Pinning "Reflection" on the Agenda: Investigating Reflection in Human-LLM Co-Creation for Creative Coding
This research addresses the problem of enhancing creative engagement in LLM-assisted coding for users, though it is incremental in exploring specific prompting conditions.
The study investigated how different prompting strategies affect reflection in human-LLM co-creation for creative coding, finding that decomposed subtask invocation leads to more frequent and strategic reflection, fostering diagnostic reasoning and goal redefinition.
Large language models (LLMs) are increasingly integrated into creative coding, yet how users reflect, and how different co-creation conditions influence reflective behavior, remains underexplored. This study investigates situated, moment-to-moment reflection in creative coding under two prompting strategies: the entire task invocation (T1) and decomposed subtask invocation (T2), to examine their effects on reflective behavior. Our mixed-method results reveal three distinct reflection types and show that T2 encourages more frequent, strategic, and generative reflection, fostering diagnostic reasoning and goal redefinition. These findings offer insights into how LLM-based tools foster deeper creative engagement through structured, behaviorally grounded reflection support.