Representing Patterns
Crafting stories with data
The inclusion and exclusion of information and its organization can have a great bearing on how the content is perceived. By determining the stories that are told and the prominence of points conveyed, designers hold great responsibility in communicating information to audiences, which must not be taken lightly. In this project, we will investigate relationships inherent in data and explore responsible approaches to communicating them to specific audiences. Representing data through graphic, written, and temporal forms, we will strive to help viewers engage in, and think critically about, the topics we present.
Communication Design Studio | Week 11 | 4th November 2024
“A designer’s greatest power lies in how they choose to deliver the story.”
As I perused the datasets presented to me, I worried about finding the “right” dataset that would help me create the ‘perfect’ story. Which dataset would reveal the most awe-inspiring insights that would cause sharp intake breaths of realisation among the general audiences and would get designers’ minds cracking about creating approaches to address the issues presented here?
I gravitated towards data related to Health, and decided to explore further about the different aspects of Health that were covered. Mental health — in specific, has always been an area that has piqued my interest, mainly because of how it is often hidden in plain sight.
The idea was to present our datapoints in the form of a cohesive and coherent story through the introduction of ‘Scrollytelling’- a method that incorporates visual and textual elements to help a reader understand a concept as they scroll through it. ‘Scrollytelling’ helps in engaging a reader to the story being communicated through enabling interactive data exploration and by breaking down stories into bite-sized, interactive pieces for easier consumption and faster comprehension.
Shortlisting the data set
Sifting through the different datasets- from exploring prevalence of mental illnesses country-wise and age-wise to exploring public perceptions towards how important it was for national government to fund research on anxiety/depression, I felt myself becoming more and more fascinated with people’s perceptions towards mental illnesses such as anxiety disorders and depression.
A Personal Connection: Mental Health in India
As someone from a country that is quite conservative and opposed to openly discussing matters of mental illnesses, I found it interesting to compare and contrast questions relating to country-wise data. By using filters, I tried to look into certain representations of data through the lens of data from India vs the United States.
Below are screenshots of some of the representations that caught my eye.
As I look at these charts (a small set), I ask myself a bigger question — How aware are people about mental illnesses? And in what ways does that affect their perception of the topic and their ability to answer questions?
Combining Human Stories with Machine Data
Some of the information design examples shown in class last Thursday gave me glimpse into how people were storytelling with data. One of my most favorite peices was an immensely moving personal journal that recorded key events of living with long COVID. The layering aspects — date, time, symptoms, intensity — helped me resonate with how stressful it might have been for the author (also the patient experiencing symptoms of long COVID).
I aspire to layer from multiple resources to help tell my story.
Some of the key questions that I am exploring at this early stage are:
- How has anxiety and depression affected people in both these countries over the last year?
- What have they done to address it?
- Perceived level of comfort speaking about anxiety/depression within community
Communication Design Studio | Week 12 | 11th November 2024
As I sift through my data, I’m also reworking on my script to better explain the story that I want to convey.
My current data points seems to be showcasing the following story:
- Self reported anxiety in individuals
- Reported that they knew someone they know with anxiety/depression
Both these data sets will help set the context to show how prevalent anxiety/depression is around the world. I’m also taking into account a single year — 2020 as my current time period to situate my problem in.
3. Ages of individuals when they first experience anxiety — based on income of countries (India vs the US)
I’m trying to explain that the prevalence of anxiety and depression doesn’t limit to borders and specific high-income vs low-income countries but rather show how early do feelings of anxiety/depression manifest irrespective of any country
4. Different ways people are dealing with anxiety and depression around the world to show a country-wise view on what specific ways/strategies of dealing with anxiety are popular across these countries
The AHA moments in my story:
5. Perceived discomfort speaking about anxiety and depression to people they know to show the hesitance of discussing mental health struggles
6. Individuals who reported that they talked to friends or family to make themselves feel better to hint at how discussing mental health challenges within your social circles can lead to successfully addressing anxiety and feeling better.
I am planning to close my story by discussing about the power of social connection, and focusing on the need to build a “culture of communication” which will help people forge better social connections, deeper relationships and combat anxiety and depression.
An in-class exercise and the weekly assigned reading of Yau (2013) helped orient us to the different ways of deconstructing and reconstructing data for organization.
After breaking up my story into individual post-its, we then tried to identify where there would be a linear structure vs an indexical structure. Using SCALES and the LATCH principles, we then identified which coordinate systems would suit our story the best.
At this moment, I feel like the Cartesian Coordinate system suits my data set the best.
Since my data mostly involves reporting “percentages of people responding”, age-wise, and country-wise distribution- I see my scales being a mix of categorical and linear/hierarchical.
Communication Design Studio | Week 13 | 18th November 2024
My current rationale behind organising my data is to mainly show the following:
In the year 2020,
- how prevalent anxiety is around the world, by either people experiencing those symptoms, or knowing others who do
- contrast of how early-on anxiety sets in irrespective of country— in US and India
- a comparison of the ways in which people deal with anxiety in US vs India
- comparison of perceived levels of discomfort speaking about it in US vs India
- contrasting data that shows how speaking about it helps people address and overcome the symptoms
By creating rough sketches for ways in which I can show my data, I’m exploring different ways my data can be presented to convey a coherent story. For the very first dataset, I want to set the stage by stating that “19% of the world’s population self reported anxiety or depression symptoms”.
The images below give an idea of the different kinds of ways I’m thinking of layering data while organising them.
Challenges so far in the process:
- identifying if the buckets in ranges need to be equally distributed or not. For eg: Around 80% of respondents in both countries overcame symptoms by speaking to others. In this case, if my range buckets are incrementing by 10% each — like 0–10%, 11–20%, etc.. — it doesn’t effectively convey the difference between the countries — since US is at 80.9% and India is at 82.8%. The finding here is too similar to be called a difference but in other areas, I would like to show a gap between them which I’m unable to achieve with equal distribution in range buckets.
Communication Design Studio | Week 14 | 25th November 2024
Speed Dating for clarity
A speed dating activity session yielded me the much needed insights and ideas on visualising different ways of representing my data. One of my reviewers and classmates, Jean Chu questioned and pointed out if the bar chart representation was really necessary especially when the actual percentage in itself was the hero. However, when placed within a scale of 100, it seemed quite small. For example: showcasing that 19% of the respondents reported they were feeling anxious as a red mark that took up less than a quarter of real estate in my graph, wasn’t exactly conveying how important the number was. Furthermore, she pointed out something that I had completely overlooked. “Your data says 32% of people knew family and friends suffering from anxiety. So they might know one person, two people, three people suffering from anxiety. These people who are suffering, have not taken this survey. So the number of the sufferers is so so much more higher than you can imagine. Am I right?”
I walked away from that session with the clarity that I needed to rework on my visual representation.
“I urge you to explore more and embrace your design flexibilities.”
-Stacie Rohrbach
Another critical feedback to come out of Monday’s speed dating session was Audrea’s question about if there was a way to represent culture through my visual representation. Since I was planning on comparing and contrasting how anxiety was addressed among India and the US, Audrea wondered if there was a way I could bring that element into my visual storytelling. While I am not clear about how I’d want to bring this into my storytelling — I am deeply appreciative of all my three peer reviewers and the feedback given by them.
“Move beyond bars and pies”
Wednesday’s session was primarily a working session with students signing up with slots to receive feedback on our work with Stacie and our TAs.
I expressed my desire to move away from using bar graphs and Stacie helped me visualise different ways of doing so. She also brought in Yau’s visual variables to help communicate the crux of my story through each frame. A picture of the feedback session with Stacie is attached below.
A quick sync-up session with out TA, Vanya was the catalyst in having me completely redo the visualisations.
Get Nikita V’s stories in your inbox
Join Medium for free to get updates from this writer.
Abandoning bar charts
“If it’s anything that can be created by a software, it shouldn’t be a part of your visualisation”.
-Vanya Rawat, Teaching Assistant for CD Studio
Vanya encouraged me to re-look at my data, and explore different ways of visualising it.
Creating a new visualisation for showcasing Anxiety in the world, in the US and India
Vanya’s conversation led me to move away from using bar charts, and instead had me focus on the number — the 19% of self-reportees, 6.8% of diagnosed patients, etc. This also leans into the feedback that was given to me by Jean about the number needing to have a higher predominance over its relativity to the absolute whole.
I decided to break up the visualisation for the first act of my story — where I discuss people with diagnosed anxiety, self reported anxiety, knowing others with anxiety and the potential people with anxiety and represent it.
While the moodboard built by me was mostly an amalgamation of dark greys and muddy greens and blues, I decided to use a red color to showcase anxiety.
The screenshot above gives a glimpse into how I had visualised my categories (diagnosed self-reported, knowing others, and potential undiagnosed, and unanswered cases) but I decided to create a legend in Illustrator and test it out with someone who had no connection or familiarity to the topic, my process and my progress on the topic so far.
Rapid user testing to finalise on visual direction
Below is a screenshot of my friend from India, who was gracious enough to hop on a call and test the illustrations out for me. I had created a simple Match game where I created the illustrations for the four categories and wrote down their descriptions, and asked her to match the written descriptions to the visual representations that made the most sense to her. Surprisingly, her feedback contradicted two of my initial assignments. I had wanted to assign a circle with reduced opacity to self reported respondents, and a dotted circle to those who knew friends and family suffering from anxiety. However, Moukthika interchanged those two when “matching” them. Her rationale is listed below:
- For depicting the respondents who were diagnosed with anxiety, she chose a dark red, solid circle because according to her that is clear and verified, and has been identified.
- For depicting the respondents who self-reported anxiety, she chose a dotted circle, because she mentioned that they might be experiencing it, reading about it, feeling it, and yet struggle without knowing if they struggle with anxiety or not. There are periods of lack of clarity, which could be conveyed through solid dots that made up the circle.
- For the respondents who knew friends and family with anxiety, she mentioned that it is one of the toughest things to watch your loved ones struggle with something that you cannot directly alleviate. So for respondents who had loved ones struggling, although they themselves did not have anxiety, they were still feeling it as an extension of the struggle of their loved ones- so they were affected as well but probably not as much as the ones feeling it.
- For the idea that there might be so many others out there who’ve gone unnoticed, undiagnosed, the suggested a ring with the red stroke — that would serve as an indicators that it could change, an “alarm” that could go off anytime anywhere, and they could speak up about their anxiety anytime. But for now their voices lie dormant.
Communication Design Studio | Week 15 | 2nd December 2024
The beginning of things feeling wrong
I’d built out all my assets in Adobe Illustrator and was ready to input them into Shorthand along with my writing. After putting placeholder text on Shorthand along with corresponding visuals, my initial draft began looking something like this.
Feedback from Stacie and Vanya
Stacie’s feedback was around the text coming up first and then being followed by the circles, which made the reader completely skip over the connection between the circles and the text. My response to that was to revise the layout of the visualisation such that the circles appeared on top of the page, and each scroll resulted in a description of what had already appeared above (converse of how it looked at the moment).
Inspite of testing this out with a potential reader, I found myself wrestling with the visual style. Stacie and Vanya’s feedback forced me to think about it some more. I’d built out most of the assets on Illustrator, just to be told that while all four categories felt and looked different, they looked too different. If these were to represent the four categories of anxiety, how were they related to each other. Was it the texture that was differentiating them from one another? Or was it the opacity? As the maker of these visuals, I myself wasn’t entirely clear about them.
Even so I pushed myself to complete most of the assets on Illustrator with the belief that things might start looking better once I built them all out. Spoiler alert: they didn’t make me feel better. If anything they made me feel even more strongly against the current direction that my visualisation style was taking.
I’d built out the remaining statistics which discuss how early people were experiencing anxiety across two countries — US vs India just to make a point in my story about how anxiety is prevalent irrespective of the income of the country.
A bigger problem
Inspite of building out these assets, for some reason, I couldn’t bring myself to like this data visualisation. As a fellow human being, who was recently diagnosed with Generalised Anxiety Disorder (GAD), I felt like I wasn’t doing justice to how anxiety felt.
How does Anxiety feel? How does Anxiety look?
Stacie and Vanya were excellent sounding boards who worked with me over the week to help me — a. identify what in my story wasn’t making sense (answer: the visual style), and b. what in specific, about my visual style turned me off.
One session with each of them led me to realise that the reason I felt demotivated and disappointed with the current direction of my visual style was because that is not how my anxiety felt. As someone who experiences anxiety attacks regularly, I knew for a fact that anxiety was not clean. Anxiety was not contained within boundaries.
This realisation was a result of a conversation with our TA Vnaya, who pointed out how when anxiety, her head was full of tangles.
THE BIG PIVOT : My “AHA” moment
As a designer, I’d always fantasised about finding the big ‘AHA’ moments. But never quite experienced it. Walking away from that conversation with Vanya, I quickly sketched out on paper a set of storyboards for the first half of my story, but instead of clear cut red dots, I made tangled balls, knots of anxiety, to represent the various kinds of anxiety that my story depicted. A deeply dense and tightly wound knot, could mean diagnosed anxiety, while a slightly looser, yet densely wound knot could mean self-reported anxiety. A loose, barely tangled knot could mean people who knew friends and family who were anxious. An almost shapeless, fluid loop could indicate potential number of people whose data we didn’t have.
When I presented my storyboard to Stacie, along with my rationale on changing the visual style I could see that she was just as excited as I was. But with only a week left before final submission, I wasn’t sure if I could re-start and finish all my illustrations in time.
“I think, you can do it. You should do it.”
Hearing Stacie say these words when I mentioned my skepticism about not being able to finish in time, gave me the required courage to just start with the illustrations. At this point, I was feeling extremely strongly about the visual direction of my story, and felt this is how I could best do justice to what I, and the innumerable people through my story felt. I spent a few hours pulling up mood boards on the visual style for this new direction that I was taking and began building out the required assets.
Initial designs
I began redoing my designs on Procreate and started decided to only use texture as a key visual variable when describing anxiety, in the first half of my story.
Stacie and Vanya both seemed to like this direction, and encouraged that I continue working on the same.
Vanya’s feedback suggested that I retain the handrawn aspect of the data points, yet to make sure that all my data points within a single category still appeared a same. For instance, she pointed out how in the third row, the first few scribbles/knots which were all part of one category (self-diagnosed anxiety) yet had different textures. It was also not advised to keep individual knots in different sizes. Since, size was a key visual variable according to Yau (2013), I needed to steer clear of using it as a component to describe one component unless it was absolutely necessary. Given how it was a data visualisation project, it still needed to maintain uniformity within its categories.
Stacie’s feedback raised some concerns for me specifically, when she mentioned that people who are experiencing anxiety (self-reported) might also know friends and family who are experiencing anxiety. Both these data sets may have overlaps within them and therefore, the current representation isn’t valid. I needed to show each of the visualisations as parts of a whole (individual percentages out of 100%).
Communication Design Studio | Week 16 | 9th December 2024
As part of the last leg of my CD studio, I completed illustrating all my assets (for the third time!) on Procreate. As I edited and changed a few illustrations from before on Shorthand, I built out additional assets that I knew I needed to set the scene. I was keen on finding a way to use Scrollpoints (a theme effect provided in Shorthand), which would allow me to zoom into specific points of an image to show detail and nuance, and zoom out to show this detail became part of the bigger picture. However, given the time crunch, I was on I was unable to complete it on time.
I prioritised on building out all the illustrations for my story, and published them on Shorthand.
This is my prototype video of my Scrollytelling on Shorthand!
Anxiety categories — represented through texture
I used texture as my key visual variable when representing the various categories of anxiety through my story. My rationale behind using texture to represent categories of anxiety stemmed from the fact that I needed to showcase how intense anxiety was for different people. As someone who has been diagnosed with anxiety, this is very real, and intense, so I tried to showcase that through dense knots. For people experiecing anxiety — self -reporting their anxiety symotoms, I wanted to give weight and honor how serious it was for them, especially without having a diagnosis for it. At the same time, given the ambiguity in data around if it was actually anxiety or sporadic moments of panic, I chose to depict the self reported anxiety category with tightly woven knots albeit not as dense as the estimated people with anxiety disorders around the world.
Countries — represented through position
I’ve used a cartesian coordinate system for all of my data visualisations, and although a part of my story does contain comparison between the US and India, I managed to showcase comparison between the countries through position, by placing the data visualisation for USA on the left (indicating east) and India (indicating west) to give viewers a relative representation of the countries through their position.
Highlighted findings — represented through density
For data points that needed more attention, for example: an instance where I compare USA vs India’s comfort speaking about anxiety, as an example to draw readers attention to one data set over the other, I drew attention to that by increasing the opacity / thickness of the data points in the visualisation.
Key themes — represented through color
Towards the end of my story, I had only 3 key categories: perceived comfort to speak about anxiety with friends and family, dealing with anxiety through seeking therapy, and speaking to friends and family to alleviate anxiety. These three categories led to my AHA moment, which would then set the stage to talk about social connection. For these three categories, I introduced color.
The key reason behind choosing color to depict these three categories was also that these three categories hint at the need for social connection. Social connection, has the potential to address and alleviate anxiety. I was trying to position it as the antidote to anxiety. So where anxiety was scribbles in black and gray, social connection was scribbles with life- pops of color. Therefore, it felt befitting to use colors to describe social connection.
Reflections
- My biggest takeaway with this project would be to lean into my instincts and sit with my discomfort. Looking back, I feel like I let go of precious days because I was anxious to identify and discover what about my previous visual style made me uneasy. I knew what was wrong, I just didn’t have the courage to face it and re-do it. I am so so happy that I took the plunge and redid the entire visualisation in a way where I could portray how living with anxiety feels like.
- The joy that I felt when I shared my work with the rest of the people in Design week. I was caught up in the moment and couldn’t take any pictures, but my I vividly remember the way people reacted to my article and the visceral reactions my illustration evoked among my audience. Our professor Bruce Hanington remarked that ‘showcasing anxiety through knots, humanised the data — which is so refreshing and unlike anything he has ever seen”. This made my day, and gave me the confidence that I couldn’t bring myself to own up. My classmate Yujin, remarked how the illustration style in itself was giving her anxiety. As a designer, I believe this is where I felt that I had done a good job with my visualisation. I was heavily inspired by Giorgia Lupi’s 1374 days with COVID — a personal opinion piece in the NY Times, which was extremely personal and equal parts poetic. I aspired to design a piece that evoked feelings within people not only through text, but through the visual design in itself. I am grateful that this project helped me do just that.
- Next steps: I am now convinced that I am capable of data visualisation. With this knowledge, I know plan to delve deeper and identify the triggers of my anxiety and bring that to the fore by conducting a self ethnographic study on myself. I have begun to document my anxiety attacks, what I do consciously to address them in the long term(through diet, exercise etc.) and in-the-moment relievers (abdominal breathing, walks, etc.). I plan to study myself for a 100 days and create and share a story around it. Hopefully, it will help people identify their triggers and find ways to address them.
Note: This writeup is part of an ongoing project of the Communication Design studio course Carnegie Mellon University. I will continue to update and refurbish the writeup on a weekly-basis as I learn new things about my topic.
References:
IHME, Global Burden of Disease (2024) — with major processing by Our World in Data. “Anxiety disorders prevalence” [dataset]. IHME, Global Burden of Disease, “Global Burden of Disease — Mental Health Prevalence” [original data]. Retrieved November 3, 2024 from https://ourworldindata.org/grapher/anxiety-disorders-prevalence
Yau, N. (2013). Data Points: Visualization That Means Something. Germany: Wiley.
