Valens, Ilona (2025) On the AI Composer Bias in Music Appreciation: A Moderation Analysis. Bachelor thesis, Psychology.
|
Text
IlonaValensBachelorThesisAIMusics5175690.pdf Download (345kB) | Preview |
Abstract
In recent years, music composed using AI has become increasingly common. This has led to the question of how believing a song is AI-generated affects a listener’s appreciation of it, and whether an AI composer bias exists. We conducted research on the effect of AI authorship information on music appreciation, with preference for lyrics as a moderating factor. We used a between subjects design consisting of three conditions: fully AI, fully human and a hybrid condition. We found that participants in the hybrid condition showed significantly lower levels of music appreciation than those in the fully human condition. We also found that participants low in preference for lyrics in both the fully AI and human condition appreciated the song significantly less compared to those in the fully human condition. Our study provides support for the AI composer bias. With the rapid development of AI and its increasing prevalence, we recommend researchers to keep investigating the AI composer bias as its effects might change. For now, we recommend artists to be careful about using AI and to not state it too explicitly when it is used to generate (parts of) a song.
| Item Type: | Thesis (Bachelor) |
|---|---|
| Supervisor name: | Meerholz, E.W. |
| Degree programme: | Psychology |
| Differentiation route: | None [Bachelor Psychology] |
| Date Deposited: | 25 Jul 2025 09:46 |
| Last Modified: | 25 Jul 2025 09:46 |
| URI: | http://gmwpublic.studenttheses.ub.rug.nl/id/eprint/5743 |
Actions (login required)
 |
View Item |