics (also called sequential art or bande dessinée). These forms of visual representation typically have different purposes—to inform or to entertain—and show up in different contexts, ranging from government reports to the comic books that formed the basis of Hollywood blockbusters.

In our work in statistics and social science, we have used data graphics for several decades in applied research and have also contributed to theory and methods linking graphical communication to statistical modeling. When it comes to cartoons, we are merely readers and fans, not creators. In learning about their history, we were struck by parallels to the history of data visualization (see Figure 1), and we also see some convergence between these two forms of narrative, now that information visualizations have become more prominent in advertisement and communication, and comics have come to be viewed not merely as a pop culture phenomenon but as a branch of literature.

Figures 1a and 1b. In one of the most famous political cartoons in history, James Gilray’s The Plumb-Pudding in Danger from 1805, William Pitt and Napoleon Bonaparte are portrayed slicing up the world (left). Florence Nightingale sliced up units of time in her compelling, although easily misinterpreted, 1858 statistical graph depicting the causes of avoidable deaths during the Crimean War (right).

Surprisingly recent histories

Cartoons and data graphics require nothing apart from ink and paper, yet only during the past two centuries did scientists and artists go beyond the basics to develop now-routine visual techniques for displaying data and stories on the page. The slow development of these media is interesting, especially considering that, unlike with cinema, for example, the basic technologies have always been accessible. Scientists, mathematicians, and accountants in earlier eras could have been understanding patterns in their data using scatterplots–but they weren’t. Artists and authors could have been combining words and drawings to vividly convey speech and action-–but they weren’t.

Time series and scatterplots seem ordinary to us, but as visual representations of information, they are highly abstract compared to such centuries-old schematic illustrations as geographical maps or anatomical drawings. Depictions of data have existed for thousands of years, whether pressed into Babylonian clay tablets or inked onto bamboo slips during China’s Qin dynasty, more than 200 years before paper was invented. Illustrations of data and mathematical concepts evolved over the centuries to exploit each historic advance in visual media—inks, papers, brushes, pens, printing, computers—as well as building on the diagramming innovations of other scientific disciplines. Michael Friendly and Daniel Denis trace the development of general-purpose data graphics in the 1700s and 1800s to earlier uses of quantitative displays in astronomy, where the positions of stars and planets in the sky can be directly mapped onto a two-dimensional space, as well as to depictions in mathematical physics.

Figures 2a and 2b. Engravings exploded in popularity in the 15th century, but were superseded by the newer technology of etchings in the 16th to 17th centuries. Jacques Callot was a master printmaker who invented techniques so that etchings looked cleaner, more elegant, and more precise , as seen in his depiction of Envy (left) in his Deadly Sins series from 1620. The intaglio printing of astronomer and mathematician Edmund Halley’s drawing of what may be the first bivariate plot from his 1686 book Philosophical Transactions (right; the vertical axis is barometric pressure and the horizontal axis is altitude) reflects the delicate and meticulous style of contemporaneous illustrations.

The connection between time series, scatterplots, and mathematical functions can be seen in graphs in which a curve is fit to go through data (see Figure 2b). It seems to have taken centuries for people to go beyond this to plot data that did not fall exactly on or close to a smooth curve; indeed, this happened at roughly the same time that Fourier and others generalized the mathematical concept of function to represent arbitrary mappings between spaces. In the later 1800s, innovators such as Charles Minard, William Playfair, and Francis Galton (Figure 3b) demonstrated the open-ended possibilities of revealing patterns in data through novel visual conceptions; since then, data graphics have been increasingly important in the natural and physical sciences. In the past twenty years or so, information visualizations have colonized popular communication as well, from New Yorker cartoons of worried executives staring at charts of declining stock prices, to time series of global warming and the flatten-the-curve graphs during the pandemic. Data graphics have followed a steady increase in abstraction of conceit and presentation, which has paradoxically allowed them to be accessible to a wider range of purposes and audiences.

Figures 3a and 3b. In Thomas Rowlandson’s 1808 cartoon The Corsican Spider in His Web (left), the geometric pattern vividly and accessibly conveys a political point. Francis Galton’s 1886 correlational diagram of the heights of parents and children (right) has a similar visual appeal but requires the reader to put in much more effort to understand the data and statistical relationship being shown. The increasing abstraction of statistical graphics allows more information to be conveyed; the subsequent establishment of graphical conventions has allowed readers to more quickly interpret the content of scatterplots and fitted distributions.

Humorous caricatures and satiric cultural commentary in simplified visual form have been found to have existed at least as far back as the ancient Romans. In the supposed Dark Ages, lively drawings that lampooned society lined the margins of illuminated manuscripts, while the first known bar graphs, drawn by Nicole Oresme in 1486, seemed to have gone largely unnoticed. Playfair reinvented them, along with conceiving the pie chart, a few hundred years later, at around the same time the concept of the cartoon was taking shape. Scholars trace modern cartoons and comics as an outgrowth of printmaking, with political humor drawn by Thomas Rowlandson, James Gillray (see Figure 4a), and George Cruikshank inspired by the French Revolution and packed with visual analogy. According to French-American cartoon historian Maurice Horn, perhaps the first to formally study the art form, “It was the universal acceptance of prints that led to the phased transition from caricature to what would later be called ‘cartoons,’ a form no longer devoted simply to cataloging external human idiosyncrasies, but one with an enlarged field of vision encompassing the whole political, social and cultural scene—indeed, the human condition itself.” These became staples of periodicals for the general reader at the tail-end of the eighteenth century and on into the nineteenth. It took a while during the first hundred years for the medium to evolve beyond successions of static images to the more fluid visual storytelling associated with turn-of-the-20th-century newspaper comics and then the longer-form stories appearing in the comic books, manga, and bandes dessinées that flourished in the mid-1900s.

Figures 4a and 4b. James Gilray’s 1793 political cartoon The Blood of the Murdered Crying for Vengeance (left) was a bestselling print in its time, as the public developed a taste for this genre; hundreds of cartoons were created during the French Revolution. William Playfair, once a spy for the French government who helped storm the Bastille, eventually settled into more sanguine vocations. He is credited with inventing the bar chart, Oresme having been long forgotten. Many of the foundations of statistical graphs were laid with Playfair’s line graphs, pie charts, and time series plots. Above, we see his graph of England’s trade balance with Denmark and Norway (right; from 1786), its artistry and annotations echoing the style of Gilray and his contemporaries.

Many consider school principal Rodolphe Töpffer to be the originator of the comic strip, having drawn cartoon-stories told in chronological series for the amusement of his students as early as the 1820s, later to be published to much acclaim. This artform was codified into box-shaped panels by Georges Columb (see Figure 5c), better known as Christoph. Töpffer and Columb were the forefathers of bande dessinée and of comic strips in general, along with other French innovators such as Emmanuel Poiré, a.k.a. Caran D’Ache (see Figure 5b) and Gaspard-Félix Tournachon, a.k.a. Nadar, as well as the German Wilhelm Busch and the American A. B. Frost. The realism and refinement that developed with those genres matured into the albums and graphic novels of today.

Figures 5a, 5b, and 5c. The Swiss-French polymath Johann Heinrich Lambert called his surprisingly modern-looking diagrams figuren, and seems to have been the first to create lines of best fit, as seen in this graph of temperatures at varying latitudes from 1779 (left). The elegance of his rendering heralds the whimsical clean lines of the pioneers of bandes dessinée, such as French satirist and illustrator Emmanuel Poiré, better known by his pseudonym Caran D’Ache. His Le Rêve de M. Emile Zola​​ (center) was published in Le Figaro in 1889. Georges Columb, known as the children’s magazine illustrator Christophe, packed rectangles with painstaking detail, creating multiple-panel stories and establishing the visual grammar of comic strips. This image from 1893 (right) was part of a recurring series called L’Idée fixe du savant Cosinus published in Le Petit Français illustré.

As with statistical graphics, we are struck by how recently some of these developments arose: just as the capacity for scatterplots was available long before they were regularly made, so there is no reason why Tintin-style storytelling with rapid transitions and speech balloons could not have been done hundreds of years earlier. 

Heinz Pagels tells the story of “a stranger, who, recognizing Picasso, asked him why he didn’t paint people ‘the way they really are.’ Picasso asked the man what he meant by ‘the way they really are,’ and the man pulled out of his wallet a snapshot of his wife and said, ‘That’s my wife.’ Picasso responded, ‘Isn’t she rather small and flat?’” The relevance to our discussion here is that scatterplots, time series, speech balloons, and other tropes of statistical graphics and comics are so familiar that readers can see through the abstractions, as it were, in the same way that the husband on the train saw the photo not as a flat artifact but as a representation of a three-dimensional person. The story dramatizes that the difference between a cubist collage and a photorealist painting is not so much the level of abstraction as the familiarity of its conventions, and indeed it can take a generation for abstractions to enter the mainstream sufficiently that they can be built upon by new creators.

Statistical graphics came to maturity as a result of the mathematical use of Cartesian coordinates to represent dimensions other than physical space (see Figure 6b), along with probability distribution for variation that allowed real-world data to be represented by non-deterministic models. The rise of sequential comics coincided with the advent of film as a popular and artistic medium. Graphs and cartoons exist for entirely separate purposes, and so there may be no direct parallel here except a recognition that in science, policy, or entertainment, developments in different media feed off each other. The effectiveness of film opened the door to dynamic forms of visual storytelling on the page and in animation. Technologies of reproduction affect the forms of popular art, from printmaking in the 1700s to mass-circulation magazines and newspapers in the 1800s and 1900s, to movies and television today. Similarly, advances in mathematics and computing have turned statistical graphics from craft work into a set of routine tools in science and communication.

Figures 6a and 6b. Winsor McCay experimented with the form of the full-page newspaper comic strip with Little Nemo in Slumberland from 1905 until 1927 (left; this example from 1905). With exquisite draftsmanship, he frequently subverted the constraints of the strip’s panels. Within the same historic time frame, our understanding of atomic numbers was usurped in Henry Moseley’s graph of High Frequency Spectra of the Elements from 1913 (right). This visualization made clear that increases in atomic mass correspond to a physical property, correctly supposed by Moseley to be the number of electrons. Its lines foretold three then-unknown elements and that electrons hold a mysterious property, later discovered to be spin.

But even as they historically evolve at what seems like a yawning parallel distance, we may notice reflections of method and design between data-oriented graphs and cartoons (and the related illustrations that preceded their inventions) depending on the era and trends in artistry, as may be observed in the comparisons in our appendix. This points not only to contemporaneous conventions, but to the similar constraints required to deliver such abstractions as mathematical concepts and humor. That which is more comfortably communicated in written or spoken form (sentences or equations) is conscripted into a visual format built from the media available at the time.

Outsiders entering the mainstream

Statistical graphics and cartoons both have the feeling of “outsider art,” with an uneasy relation to more accepted forms of data analysis or storytelling. This may perhaps be most apparent when considering the visual outputs of such outsiders to the mainstream as sociologist W. E. B. Du Bois (see Figure 7A) and the Creole artist George Herriman (see Figure 7b), with his aslant artistry and humor that featured a genderfluid cat.

Figures 7a and 7b. W. E.B. Du Bois, who established the first American school of sociology at Atlanta University, created a series of boldly colorful and geometric graphs depicting a social study of Black life in the U.S., exhibited in 1900. The above example (left) depicts the “proportion of almshouse paupers in every 100,000” Black citizens. George Herriman’s Krazy Kat (right), which ran from 1913 to 1944 (this example is from 1942), was groundbreaking not only in its audacious design and narrative, but also in that Herriman was a Creole artist of national importance, and that his character Krazy was unequivocally genderfluid.

The meat of a scientific analysis or policy report will typically involve some mathematical modeling, with graphics being used for exploration or communication. There is a general recognition that exploration is a crucial component of learning from data, and communication is necessary in all areas of science, technology, and decision making—but graphics have traditionally been seen as less of a science and, at best, a form of practical art. Only recently have exploratory data analysis and visualization been formalized as part of statistical workflow; this has come during a period in which statistics has combined with data science and machine learning into a field in which computing is as important as mathematics. Visualization has moved closer to the mainstream of science.

Meanwhile, the role of comics in popular and literary culture has changed several times since 1900, moving from disposable newspaper strips, to wildly popular entertainment for children in the form of comic-book and television animations, to become a form of genre literature and, more recently, source material for popular movies. Commercially this has been a series of ups and downs, but from a cultural perspective, comics have followed the paths of crime fiction and science fiction into literary respectability. As with these other genres, comics retains its own insular culture along with some outlaw mentality.

In their modern forms, comics and statistical graphics both lean on conventions, some of which have become so familiar that they feel nearly invisible. For example, we take it for granted in Western culture that a time series runs from left to right, that comics run from left to right and from top to bottom of the page (except when they don’t), that the horizontal axis on a scatterplot represents a predictor and the vertical axis represents the outcome, that the wedges in a pie chart add up to 100%, that a “pow!” exploding with stars conveys a painful punch in the face and that overlapping speech balloons convey interruption, and so on (see Figures 8a and 8b). These conventions can sometimes overwhelm legibility, as with the popular but notoriously difficult-to-read parallel-y-axis plot or baroquely hyperkinetic superhero fight scenes. As with genre literature, reliance on conventions facilitates new developments for insiders that can baffle readers who are unfamiliar with the form, which in turn motivates the sorts of swings between sophistication and simplicity that are characteristic of the history of popular music.

Figures 8a and 8b. We understand the motion and pain of the frog from the conventions of simple lines, swirls, and stars in this 1945 edition of Ernie Bushmiller’s comic strip Nancy (left). Likewise, the spare, unadorned presentation of points and lines conveys the covariation in the 1958 plot by Alban William Phillips (right), which efficiently depicts a historical relationship between inflation and unemployment in a now-familiar format of data and fitted curve, while at the same time arguably being misleading due to the convention that a scatterplot represents a causal relationship.

Mathematics, too, has advanced through the use of conventions, such as Leibiniz’s notation in differential calculus, or even more basic ideas such as the expression of mathematical reasoning in equations rather than words. Just as we can read an English sentence without needing to be aware of the individual letters in the words, and we can follow basic algebraic expressions without needing to puzzle over the meaning of the equals sign, we are accustomed to time series plots and sequential panels speech balloons and can see through these forms directly to the stories and data being conveyed.

Differences between these two modes of visual storytelling

Comics have been used to teach statistics, and data graphics have been used within comic strips; quantitative visualizations can be beautiful and comics can be informative. But these two forms of expression are generally used in different places and with different goals: explanation and mathematical understanding in one case, art and entertainment in the other. And yet, both have in common a mission of delivering an abstract concept efficiently within the constraints of their inherent structures, requiring such conventions as economy of line and messaging that registers intuitively for the reader. 

Different goals lead to different visual priorities: clarity in data graphics is absolutely necessary if any useful information is to be conveyed, whereas ambiguity in comics can help create suspense, point of view, and other dramatic effects. Comics, as with purely literary stories, typically follow a narrative structure—or, if not, are consciously operating in opposition to conventional narrative. In contrast, statistical graphics on the page are often static, taking the form of a single display rather than a sequence.

The content of single-panel cartoons and statistical graphs require a short but concentrated effort by the reader, while the sequentiality of comics leads to a much different reading experience. Most comics, like most films and works of literature, offer a guided reading experience, a sort of theme park ride in which the reader follows a story through a sequence of panels: in addition to providing the words and images, the authors dictate the structure and pace of the narrative. In contrast, when reading a time series or scatterplot, we perceive a general pattern and then can then focus on individual segments or points. When a graph is constructed as a trellis, or grid of small multiples, this just adds one more level for the reader, who can now slide up and down between individual points, subgraphs, and the entire picture. Indeed, we would argue that the sequentiality of comics and the all-at-onceness of statistical graphs are fundamental characteristics of these forms.

An early example of a small-multiples graph is Francis Amasa Walker’s state-by-state “gainful occupations” grid of 1874 (see Figure 9a), which appeared a century after William Hogarth’s groundbreaking series of prints of the Rake’s Progress. To the extent that each of Hogarth’s scenes is itself a detailed storyboard, the sequence as a whole feels less like a comic strip or bande dessinée and more like a sequence of static images.

Figures 9a and 9b. Francis Amasa Walker’s 1874 small-multiples graph lays out the ratios of those above the age of 10 who were employed or in school in the U.S., with each box representing a state (left). Hergé’s 1932 Tintin en Amérique (right) is similarly divided into discrete panels but, unlike the statistical graph, is intended to be read in order so that it forms a narrative..

A modern comic can be drawn beautifully, but its individual panels are directly read as part of a story rather than as individual tableaus. With statistical graphics, it is the opposite. News organizations now sometimes construct interactive data visualizations that explicitly guide the viewer, but to the extent that graphics support exploratory data analysis, it is often essential that the reading experience be open-ended and not directed by the creator of the graph.

Somewhere in between are dynamic scatterplots such as those developed and popularized by Hans Rosling, in which each circle represents a country and the graph refreshes for each year, with movement of the circles showing changes over time (see Figure 10a). From the audience’s perspective, this sort of “movie” is more of a guided tour than an open-ended exploration. It becomes an exploratory tool when the user is given the power to stop the motion of the image and look around, and to select what variables to display. The creation of animated graphs in open-source software such as R or Python facilitates both analysis and presentation when it comes to machine learning and is becoming standard with the younger generation of data-crunchers, and flowing geometries can be beautiful.

As discussed earlier, data visualization and comics both rely on conventions that serve as shortcuts to legibility. The establishment of conventions also gives the opportunity to push back against expectations, whether it be poems that don’t rhyme, machine-made art, neo-noir film, countercultural science fiction, or comic books and bandes dessinées such as Maus, Watchmen, and the Spirou of Émile Bravo that use traditionally genre materials to tell more serious stories. We see less of this sort of reaction in statistical graphics (setting aside jokes such as pie charts representing actual slices of pie or gimmick graphs such as bar plots showing the heights of buildings).

Figures 10a and 10b. The Swedish physician Hans Rosling developed and popularized the Trendalyzer software system that facilitates dynamic scatterplots that animate sequentially across time (left). Since the late 20th century, Chris Ware has been innovating comic strips and graphic storytelling with designs that sometimes resemble charts or technical drafting, as in this example from 2010 (right). He often bucks the conventions of temporal order with narratives that are chronologically shattered.

The arrow of time

A detective story will typically involve two time sequences: the forward sequence of (a) the motivation for the crime, (b) the planning of the crime, (c) the crime itself, (d) the aftermath, (e) the arrival of the detective, (f) the collection of clues, (g) the discovery of the solution of the crime, and (h) the unmasking and punishment or escape of the criminal. But this is not quite the sequence given in the story, which will typically follow an order such as d, e, f, c, g, b, a, h. These two different sequences roughly correspond to the processes of data generation and inference in statistics. Data generation goes forward in chronological time, while inference starts in the middle and goes back and forth in time.

The strict ordering that is typical of comics (setting aside experimental work such as that of Chris Ware; see Figure 10b) implies that some decisions need to be made about the sequence by which the story is experienced by the reader. In contrast, a static graph that appears all at once can imply different stories, depending on the order with which it is read. The title and caption of a graph can thus have a strong effect on its meaning, in the same way that point of view is important in storytelling.

Looking forward

It took a while for the methods of data visualization to detach from their original sources in mapping, astronomy, and economic and demographic time series; similarly, sequential art was slow to move into new domains beyond reportage and humor.

Both fields feature a series of technical developments that have facilitated communication through juxtaposition. A time-series plot contains no more information than a series of numbers, and a scatterplot is just a way of displaying a two-column table—but graphics allow visual comparisons in a way that the numbers do not. Similarly, a political cartoon or a single-panel gag employs a discrete, often uncomplicated tableau of squiggly ink lines and perhaps a splash of color to communicate the many layers of meaning that make up a joke or a sharp commentary. A sequential cartoon, in contrast, can be thought of as an annotated series of images or as illustrated prose, but it is more than either of these. In a graphic narrative, the forward progress of the story is governed by the architecture of the content flowing panel to panel. Advances in statistical graphics and comics have come from ever-evolving conventions such as grids of scatterplots and strips of panels, which represent conceptual leaps and in turn open the door to further developments.

At the same time, historical contingencies and the imperatives of commerce can lead to developments that are inherently unpredictable. To think of comics as a set of variations on the superhero form would be as limiting as to consider pie charts and histograms as the building blocks of statistical graphics. Superheroes, pie charts, and the cozy detective story are examples of subgenres that have taken up too large a space of their genres in the popular imagination, motivating strong reactions against these forms among authors and designers. When the goal is communication–whether to convey information or to tell a story-–there is a tension between the convenience of existing popular forms and the need to innovate to shake readers out of existing modes of thinking.

It has taken applied researchers a long time to realize that graphical visualizations of data and models are not just decorations to be added to make statistical results more accessible to lay readers; rather, they are a necessary part of any serious quantitative analysis. Similarly, the techniques and conventions of cartoons and comics are not just a way to make jokes or stories more accessible to children, any more than movies are just filmed books. In the famous words of Marshall McLuhan, the medium is the message.

Recognizing these historical parallels can point to potential future developments. We are in no position to say where comics and bandes dessinées, or literary or visual art more generally, could or should go next. But we can comment on something that statistical graphics can learn from comics, which is how to add some structure to the viewing experience. It should be possible to design graphs to support discovery of the unexpected without entirely leaving readers on their own during the process. This is especially a concern with big data: when the dataset is large and complicated enough, even an attempt to visualize all the data at once will require some choices. One way to approach this is to construct a sequence of graphs, starting with the big picture and then focusing on details. It can also help to accompany a graph with text suggesting how it is to be read, perhaps with further explanation using a sequence of images or a video. Shneiderman offers similar suggestions for computer-user interfaces, which is what data graphics are nowadays. A certain amount of storytelling or imposed structure can be necessary in the interpretation of data, just as we often need to embed real-world events into narratives in order to understand them.

The post Statistical Graphics and Comics: Parallel Histories of Visual Storytelling appeared first on Nightingale.

We both work full-time and raise two young children, so our evenings are the only moments left for connection. He is the hardest-working person I know—relentless in chasing a dream that has recently taken shape as a tech startup. His quietest, most productive hours begin when the kids sleep. I, on the other hand, go to bed alone.

At first, I texted him, “Good night.” He would reply hours later, sometimes with a simple “coming to bed now.” But after two months, I realised something: I had data. It’s not formal, tidy data- casual timestamps, scattered messages, and a discernible pattern.

So, I started tracking it deliberately. A spreadsheet. One column for the time I went to sleep, one for him. If I forgot to log it, I would ask him in the morning: “When did you come to bed?” He always knew I was tracking his bedtime time. It comforts me. Data gives shape to ambiguity. It holds stories quietly, waiting to be told.

There is a particular calm that comes with collecting data. For me, it is a form of emotional processing. Tracking our sleep was something small I could hold onto. A private system of meaning. Each entry felt like a whisper to myself: “You are paying attention”.

As the dataset grew, so did the emotional tension. We argued. Not because I didn’t support him, but because I missed him. Missed us. One night, I added a third column: fights. Eventually, a pattern emerged. We always argued at the start of the week after the weekend reset. I brought it up, along with the spreadsheet, and together, we coined the Sunday-In-Sync Rule. Once a week, we would meet in the middle. He would wrap up earlier, and I would stay up later—a small act of reconnection in a sea of drift.

At one point, I realised that the actual times we went to sleep didn’t matter as much as I thought. What mattered was the delta—the difference between them. That delta became the emotional signal, more about how far apart we were. From that moment on, I shifted my perspective on the data and how I wanted to visualise it. Each petal would represent a single day. My sleep time became the baseline. His would be expressed as a distance—the space between us.

Five months in, the dataset had become too complex to keep in rows and columns. I started sketching. One flower for each week. Some flowers bloomed, representing nights spent together, even if late. Some wilted, marking long gaps between bedtimes. A visual garden of our sleep patterns emerged. A bouquet of data is carefully drawn.

Why flowers? Because they are the most classic romantic gesture. A universal symbol of affection, apology, and devotion. I was not just visualising data; I was creating a love letter. One that said, “I see you. I miss you. I’m with you”. Presenting the data as a bouquet felt right. It framed the tension with tenderness.

Drawing it brought another kind of clarity. Unlike the spreadsheet, which was linear and clinical, the floral format made space for nuance and softness. It became a way to honour the emotional weight of these minor, repetitive, daily differences. A quiet ritual that helped me come to terms with what I could not control.

When I showed him the finished visual, he was speechless. Then smiling. Then laughing. He told me I was the most supportive wife he could ask for. Later, I found out he had shared the piece with his friends. That meant everything to me. He saw what I saw.

Is it a visual love letter? A quiet protest? Maybe both.

What I know is this: in tracking our distance, I found a way to feel closer. I supported him in silence, witnessed his effort, and honoured the rhythm of our parallel dreams.

The post A Garden of Sleep: Tracking the Emotional Distance Between Two Bedtimes appeared first on Nightingale.