Built an LLM-powered web prototype and designed a study of personalized background music, attention, memory, and planning.
I developed a web-based prototype to investigate how personalized background music relates to cognitive performance. Built with React, TypeScript, and LLM APIs, the system connects music personalization with structured cognitive tasks.
I designed a study with three music conditions: LLM-generated personalized music, self-selected music, and fixed baseline music. Comparing these conditions provides a way to investigate personalization alongside participant choice and a common reference condition.
I implemented three cognitive tasks: reaction time for attention, paired-associate memory, and Tower of London planning. I also built an LLM follow-up chat to generate task-specific music tracks based on participant preferences and information.
The work was informed by a review of 35 academic papers covering background music, attention, memory, mind-wandering, mood, and arousal. The prototype and study were prepared for future participant data collection.