Neuroscientists Study Real-Time Learning in Classroom Lab
Ramsey Itani and a handful of other students from Pullman District 267 are wired, in every sense of the word.
While their class at first appears to be just a basic row of computer desks, the students in it look like extras from a mad-science movie. As 1st grader Ramsey puzzles over a computer-coding exercise, a Karate Kid-style headband over his forehead records his brain activity, triggering flares of yellow and orange splotches on a purple band on a computer monitor behind him. His classroom neighbor adjusts a blood pressure sensor on her wrist as she types, and in the next row, a 3rd grader peers at her work from beneath a set of eye-tracking goggles. Pullman special-assignment teacher Laura Grant comes over to help, wearing a sensor on her cheek that monitors her stress levels.
The classroom is part of the Educational Neuropsychology Laboratory at Washington State University here. The brainchild of an interdisciplinary team supported by the National Science Foundation, the $200,000 lab—believed to be the only one of its kind in the United States—is designed to interweave behavioral, academic, social, and neurological data during live class sessions, to create a more complete picture of what's going on in the interplay between teachers and students in the classroom.
"The idea is that they all overlap in student learning," said Richard Lamb, an assistant professor of science education measurement at Washington State and the director of the lab. "We want to embed as much of the data collection in as natural an environment as possible. We have these tools, we know what they can do, so how do we translate what we find in these tools to what we can do in the classroom?"
For perhaps as long as there have been education researchers, there have been educators frustrated that the careful confines of the lab don't produce the same results as the messy real world of the classroom.
Nowhere has that divide been greater than in developmental neuroscience. To guide classroom practice, the field must solve what University of Virginia psychologist Daniel Willingham calls the "horizontal" problem—matching neurological patterns to student behavior—and the "vertical" problem of interweaving many facets of learning typically studied by totally different disciplines.
"The information that education researchers most often try to import from neuroscience concerns a single cognitive process in isolation, but the interactions with other systems will be part of the educational context," Willingham said. "Neuroscientists usually cannot characterize these interactions."
Researchers from different disciplines often don't talk to each other and may even be testing the same questions on different groups of people.
In response, Lamb and his colleagues take a kitchen-sink approach to studying learning. Video cameras in the lab record speech and body language, while computer software collects students' work, and an array of instruments track what's happening in the bodies and brains of students and teachers. By recording so many types of information, researchers hope to bridge brain and behavior.
For example, Lamb is using brain imaging of teachers and students to evaluate the cognitive difficulty of math concepts in the Common Core State Standards for elementary mathematics. "We believe there's a disconnect between the cognitive demand the experts say these concepts require and the cognitive demand students experience learning them and teachers experience teaching them," Lamb said.
In a preliminary test, 60 4th graders were scanned using a functional near-infrared spectroscope—which measures changes in blood-oxygen levels in the brain associated with thinking—while completing a 13-item math test aligned to the common-core math standards for their grade. The fNIR readings showed the questions rated as difficult were indeed mentally demanding, but they activated phonological systems in the brain, not those associated with processing math operations. That suggests "the problems are hard because of the reading, not the numbers," Lamb said. He is now planning a larger study to understand how different tasks within test questions and relationships with teachers affect students' cognitive load.
More than 60 active studies have started in the lab since its launch this spring, with many of them focused on practical problems in schools. Of the lab's nine postgraduate researchers, four are elementary preservice teachers who plan to go into teaching after graduating.
"It's a lot of data to be collecting on children, but do I think it's powerful? Oh, my gosh, [it's] amazing information to be able to share with parents," said Grant, a 5th grade teacher and a math and science education postdoctoral student at Washington State.
In a separate study, Joshua Premo, a biology doctoral researcher at Washington State and a former New York teacher, became interested in how classroom dynamics affect cognitive load after reading a study on species of ants that communicate via touch; small breakdowns in communication lead to insects moving wildly off course.
"It's really analogous to classrooms," Premo said. "You have a teacher who thinks he's giving correct information to students. The moment you don't have that correct information, and you start going on with the lesson, you've lost the students. It's happening with individual students and the teacher, and in the class as a whole. If you could see exactly where that breakdown was occurring, you could essentially halt class, bring everybody back together, and keep going."
Bob Maxwell, an assistant superintendent in the Pullman district, said such studies could help the district plan pedagogy. "It would be really valuable to [use the fNIR technology on a class] and have teachers use different questioning strategies to see which are more engaging them in cognitive thinking," Maxwell said. "If you were teaching and you could see that visual of the [student's] thinking starting to shut down, you could change your teaching strategy."
Lamb previously helped develop the Student Task and Cognition Model, an "artificial neural network," or a type of artificial-intelligence system that mimics human learning and pattern recognition based on thousands of high school students. The lab still uses an AI program that collects and analyzes information at every student work station.
The university is planning a $1.9 million version of the lab in a building set for completion in 2018, but Lamb said it's unlikely similar technology will be used in regular classrooms anytime soon.
It takes a huge amount of computer-processing power to pull together student data from so many sources; a few minutes of fNIR data more than fills a 32-gig thumb drive. The lab also runs extensive security protocols to ensure students' health and education data can't be released or traced to them accidentally.
Alexis Grow, a 6th grader working in the lab, initially said it was "creepy" to see her own brain waves visualized by the electroencephalogram, or EEG, she wore, but by the end of the coding exercise, she "got used to it. The brain-thingy on my head, it just felt like a baseball cap."
Vol. 35, Issue 23, Pages 1, 12Published in Print: March 9, 2016, as Classroom Lab Helps Researchers Study Learning in Natural Setting