Principals, teachers, tech leaders, investors, and parents are constantly bombarded by a barrage of tech products all claiming to be “adaptive” and promising a “personalized” learning experience for students.
As far as Edtech jargon goes, the term “adaptive” is perhaps one of the hottest in the industry. The phrase goes hand in hand with other Edtech buzzwords, such as “personalized,” “differentiated” and “individualized,” while simultaneously hinting at the mystique and allure of AI and machine learning, casting the image of an intelligent robot tutor answering students’ questions and helping them with their homework.
However, as a recent study completed by EdSurge and Pearson suggests, there is surprisingly little agreement on what the term “adaptive learning” actually means, and few tech products that actually provide a genuinely adaptive learning experience to students. Based on stakeholder interviews with teachers, administrators, data scientists and researchers, and reviews of dozens of adaptive learning tools, the report provides a comprehensive definition of the phrase “adaptive learning tool” and demystifies the often overly convoluted explanations of how various tech products actually “adapt” to student learning.
Here are the main takeaways from the report.
What is Adaptive Learning?
Putting technology aside for a bit, educators define adaptive learning as an instructional strategy, where the teacher adjusts his or her teaching method (e.g. providing different resources, changing lesson focus, etc.) in response to student needs.
Throwing technology back into the mix makes the definition of an adaptive learning tool a bit more complicated, as it raises several questions. In order for a tool to be classified as adaptive, how complex should its algorithms be? What part of the tool needs to adapt? How much “adaptation” needs to happen?
To address these concerns, the report provides a simple yet satisfactory definition for adaptive learning tools:
Education technologies that can respond to a student’s interactions in real-time by automatically providing the student with individual support.
A tool that provides real-time assessment for the teacher to adjust their teaching method may lend to adaptive learning, but does not classify as an adaptive learning tool. Similarly, a product that provides generic hints once the student gets stuck on a question may seem to be adaptive, but if the feedback is not individualized then the product does not make the cut.
So now that we have defined what an adaptive learning tool actually is, what makes these tools actually adaptive? What goes on in the computer’s mind as it takes input from the students and provides individualized support?
The report decodes adaptive learning tools by identifying where the adaptivity occur within the software: Content, Assessment and Sequence.
Adaptive content tools receive the student’s input when answering a question, and based on that, suggest unique hints, feedback or additional resources. So, if a student struggles with a question, the program might suggest a video explaining the concept, or drop a hint pushing them towards the answer.
The really clever tools include the feature of scaffolding or branching, identifying the supporting skills needed to complete a specific assignment. For example, if a student is struggling with a certain task, the tool would break the question down into the more basic skills the student needs to answer the question.
Adaptive assessment tools adjust depending on whether or not the student answers a question correctly. If she answers an easy question correctly, for example, the tool will respond by presenting a more difficult question, and so on.
Typically, these assessments serve two purposes. Firstly, as practice engine, where the student practices the material she just learned. Secondly, as a tool for monitoring student progress and benchmarking the student’s performance.
The most complicated of three, both in terms of explanation as well as the algorithms and predictive analytics that make it run. Adaptive sequencing analyzes the student’s behaviors and performance when completing an exercise and then suggests a new set of skills based on the this analysis.
In adaptive content, the journey the student takes is not affected by their behavior or performance. Once the student has mastered a specific skill, she is moved on to an entirely different one, maintaining the same sequence as the rest of the students.
With adaptive sequence, the student is moved onto a different skill based on their performance and behavior. The learning journey she takes might end up being very different from the one her classmate takes.
The Greatest Adaptive Mechanism is the Classroom Teacher
While the ed tech tools are getting smarter and more complex, the role of the teacher in the classroom is as crucial as ever. The social and personal aspect of learning is essential for a holistic and impactful student experience.
While students can use these tools independently, the most successful use cases are when these tools are not used to replace the teacher, but rather as an “ever-present teacher’s assistant”, where the data collected is used to inform and enrich classroom instruction. Impactful tools and the data they provide must therefore be as teacher-friendly as they are student-friendly.