Intelligent Tutoring Systems (ITS) at the University of Memphis

Student success in learning and retention is governed by four easily observable emotions.

Intuitive faculty can spot these emotions by working one-on-one with a student. Unfortunately, individual attention by a human can’t scale to an entire classroom, much less a large virtual one.

Fortunately, scaling up is something that computers using AI and machine learning do very well.

Dr. Art Graesser, professor in the Department of Psychology and the Institute for Intelligent Systems at the University of Memphis, joined the Enrollment Growth University podcast to explore the lessons all learning institutions can glean from Memphis’ AutoTutor program.

Intelligent Tutoring Systems (ITS)

Intelligent Tutoring Systems (ITS) track the knowledge, skills, and strategies of learners at a very fine-grain level, and then formulate tasks and conversations that will help them learn.

Now ITS are so evolved that they’re reading students’ emotions.

ITS’ emotional awareness evolution has suddenly allowed new learning ecosystems to flourish that will help more students of all ages learn and succeed.

Graesser’s team includes interdisciplinary with experts from psychology, computer science, engineering, English, physics, and more specialities all applying their insights to build the most effective automated tutor.

Intelligent Tutoring Systems Have Three Benefits

  1. Subject Matter Teaching 24/7 Everywhere: In economically depressed or remote areas, anyone can now take a university-level interactive physics course. This can scale from one student to millions.
  2. Tracking at a Fine-Grain Level: These systems track interactions down to the microsecond to spot areas of uncertainty.
  3. They’re Conversational: Because of the conversational nature of these programs, student gains are about the same as with human tutors.

ITS Secret Sauce: Creating Cognitive Disequilibrium

To get students to learn on a deeper level, you have to push them off-balance from their surface knowledge. This makes them confront contrasting information that subsequently forces them to fill in the gaps of their understanding.

ITS systems can see where each student needs a deeper understanding and create challenging interactive exercises that will force them to work harder to fully grasp the concept.

The Four Emotions That Govern Learning & Student Retention

  • Frustration: Games can be frustrating but if students can conquer that frustration, it’s a real high. So, like in games, the ITS system creates challenges that allow the students to gamify their learning, and make it just frustrating enough to keep them interested. Rewarded for their victories with little dopamine hits, students keep trying. If they are too frustrated, the system will give them a good hint and direct them on a better trajectory.
  • Confusion: You want students in the optimal zone of confusion. Some students don’t know enough about the subject matter to be confused. That’s where interactive agents create useful confusion. You give the students contradictions or surprising things that contradict what they believe or an argument back and forth. These challenging agents create some cognitive disequilibrium.
  • Boredom/Disengagement: If they’re bored, you’ve got to shake students up a little bit. You increase the razzle-dazzle or make the problems a little more challenging.
  • Flow: A flow experience is when they’re just thinking and learning at the right pace and learning. You want the flow, where students are so absorbed that time, space, and fatigue all disappear.

ITS Systems Have Evolved Ears and Eyes

To detect these four emotions, ITS systems track natural language interaction, speech intensity, and computer interaction timing.

But the big advance is adding cameras to the mix. These systems can now read facial expressions and even a person’s posture.

“You can learn a lot from posture,” says Graesser. “How people are sitting or how close their nose is to the material. If it’s very close, they’re more engaged. If it’s far away, they’re less engaged.”These systems can read facial expressions. For example, when you’re confused, part of your eyebrows go up and part of the forehead wrinkles. Even the cheapest computer cameras can reliably detect confusion.

Who Benefits Most From ITS Systems?

According to Graesser’s research, the lower ability learners and lower socioeconomic status learners find the conversational agents fun and motivating which helps them learn.

Another segment is struggling adult readers. There are many adults who don’t read at a level that will get them a decent job.

Graesser estimates that 38% of college students don’t read at an eighth-grade level. The student numeracy deficit is even higher.

ITS can help with student remediation privately with no stigma.

“The big exciting thing is this personalized matching of the right learning environment on a topic for a student, ” says Graesser.

This post is based on a podcast interview with Dr. Art Graesser, professor in the Department of Psychology and the Institute for Intelligent Systems at the University of Memphis where their AutoTutor program is advancing the frontiers of interactive learning. To hear this episode, and many more like it, you can subscribe to Enrollment Growth University.

If you don’t use iTunes, you can listen to every episode here.