Analysis
Adaptive Learning: Will we reach its potential in 2015?

The Year 2015 will bring many new opportunities for learning specialists to contribute to their profession. But, given such market forces as the convergence of new technologies, advances in neurobiology, changing learner demographics, and escalating client expectation (to name a few), it will also provide a number of new challenges.

by James J. Goldsmith, Learning Architect/ Business Advisor at Accenture (Chicago, IL)

 

(@americalearning) Drilling down a bit, several specific challenges come to mind, including the appropriate use of:  collaborative work environments, serious games and gamification, Personal Learning Networks (PLNs), Personal Knowledge Management (PKM) systems, mLearning application, BYOD integration, MOOCs, VUCA, etc. They are all interesting topics (and, doubtless, there will be many articles and blogs devoted to them), but I would like to focus on one area that I think is particularly challenging and one that also offers great rewards if its challenges can be met: Adaptive Learning.

 

Adaptive Learning seeks to achieve the same level of flexibility and real-time support found in the relationship between an engaged learner and an active learning mentor. In Adaptive Learning, the learner can dynamically adjust the form and/or content of the instruction to personalize the learning experience. Importantly, the learner can tailor the event to focus on unique learning needs and, also, the learner’s questions are addressed specifically and in real-time.  While often associated with computer-based instruction, Adaptive Learning can be applied more broadly to any learning delivery format but, unfortunately, not without difficulty.

 

Among its benefits, Adaptive Learning can (when integrated successfully into a learning event):

  • Increase learner motivation and engagement
  • Decrease learner time to mastery
  • Decrease the time (and, therefore, the cost) needed to run a learning event
  • Reduce the time learning mentors  need to spend on routine, repeated issues (giving them more time to work on important things)
  • Make effective learning truly scalable at a reasonable cost

 

Adaptive learning is not a new idea (with initial attempts starting in the 1970’s), but achieving its goals has been elusive. Let’s examine why…

 

As mentioned, it can be said that Adaptive Learning seeks to emulate the virtues of a successful mentor/learner relationship. Though labor-intensive and expensive (sometimes extremely so), when done well, the mentor/learner model has several advantages, including:

  • The learning approach is holistic, involving cognitive, social and emotional elements. Under the right circumstances, the relationship between the mentor and learner provides the structure, resources and motivation to accelerate high-quality learning.  This relationship is a key (if not THE key) component in this learning model.
  • The mentor is able to continually respond to the learner’s needs, adjusting the learning approach and/or content shared to the right level. Of course, the learner is also an active participant in this process and can request or make adjustments, as needed.
  • Elaborating on the point above, the mentor can draw on an unlimited number of examples (often using a story format) to make points in context that will help the learner achieve targeted learning goals quickly and efficiently.
  • Further elaborating, the learner can ask questions about anything at any time, and can expect to get helpful, real-time responses.

 

Because the direction that human communication can take is infinite and unpredictable, this is a very complex model – one not easy duplicated when the live mentor / learner relationship is removed or modified. When compared against this model, computer-based learning/eLearning has been most successful in addressing learners’ cognitive needs but has been less successful in addressing their social and emotional needs. Though attempts have been made to:

1) anticipate different responses through branching and

2) provide a human presence using a recorded on-line coach, etc., the results sometimes are too basic or too artificial to be effective and, ultimately, disappoint. Also, computers today continue to perform faster and faster but their application in a learning context is restricted. Anyone who has flummoxed Siri with a question easily answered by a five year-old would probably agree that there is room for improvement. Having said this, advances in “cognitive scaffolding” (i.e., a system’s ability to assess cognitive status and adjust to address gaps rather than areas already mastered) show great promise and merit further study.

 

So, how can we, in 2015, address our current challenge to maximize the potential of Adaptive Learning? There are many possible actions but I’d like to offer one short-term and one long-term suggestion:

  • Short Term: Blended Learning. As is true for many learning design problem, a viable approach could be the synergistic blending of a variety of learning design and delivery approaches. For example, on a project that I recently completed, we chose to cover the program’s general awareness-type material using an online self-study format.  This approach enabled us to develop (and ensure) the realization of basic core competencies for all participants. It also helped us to prepare them for meaningful participation in the program’s more advanced studies.  For these studies (case-based requiring critical thinking), we provided opportunities (both live and virtual) for real-time dialog between the participants and the course facilitators, as well as peer-to-peer discussions.  While not as seamless or dynamic as a comparable live mentor/learner model (which offers continuous opportunities for dialog in real time), this approach did provide opportunities for meaningful discussions with experts and, combined with the self-study segment, was a cost-effective solution.
  • Long Term: Employ a computer system that is both smart AND compassionate (a combination of HAL-9000 and Wall-E, as it were).   Of course, the goal is to develop an intelligent system that closely mimics the dynamic relationship of a live mentor/learner and which provides balanced support for the cognitive, social and emotional needs of the learner.  In addition, this system needs to be safe, affordable and reliable.  Achieving this in 2015 is, well, a stretch goal.

 

Currently, segments of the learning industry have embraced achieving Adaptive Learning as a worthwhile goal.  Academic studies continue and several vendor products designed to contribute to Adaptive Learning outcomes are currently available. Examples include Carnegie Learning’s Cognitive Tutor, Desire2Learn’s Knowillage LeaP, Sherston Software’s PlanetSherston, and 30 or so more.  Likely, they will be joined by competing products as the underlying technologies improve and more business opportunities emerge. Though these products are promising, for now a well-conceived blended solution that includes live mentors as well as online adaptive learning systems may continue to be the most prudent path for many learning challenges.

 

America Learning Media, December 2015