WPI Computer Science - Colloquium Announcement

Towards Educational Software that Detects and Responds to All the Ways Students Choose to Use It

Ryan S. J. d. Baker
Carnegie Mellon University

Friday, September 14^th, 2007
11:00 a.m. - 12:00 p.m.
Fuller Labs 320

Students use educational software in a considerable variety of ways. In this talk, I will present research towards developing learning environments that can automatically detect and adapt when a student is engaging in behaviors that result in poorer learning.

In specific, I will present systems that can detect off-task behavior (such as talking to another student about subjects unrelated to the classroom

material) and "gaming the system", attempting to succeed in a learning environment by exploiting properties of the system rather than by learning the material and trying to use that knowledge to answer correctly. I will discuss our efforts to validate these detectors' effectiveness, and to generalize them for use in different contexts within a year-long mathematics curriculum.

I will discuss how these systems can be used to develop educational software which adapts to how student choose to use it. In particular, I will present a software agent built on top of the gaming detector, which responds immediately to students' gaming behavior. This agent has been shown in a controlled experiment to reduce gaming and improve gaming students'

learning.

_____

Ryan S. J. d. Baker is a Post-Doctoral Fellow in the Human-Computer Interaction Institute and Pittsburgh Science of Learning Center, at Carnegie Mellon University. He graduated from Carnegie Mellon University in 2005, with a Ph.D. in Human-Computer Interaction, and was research faculty for a year at the University of Nottingham before returning to Carnegie Mellon. His research is at the intersection of human-computer interaction, educational data mining, machine learning, and educational psychology.

Host: Prof. Neil Heffernan

Refreshments will be served.