The Knewton Blog

Our monthly newsletter features edtech and product updates, with a healthy dose of fun Knerd news.

Why Adaptive Learning Is a Sound Investment

Posted in Adaptive Learning on January 26, 2015 by

Education is undergoing a massive shift from print to digital learning tools. Perhaps the most important consequence of digital learning is bringing high-quality education to more students. Given a tablet and an internet connection, any child can access the world’s most powerful education technology. Digital tools can also improve the quality of supplemental learning materials and expand access to specialized subjects and expert teachers.

Most everyone agrees that developing digital tools, and improving them to better support teachers and students, is good for the world.

But for those creating these new tools, things are more complicated. For traditional print publishers, moving to digital requires a dramatic shift in business models and a large investment of human and financial capital. It also involves significant risk. A successful print product doesn’t guarantee a successful digital product. In a changing market, it can be hard to pinpoint what students and teachers want right now, and how this will evolve.

Integrating Knewton adaptive learning into digital products helps learning companies de-risk this transition in three main ways.

1) Ensuring products improve outcomes

Today, a learning product’s success is determined by a medley of factors: how well it’s marketed, how influential its brand is, how much students like it, how much teachers like it, how engaging the user experience is… and so on.

Soon, the success of a product will be determined primarily by one thing: effectiveness. In a digital world, analyzing information about student interactions with content makes it possible to measure much more accurately how well a given tool drives student understanding.

Knewton’s products increase transparency into student progress and content performance to improve outcomes. Students follow personalized recommendations for what to study next, helping to address skills gaps. Teachers get actionable learning analytics to guide interventions. Content creators can see how well individual pieces of content are working, for which students, in which contexts. (This helps them focus revisions to improve the student experience and more efficiently manage internal spend — driving down costs and expanding margins.)

Multiple internal and external efficacy studies prove that Knewton demonstrably improves student learning. For example, one recent report highlighted an average 12 point increase in exam scores for students using Pearson’s MyMathLab at Reading Area Community College and Northeastern Illinois University.

2) Streamlining product focus

There’s a lot to consider in building a digital learning product: user interface, single sign-on, engaging content, CMSs, and more. Many traditional publishers have terrific editorial, content, sales and distribution, and product teams, but haven’t built out a data science team. Integrating with Knewton allows partners to focus on their teams’ existing areas of expertise, while still providing powerful adaptive learning to teachers and students.

Each partner’s content and product goals ultimately determine implementation details. While Knewton is known for working with large multinational learning companies, our platform is also frequently used by smaller companies and app developers. We are continually introducing new services and processes to reduce the friction and cost of integration.

3) Future-proofing products

Not all adaptive learning is created equal. Many of today’s “adaptive” solutions consist of a limited content pool with a rules-based decision tree, sometimes made up by a single content expert, that uses a few checkpoints to determine a student’s path. For example, the decision tree might specify, “Students who get 7 out of 10 questions right on this fraction division exercise can move on; otherwise give them more questions on fraction division.”

Decision trees are a relatively straightforward way to make products that seem adaptive. But they essentially result in predetermined adaptive learning — a contradiction in terms. While students move through courses in different ways, their pathways are based on static views of content and student understanding.

Some organizations have decided that this is good enough. But their products won’t survive in the marketplace. Given the increased transparency that accompanies digital learning, schools will soon be able to see that products with decision trees don’t work nearly as well as those with robust adaptivity. Then they’ll stop using them.

Knewton has taken a fundamentally different approach to adaptive learning. Knewton uses anonymized student interactions to gain a nuanced understanding of student learning and the content in the application. We build on this information to provide personalized recommendations for students, actionable analytics to instructors and students, and content insights to content creators. When a student struggles with a particular question, Knewton figures out the student’s exact misunderstandings, and delivers material designed to increase the student’s skills in those areas. Knewton’s back-end infrastructure makes this process automatic as students move through course material.

Learning companies across the world are embracing the transition from print to digital. Those companies that simultaneously incorporate adaptive learning will not only build the best products, but also realize the greatest returns.