By William Cipolli (Colgate University)
Information
We will discuss the creation and exploration of a dataset containing musical and lyrical features for tracks from three rock bands -- Manchester Orchestra, The Front Bottoms, and All Get Out -- who contributed to a collaborative track, ``Allentown.'' We use the goal of disentangling the different contributions to ``Allentown'' to motivate learning logistic and multinomial regression and how to implement these methods using R and R Shiny. Furthermore, collecting data about what each band sounds like (using Essentia) and what the lyrics of each band read like (using Linguistic Inquiry and Word Count software and the Bing Lexicon) provides opportunities for students to work with various technologies to collect data and experience cleaning and creating features from data. These data are approachable to students because of the context, and the overarching results match public comments about the collaboration, providing a real-world connection. We have successfully integrated these data into various undergraduate and master's level courses to primarily mathematics and statistics students at small and medium private liberal arts universities through classroom activities, assignments, and assessments. Student surveys indicate the use of the data provided engaging and positive experiences that helped students better understand data collection, coding, and statistical concepts.