How adaptive pathways make digital learning more elastic
By Nicholas Murphy
One of the key challenges for digital learning design is creating solutions that meet the needs of all learners. Risk often drives decision-making when it comes to content: if we don’t know how much people already know, we create content that tries to teach everybody everything, regardless of their level of expertise. This is particularly true for any training that is driven by a regulatory or compliance motivation.
Challenging this approach has become a key driver for us at Lumesse in moving, with our clients, towards a more personalised, learner-centric dynamic.
Typically, courses teach and then test: it’s the foundation of most e-learning. But that model is founded on an assumption that the audience will have a low baseline of understanding. The reality, however, is that most learner populations will already know quite a bit about a given topic (even if some of that ‘knowledge’ comes from hearsay, myth or legend!).
One sure way of making the learner switch off is to make them sit through a lot of material they know already. So reversing the teach-test structure and running an initial diagnostic has been a principle in learning for some time. Test me first, teach me what I don’t know, and then test me again.
However, both of these approaches are limited. They work for content you need to remember, but much less well for behavioural competency, where we need to feed the subconscious to drive behavioural changes.
Increasingly, we are beginning to use adaptive learning paths to increase the effectiveness of digital learning. Here’s how it works.
Making digital learning more elastic
Terms such as ‘adaptive learning’ cover an area of hotly disputed territory and can confuse people, but I like to think of this as a way of creating more ‘elastic’ learning.
The approach is constantly to test the learner and to adapt the programme according to that feedback as the programme goes along, better meeting the learner’s individual need. ‘Elastic learning’, as I think of it, never supposes that any final level of perfection is achieved. Instead, it supposes that a learner only ever makes marginal gains in understanding, and that if they go off track at any point they should have the relevant message or outcome reinforced.
Intelligent, ‘elastic’ learning is never rigid. As I engage with the content as a learner, an underlying categorical structure is constantly in evaluation. I may be great at fractions, for example, but if I veer off-track at multiplication, I’m given more opportunity to practice, even if I don’t realise it. This is an approach that’s particularly well suited to both game-led and observational learning. If I am observing a video take place on screen, and the narrative that plays out is designed to highlight certain key themes, my ability to identify those themes is assessed. A learner’s observation, lack of clarity or misunderstanding on any category can be identified during this process.
Visually, let’s see how this works.
As you can see, content relating to Category A (where I evidence the highest level of expertise on average) I see the least of during my journey through the learning, only five pieces of media related to it during my use of the module. However, my performance is lowest in Category C and therefore more content related to Category C is added to my journey as I progress (I see eight content elements in this example). The principle that we value people’s time and experience level is retained throughout. The learner follows a semi-linear direction, expertise is identified and valued, and ultimately the less I know about something, the more I am supported in my development.
This approach is particularly suited to media rich activities that don’t in themselves take a lot of time: observing a video, for instance, listening to a phone call, answering questions or playing a game.
In the quest to drive improvements in learner experience we must respect our learner’s time and expertise in equal measure and adapt our learning to their needs in real time. Elastic learning principles such as those described in the example above are a small step in the right direction.