As proposed by psychologist Mihaly Csikszentmihalyi, "flow" is an ideal psychological state in which we are engaged enough by a task not to find it boring, and yet not so challenged by it that we get discouraged. When learning new subjects, however, students often end up falling at one end or the other of that scale. Now, a new computerized tutoring system has been developed to keep students in the "flow" zone. It does so by monitoring their emotional state, then adjusting its teaching method to steer them away from boredom or frustration.

Called AutoTutor, the system was created by University of Notre Dame Assistant Professor of Psychology Sidney D'Mello, Art Graesser from the University of Memphis, and a colleague from MIT. It teaches complex concepts in Newtonian physics, computer literacy and critical thinking.

AutoTutor (along with its sister program, Affective AutoTutor) works by holding a natural-language conversation with the student, in which it asks them questions about the subject. By analyzing their responses, it is able to assess their knowledge level, and can subsequently recognize and address misconceptions that they may have regarding the material. Additionally, it is able to respond to their own questions or criticisms, and is designed to maintain their interest with images, animations and simulations.

More importantly, however, it also monitors their facial features, body language and conversational cues, in order to gauge their psychological state. Should it detect negative emotions, it will subtly alter the pace, direction and complexity of the learning task, to improve the situation. It can also alter its own "emotions" accordingly, by changing the content and intonation of its verbal responses, and the facial expressions of the on-screen animated teacher.

"Most of the 20th-century systems required humans to communicate with computers through windows, icons, menus and pointing devices," said D'Mello. "But humans have always communicated with each other through speech and a host of nonverbal cues such as facial expressions, eye contact, posture and gesture. In addition to enhancing the content of the message, the new technology provides information regarding the cognitive states, motivation levels and social dynamics of the students."

In tests on over 1,000 students, AutoTutor reportedly resulted in a learning gain of approximately one letter grade. According to the researchers, this is better than what can be expected from a novice human tutor, and approaches the proficiency of an expert tutor.

A paper on the technology is about to be published in the journal ACM Transactions on Interactive Intelligent Systems.

Source: University of Notre Dame