Online tutoring sites have been around for a while. But recent advances are taking virtual tutors to a whole new level of sophistication: Imagine a virtual tutor with a computer generated face, a gender, a voice, and, most strikingly, one that responds to the emotional cues of the student The New York Times recently reported on remarkable advances in affective computing – computers that monitor and respond to the emotional cues of the students. Maggie Jones writes about her experience with a virtual tutor, named Isabel:
On a summer afternoon, Isabel, a math tutor with long chestnut-colored hair and hoop earrings, sat in the lower-right corner of my computer screen as I wrestled with geometry problems. When I answered correctly, Isabel gave me a quick congratulatory smile. When I rushed, randomly guessing at perimeters of triangles and rectangles (geometry was never my favorite), Isabel, inferring from the speed of my keystrokes, wanted to know if I was bored. Was it because of the last problem? Did I want to choose the level of the next problem? “I think that more important than getting the answer right,” she said in words reminiscent of many a high-school teacher, “is putting in the effort and that we can all be good in math if we try.”
This fall, hundreds of students will experience Isabel and her digital counterparts as part of an online tutoring program, Wayang Outpost. This program uses virtual tutors, or “affective pedagogical agents,” via a game-like interface to read students’ emotional cues, like boredom, frustration, anxiety and nervousness. The students are hooked up to sensors monitoring sweat, pressure placed on the mouse, and fidgeting. A small camera monitors facial expressions. This information is then used to cue the tutor’s responses, whether offering hints and explanation where needed or finding various ways to keep middle and high school students engaged. Wayang Outpost is not just limited to student interaction; the program provides several teacher tools that allow classroom educators to create new classes, assign lessons for certain days, and see reporting on students’ progress.
Programs like Wayang Outpost are posting positive results. That’s not surprising. We’ve long argued the importance of differentiating instruction and targeting students as individual learners. Affective computing offers the promise of individualized instruction, moving beyond differentiating for small groups toward tailoring instruction for each individual learner based on his unique struggles and frustrations. We’ve written before on the next generation of assessments, assessments that blur the line between assessment and instruction and respond to individual learner strengths and weaknesses. As Malbert Smith writes in Next-Generation Assessments:
Research suggests that a novice develops into an expert through an intricate process that includes the following components (Glaser, 1996; Kellogg, 2006; Shea & Paul, 1996; Wagner & Stanovich, 1996):
- Targeted practice in which one is engaged in developmentally appropriate activities;
- Real-time corrective feedback that is based on one’s own performance;
- Intensive practice on a daily basis that provides results that monitor current ability;
- Distributed practice that provides appropriate activities over a long period of time, which allows for monitoring growth towards expert performance; and
- Self-directed practice for those times when a coach, mentor, or teacher is not available.
An important question for teachers and policy makers to address is: How can this intricate process be applied in the classroom to promote the development of expertise in reading, writing, and mathematics?
Personalized learning programs, like Wayang Outposts, offer an important step in the right direction. And though it’s unlikely that even the most sophisticated affective computing programs could ever replace a live classroom teacher, they certainly offer another tool to help supplement instruction for each individual learner.