Skip to navigation Skip to content Skip to footer

Digital Promise

Menu

Blog

AI and Emerging Tech

Collaborative Learning in the Age of Social-Distancing


As schools rapidly shift instruction to a learn-from-home situation, we’re thinking about how to integrate core learning sciences principles in this new context: What do we know improves learning? How can we implement these strategies with students at home? And most importantly, how can we do this with attention to the equity issues that will arise?

Collaborative learning is a pedagogy backed by long-standing, proven learning science findings that can easily be implemented in a remote learning context. Most teachers already know about this principle, which includes the familiar “think-pair-share” routine. How can they effectively promote collaborative learning in today’s socially distanced teaching environment?

Educators know that collaborative learning is not simply saying, “Students, work together.” It takes careful planning and structure to work well. One way to structure collaborative learning is for students to take on distinct roles, like solver and checker. For example, take a set of 12 math problems and provide both an answer key and some discussion questions. Two students can divvy up the problems, one answering “odds” and the other “evens.” They can each send their answers to their buddy, who can check them against the answer key. Then, the students can get together on a phone call to discuss what they can learn from their mistakes. Finally, students can write back to the teacher with the two things they learned from doing this.

A key pair of principles to keep in mind is “positive interdependence” and “group accountability.” Positive interdependence means that the assigned task requires work for each student, or it won’t get done; students need each other in order to succeed. Group accountability means that the outcome, or grade, for the strongest student in a group depends on the outcome for the weakest student in the group. When students realize they are really in it together, there’s an incentive for peer teaching. To extend the example above, both students would only get credit for the work if each provided answers for their half of the problems. Now suppose the teacher adds, “I’m going to make a video of me dumping a bucket of popcorn on my head if you can help each other so that every student gets at least 3 of 4 additional questions right on Friday’s quiz. Can you help each other?” That’s group accountability.

In a research study, we used principles like this with mobile, handheld technology. Students answered multiple choice questions individually on their handsets. They were not told whether their answers were right or wrong, but rather whether they agreed or disagreed with their teammates (three students were in each team). They were told to discuss the questions they disagreed on, and then asked to submit one answer as a group for a response from the teacher. We found that students who did this collaboratively learned more than students who worked alone and received instant feedback from the computer.

Collaborative learning can also be effective across different age groups. “The best way to learn something is to teach it” applies to an eighth grader helping a sixth grader, not just to teachers and students. With students at home there are now more near-peer opportunities. Just make sure students know the ground rules for giving help, like not doing the work or just giving the answer to another student. Instead of taking over, they can help via hints, sharing examples of similar problems, or suggesting a strategy to try.

For teachers, it’s important to set the tone that collaborative learning is not just about right answers, but also about students reflecting on what and how they learned. Teachers can ask students to not only share their answers, but also something they learned from a peer. Beyond the content, this learning might be about self-regulation (e.g., how a peer helped when a student experienced frustration). Teachers also need to know when to step in to provide their expertise. In collaborative learning, it’s often best if students explore and discuss a problem first, and then get an expert perspective from the teacher. After the collaboration occurs, a teacher can provide direct instruction that either corrects remaining misconceptions or summarizes the key ideas in a more thorough and well-structured manner.

For more information on the research behind collaborative learning, check out the CIRCL Primer and visit our library of free online learning resources for educators.

Authors

Jeremy Roschelle

Co-Executive Director, Learning Sciences Research

Read bio

Jeremy Roschelle, Ph.D., serves as co-executive director of Digital Promise’s Center for Learning Sciences Research (CLSR) where he leads research on the future of learning with technology. He co-leads the NSF Engage AI Institute and leads the U-GAIN Reading R&D Center, both of which investigate advanced uses of AI in education. Dr. Roschelle is a Fellow of the International Society of the Learning Sciences, and is internationally known for his work on artificial intelligence, computer-supported collaborative learning, learning mathematics with technology, and on scaling up research-based technologies. Over his 30+ year career, he has authored about 200 publications. His publications have about 28,000 citations and he holds 10 patents.

At Digital Promise, Dr. Roschelle oversees a dedicated group of researchers and learning scientists working to improve teaching and learning across preschool, PreK-12, and postsecondary programs. His focus is particularly on improving the infrastructure for generating research evidence and creating an infrastructure for learning-sciences-based applications of AI. Programs including SEERNet, AIMS EduData, and SafeInsights are creating state-of-the-art approaches to generating evidence on digital learning platforms. The K-12 AI Infrastructure Program is funding public goods, cultivating datasets, benchmarks and models that are grounded in the learning sciences.

Prior to joining Digital Promise, Dr. Roschelle led some of the most important research studies on the efficacy of technologies for learning mathematics. He is also well-known for his contributions to collaborative learning and his leadership of several national research hubs related to educational technology. Over the course of his career, he has secured and managed tens of millions of dollars in highly competitive external funding from federal agencies and global foundations. His research has contributed to policy, products, and into practical applications in schools. He also communicates about learning sciences research and AI directly to the public through his podcast, Future Fluent, and through his large following on LinkedIn.

Dr. Roschelle earned an bachelor’s degree in Computer Science  from the Massachusetts Institute of Technology and a Ph.D. in learning sciences from the University of California, Berkeley. He enjoys jazz, backpacking and photography.