An empirical investigation into music listening behaviour in the presence of the network effect
Permanent link to Research Commons versionhttps://hdl.handle.net/10289/16062
The rapid expansion of online platforms has revolutionised the digital media industry, transforming the way people consume digital content and interact with each other and the platforms. The network effects (NEs) play a vital role in the success of online platforms, fostering user collaboration and interest exchange, thereby creating a positive feedback loop that influences user behaviours and contributes to a platform’s success. However, initial studies exploring the NEs phenomenon primarily focused on network size, predating the widespread adoption of online platforms, and thus providing little insight into the application of NEs in the online platform context. Furthermore, despite extensive studies on online platforms, established theoretical constructs and practical frameworks that integrate other variables contributing to NEs in online platforms are lacking. The thesis consists of three research chapters that significantly contribute to the study of NEs and their influence on users’ online music listening behaviours. In the first study, a systematic and rigorous approach was adopted to develop an NEs measurement scale. Drawing on social network and social action theories, we developed a novel NEs model with two subconstructs: social network structure and social action. An empirical research design was applied using the data of 200 Last.fm users. We employed a combination of partial least squares (PLS) path modelling and an expert focus group to validate the model. The results supported the validity and reliability of the developed NEs model. The second study addressed the scarcity of longitudinal analysis related to the evolving nature of NEs and the lack of empirical research to measure the impact of NEs on online music listening behaviours. We examined the NEs construct from our first study to show the impact of NEs on Last.fm users’ music listening behaviours cross-sectionally and longitudinally. The research method used was partial least square-structural equation modelling (PLS-SEM) of data obtained from Last.fm within two time intervals, targeting 1,708 users. Our study found that NEs positively influence users’ music listening behaviours, including the quantity, variety, and novelty of their music consumption. Specifically, the multigroup analysis revealed that the positive impact of NEs on users’ music listening behaviours becomes stronger over time. Furthermore, as the social network structure strengthens and users engage in more social actions, there is a carryover effect on NEs at subsequent times. The third study explored the impact of COVID-19 on online music listening behaviours in relation to listeners’ social interactions. We analysed the online music listening behaviours and social interactions of 37,328 Last.fm users in 45 countries before and after the first wave of confinement, using robust causal inference methods: difference in differences (DiD) and two-way fixed effects (TWFE). The results revealed that, in response to COVID-19, there was a decline in the quantity, variety, and novelty of music consumption, with a shift towards mainstream artists. However, our analysis also found that users with more online social connections and communications exhibited the opposite behaviour. This study provides guidance for the development of innovative design strategies for digital media, including music, movies, and games.
The University of Waikato
All items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.
- Higher Degree Theses