AI for evaluation capacity: A chatbot approach to teaching logic models in government settings
Demonstration | Online
-
Organized by:
University of Illinois Urbana-Champaign
About the Event
This session demonstrates a scenario-based AI chatbot developed with the City of Chicago to teach logic models to public-sector staff. Adapting the “My Rich Uncle” scenario into an interactive learning experience, the chatbot helps users work through inputs, activities, outputs, and outcomes in a low-stakes format. The session examines its value for evaluation capacity building in government, along with practical considerations for responsible AI use, including transparency, data sensitivity, and human judgment.
Speakers
| Nome | Título | Biography |
|---|---|---|
| Xinru Yan | PhD Student, University of Illinois Urbana-Champaign | Xinru Yan is a PhD student in Educational Psychology at the University of Illinois Urbana-Champaign. Her work focuses on AI literacy in evaluation and public-sector applications. She developed an AI chatbot to support evaluation capacity building in the City of Chicago. |
Resumo
This session discussed how a GenAI chatbot can support learning about logic models through an interactive “Rich Uncle” activity. Instead of starting with a blank template, learners use a familiar scenario to identify inputs, activities, outputs, and outcomes. The chatbot helps clarify vague ideas, distinguish outputs from outcomes, and connect everyday reasoning to evaluation language.
We learned that AI can support evaluation capacity building when it helps people think, not when it replaces judgment. The City of Chicago example showed that the tool worked best with human facilitation and group discussion. Next steps include sharing the chatbot link and slides, collecting user feedback, refining unclear prompts, and continuing to frame the chatbot as a learning tool.