Computer-driven adaptive learning has been around for decades; in its most basic form, it is simply the computer program branching the learning path based upon responses the student makes. Some learners may be best served by materials delivered in a different format -- for example, case studies rather than theoretical study. Others may need refresher learning for underpinning skills, principles and theories upon which more advanced learning is built.
Computer-assisted instruction (CAI) has enabled this kind of learning program for half a century and more. I recall working with others in the 1970s as we programmed simple lessons that would quiz students and branch their learning path based on right answers as well as wrong answers. Simple coding in the PLATO TUTOR language would allow programmers to branch to different review or new materials based upon which answer was selected.
More sophisticated adaptive-learning programs that have been developed recently aggregate much more data from the learner to better adapt the learning path. These data can include stored prior learning experiences and performances; self-expressed student preferences in modes of delivery; analytical prediction of likelihood of success for the individual student through different modes of delivery; and much more.
For the past half dozen years, Khan Academy has developed and enhanced their flow-of-learning model. These and other like programs can more finely and accurately identify and address gaps in learning. Coupled with effective support modules, they can fill in the gaps on an individualized basis. “Particularly in high-enrollment classes, adaptive learning can provide tailored support and guidance to all students,” says this primer from Educause. Adaptive learning has effectively been used by many publishers for their online homework and supplementary materials.
Adaptive learning, while it has provided an important step forward in helping to assure that all learners get the material that they need to achieve learning outcomes, has fallen short in cultivating full engagement with the individual student. That's where personalized learning takes the next step. It is defined by the Glossary for Education Reform as:
The term personalized learning, or personalization, refers to a diverse variety of educational programs, learning experiences, instructional approaches, and academic-support strategies that are intended to address the distinct learning needs, interests, aspirations, or cultural backgrounds of individual students. Personalized learning is generally seen as an alternative to so-called “one-size-fits-all” approaches to schooling in which teachers may, for example, provide all students in a given course with the same type of instruction, the same assignments, and the same assessments with little variation or modification from student to student.
This takes student-centered learning to the next level. It goes beyond simply responding to requests from students. Instead, students become part of the process of defining the learning outcomes, pedagogy and practices of the learning experience. Until recently, it has not seemed feasible to meet student needs in this way. To customize learning for each of 30 or 40 students in a class, monitor their individual progress and provide meaningful feedback just is too time-consuming.
Now, machine learning can synthesize the huge volume of data needed to more fully deliver student-centered learning. It can assemble the background, take input from the individual learner regarding their self-determined needs and expectations, identify learning deficits and needs, and produce and present the learning path to best accomplish those goals.
In this case, the role of the faculty member shifts from directly delivering materials and grading based on a single syllabus to advising, assisting and assessing personalized learning that meets the needs of both the individual and the prescribed outcomes of the program. Certainly, this is a change for the faculty member. It is no longer administering a one-size-fits-all class. Instead it is a much more personal, individualized mentoring of each of the students while AI assembles the learning stack for each student.
Have you incorporated any of the adaptive-learning tools in your classes?
Are you preparing for the next step of personalized learning?
Are you preparing your faculty colleagues for this process?