Analyzing Music and Mental Health Dataset Through Correspondence Analysis and Logistic Regression Models

Abstract

Mental Health problems are now very common in modern communities today, and music is often cited as an impactful tool for mental health improvement. Therefore, this study uses descriptive analysis, correspondence analysis (CA) and ordinal logistic regression to investigates the relationship between music and mental health preference. The aims of this study are to identify the most preferred music genre among respondents, uncover patterns between favourite music genre and mental health variables and investigating the association between music’s effect and favourite music genre alongside with other factors. The dataset that has been used consists of age, primary streaming service, hours per day, while working, instrumentalist, composer, favourite genre, exploratory, foreign language, frequency of listening for each genre, anxiety depression, insomnia, Obsessive-Compulsive Disorder (OCD) and music effect. Descriptive analysis indicates that Rock is the most favourite music genre based on the highest frequency of the respondents which is 188 respondents. The biplot of correspondence analysis reveals that respondents with Depression and OCD preferred Electronic Dance Music (EDM) genre, while Insomnia and Anxiety have an inconsistent genre either Metal or Classical for Insomnia and Rock or Pop for Anxiety. Ordinal logistic regression found that Gospel is the most consistent genre in the model with Odds Ratio of 13100000 and  p-value almost zero.  This study highlights the complex relationship between music preferences and mental health, suggesting that certain genres may align with particular mental health conditions. At some point, music can have potential for mental health support and therapy