Vary and assess the resolution of simulated stakeholder dialogues

Develop and evaluate methods for varying the resolution and fidelity of simulated stakeholder conversations and dialogues within LLM-driven synthetic participatory planning workflows, and determine how conversational granularity affects identified issues, objectives, performance metrics, and generated SAEMS alternatives.

Background

In the case paper, participatory methods (e.g., Delphi, free-style brainstorming) are treated as parameters rather than fully simulated conversations. The authors recognize that dialogue resolution may influence planning outputs.

A systematic approach to controlling and studying conversational fidelity is needed to assess its impact on process inclusivity, decision quality, and reproducibility.