INTRODUCTION

In stressful times, people often listen to "coping songs" that help them reach emotional well-being goals. This paper is a first attempt to map the connection between an individual's well-being goals and their chosen coping song.

METHODS

We assembled a large-scale dataset of 2,804 coping songs chosen by individuals from 11 countries during COVID-19 lockdown. Individuals reported their well-being goals and also named their coping song. We applied an unsupervised topic-modeling approach to identify 15 self-emerging topics from the song lyrics, and connected them to well-being goals.

RESULTS

We found significant association between certain lyrics' topics and specific well-being goals. This association weakened for participants for which music is highly important. No significant patterns were found for the songs' acoustic features.

DISCUSSION

This paper posits that song lyrics, despite their brevity and presumed simplicity, can be meaningful for self-regulation of emotional states, and should receive more attention by researchers and streaming services alike.