AutoMentor: Virtual Mentoring and Assessment in Computer Games for STEM Learning

The AutoMentor project investigates a critical research question about technology-based STEM learning: Can we automate professional mentoring in a STEM learning game?

The project will develop an automated mentoring technology, AutoMentor, within the context of a specific STEM discipline (the study of ecology and the development of systems thinking more broadly) and within the context of a specific STEM computer game for middle school students.

The goal of this project is to develop, through this specific example, the principles and techniques that will make a generalizable system for providing professional STEM mentoring within the context of STEM learning games.

The epistemic game we are using for this project, Urban Science, is a computer-based game in which late elementary, middle, and high school students learn ecological thinking by role-playing as members of an urban planning firm dealing with land use issues in ecologically sensitive areas.

In the game, players interact with non-player-characters in the form of stakeholders in the community and other planners in the firm. These computer-generated characters represent different interest groups with competing agendas, as well as supervisors and other members of the firm who provide professional resources, information about ecological issues, and advice about the planning process.

A key component of the game is that players interact with professional mentors: undergraduate and graduate students playing the role of more senior urban planners in the firm. These mentors help players in the game take action as urban planners to deal with ecological issues in the land use problems they are solving; but even more important, they help players reflect on their actions in the game.

Previous research on Urban Science has shown that (a) the game is effective in developing ecological understanding for students, and (b) the time players spend reflecting with mentors is a key part of that process.


David Shaffer


Completed on August 31, 2015