Generative AI as a Research Confederate: The LUCID Methodological Framework and Toolkit for Human-AI Interactions Research – MSI

Working Paper

Generative AI as a Research Confederate: The LUCID Methodological Framework and Toolkit for Human-AI Interactions Research

Aaron M. Garvey and Simon J. Blanchard

Jun 25, 2025

Download

Illustrates a framework for prompting LLMs to act as “research confederates” in experiments, enabling researchers to configure chatbot behavior, randomize experimental conditions, capture full conversational transcripts, and implement ongoing behavioral manipulations.

Topics

research recap

Improving Human-Chatbot Interactions: GenAI Agents as Research Confederates

As companies increasingly rely on AI chatbots to interact with customers, it is essential that the information they provide be relevant and accurate and delivered in an appropriate manner. Prompting LLMs to act as research confederates can help design experiments managers can use to safely pre-test chatbots before deploying them.

By using MSI.org you agree to our use of cookies as identifiers and for other features of the site as described in our Privacy Policy.