Dataset for 'Harnessing GenAI for Higher Education': A Mixed-Methods Study of Prof. Leodar, a Custom RAG Chatbot
Description
This dataset accompanies the manuscript "Harnessing GenAI for Higher Education: A Study of a Retrieval Augmented Generation Chatbot's Impact on Learning". It contains the following components: Survey results and analytics: Raw data and analysis from the structured surveys administered to students who used Prof. Leodar, the custom-built Retrieval Augmented Generation (RAG) chatbot. Usage analytics: Data on the daily and weekly usage patterns of Prof. Leodar throughout the course duration, including query volumes and timing. Code for Prof. Leodar: The implementation code for the RAG chatbot, including the architecture and integration with course materials. List of questions asked by learners: A complete list of queries given to Prof. Leodar during its deployment over a single semester. This dataset provides comprehensive information for researchers interested in the development, implementation, and evaluation of AI-powered educational tools in higher education. It supports the replication of our study and enables further analysis of the impact of GenAI chatbots on student learning and engagement. Note: All data has been anonymized to protect student privacy, in accordance with the ethical guidelines outlined in the manuscript.