When Justina Cheng, a graduating MS Sustainability (MSS) student, signed up for UEP’s Spatial Statistics course Spring of 2022, she had no idea the decision would take her to a national transportation conference in the US capital. During an interview, Justina shared the story.
Cheng started her UEP journey in the Graduate Certificate in Spatial Data Analytics program. While getting her certificate, she applied for UEP’s MSS program and is now graduating. A required course for the certificate tract is Spatial Statistics, taught by UEP Senior Lecturer and accomplished researcher Sumeeta Srinivasan.
In Spatial Statistics, students are tasked with designing a study for their final projects. Projects involve collating a spatial dataset and developing a statistical methodology that uses software introduced in class. With a background in transportation research, Cheng chose to investigate the relationship between essential workers and the Massachusetts Bay Transit Authority (MBTA) public transit.
Cheng’s original project was a story map, Essential Workers and Public Transit: A spatial statistical analysis of bus ridership retention and Socio-demographics in the Boston Metro Area during the COVID-19 pandemic.
“I was trying to figure out the link between essential workers and ridership retention of MBTA buses during COVID,” Cheng explained in our conversation. “It’s groundwork basically entailed a big lit review to find factors that constitute an essential worker.”
Cheng’s literature review showed that essential workers, who rely on mass transit, are primarily from low-income, black, indigenous and people of color (BIPOC), and marginalized communities. “Things like minority percentage, occupation, age, and race,” she said. “Using census data, I pulled out those characteristics to create a proxy for essential workers and compared that to the percentage of ridership for pre and during COVID using data available through the MBTA.”
Cheng’s literature review also highlighted “captive riders,” those with no alternate means of transportation. She found that captive riders tend to have lower access to vehicles, live in single-adult households, and report public transit as their most frequent mode of transit.
“We intuitively know that essential workers were on the go during lockdown and white-collar workers were not. We know that. But it’s hard to prove.” Cheng explained. “So, I had to find a way to determine whether a relationship existed indirectly.”
Cheng explained that creating the essential worker proxy variable using sociodemographic data was the main challenge in proving the relationship between essential workers and public transit ridership during COVID-19. Since individual rider characteristics are not easily available with data on public transit ridership, researchers have used area sociodemographic variables or conducted their own surveys to analyze transit use. Cheng said surveys would be the most accurate but they take a ton of time and resources, so using sociodemographic data was the clear choice. “The problem is that there is so much inter-correlation between [sociodemographic] variables, you can’t necessarily tease out individual attributes,” she explained. “But you can distill them into simpler components.”
She refined her variables using a statistical technique called Principal Components Analysis (PCA). “Say a variable is positively related to people of color, positively related to low-income people, and positively related to the number of essential workers. You can say, ‘This is my essential worker component.’ Once you have that, you run regression analyses on their relationship with ridership retention.” Cheng explained.
Cheng’s research focus and inventive methodology caught Srinivasan’s attention.
Srinivasan brought Cheng’s project to Associate Professor at UEP, Shomon Shamsuddin. Sensing potential, both researchers approached Cheng with a proposition to expand her study into a research paper. “Sumeeta and Shomon were like, ‘This is cool. Can we make this into a paper?’ and I was like, ‘Sure.’” Cheng playfully paraphrased as she recalled the interaction.
Using Cheng’s original project’s core idea and principal components, Srinivasan and Shamsuddin expanded the study by analyzing data from additional years and spatial levels and using spatial statistical tools she hadn’t yet learned about. “With their knowledge, they were able to take the analyses a lot further than I could have on my own.” Cheng expressed.
As expected, they found clear relationships between bus ridership retention during the COVID-19 pandemic, essential workers, and vulnerable populations. “What’s really interesting though,” Cheng said, “is where the highest correlations were found.”
MBTA reduced service overall during COVID because there was less need, and they wanted to promote social distancing. However, they maintained service in areas known to have a lot of essential workers that rely on public transit. “Places known to have essential workers – Dorchester, Roslindale, Everett, and Chelsea, for example – these are very front of mind for us.”
However, it turns out other areas reliant on transit weren’t on MBTA’s radar during COVID. “Our study found additional places within the MBTA bus shed with high essential workers ridership that we don’t typically think of as vulnerable populations, and who weren’t being serviced by the MBTA.”
Srinivasan and Shamsuddin encouraged Cheng to submit the paper to the Transportation Research Board (TRB), a National Academies of Sciences, Engineering, and Medicine division focused on promoting transportation research and collaboration. TRB accepted Cheng, Srinivasan, and Shamsuddin’s coauthored study for a poster presentation at the 102nd TRB Annual Meeting in Washington, DC, in January 2023.
“[The TRB Annual Meeting] is this huge gathering of people across the transportation industry who converge on Washington, DC,” Cheng described.
According to their website, the TRB Annual Meeting is the largest transportation gathering in the world. It attracts over 13,000 attendees, including policymakers, administrators, practitioners, researchers, and government, industry, and academia representatives from the US and abroad.
Cheng’s presentation was in the morning, she recalled. “It was a poster session in this giant cavern with rows upon rows of posters.”
The study was specific to the MBTA region, but Cheng said her poster presentation was applicable to the international audience of the conference because neglected vulnerable populations likely exist in other regions too. The study demonstrates how essential workers can be identified in other metro areas based on their socio-demographic characteristics and how different regions or transit organizations can more effectively target services.
Cheng reflected on how her project ties into UEP’s Practical Visionaries theme. She said, “UEP offers a lot of technical classes now that are still rooted in the human element of cities. That’s one thing that I really appreciate about people like Sumeeta who are so knowledgeable about data analysis but apply those skills to improve people’s lives and make cities more livable and enjoyable.”
In her final message, Cheng said, “I’m appreciative to Sumeeta and Shomon for wanting to do this with me and for pushing and encouraging me. I was hesitant, but they saw potential and convinced me to go for it. I’m thankful for their support and the opportunity to work with them and present in DC. They’ve created a really welcoming, nurturing environment that allows people to flourish, and I really appreciate that. “
Curious about Cheng’s methodology, Srinivasan and Shamsuddin’s contributions, and what their policy recommendations were? Check out Cheng’s poster at this link to learn more and let us know what you think in the comments.