Learn how learning technologies can facilitate deeper forms of learning.
Learning technology has proven itself indispensable in many ways during COVID-19 school closures. It’s been far from perfect, of course, but platforms like Google Classroom and Canvas allowed teachers to communicate with their students, and assign and grade work.
Meanwhile, math homework platforms from ASSISTments and Carnegie Learning helped many students learn and make progress at home; CommonLit helped students progress in ELA with reading passages, quizzes, and other resources; and TalkingPoints kept non-native English speaking families informed about their students’ education. The discipline of learning engineering, which involves leveraging learning technology to study learning and iteratively improve learning platforms, has contributed to many of these successes.
But, during the pandemic, the technology has only gone so far in recreating a rich educational experience, even in classes where teachers have made good decisions about which technologies to adopt. Many teachers, students, and parents have been frustrated because complex skills — like critical thinking, collaboration, and deeper forms of written and oral communication — have been hard to teach during COVID. Clearly, learning technology is mostly not yet suited to facilitate this kind of instruction.
But is this limitation simply inherent, or can learning technology rise to the challenge? We’d argue that there is a lot of potential for learning technologies that facilitate these deeper forms of learning. Natural language processing and machine learning, in particular, have made whole new kinds of instruction, assignments, and assessment possible.
The Importance of “Soft Skills”
Let’s back up and talk a bit about why these skills are so important and the challenges to teaching them. Skills like critical thinking and the ability to work collaboratively are prized by employers. Three out of four employers surveyed said these soft skills were lacking in their applicants. These skills are absolutely crucial in a rapidly changing technological and economic environment, where young people will be expected to reason through complicated problems, work in teams, and adapt quickly to changing circumstances.
But there’s good evidence that the educational system is simply not up to the task. Studies suggest that even college graduates leave school without the deeper thinking and writing skills that we’d hope to see. And, at the K-12 level, teachers routinely cite curricular requirements and testing regimes as obstacles to spending needed time on deeper thinking.
The Role of Learning Technology
Many would argue that learning technology and innovations are part of the problem here, and there is some reason for thinking that. Studies of learning technology tend to focus on short-term learning outcomes, since these are easier to measure. Learning engineering that focuses on rapid improvement risks focusing only on improvements that are easy to measure and where students can show improvement quickly. That’s not necessarily a bad thing, of course — at least in moderation. There’s a lot of good that can come from easy-to-measure improvements. But it’s a problem if it becomes a single-minded focus. Not enough attention is paid to studying long-term effects, for one thing. Furthermore, policymakers and edtech developers must be sure not to neglect the harder-to-measure, often deeper learning that employers — and society as a whole — rightly prize.
But learning technology and learning engineering do not need to be barriers here: far from it. Technological advancements in data collection and analytics have made it possible to develop new tools to both measure these hard-to-measure skills and to help students better develop them. Among these technologies are intelligent tutors, game-based assessments and simulations, natural language processing, and machine learning.
Learning engineering that focuses on rapid improvement risks focusing only on improvements that are easy to measure and where students can show improvement quickly.
Learning Technology for Collaboration, Communication, and Critical Thinking
Most advanced learning techniques, curricula, and assessment tools are designed with individual learning in mind. The standard educational model sees learning as basically an individual process. But the truth is more complicated. Even when students are using systems designed for individual use, their learning often depends on communication and collaboration.
Of course, in a traditional classroom environment, it’s difficult (though not impossible) to understand and assess collaborative learning. But edtech platforms offer an opportunity to make this kind of learning more visible to both students and teachers. Tools that analyze discussion forum data, for example, could help teachers gauge engagement.
Research here is still very much in its infancy, but one study suggests that social presence in discussion forums can be a good predictor of overall course success. Natural language processing could be used to identify and evaluate collaboration in forums, and to identify the level at which students are engaging with material. Other studies have looked into the possibility of using AI-based simulations to assess higher level skills like critical thinking.
Advances in natural language processing also represent an important opportunity when it comes to improving student writing skills, which are a vital part of critical thinking. Natural language processing technology now allows for formative feedback that directs students to engage in high-level revision around essay organization and argumentation.
One study, for example, found that students’ revising behaviors can help inform adaptive feedback to promote better writing and better thinking. More needs to be done to test and develop these tools and get them in the hands of teachers and students.
The technology is increasingly possible. What’s needed now is accelerated research and development to leverage the technology to turn the possibilities of skill assessment and development into a reality.
The standard educational model sees learning as basically an individual process. But the truth is more complicated. Even when students are using systems designed for individual use, their learning often depends on communication and collaboration.
To learn more, see our report “High-Leverage Opportunities for Learning Engineering”.
– Ryan S. Baker, Ulrich Boser