↞ My work

Visualizing Pittsburgh’s coffee scene

Exploring of the use of independent coffee shops as a metric for measuring gentrification, visualizing the independent coffee scene in Pittsburgh in a coffee table book.

Context: CMU M.Des. Communication Design Studio; advised by Stacie Rohrbach

Time: 5 weeks, part time

Team: individual

Tools: R, Adobe Illustrator, Adobe InDesign

Background

Our professor, Stacie Rohrbach prompted the class to explore data relevant to Pittsburgh. For this project, I wanted to continue my exploration of gentrification.

In a previous project, I visualized the distribution of vegan restaurants compared to the gentrified neighborhoods of San Francisco. One metric of gentrification used used in one particular research appearance of coffee shops [1]. The researchers asserted that coffee shops are one aspect of gentrifiers culture, possibly serving as a “real time” metric of the process gentrification as opposed to the infrequent collection of demographics in by the United States census [1].

In my project, I aimed to interrogate whether coffee shops are indeed a component of gentrifier culture. As coffee lover, I found this a perfect opportunity to explore the implications of the kind of culture I participate in.

Are Pittsburgh’s independent coffee shops more prevalent in neighborhoods that have residents with higher levels of income, education, and whiteness?

Getting the Data

I used Pittsburgh census data on education, income, and race for each neighborhood. I used Yelp’s API to aggregate listings under its “Coffee & Tea” category. Then, I distilled these listings into independent coffee shops using the following criteria:

  • Coffee is the focus on the business, not food
  • Fewer than 4 local chain locations

Understanding the Data

In the abstract

Central to my methodology in this project was our professor’s insistence of using large sheets of paper to write and understand each metric within our data sets. We used several ways of slicing the data (scales, scope, LATCH) to explore, in the abstract, what our data contained:

Using this large physical paper to consider my data set was a powerful tool in distancing myself from a digital (and possibly myopic) view of the data.

Critical findings in this phase included:

  • Finding the median across each demographic measure (income, education, and whiteness) for all neighborhoods
  • Organize my exploration by coffee shop (as opposed to neighborhoods), to effectively show the coffee scene. In this way, I could show the neighborhood characteristics each coffee shop was placed in (rather than comparing the demographics of neighborhoods with large numbers of coffee shops)

Through visualization

Using R, I created several summary graphs (beeswarm + quantiles) for each aspect of the data: income, education, and whiteness. Critical to this exploration was showing medians across all neighborhoods compared to medians across all coffee shops. By comparing medians, the typical kind of neighborhoods in which coffee shops are located could be shown:

In sum, coffee shops are typically in neighborhoods with higher income, higher levels of education, and greater numbers of white residents.

Communicating the data

Designing a coffee table book?

Because of my background in interactive data visualization, I earnestly pushed myself out of my comfort zone in this project, by choosing to be one of the few students in the class to present my final deliverable as a physical artifact.

Further, as I thought about the possible audiences for the resulting data visualization, the imperative for a physical artifact grew stronger. Because my project was aimed at an interrogation of the coffee culture I participate in, I figured that a form factor found at coffee shops, a coffee table book, might be an appropriate choice. Thinking of Oldenberg’s romantic notion of the Third Place, and considering the idea of coffee shops as the gathering places for intellectual reflection, I decided an analysis of a reader’s coffee-drinking behavior might be suited to the form of a coffee table book read within a coffee shop. It was interesting to design my physical artifact as a book that might be found as reading material in this context.

Design decisions

In exploring the affordances of visualization in print, I made several consequential design decisions:

1. Cognitive connections

In which ways could I include design details that reinforced the data they represent? In other words, I decided to reference coffee with my visual. Across the board, icons were used to represent mugs. I also used mugs filled with liquid to show percentages. I used brown as a highlight color, reflecting the color of coffee.

2. Linear narrative

A coffee table book lends itself to a linear narrative. Indeed, given the specific trends within my data I wanted to communicate, I decided to show each separately. Furthermore, given my experience designing interactive visualizations that leveraged an indexical narrative structure, I opted to challenge myself by communicating my data linearly.

3. Translucent paper

One of the most interesting affordances of print material was the unique way translucent sheets of paper layer visually, hiding printed material beneath. Because I was committed to a linear narrative, I paired sheets of translucent vellum with opaque sheets:

>> initial test of acetate. The material felt smooth, yellowy, and low quality for the feel of the final book. 

 

>> final result with translucent, matte paper

Sources

  1. More Coffee, Less Crime? The Relationship between Gentrification and Neighborhood Crime Rates in Chicago, 1991 to 2005