A new study published in the Journal of Science looked at active learning.
“The notion of active learning has been receiving a lot of attention recently,” study author Nesra Yannier told us. « Especially with COVID-19, the importance of engaging students with novel ways of learning has become even more apparent. While schools and teachers are trying to adapt by incorporating new techniques and technologies, quarantine has also been having psychological effects on students such as social isolation, restlessness and attention problems.”
It is becoming clearer that the traditional approaches of education may not be the best way to teach and engage this generation of students. But also, some important questions about education are coming to surface again: What are the critical components of active learning? When is hands-on also minds-on? How can technology be integrated into classroom practice without isolating students from the physical or social world?
“Our collection paper summarizes recent learning experiments, involving all the way from preschool, early elementary children to college students in formal and informal learning settings,” Yannier told us. “The experiments in this collection paper demonstrate how and when different approaches of active learning can be effective and engaging.”
More specifically, the research is about developing a new genre of mixed-reality Intelligent Science Stations (norilla.org) that combines physical and virtual worlds to improve children’s inquiry-based STEM learning in a more enjoyable and collaborative way, fostering their curiosity and 21st century skills like critical thinking and persistence.
“Our goal with our studies was to investigate if having a mixed-reality system combining physical experimentation and AI-based inquiry guidance can improve children’s learning and enjoyment compared to solely screen-based technologies such as computer games or tablet games that are widely used today,” Yannier told us. “We also wanted to see if having the Artificial Intelligence (AI) guidance on top of physical experimentation (as offered in our Intelligent Science Stations) can improve learning and engagement compared to physical exploration alone (which is similar to what many hands-on maker spaces or museum exhibits offer currently).”
Yannier see areas where education can be improved. Even though children are curious by nature, according to National Center for STEM Elementary Education, a third of children have lost interest in science by the fourth grade.
“According to our research with teachers, many teachers in early education want to teach science but don’t have a science background and the tools/curriculum to teach science from an inquiry perspective,” Yannier told us. “Another big problem in today’s world is that even though there are a lot of technologies out there, most of these technologies are also making kids more socially isolated from their physical environment.”
The researchers wanted to see if they can create a new type of AI mixed-reality system that combines physical and virtual worlds to support students as well as teachers in learning/teaching STEM concepts in a more enjoyable and collaborative way.
To test their theory and see if having AI-supported physical interactions improve children’s learning compared to screen-based computer or tablet games, they conducted a controlled experiment where they compared learning and enjoyment outcomes (using pre and post-tests) of children learning from our mixed-reality system compared to a tightly matched computer or tablet game that is on a screen.
Similarly, to test their theory if having an AI layer on top of physical experimentation improves learning compared to hands-on exploration/construction, they conducted a controlled experiment where they compared the AI-enhanced Intelligent Science Station with a matched control with hands-on exploration alone by turning the AI on and off.
“Our experiments demonstrated that students learn five times more with these AI-supported mixed-reality physical interactions compared to tightly matched screen-based interactions on a computer or tablet screen,” Yannier told us. “Children also enjoy these AI-supported physical interactions significantly more compared to solely-screen based interactions.”
Researchers also saw that not any type of active learning works. When they turned off the system’s intelligent guidance so that students were freely performing tower building activities on their own (similar to most current museum exhibits and maker spaces), they still enjoyed it, but they learned far less. When they have this Artificial Intelligence guidance on top of physical experimentation they learn and understand the scientific concepts significantly better and can apply them to a real-world building and engineering task as well.
“The most surprising result was that even though children were doing a lot of more building in the explore/construct condition (where they were building towers of their own without AI inquiry guidance) compared to the AI-guidance condition (where they were doing a combination of guided discovery and hands-on construction activities with AI guidance), their hands-on building skills did not improve much in this explore/construct condition,” Yannier told us. « In other words, an AI enhanced guided discovery and construction condition led to better learning of the very hands-on construction skills that are the sole focus of the hands-on constructive learning condition (>10 times more pre-to-post improvement).”
So, children were not only learning the scientific concepts better, but they could also apply them to a constructive problem-solving task in the real world with the AI-enhanced mixed-reality system. These results suggest ways to achieve powerful active learning of science and engineering that go beyond the widespread temptation to equate hands-on activity with effective learning (especially given that the explore-construct condition was very similar to how most hands-on museum exhibits, physical products and maker spaces are currently).
“That kids learned much more from the 3D mixed reality version than from an otherwise-identical 2D flatscreen version was surprising,” study author Ken Koedinger told us. “This result is particularly surprising given the everyday impression that academic learning results from the words students hear or read. These two versions involved exactly the same verbal instruction, yet the interaction in the physical world produces much more learning.”
Similarly surprising, Koedinger explained, are the results that walking improves thinking and that college students think they learn from passive listening to faculty lectures when they actually learn more from interaction through answering questions and getting feedback from faculty.