Mining Policy Evaluations in Brazil’s Dual System with AI: Contrasting Accountability and Learning Logics

• In Event: Technology: Uses and Responses

Brief Overview

Analyzing 125 evaluation executive reports from Brazil’s TCU and CMAP, this study applies AI-assisted content analysis to reveal contrasting logics of accountability and learning, demonstrating how LLM methods enhance comparative evaluation research.

Author

• Rodolfo de Camargo Lima, University of Sao Paulo