Analysis: Knewton’s new Adaptive Learning Platform

“After seven years, $100 million, and the dedication of hundreds of Knewton employees, we’re proud to unveil Knewton.com — the world’s first and only open adaptive learning platform. We think of it as a genius robot tutor in the cloud. It takes the best traits of a human tutor and makes it free, unlimited, on-demand, personalized and available to the world.”

That is the introduction of a new beta version of Adaptive Learning Platform by Jose Ferreira.

With all due respect, let us take a critical look at what Knewton really offers:

1. Authors cannot create a new course, category, topic, subtopic, … of content. Everything is already predefined by Knewton from a pre-published source of content (like a textbook or taskbook)
2. Authors can select topics, but they are not masters of content anymore, they are rather slaves of Knewton’s automation
3. There is no way to define learning Goal/Objectives by Authors. These fundamental instructional concepts do not exist in Knewton’s vocabulary. Authors are only allowed to add and specify presenting and testing items, but not target nor connect them
4. The assignment (for clarity I would name it a lesson) content structure is completely hidden from authors (authors are more like constructors, not architects. No room for human architects of content anymore, they are replaced with robots)
5. Assignment (lesson) content is amorphous, no explicit structure, no hierarchy, no generalization, no holistic view.
6. Content interconnections are defined behind the scenes automatically with a kind of keywords-based distance metrics.
7. No content quality assurance, which translates to No learning quality assurance
8. No root-cause diagnosing, authors can define only shallow feedback on incorrect answers. It can cause dead-loops in learning.9. In general, Knewton keeps the teachers/educators out of the course creation process. Instead, it “creates” drill & practice courses automatically from the pre-published textual material.

1. Learners can select topics of study, but can do nothing about how to learn it.
2. The content organization and logic of learning is completely hidden and is not clear for learners. So it is a kind of a blind drill without a purpose in mind.
3. Assignments (“Your” and “Suggested” ) are very coarse (course or lesson level), which means coarse, not precisely focused, supervising of learning process
4. Objectives, Purposes of learning are not communicated to, and unknown by, learners. For learners, learning is just a Path to nowhere
5. The progress of learning is represented with a moving (forth and back) bar. So, probably the Learning Objective is to hit the predefined threshold on that bar, but it is not specified enough to be meaningful.
6. It does not teach learners how to learn better on your own
7. No self-driven learning on your own, there is no room for that luxury
8. It is a drill & practice robot for passing tests on low-level robotic kind of skills. The human high-level analysis and creativity skills are out of consideration at all.

1. Teachers are not helped much. The progress of learning represented by moving bar is too superficial for teachers. The bar (dynamic score) is not informative enough for making constructive teaching decisions. Analytics is beautiful but shallow. So, the technology does not really help teacher to make or recommend meaningful decisions, teachers are supposed to do some extra manual research on their own to make them
2. The technology does not improve existing teaching towards “an ideal teaching model”, because it is based on the same traditional teachers’ manual decision-making supported with extra but shallow analytic data.
3. No trace of ANY pedagogy behind!? It means no ground for implementation
4. It is not even a competency-based education, because there is no room for competencies to define

In general, it is not a proper educational technology, it is a nice technology but designed without proper pedagogical foundation, vision and goal.

May be that is why David Kuntz, chief research officer of Knewton, said that “the company regularly faces resistance from professors who try the approach. … [This] instruction is not the way they’ve always done things.”

Take also a look at “Yes, I did say that Knewton is “selling snake oil” by Michael Feldstein.

Fortunately, our Adaptive Learning-Authoring Platform CLARITY is quite different. It clarifies and fixes all the above deficiencies and designed for humans. READ MORE.