Musical Preferences as Indicators of Social Attitudes and Personality Traits: A Statistical and Machine Learning Analysis

This thesis investigates how psychological traits and social attitudes influence musical preferences in an Italian population, combining statistical analyses and machine learning approaches. Data were collected via an online questionnaire covering Big Five personality traits, social biases—misogyny, racism, homophobia, and social anger—and self-reported music preferences. Specific hypotheses about associations between social attitudes and genre preferences were tested. Statistical analyses revealed mixed results, confirming some expected relationships while highlighting cultural nuances in others. Machine learning models were tested to predict musical tastes based on psychological variables, achieving moderate predictive accuracy. Findings demonstrate the complexity of musical preferences as shaped by psychological and cultural factors and suggest potential applications in personalized content recommendations.