爱漫岛

• An economics professor at the University of Delaware uses information about Taylor Swift’s Spotify streams and album sales to introduce undergraduates to data visualization methods and programming languages.
• In each session, students gain hands-on experience with tools such as Stata, Tableau, and Python that they can later apply in advanced economics courses and research projects.
• Student mentors help design and lead the workshops, deepening their own knowledge and enhancing their teaching abilities as they share their expertise with less-experienced peers.

 
What do Taylor Swift and data analysis have in common? Surprisingly, when combined, they help students gain valuable, real-world data skills that will enrich their understanding of economics.

Proving this point is a workshop series designed by Kathryn Bender, assistant professor of economics at the Alfred Lerner College of Business and Economics at the University of Delaware (UD) in Newark. Most economics classes focus more on theory than real-world applications, Bender explains. She wanted to find a way to supplement that education to give students practical coding skills that would stick.

The career of pop music icon Taylor Swift, she realized, would provide the perfect vehicle to make that happen.

From Concerts to Coding

Bender made this discovery in October 2023 while teaching her Introduction to Microeconomics undergraduate course. Students were discussing the fact that Taylor Swift had attended a Kansas City Chiefs football game, when their conversation evolved into a lively discussion about ticket supply and demand.

Their engagement with the topic gave her an idea. Bender had already identified a need to add a data workshop to the curriculum. Using a that she had obtained through UD’s Academic Technology Services department, Bender created Data Enchanted: Transforming Numbers into Knowledge. This series of three 90-minute data visualization workshops tied Taylor Swift’s popularity to fundamental data skills.

Kathryn Bender, assistant professor of economics, teaches a workshop that is part of the Data Enchanted series at the University of Delaware.

In their inaugural offering in fall of 2023, the workshops were held in the evenings, once each month. They covered three topics, each named after a popular Swift song:

• : Introduction to Stata (a statistical software program)
• : Building and Structuring Data for Analysis
• : Transforming and Cleaning Data for Analysis

Using real-world data from Swift’s Spotify streams and album sales, the workshops guided students through practical exercises in data cleaning, visualization, and analysis. Student interest in the series was immediate: More than 60 students signed up, although Bender could accept just 32 due to space limitations.

Expansion and Growth

Due to its popularity, Bender decided to expand the program in the Spring 2024 semester, running the original three workshops again and adding five new ones. The spring lineup continued the theme:

• : Presenting Data Efficiently and Effectively
• : Stitching Together Stata Knowledge
• : Introduction to Tableau
• : Learn Python!
• : Conducting Quantitative Analysis

At the request of the students, the eight workshops were offered on a weekly schedule. This time, 80 students applied for the 32 available slots. Eighteen participants earned certificates of completion by attending at least six of the eight workshops, including all of the first three.

The first three workshops were based on Spotify data, while the fourth and fifth focused on data from Swift’s global album sales, building on earlier content and providing students with opportunities to apply what they had learned, Bender explains. For example, in Long Story Short, students were asked to use a visual to drive home a point or summarize a topic.

In Glitch, they used their new knowledge about Stata to consider a wider range of factors while interpreting data sets. They not only studied the global sales of Swift’s albums, but also conducted a comparative analysis of those sales according to release date and geographical location. For example, when students discovered that Swift’s Red album had lower sales than her Reputation album, they speculated that the reason might be that Red was released earlier in Swift’s career. The objective, Bender explains, was for students to try to “understand the story behind the data.” 

Using real-world data from Swift’s Spotify streams and album sales, the workshops guided students through practical exercises in data cleaning, visualization, and analysis.

At this stage, participants also could choose data sets to study. “My requirement was that the data set could be any topic with data that could be easily understood. Who better to pick a data set than the undergraduate students themselves?” Bender says.

In addition to receiving guidance from Bender, participants had the support of Data Mentors—a team of undergraduate students with advanced data skills. “We explored a variety of topics in the latter half of the workshop that reflected the interests of the Data Mentors and participants.” 

The sixth and seventh workshops explored video game production and Netflix data as participants were introduced to the data visualization program Tableau and the coding language Python.

The title of the final workshop, Mastermind, implied that students now were themselves masterminds of analysis. In this session, they chose to analyze one of four pre-selected data sets that covered a range of topics, based on their individual interests, applying the tools of their choice. The data sets related to Swift lyrics, the game stats for Argentinian footballer Lionel Messi, the number of arrests at NFL games, and streams of Beyonce’s music.

Once again with the help of Bender, the mentors, and their peers, participants worked with their chosen data sets to summarize, clean, and transform variables of interest before creating their own data visualizations. “I wanted participants to produce their own work that could be included in a portfolio as evidence of their skills for future employers,” Bender says.

Student Data Mentors Step Up

As mentioned above, Bender didn’t run these workshops alone. She was supported by the Data Mentors, who were each responsible for designing and leading a specific workshop. During the sessions, they assisted their peers by answering questions, identifying coding errors, and helping them choose which data sets they wanted to analyze. Data Mentors also helped identify and select data sets that would be of interest to other undergraduate students.

Kathryn Bender, lower left, with five Data Mentors who helped her design and lead her workshops. The mentors wear T-shirts that reference Taylor Swift’s song “Anti-Hero.”

In fact, Bender could not teach the fifth workshop because she was scheduled to undergo ankle surgery. The mentors “really stepped up to the plate and ran it completely on their own,” she says. “And they did an amazing job.”

Bender noted that the workshops not only helped students learn data analysis, but also allowed the Data Mentors to develop their skills through creating and running the sessions.

“During the Python workshop, the mentor probably knew more about Python than I did,” Bender says. “It was absolutely amazing to see the knowledge he had and how he structured the 90-minute workshop to introduce a complex programming language.”

In this way, she adds, the series ended up helping the Data Mentors delve more deeply into the material as they considered how best to present the concepts to other students who had no prior knowledge of the topic. They “definitely gained an appreciation for the whole process.”

Jenna Demaio, a senior majoring in economics, reinforces that sentiment. Being a Data Mentor was “an amazing learning experience,” she says. “I loved having the opportunity of developing my own workshop. This allowed me to use my skills and get some insight into how much work professors put into making classes happen every day.”

Benefits Beyond the Classroom

The workshop series introduced students to complex concepts and software tools that would be challenging to cover in a single-semester course. The sessions gave space for students to master valuable skills that they could use for their projects in later classes, including Bender’s own econometrics course.

“They have to learn the concepts, figure out the coding, and bring it all together for a project,” Bender says. “It is challenging for any student, but especially ones who have never been exposed to coding before.” Students who learned to work with data and use Stata ahead of time could focus on key concepts in the classroom.

“My goal with this series was to give students a chance to learn the data skills separate from econometric concepts and be comfortable with those skills as they go into more advanced data analytics courses,” Bender says.

Oliver Yao, dean of the Lerner College of Business and Economics, makes a friendship bracelet alongside students and other staff members as part of a Data Enchanted networking event. (Photo by Suresh Sundaram)

Most professors care more about students’ understanding the underlying concepts of data analysis than they do about which software programs the students use to conduct that analysis, Bender notes. The workshops provided students with the foundational knowledge and confidence to continue to develop their data skills and experiment with other data analysis tools in the future.

Not only that, the workshops incorporated a relationship-building component. Participants were invited to attend informal networking events where they could interact with faculty members, administrators (including the dean), and industry partners, as well as each other.

At these gatherings, which featured food and Taylor Swift music, everyone in attendance took part in shared activities such as making friendship bracelets—. Through these interactions, participants formed connections that they could draw on for support, whether they were struggling with a concept in the formal workshops or simply needed another set of eyes to look over their code.

A Valuable Learning Experience

Sophomore economics major Diego Berrizbeitia found the workshop series especially helpful to students like him, who had never worked with data analytics before. He describes it as “a very beneficial experience … that encouraged me to keep learning about data and all the different tools that can be employed on its analysis.”

Emma Abrams, a junior majoring in environmental and natural resource economics, emphasizes the need for students to develop these foundational skills in low-stakes settings. “The further I get in my coursework, the more I realize the need for at least basic-level knowledge of coding and data analysis applications in economics,” she says. “The workshop series was a great way to gain exposure to several different analysis tools.”

Because this experience enables students to master valuable skills before they enroll in data-dependent courses, other professors don’t have to spend class time teaching students how to use software platforms or programming languages.

For Abrams, the workshops were especially useful in her work as a research assistant in UD’s Center for Experimental and Applied Economics. In that role, she recently worked on a project studying coastal home buyback programs.

“I had a lot of data to analyze and visualize,” she explains. “What I learned in Data Enchanted was a great starting point for me in this process to understand what is possible when it comes to working with economics data.”

From Idea to Lasting Impact

Bender views the workshop series as the perfect complement to the undergraduate economics curriculum. Because it enables students to master valuable skills before they enroll in data-dependent courses, other professors don’t have to spend class time teaching students how to use software platforms or programming languages. For instance, students who know how to use Stata will be able to apply the tool in any future data-focused projects.

Bender appreciates how what began as a casual class discussion has evolved into an eight-workshop series that has strongly resonated with Lerner students. The workshop took a lot of work to implement, she says, but seeing how well student participants and data mentors have reacted has been “a fun ride.”

“It’s great to see and hear how much the students have enjoyed it,” Bender adds. In a short time, the extracurricular offering has built a reputation among the student body and has even . She is planning to offer the series again in the Spring 2025 semester, with future workshops depending on the availability of funding. The good news, Bender notes, is that the company that developed the Stata software recently invested in the learning initiative.

Bender offers encouragement to other professors who might have an idea for a similar offering. “Your enthusiasm is contagious, no matter what the topic is,” she says. “If you go all in, I think the students appreciate that, and they will build on that excitement. It’s important to harness that excitement, nurture it, and create something truly beneficial for the students.”