Live vs. recorded music engagement: A large-scale Bayesian and frequentist analysis of valence and arousal

AbstractBackground: Live music provides a multisensory and socially engaging experience that may amplify emotional responses compared with passive listening. However, extant studies frequently rely on small samples and utilize limited statistical approaches, thereby constraining the generalizability of findings. Objective: The present study sought to compare valence (pleasantness) and arousal (activation) between live music attendance and passive listening using a large, ecologically valid dataset and employing both Bayesian and frequentist methodologies. Methods: A total of 17,498 observations were drawn from the Open Science Framework “Moods and Activities in Music” dataset. Following data cleaning, 15,007 valid cases for valence and 9,552 for arousal remained. Ratings of valence and arousal were recoded on a continuous scale ranging from −1 to +1. Independent-samples Welch t-tests and analyses of variance (ANOVAs) were conducted, complemented by Bayesian t-tests and ANOVAs (Cauchy prior r = 0.707) to quantify evidence strength. Results: Live music attendance yielded significantly higher valence (M = 0.121, SD = 0.418) than passive listening (M = 0.057, SD = 0.394), t(15,005) = −8.432, p < .001, d = −0.161, BF10 = 5.19 × 10¹³. Arousal was also significantly higher for live music (M = 0.130, SD = 0.528) versus passive listening (M = −0.011, SD = 0.448), t(9,550) = −12.645, p < .001, d = −0.301, BF10 = 267.23. Classical (frequentist) and Bayesian analyses demonstrated convergent results. Conclusions: The findings support the hypothesis that engagement with live music intensifies emotional responses, plausibly via enhanced dopaminergic reward processing and collectively synchronized neural activity. The results underscore the value of integrating Bayesian and frequentist approaches and advocate for future research employing physiological and neuroimaging measures to elucidate causal mechanisms underlying these effects. Keywords: live music; valence; arousal; Bayesian analysis; emotion; music psychology