Identifying Factors that Influence Musical Inclination Using Regularized Regression Methods

Abstract

Musical Inclination is a term to describe the propensity of a population to listen to music and especially to play musical instruments. This thesis will examine the factors that describe the U.S. population’s Musical Inclination through regularized regression methods, specifically LASSO and Elastic Net regression analysis. The process will involve determining many factors that shape the population’s Musical Inclination, including, but not exclusive to, participation in leisure activities, participation in artistic activities, attending live performances, music listening preferences, musical instrument sales, and the music industry size. The dependent variable, in this case Musical Inclination, will be observed over several decades through these observed variables, with the goals of better understanding what trends in society affect Musical Inclination and identifying patterns in the relationship between humans and music. We create a regression factor model for our dependent variable using LASSO and Elastic Net regression analysis to provide the necessary statistical methodology as a means of achieving these goals.