Adaptive Learning

Adaptive learning is a teaching method which uses computer algorithms to present custom lessons and activities for students based on their specific needs. There are a large number of potential benefits to using adaptive learning, such as helping manage diverse classes, and free up time for teachers to by taking on tasks such as grading.

Schools such as Arizona State University have made adaptive learning a centerpiece of many of their courses. In college algebra, this approach has transformed their offerings by eliminating the need for a remedial math offering. ASU leads the charge for adaptive education

ALEKS, made by McGraw Hil, is a well known adaptive learning platform which starts with adaptive testing, to come up with a psychometric profile of student’s mastery of the topic. Student’s are given an adaptive test, during which, depending on what questions they get right or wrong, they are given harder or easier questions, until a psychometric map can analyze what topics they have mastered and which need more work. Connect Master (also by McGraw Hill) requires students to show their work and uses machine learning to analyze each step the student took, to better understand where they took a wrong step. It is believed that this type of analysis will better prepare students to be critical thinkers.

Knewton was a well-known platform which used the approach of developing a map of prerequisite skills which are the components of a body of knowledge or concept. It analyzes how each student interacts with a concept, and then, using machine-learning algorithms, finds which of the components need more work and re-presents the content in an order which best satisfies the student’s needs. This creates adaptive customized learning plans to help students achieve the goals set by teachers.

Of course, there are also several challenges in implementing technology assisted adaptive learning. For example, as of May 2019, Person has phased out Knewton and it is now no more. Critics of Knewton said its measurement of student biometrics to tack attention is not a good proxy for understanding. Critics also perceived the “benefit” of freeing up teacher’s time, to actually diverting them away from teaching and building lesson plans, and focusing them more into the role of administrator or tutor. (Knewton Is Gone: The Larger Threat Remains) Last but not least, there are always privacy concerns about the massive psychometric profiles developed on students, and the potential harm if this data were misused or breached by bad actors. (The Textbook is Dying)

Models which integrate both an online adaptive learning method along with a face-to-face component show a lot of promise, such the case of Arizona State University’s use of CogBooks adaptive coursewar as away to build foundational skills, which are then applied and assimilated in a face to face course.

ASU leads the charge for adaptive education

Another example is Sheela Whelan’s pre-algebra class at Westchester College in Valhalla, New York, in which students work independently using ALEKS in the classroom, while teachers and teaching assistants move from student to student trouble shooting comprehension problems with personal discussions. In addition, if a group of students are having the same issues, the teacher can form an impromptu discussion group to help them work through the issue. (The Textbook is Dying)