Adaptive learning is a methodology for teaching and learning that strives to personalize lessons, readings, practice activities, and assessments for individual students based on their current skills and performance.
Adaptive learning software personalizes learning by using artificial intelligence and machine learning techniques to “adapt” the learning path offered to an individual student in real time. Later, the data captured by adaptive learning software can be analyzed by faculty and administrators to review the needs of individual students or groups of students in a course. They can then adapt instruction to those needs during a term, and they can revise a course between terms. Similarly, students can use data about their skills and performance to adapt their learning practices.
Adaptive learning software often refers to the underlying technology or to a platform that multiple courses can be built on. Adaptive courseware often refers to a set of lessons and practice activities in a digital format for a single course using an underlying adaptive learning technology. Publishers of adaptive courseware on a given subject often develop a library of material, lessons, practice activities, and assessments that instructors can select, sequence, and customize for their particular course.
Every Learner Everywhere and its network partners believe adaptive learning has the potential to promote equity in higher education by improving course outcomes and lowering the cost of course materials, particularly for Black, Latinx, Indigenous, and poverty-affected students.
That potential, however, comes with a number of caveats. As we discuss below, educators are being offered “faux adaptive” software that doesn’t live up to the promise, and algorithmic bias can make software a tool to reinforce equity gaps rather than reduce them. Even with good software products, positive results with adaptive learning depend on many factors such as student-centered and equity-focused instructional design and a research-driven implementation.
How does adaptive learning work?
At the most basic level, adaptive learning software assesses a student’s mastery of a concept or skill in real time and then dynamically adjusts the next lesson or practice activity presented to that student, resulting in improved learning.
For example, imagine a unit of a college course that presents 10 new concepts, concluding with a quiz before the class moves on to the next unit. Suppose this class has 100 students and the average grade historically on that quiz is 70 percent.
Experienced educators know that while there is an average score, there is no average reason that 30 percent of the quiz questions are missed. It may be that:
- student A struggled with the final three concepts in the unit;
- student B struggled with three different concepts;
- for student C, there may be a single critical concept that impacted 30 percent of their result;
- for student D, the “new” concepts weren’t new at all, so they were disengaged; and
- for student E, the issue may not be any difficulty with the material but that family or work responsibilities caused them to miss critical lessons.
A college instructor committed to personalized learning is challenged to understand and respond to the needs of each of those students before the unit quiz is completed and the class moves on to the next unit.
Adaptive learning software works by identifying in real-time the particular concepts or skills that matter for each of these students’ progress. Some software products also evaluate how students worked through the material, distinguishing between “performance data” and “engagement” data such as logins and time on task. Using artificial intelligence and machine learning techniques to evaluate data on the learning paths and performance of previous students, the adaptive learning software then presents what it determines is the appropriate review or practice activity for each current student.
As a result, individual students using adaptive learning courseware will have unique and nonlinear paths through the material. One student will be presented the original lesson on a concept; another will be directed to additional resources for a separate concept; and another will be shown a reminder and a small number of new practice activities from a bank of quiz questions.
In our hypothetical example of a unit quiz, the adaptive learning software has the potential to improve the average quiz grade above the historical benchmark. In that case, more students should master more of the concepts, which should compound into improved learning outcomes at the end of the term and throughout the degree path.
Meanwhile, the adaptive learning software provides data that the instructor can use to revise their lessons before moving on to the next unit and that each individual student can use to understand and direct their own progress.
What are the benefits of adaptive learning?
Used effectively, adaptive learning is a tool to enable personalized learning practices such as scaffolding, flexibility, and targeted instruction and resources. In a study of adaptive learning pilot projects supported by Every Learner Everywhere, faculty reported using the technology to support evidence-based teaching practices such as immediate feedback on practice activities, clear statements of learning objectives, information on how to improve, and low-stakes formative assessments.
Adaptive learning has several other potential benefits:
- It may allow students to be more successful and self directed. When students see data from adaptive learning software about what skills they have and haven’t mastered, they understand where to direct their efforts.
- It allows students to work at their own pace, bypassing concepts that are already mastered while reviewing and practicing other concepts more slowly.
- It potentially improves student engagement by providing lessons and activities that closely match their needs and readiness.
- It can be used as a cost-effective replacement for expensive textbooks.
- It provides a structure that keeps course objectives, lessons, practice activities, and assessments in alignment and shows students how each element of the course relates to the course objectives. Likewise, when students are having trouble mastering a concept, faculty can review if some instructional elements aren’t well aligned with the objectives.
- It provides relevant and timely data that faculty and administrators can use to identify how targeted subpopulations in a course are doing. As this article on disaggregating learning data explains, this is potentially a powerful tool for identifying and confronting barriers to equity for minoritized and poverty-affected students.
- It enables faculty and administrators to provide timely and targeted support by identifying individual students, or even particular sections in a multi-section course, that need attention.
- It enables faculty and administrators to make continuous improvement by comparing data across semesters.
What are some examples of adaptive learning?
Adaptive learning can be used in face-to-face classes, in both synchronous and asynchronous modes of online classes, or in hybrid formats. In higher education it is often used in large enrollment gateway courses that introduce a course of study.
The colleges and universities that partnered with Every Learner Everywhere starting in 2019 to pilot the use of adaptive learning demonstrate several examples of how the technology can be used:
- The University of Texas at El Paso used adaptive learning to redesign several courses in the biology department to integrate digital learning technologies.
- The University of Central Florida used adaptive learning software to implement more open education resources in online courses in the foreign languages department.
- The University of Texas Rio Grande Valley used adaptive learning to increase active learning practices in math courses and to track improvement over time for minoritized and poverty-affected students.
- Cuyahoga Community College used adaptive learning to align the work of eight adjunct faculty teaching the same business course across two campuses.
- Cleveland State University used dashboard data from adaptive learning software in math courses to make weekly adjustments to lessons and quizzes and to improve pass rates.
- Broward College used adaptive learning to coordinate the expertise and activities of the Center for Teaching Excellence and Learning and the Instructional Design Department with math and English faculty.
- Houston Community College provided just-in-time individual student data from adaptive learning software to academic tutors in 14 campus locations so tutors and students could engage in targeted sessions.
- Lorain County Community College used quiz results data from adaptive learning software to personalize the post-quiz communication to students in all statistics courses.
Selecting adaptive learning software
Every Learner Everywhere and its network partners are vendor neutral and don’t recommend a particular adaptive learning software product.
Colleges and universities searching for adaptive learning software should be aware that not everything advertised as adaptive actually personalizes and adapts learning paths for individual students or uses artificial intelligence technology. For example, in a survey of what students want from adaptive courseware one finding was that some products conflate randomized quiz questions with personalized learning. Other products advertised as adaptive learning are little more than digitized textbooks.
The selection of an adaptive learning software product is a complex process that includes identifying goals, coordinating with key stakeholders across the campus, defining evaluation criteria, negotiating vendor contracts, and investing in professional development.
A college or university will also need to consider where the adaptive learning software will be used. Some products are platforms designed to be customized for a range of academic disciplines, and some products are courseware written for a specific discipline.
Many colleges and universities in past years have used the Courseware in Context Framework developed by the Online Learning Consortium, Tyton Partners, and the Bill & Melinda Gates Foundation to guide their process for evaluating and selecting adaptive learning products. Recently, that work was updated and expanded in the interactive CourseGateway, an interactive guide to discover and compare courseware products.
If you are interested in consulting with the Every Learner Everywhere network about selecting and implementing adaptive learning, we invite you to contact us.
Adaptive learning resources
Every Learner Everywhere has worked with network partners to publish several practical resources for colleges and universities using adaptive learning and other digital learning technologies. All of the following resources are free to download:
- Adaptive Courseware Implementation Guide
- Optimizing High-Quality Digital Learning Experiences: A Playbook for Faculty
- Planning for a Blended Future: A Research-Driven Guide for Educators
- Improving Critical Courses Using Digital Learning & Evidence-based Pedagogy
- Teaching Practices of Faculty Adopting Adaptive Courseware
- Lessons Learned: A Toolkit for Post-Pandemic Higher Education with Equity and Student Care at the Center
- Getting Started with Equity: A Guide for Academic Department Leaders
We also recommend the research review Educational Technologies and Their Impact on Student Success for Certain Racial and Ethnic Groups published by The National Research Center for Distance Education and Technological Advancements and WCET – the WICHE Cooperative for Educational Technologies.
Lastly, Every Learner has a growing library of resources for equity-centered digital learning. It includes several toolkits for college and university educators using adaptive learning.
Consider and implement adaptive learning carefully
Adaptive learning technology isn’t a magic pill. It has the potential to close equity gaps and improve learning outcomes for minoritized and poverty-affected students, but that potential won’t be achieved by uncritically deploying the technology. In fact, that will tend to reinforce equity gaps. If adaptive learning courseware is simply used as a “supplement” without closely coordinating it with other course activities, it will tend to increase costs and complexity and add to student disengagement.
And poorly designed products can reproduce and amplify existing biases, a phenomenon known as algorithmic bias. That may happen if, for example, adaptive courseware extends the reach of a textbook that hasn’t been updated with an equity perspective to include culturally relevant teaching resources and a wider representation of students in the discipline. In other cases, courseware can actually be less equitable than the textbook it replaces if, in the interest of cost savings and getting to market quickly, it relies on century-old readings available in the public domain rather than licensing current material. Adaptive learning courseware must be vetted at least as carefully as any textbook to ensure it supports an institution’s equity goals.
Realizing the potential benefits of adaptive learning requires research-based approaches, thoughtful instructional design, strategic implementation, processes for continuous improvement, and a commitment to equity and student care. If your department, program, or institution are considering ways to implement adaptive learning to promote equity, we invite you to be in touch to begin a dialogue.