Communication Studio Project 3
After looking over the data, I’m seeing an opportunities to ask questions about food access as compared to neighborhood characteristics, such as percentage of renters, age distribution, income level, and health indicators. There could be interesting insights into how food access is different between rural and urban areas, and how having a vehicle changes food access or food security?
What interests me most, however, is how the data might misrepresent the reality of food issues? We tend to ascribe a lot of authority to data, often without investigating its origins or biases. There is also a tendency to discount the lived experience of people as “anecdotal” and thus lacking authority. I wonder how we might tell stories with data that front load the human lives that make up each data point. Especially because food insecurity is an issue that’s often unseen, how might make visible the human toll of food insecurity?
As a team, here are the research questions we developed:
- What is the division of food sources in different locations?
- What is the demographics of these locations?
- How many low-income groups are there in urban and rural ares?
- How food accessible are urban v/s rural areas?
- Is there a relationship between low-income groups and concentration of particular food assets in Allegheny County?
- How many people are food insecure in Pittsburgh?
- How many people access food assistance?
- How many people are eligible for food assistance?
- What are the social characteristics of food deserts?
- How diverse are low-income areas? How does the diversity differ from non-low-income areas?
- Does low-income area automatically mean low accessibility to food?
- What is the geographic and demographic division of WIC families?
- Is enough food accessible to WIC families?
- How many people are deemed food insecure?
- How many people who are food insecure are elligible for federal food assistance?
- How many people food assistance?
- Who does the data undercount or underrepresent?
I’m going to organize my data around the a question: How might we paint a better picture of food insecurity in the United States? Implicit in this question is the idea that the current picture of food insecure is somehow incomplete or insufficient. The “better picture” that I’m aiming to create is not a more accurate metric to track food insecurity in the US. That’s a matter better left to statisticians and social scientists. The improvement I’m aiming to make to the perception of food insecurity is to combine data with the lived experience of food insecurity. I’m curious about how I can illustrate the lived experience that lies behind a number like “29 million Americans are food insecure.” How do we contextualize that data? How do we illustrate the heartbreaking experience of food insecurity for each one of those 29 million lives?
Furthermore, I’m curious how I might illustrate gaps in the data, such as the undercounting of undocumented immigrants. All else being equal, measures of food insecurity will undercount the actual rate of hunger in latinx neighborhoods because of a fear that providing Census officials with personal information will lead to harmful outcomes. Additionally, rougher measures of food insecurity with small sample sizes are likely distorted by people who are uncomfortable telling a Census representative that they are struggling to feed their children. Admitting that you cannot feed yourself or your children must be an immensely difficult thing to admit, and it likely contributes to an undercounting of hunger in the United States. I’d like to explore how an experience with data can highlight these potential issues in the data that is being presenting. After all, understanding how data can misrepresent the truth is crucial to becoming data literate.
I’ve broken the target audience for this experience into two groups: primary and secondary target users. the primary audience are those who are who have the means or capability to help contribute to solving hunger in America, so potential donors or volunteers. Secondary users could be fundraisers or PR teams at food pantries of food banks who need data to help solicit donations or federal grants.
As far as grouping the data, I think it would be interesting to have two groupings. The first would present the aggregate data in a traditional data visualization, breaking down the total number of food insecure Americans and the different categories within that section (such as those experiencing very low food insecurity and those experiencing food insecurity with children). This may be best visualized with a series of pie charts or a tree plot, depending on the number of categories that I end up including. The second part of the experience would map statistics to personal anecdotes, illustrating how the data relates to personal experiences of food insecurity.
For class, I put together a top-level user flow for the digital data tool that I’m aiming to create. It will attempt to combine macro-level data on food insecurity from the Census with personal stories about food insecurity, pulled from Feeding America and local food banks. First, the user will encounter the macro level data, then an interaction will trigger the personal stories to come to the surface. Finally, the user will be able to see the details of the story and, with it, statistics and pieces of data that relate to it. For example, if the story concerns a person on SNAP, there may be a statistic about the number of Americans who receive SNAP benefits.
After creating this rough outline of the user experience, I began looking at visualizations for the Census data that I’ll be working with. First using paper, and then in Figma. In Figma, I stuck to traditional data viz techniques, without trying to integrate forms that signal the topic — for example, a pie chart that resembles a piece of fruit. I was only able to explore these connections on paper, so far.
As a group, we’ve coalesced around a few key metrics. For each Census tract in the Pittsburgh metro area, we’ll be looking at food accessibility (percentage of people who are 1/2 mile or 1 mile from a food source in an urban area) and food accessibility considering vehicle access. For each geographic area that users view, they will also be able to filter rate of food accessibility amongst certain populations including total population, seniors, children (>18 years old), and SNAP recipients.
This will form the common basis of our project and then we each may add on other elements depending on our interesting. We made a quick whiteboard sketch of interface.
As for the data that I plan to look at independently, I imagine it will exist as a separate but related prototype which pairs nationwide statistics on food insecurity with individual stories of people’s lived experience with hunger (similar to the user flow shared in my last post). While these two projects seem to becoming more separate as we move forward, they are each informing one another in interesting ways. For example, our discussions in class about trying to find a nice fit between form and information have helped me to narrow in on key metrics in my independent project. More on that to come…
Ah, where to begin since last we spoke. All the progress I’ve made has been on my individual project, so I’ll only be discussing that in this post. I’ve decided to look at eight stories of food insecurity pulled from various food banks across the country. For each story, I’ll visualize three pieces of data for the Census tract that the subject of the story belongs to. So, for example, if the story was of a person in Pittsburgh, the data would be related to the census tract in Pittsburgh. The three pieces of data for each story are (1) prevalence of food insecurity, (2) percentage of food insecure individuals who do not qualify for federal nutrition aid because their income is too high, and (3) percentage of food insecurity amongst children.
And now I’ll turn to the much larger challenge: How to visualize this data. In order to work through some ideas, I made two low fidelity animations to think about how we might visualize the data when user’s hovers over a statistic. I thought that there could be some kind of ambient animation in the lower left of the page using blurred gradients (to get at the idea of hidden or unseen reality). When the user hovers over a statistic, the animation would alter to visualize the statistic. For example, when the user hover over the number 27% the animation would visualize 27% and when the user is not hovering over a statistic, it would go back to the ambient blurred state. The colors for both animations are pulled from photographs of the food insecurity stories I’ll be sharing.
Over break, I explored some additional options for the four states of the data visualization. I’m still planning on using the layout discussed in my last post but wanted to explore other ways to visualize each of the three categories of data when a user hovers over the statistic.
After looking at each, I’ve decided that illustrative data visualizations, such as the orange shape on the left, are ill-suited to my topic. Since the tool is talking about people’s lived experience with hunger, a data vis that mimics the shape of an orange seems out of place (at best its random and, at worst, its insensitive).
With that in mind, I looked to the other two options, one which used shape and the other which used shape and pattern, to distinguish between the three different types of data. To me, the pattern and shape structure is the most successful. It does a nice job communicating the difference between different categories and doesn’t distract away from the powerful photos of people experiencing food insecurity. The downside to these narratives is that they don’t have any kind of cognitive link to the topic (a dot pattern doesn’t immediately make you think of food insecurity rates). When I spoke with Cat briefly outside of class, she suggested that a small key sitting along the top left of the screen would go a long way to remedying this problem
Exploring Pattern and Meaning
In class on Tuesday, Stacie pushed me to explore connections between the visualization and the topic of food insecurity. Food insecurity brings to mind feelings uncertainty, unknowing, shame, worry, and (obviously) insecurity. Stacie mentioned that to accurately represent these ideas, it may be helpful to think about visualizing absence instead of presence. How might the visuals prompt the user to think about absence the lack of something?
Well, to start, we thought about making the population of food secure households more prominent, and the population of food insecure households less prominent, so that users understand that they are largely hidden from our view. Secondly, I thought about creating a pattern to represent elements of food insecurity that gets at the idea of uncertainty and insecurity by playing around with opacity and negative space. Below are some iterations on this idea.
Beyond the pattern, however, I wanted to push on the idea of absence or unseen realities. To do this, I thought a particle effect, where small fragments of a whole come together, might help communicate this idea. Using motion in this way might also help keep users attention and allow information to be communicated over time, instead of all at once. Below of a draft of how this motion might work when a user hovers over a statistic related to food insecurity, such as “44% of food insecure households in Pittsburgh do not qualify for federal food assistance.”
The delivery of the project is the project. I had the experience this week of two final presentations for CD and IxD Studios. For IxD, we took a generous amount of time as we wrapped up the project to prepare the presentation and our remarks for the presentation. For CD I was, up until the last minute, fine tuning the visuals and the ideas of the project, continually pushing and pushing the presentation back further. This meant that I had two very different presentation experiences this week, and I strongly prefer the IxD version. With the more thoughtfully prepared presentation, I felt that it truly reflected the time and effort I had put into the project. With the CD presentation, I felt as though the presentation did a disservice to the work that I had done during the project. I need to make the mental shift toward thinking about how the work will be presentation at a much earlier stage. Fair or not, the presentation of the work is the biggest factor in how we evaluate our design. I need to give it the thought and time it deserves.
(Admittedly, the CD presentation also suffered from some technical and timing challenges which were out of my hands)
Start with constraints, let the creativity come later. Another lesson that I should probably tattoo on my hand so I don’t forget. I need to let the constraints guide be in the earlier stages. I’ve started every project in this class with some pretty wild ideas and then, when it comes to bringing them into the world, I get myopic about a single detail or aspect of them to the detriment of the entire project. I think this kind of attention to detail is important, but I need to scope the projects so I can actually put that attention into the entire thing. It would make a huge difference if I started projects by putting some constraints on myself (outside of the ones offered by the project brief) around what is attainable, what is absolutely necessary, and what structure can I set up early, so I can focus on the details later on.