North Carolina

Bayesian Pedagogical Agents for Dynamic High-Performance Inquiry-Based Science Learning Environments

Principal Investigator: 
Project Overview
Background & Purpose: 

Pedagogical agents are embodied software agents that have emerged as a promising vehicle for promoting effective learning. This project focuses on the development of a full suite of Bayesian pedagogical agent technologies for inquiry-based science learning. It will provide a comprehensive account of the cognitive processes and results of interacting with Bayesian pedagogical agents.


Each year approximately 2-3 studies are conducted with students in urban and rural settings. Studies typically include between 50 and 190 students.

Research Design: 

This project is designed to generate causal [experimental] evidence. Original data is collected through survey research [self-completed questionnaires, semi-structured interviews] and assessments of learning/achievement tests.

This project primarily uses domain-specific instruments for learning gains and standard instruments as well as customized instruments for self-regulated learning. A secondary focus involves studies of affective impact, engagement, and empathy. Standard analyses are used for impact of conditions on learning gain, self-regulated learning.

Data will be made available to members of the research community after the project’s analyses have been completed.


Years 1 and 2 of the project have focused primarily on the design, implementation, and preliminary study of the Crystal Island game-based learning environment for 8th grade students, the pedagogical agents that inhabit it, and the student modeling and scaffolding mechanisms that will be empirically investigated in Year 3.

Publications & Presentations: 

S. McQuiggan, J. Rowe, and J. Lester. The Effects of Empathetic Virtual Characters on Presence in Narrative-Centered Learning Environments. Proceedings of the 2008 SIGCHI Conference on Human Factors in Computing Systems (CHI-08), pp. 1511–1520, Florence, Italy, 2008.

S. McQuiggan, J. Robison, R. Phillips, and J. Lester. Modeling Parallel and Reactive Empathy in Virtual Agents: An Inductive Approach. Proceedings of the Seventh International Conference on Autonomous Agents and Multiagent Systems (AAMAS-08), pp. 167–174, Estoril, Portugal, 2008.

S. McQuiggan, J. Robison, and J. Lester. Affective Transitions in Narrative-Centered Learning Environments. Proceedings of the Ninth International Conference on Intelligent Tutoring Systems (ITS-08), pp. 490–499, Montreal, 2008.

S. McQuiggan, J. Goth, E. Ha, J. Rowe, and J. Lester. Student Note-Taking in Narrative-Centered Learning Environments: Individual Differences and Learning. Proceedings of the Ninth International Conference on Intelligent Tutoring Systems (ITS-08), pp. 510–519, Montreal, 2008.

S. McQuiggan, J. Rowe, S. Lee, and J. Lester. Story-based Learning: The Impact of Narrative on Learning Experiences and Outcomes. Proceedings of the Ninth International Conference on Intelligent Tutoring Systems (ITS-08), pp. 530–539, Montreal, 2008.

J. Rowe, E. Ha, and J. Lester. Archetype-Driven Character Dialogue Generation for Interactive Narrative. Proceedings of the Eighth International Conference on Intelligent Virtual Agents, pp. 45–58, Tokyo, Japan, 2008.

Scott McQuiggan, Bradford Mott, and James Lester. Modeling Self-Efficacy in Intelligent Tutoring Systems: An Inductive Approach. User Modeling and User-Adapted Interaction, 18(1-2), 81-123, 2008.

Jonathan Rowe, Scott McQuiggan, and James Lester. Narrative Presence in Intelligent Learning Environments. In Working Notes of the 2007 AAAI Fall Symposium on Intelligent Narrative Technologies, Washington D.C., pp. 126-133, 2007.

Sunyoung Lee, Scott McQuiggan, and James Lester. Inducing User Affect Recognition Models for Task-Oriented Environments. In Proceedings of the Eleventh International Conference on User Modeling, Corfu, Greece, pp. 380-384, 2007.

Scott McQuiggan, and James Lester. Modeling and Evaluating Empathy in Embodied Companion Agents. International Journal of Human-Computer Studies, 65(4), 348-360, 2007.

Other Products: 

There are no project products at this time but the work is expected to yield technology innovations for intelligent tutoring systems and pedagogical agents.

Target Population: 
Research Design: 


Subscribe to RSS - North Carolina