Despite advances in neuroscience and effective education interventions, researchers still face one big problem in studying student learning: a lack of collaboration.
Research on student learning is scattered across disciplinary silos including psychology, neuroscience and economics. While researchers in each discipline have unique student learning insights, they don't share or collaborate with their peers in the other disciplines.
One researcher might study cognitive issues about how people think. Others may study perception — how people view and encode the world. And still others would look at social psychology — the way student and teacher interactions affect learning. If these researchers don't work together, they will have a harder time identifying a comprehensive, effective approach to student learning.
"All these things contribute to learning, and really effective instructional systems will really have to address these aspects of human behavior," said David Klahr, professor of cognitive development and education sciences as well as director of the Program in Interdisciplinary Education Research at Carnegie Mellon University.
Universities including MIT, Carnegie Mellon, Johns Hopkins and Harvard are separately looking into breaking down these research silos to better understand student learning.
MIT is tackling this problem with new digital learning initiatives announced in early February. In particular, the MIT Integrated Learning Initiative will bring together researchers from disparate disciplines to study the learning process and the education ecosystem. Once they've done their research, they can redesign learning systems that work better in pre-K-12, higher education and professional learning.
"It is a daunting challenge without a doubt, but that's why we want to make it so interdisciplinary," said MIT Professor Sanjay Sarma, who will oversee these initiatives as the new vice president for open learning. "We want to take it on in all its glory."
Throughout their interdisciplinary research, MIT researchers will tackle two huge challenges: asking fundamental questions and figuring out how to answer them. Some of the questions include asking how people learn differently at different ages, how to detect and tackle learning challenges including dyslexia, and how to help students stay focused and motivated. With these questions as a starting point, researchers can then investigate education systems and technology that could help.
But even if researchers are working together across disciplines, they still have to agree on what they're actually researching. Researchers produce conflicting results about which approaches to learning work best, such as inquiry science, discovery learning or direct instruction. The issue is that everyone has a different definition of what these terms mean, so there's no way to truly compare the results of studies.
In fact, Klahr said, it's more complex to have teachers use a new curriculum than it is to launch a rocket to the moon, because no one agrees on the details of how the curriculum will be used. Even scientists agree on the type of propulsion system they'll use and other details of how to launch a rocket.
Instead of talking in buzzwords, Klahr recommends describing actions in detail. Educators would share what they do, how long they do it for, how the learner responds and how a computer or human teaching system analyzes student responses.
"Educational recipes have to be extremely carefully articulated before you can tell whether they're effective or not," Klahr said.
Intelligent tutoring systems already produce plenty of these details that will be helpful in the future. They just need to be fine-tuned.
In the LearnLab, researchers from Carnegie Mellon University and the University of Pittsburgh design experiments with advanced technology to study the ideal conditions for good student learning. Across the country, tens of thousands of students use an intelligent tutoring system to do algebra and geometry problems in math classes, and their data goes into a large repository for analysis.
The lab researchers analyze data points including what problem the system presented, how long the student took to respond, whether the child made a mistake and whether students accepted hints from the system. In some cases, the system presents different variations of a lesson to different sets of students, compares the results, and then modifies the lesson for the future based on those results.
This kind of specific, detail-oriented research will help researchers across disciplines accurately study how students learn.