Adaptive learning is an educational method which uses computers as interactive teaching devices to adapt course content to meet the unique needs of each learner. Data is collected as the learner engages and that data informs learner needs. The more data the system has access to, the more precise the tailoring of content for the learner.

     The adaptive learning system depends on predictive analytics, which assumes that you can predict the future by using data from the past, and upon machine learning, an artificial intelligence model by which systems automatically learn and improve performance from experience rather than by being directed by programming. Therefore, you have a system that can manipulate and categorize data to consistently improve itself as it analyzes and predicts the best path forward for the learner as they approach mastery.

     In 1959, Arthur Samuel first suggested the possibility of artificial intelligence, that a computer can learn for itself, but at that time the technology was not at a level to support machine learning. As technology continues to advance, we move closer to the possibility of using AI with ever more efficiency. 

     The adaptive learning model depends upon four theories of how human beings learn. Theory 1 is Deliberate Practice which focuses on learning precisely where it is needed with intentional focus. Theory 2 is the Ebbinghaus Forgetting Curve, which says that the best time to commit learning to long-term memory is just before we forget it. The benefit of moving from short-term memory to automaticity is that it frees the short-term sphere or makes more room for new learning. The third theory is called the Theory of Metacognition which states that learners learn best when they know what they know, and they know what they don’t know. This simply means that a learner who is aware of what they can do and what they are ready to learn to do will be best prepared to, with intentional focus, work to achieve mastery. Theory 4 is the Theory of Fun for Game Design which states that learners are most engaged when challenged according to the Goldilocks rubric of not too little, not too much, but just the right amount of challenge.

     To use this adaptive learning model, course content needs to be broken down into the most basic and smallest series of performance objectives that are each clearly aligned to assessment items. In the six-part video series for e-Literate TV conducted by Michael Feldstein, Chief Digital Officer Stephen Laster, VP of Research and Development Alfred Essa, and VP of Customer Success Matt Haldeman all mentioned the granularization of content. The granularization of content means to bring content to its smallest pieces aligned to specific measurements. Granularization is the first step in shifting traditional course content into an adaptive model. Then, through adaptive testing, and using as few questions as possible for each objective, learner proficiency and skill level can be determined. From that level the adaptive model can map the best path forward for the learner.

     For my signature assignment, I detailed 3 tasks for the student to master in order to reach the goal of being able to write an original 5-7-5 Haiku poem. If I were to apply the adaptive learning model, I would need to break down those 3 tasks even further into ever more specific objectives. And all those tasks would need to align to specific criteria for measurement. I think adaptive learning is amazing, but I’m still not there yet as far as being able to tailor content to that level of specificity. I think you will all agree that this week’s module was super packed in comparison to the other 7 weeks. I wanted to read absolutely every article at every link, but it was taking me too long. So, I’ve arranged a list for myself to revisit when and as I can. Perhaps, after reading all of those I would have a better idea of how to shift my signature assignment module into an adaptive model.