Undergraduate Honors Theses

Thesis Defended

Spring 2015

Document Type

Thesis

Type of Thesis

Departmental Honors

First Advisor

Francisca Antman

Abstract

Music and arts programs have diminished in importance since the passage of No Child Left Behind, upon which accountability standards have focused on progress in STEM subjects as the sole metrics of success (Beveridge 2010). Research on the importance of music and arts suggests standards need to be reset and success redefined. This study looks at how music participation affects academic achievement for individuals participating in high school band. The hypothesis being tested expects for music to facilitate higher achievement; the null predicting music has no effect on achievement. Using data from the National Longitudinal Study of Adolescent to Adult Health or Add Health, Ordinary Least Squares (OLS) and Fixed Effects Models (FEM) reveal contrasting results when analyzing academic achievement, GPA. Past research reflects the contrasting results in this study, where researchers may manipulate methodology in order to get desired results. This research topic, as raised in prior researched, is subject to concerns of endogeneity and selection bias. To address these concerns, I compare simple OLS Models and FEM, the latter of which controls for unobserved heterogeneity or omitted bias from time invariant influences. Research from Miksza (2010) and Elpus (2013) employ OLS and FEM, respectively, in which the disparate results reflect inadequate analytic techniques using panel data. The preferable model for this longitudinal data is the FEM, which shows no significant results for music participation. The nature of the Add Health data is limited for education research so the sample I draw from consists of n=11,172 observations across three cohorts, with n=356 music participants. These cohorts describe the overlapping of two waves of data collection within four years of high school transcript data, drawn from a study called the Adolescent Health and Academic Achievement transcript study (AHAA) conducted during the third wave of Add Health. After implementing both models, the coefficient estimates for marching band participation are significant in the OLS model but not in the FEM. Using a unique data set to this research topic, the results confirm past research.

Share

COinS