Study Tracks When Students Zone Out

Researchers from the University of Colorado Boulder have developed a way to detect when students using personalized learning software start to daydream. By using machine-learning algorithms on recordings of student eye movement, the researchers were able to figure out which eye patterns were associated with the mind wandering.

The study found that when students’ eyes matched “zoning out” patterns, they were less focused on the work than those students who showed “not zoning out” patterns. It also noted that when students were paying attention, their eyes bounced around the screen more.

“When you’re zoning out, you’re just fixating,” explained Sidney D’Mello, leader of the University of Colorado research team. “You’re not moving on.”

The study could lead to instructional software that monitors mind wandering in real time. That troubles Jill Barshay, a contributing editor for The Hechinger Report who writes about education research and data.

“Do we really want to curb mind wandering?” she asked. “It’s associated with creativity, and perhaps a bit of mind wandering is needed to come up with big thoughts.”

Barshay suggested the result might be used better to point out the places where the computerized learning bores students, instead of creating prompts to keep them on track.

“But what I find fascinating about this research is how data scientists have come to a conclusion that contradicts human intuition,” she wrote. “You often hear teachers say that they don’t need data to tell them what their students know. Well, this research points out that it’s hard for teachers to know when students are really absorbing something just by looking at their faces.”

AI May Help Screen for Dyslexia

Despite all of technology’s advances and the many data dashboards available to track classroom performance, screening children for dyslexia is still typically conducted using paper tests, whose evaluation can fall prey to subjectivity on the part of teachers.

Lexplore, a Swedish company operating in the U.S. out of Naperville, IL, hopes to employ eye-tracking cameras on computers with artificial intelligence and special algorithms to identify more students with dyslexia who might be missed by the current outmoded, time-consuming method.

The company’s tools analyze patterns in how a subject’s eyes follow words in sequential or nonsequential order as they read. Those at high risk for dyslexia make more right-to-left movements—vs. the more normal left-to-right—and take fewer or no regular pauses during reading.

Although its tech and algorithms are new, Lexplore’s underlying ideas draw on research dating back for decades that indicates tracking eye movements is one of the best ways to gauge reading ability. A 2015 study built a statistical model using eye-tracking that could identify dyslexia with more than 80% accuracy.

Two of Lexplore’s co-founders published their own study in 2016 that claimed an even higher success rate of 95.3% using their own technology, which is now in use across the city of Stockholm’s municipal education board.

In the U.S., Lexplore is still fine-tuning its business model and has so far only been tested in a handful of private schools in the Atlanta area.