Adaptive Learning: Understanding Experience from the U.S.
Flexibility and personalization are universally recognized as the key hallmarks of 21st century learning, but how do we go about implementing these principles in practice? At Arizona State University (ASU), which ranks among top U.S. institutions by technology and innovation, it is an advanced AI framework of adaptive learning that reigns supreme, enabling student-centric strategies that are best tailored to one’s personality traits, aptitudes and educational needs. IOE Head Isak Froumin has talked to Dale Johnson, EdPlus Manager at ASU, about what makes this adaptive learning system tick and how it contributes to effective learning.
The ideal model of personalized learning, where an individual tutor is assigned to a student to guide them through each and every aspect of the chosen track, is barely possible in the modern-day educational reality of mass-enrollment, funding constraints and various other limitations. At ASU, it is advanced IT that steps in to bridge this gap by offering an adaptive, smart learning environment that can deliver the right lesson to the right person at the right time.
Embracing the latest developments in machine learning, educational psychology and other areas, ASU’s Adaptive Learning System (ALS) is a sophisticated, cutting-edge solution in computer-assisted training and evaluation that marks a huge leap ahead against conventional linear-logic AI educational systems. By factoring in a variety of individual social-psychological attributes, prior academic background and performance in specific areas of the university program, learning expectations, etc., ALS is able to distinguish between a wide multiplicity of ways in which different student cohorts learn and behave. This enables effectively adapting the coursework and courseware to best dovetail with the specific current and prospective needs of every learner, where an appropriate balance is stricken between such educational facets as time, academic load and the required amount of personal tutoring.
It takes a great deal of brainpower and AI effort to delve into what is missing from a student’s learning that prevents him or her from mastering a particular curriculum item and, more importantly, why this element is missing, Dale Johnson comments. ALS adopts an advanced, multifaceted operational mechanism that builds upon the following four core components: algorithm, assessment, association, and agency.
“Algorithm is an element that is most widely known but the most difficult to achieve, as it takes thousands and thousands of trials to ensure a high degree of reliability. Assessment combines the ideas of the item response theory, which is about how each student responds to various assessment items, and the knowledge space theory, which states there is a particular knowledge space that every student has and what we do here is probing the limits of this space. In the nutshell, association is all about expanding the frontiers of this individual knowledge space, and it is important that the concepts being taught are associated in a proper sequence. The newest addition to adaptive learning techniques, agency refers to factoring in students’ own opinions of themselves as learners; students are encouraged to engage in metacognitive self-analysis when evaluating how well they understand a particular lesson on a five-point scale,” Dale explains.
Watch this video to learn more about how ALS works and in which ways it can facilitate personalized learning.