‘Digital Tutor’ Adapts to Students

Published August 22, 2012

A company has created a digital tutor incorporating what Michael Horn calls the “holy grail” of education technology: the ability to tailor itself to individual students as they use it. It’s part of a wave of adaptive technology, said Horn, executive director of education at the Innosight Institute.

Publishing giant McGraw-Hill Education distributes the program, called LearnSmart, as a separately purchased aid to more than 100 of its college textbooks.

“A lot of companies have emerged under the idea that if Netflix can help us find movies that we’d most like based on our prior preferences and Amazon can help us shop, we can figure out what’s most likely to help students learn best,” Horn said.

Students using LearnSmart answer multiple-choice or fill-in-the blank questions, noting their confidence in each answer. The program uses an algorithm to determine the next question, gearing the students toward studying material they don’t know as well.

Students get a customized report of their strengths and weaknesses, along with recommendations for specific material to study. LearnSmart costs $24.99 per course.

Efficient, Personalized
“It’s highly personalized. It lets students study in a way that is as efficient as possible. Making studying more efficient for students encourages them to study more … and it more closely connects students to the learning process,” said Brian Belardi, director of media relations for McGraw-Hill Education. “They know at all times where they’re at.”

Frank Wray, a professor of biology at the University of Cincinnati, has been requiring his online students to use LearnSmart for several years. He’s compared the grades of students who use the program with those who don’t.

“I find a very direct correlation between students who complete a [LearnSmart] module and how well they do in the class, so I think for the majority of students, it’s a real benefit to their learning,” he said.

Early Stages of Adaptive Technology
Adaptive technology is in its early stages, Horn said, while teaching is more nuanced and requires more data than predicting book and movie preferences.

“The technology is a] super exciting horizon and potential, but very early in terms of collecting the amount of data we need … and being able to better target learning objects for each student,” Horn said.

In Wray’s classes, completing LearnSmart comprises about eight to 10 percent of students’ total class grade, in addition to homework and other assessments.

“The vast majority of the students really like it, and the students that don’t are the students that are not as motivated,” he said.

More Effective at Basics
Generally, LearnSmart works best in quantitative subjects, like math and science, Belardi said.

Adaptive technology will likely work better in more objective subjects and lower-level classes where students are building basic knowledge rather than higher-level and discussion-heavy subjects, Horn said.

Belardi said K-12 LearnSmart programs are “nothing I can rule out.”

Image by CityYear.