Digital humanists with an interest in environment 🍃

We are 6 UCLA students brought together by our interest in data and its story-telling abilities, and the story we chose to investigate was walkability — living in a walkable college town made us realize how crucial it was to our livelihoods, thus we wanted to take a deeper dive.

We set our interest to investigate what truly makes a walkable city successful, and how the narratives we create regarding walkability could impact governments to make a bigger push towards human-centered urban planning. We also wished to explore what demographics of people are more or less likely to have access to higher quality of life through walkable communities, and if there are any patterns or systemic barriers to such access. 

What steps does a Digital Humanities project entail? 📐

Source Selection

To investigate walkability and its associations with other factors, our team chose to work with the Nationality Walkability Index and Smart Location Database (SLD). In 2014, the SLD was released and recorded 106 variables ranging from Area, Demographics, and Employment, to Density, Diversity, Design, Transit, and Destination Accessibility. In 2021, the National Walkability Index was released as a byproduct of SLD by calculating a National Walkability Score across the US using four variables of the SLD (intersection density, proximity to transit stops employment mix, and employment and household mix); thus in essence, the two are extremely similar with the National Walkability Index having one additional walkability score. 

Since the National Walkability Score is only based on four variables, we wished further to explore its connections with other variables in the SLD, thus we employed several measures previously unrelated in the calculation process of the score. The SLD is recorded in units of Consensus Data Groups (CBG), the smallest geographical unit for which the bureau publishes sample data. With all 50 states and the other USA territories, a total of 220,740 CBGs made up the enormous databases we worked with. 

Although extensive, the SLD was still missing a few factors wished to explore, such as mental health, BMI index, green space measures, etc. After confirming the additional topics needed, each member researched for supplemental empirical research and articles; then, we reconvened to look for a consistent narrative and decided on the emphasis of factors in our representation.

Data Processing

The Environmental Protection Agency (EPA) provided a mostly clean data set that had minimum missing and null values. Although they were mostly not of interest to our research questions, we completed some further cleaning with Excel for data involving cities on the state borders or when creating visualizations with an emphasis on California regions. We employed Tableau as our main method of data visualization since it required minimal CS experience and was still capable of providing interactive and comprehensive results — although all of the members here have been previously unfamiliar with the software, we gained experience in class and showcased such skill through the graphs we’ve created for this project. When further data manipulation was required to extract additional information from the data, we used other software such as R and Python sparingly due to the more customization, faster running ability, and stronger statistical analysis power they had. Each visualization was created to best represent the narratives being told, while we kept in mind the structure of the variables to select the best type of visualization. 

Narrative Presentation

Our team created this website using WordPress and hosted it through UCLA’s HumSpace Portal, a website hosting service aimed to assist UCLA students and faculties in the humanities department. We landed on the decision because of the creative capacity and abundant functionality from third-party plugins that were available to use through WordPress when the majority of us had no experience with website designing. As a website-building platform, WordPress is maturely developed and provides numerous features in organized formats (such as themes, block-design systems, easy typography adjustments, various device previews, etc) all through a beginner-friendly user interface. With the assistance of the UCLA Disabilities and Computing Program, our team understood the necessary elements to make web designs accessible to all, thus we made sure to use accessible color palettes and hierarchical typography for easier legibility and reduced cognitive load across devices and platforms. 

Since WordPress was extremely versatile when it came to linking external functions using “shortcodes” or HTML codes, we were able to take advantage of several third-party functions as well as their in-house plug-ins. For example, we employed the plugin Ultimate Blocks to create the dropdown annotated bibliographies. We also used Timeline.js to present our timeline and bring the historical factors of walkability in California throughout the years along with major events that could’ve conflicted such scores into life. The extension tool was beginner-friendly for our website designers, and we successfully visualized historical data with images and interactive features to foster an engaging user experience for viewers. 

When deciding on the visualization formats for the findings we have discovered in our datasets, we heavily referenced the website “Which chart or graph is right for you?” (Hardin et al.) and decided on the types of graphs to employ for different types of data; then, we published such visualizations to Tableau Public and utilized their respective HTML code to present our graphs with interactive features such that viewers could view multiple measures within each data visualization and clarify our team’s thought process for the conclusion with visualized evidence.
Lastly, our vision for the website’s aesthetic was also considered through factors including color palette, typography, and symbolic representations for specific topics. In terms of color, we landed on an overall neutral palette of black, white, and gray for the hierarchical texts, and we occasionally used teal for certain small elements as it is gender-neutral, associates well with nature and walkability, and provided pops of color every so often. As our investigation could get data-heavy and complicated, eligibility and minimal cognitive load when reading the content are extremely important for our readers, thus we employed a simple yet classic sans serif font family. Lastly, to efficiently convey content and express our narratives, we often used symbolic visual elements, whether it be the walking montage on our landing page or emojis found next to titles or key points throughout the website. Not only did those design elements aid the website aesthetics to match our research, but they also ensured user-friendliness and accessibility for all.