Reproduce “ISCA: A Framework for Interview-Style Conversational Agents” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Charles Welch, Allison Lahnala, Vasudha Varadarajan, Lucie Flek, Rada Mihalcea, J. Lomax Boyd, João Sedoc
ISCA is a rule-based, non-generative framework that empowers researchers to create and deploy interview-style conversational agents for social science and health studies without coding. It facilitates controlled, standardized interactions, demonstrating effectiveness in reducing COVID-19 stress and gathering nuanced public opinion on neurotechnology.
Key insights
- A non-generative, rule-based design offers essential control, predictability, and ethical safety for conversational agents, making them suitable for rigorous research in sensitive domains where LLMs fall short.
- Providing a no-code administrative interface democratizes access to sophisticated conversational agent development, enabling researchers without programming skills to conduct complex studies.
- Structured, adaptive interview designs can effectively elicit rich, open-ended qualitative data while maintaining the standardization crucial for robust comparative analysis across participant groups.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.