Adaptive Learning: Creating Impact at Scale

Adaptive learning is inherently interactive and challenges learners in a different, more effective way.

People are different and, as a result, that creates a lot of challenges for human resources and for learning professionals. 

As leaders in these particular areas know, one of those challenges is making sure the learning strategy and all of its components can be used successfully by employees to drive continued skills development.  In some cases though, it’s not just about the skills the organization or employee believes they need to be successful – it’s about the development of those skills they are unaware of and need for continued growth and success.

That’s where adaptive learning can be useful.

Adaptive Learning

The Problem with People

According to Nick Howe, Chief Learning Officer for Area9 Lyceum, in order to understand the human aspect of this problem we have to look at the Dunning-Kruger Effect. 

Essentially, David Dunning and Justin Kruger wanted to understand people’s perception of their own capabilities. The two ran a series of tests on a range of different people regarding various topics. When they plotted out how well people performed, some performed well and others not so well. 

Dunning and Kruger followed up this research with another question. They asked participants to rate how well they believed they performed.

What they found was that those who performed really well generally underestimated their performance. The challenge, however, is with those that underperformed. Those individuals believe they performed much better than they actually did.

“It’s a universal human trait,” Howe said. “The worse we are at something, the more we think we are good at it.”

That’s a critical point to understand; that some 10-40% of people are unconsciously incompetent.

Another problem with people, according to Howe, is most already have pre-existing knowledge or skills. As a result, learning professionals often spend time teaching concepts or skills their employees already possess thus wasting time; some 10-60%, and negatively impacting productivity.

Thirdly, research shows that humans are really good at forgetting things. In fact, 70 percent of knowledge obtained is forgotten in one day; 90 percent is forgotten within two weeks. The way to prevent this is through re-enforcement.  As a result, humans will retain more knowledge over a longer period of time.

The Solution

Howe said another piece of crucial yet often overlooked research on learning was done by Benjamin Bloom in 1984.

2 Sigma study visual

Bloom studied how classroom learning might be improved. The graph above shows test results from students – the horizontal axis is how well they did. The lower curve is a test administered after a regular classroom session. Unsurprisingly, some did very well, others quite badly, with the majority somewhere in the middle.

The second curve shows the results when a mastery-based teaching approach was introduced to the classroom – the instructor asked questions in class to gauge how well students were learning and adjusted accordingly. You can see that the test results were better using this approach. You will notice that the results are skewed highly to the right and are tightly grouped, meaning that most people scored highly. 

It is called the two sigma problem, because the median score of the students who were tutored is two standard deviations, or two sigma, above the median score for students in a regular class. Or to put it in real terms, the average performance of tutored students was better than 98% of the students in class.

Adaptive learning uses software to recreate at scale the results you would get if you could tutor each student. It levels the playing field and lets every student achieve their maximum potential – improving proficiency across the board and especially helping those who need the most support.  Essentially, adaptive learning marries learning science and computer science to deliver personalized learning outcomes that adjust in real-time.

To best demonstrate this, Howe gave the following analogy:

“The best analogy I have found for adaptive learning is Google Maps. Paper maps were great. We used them for centuries to navigate the world and were indispensable -- until Google Maps came along. AI-based electronic maps know where you are and can adjust in real-time if you get lost. They’re designed to get you to your destination, rather than leave it up to you to find your own way.”

Traditional learning is like paper maps.  It served a purpose, but needs to be relegated to the past. Adaptive learning is as disruptive to learning as AI mapping technologies were to navigation.

Most training is built on a single idea – people lack competence and learning professionals are responsible for creating the strategy to make them competent.

But this misses an important dimension, that of awareness. Take a look at the graph below:

metacognition graph

The green dots represent topics where a person will likely know some aspects and are aware of that knowledge. This is conscious competence. Learning professionals will waste time if they try to teach this since the worker already knows it.

The orange dots represent information the person doesn't know. This is information that needs to be taught and is often the focus of traditional training.

The yellow dots represent knowledge a person knows, but doesn’t realize they know well. Also called unconscious competence, learning professionals need to help employees build confidence in this knowledge.

The real problem, according to Howe, is the red area.  This is unconscious incompetence – workers not knowing that they can’t do something. Data from Area9 Lyceum shows almost everyone is unconsciously incompetent to some degree, typically ranging from 15 to as high as 40 percent of content being taught. Howe says many workplace mistakes and poor performance is caused not because someone doesn’t know something, but because they think they do.

Area 9 Lyceum’s Rhapsode helps solve this problem.  Learn more by watching the video below.

In Summation

Adaptive learning is inherently interactive. Minute by minute -- learners are challenged and must interact with the system. Howe says learners like this way of learning and find it very satisfying, even for dry subjects like compliance. If you want to hear more from Howe on this topic, click here to watch his presentation entitled Creating Impact at Scale with Adaptive Learning:  How Adaptive Learning is Changing How We Think about Online Learning.