Organizations hold up Amazon’s customer experience as the Holy Grail for adaptive learning, but relatively few have developed a similar experience for their employees. Adaptive learning is an educational method that uses interactive teaching devices and orchestrates the allocation of human and mediated resources according to the unique needs of each learner.
Integrating adaptive learning is more of a cultural challenge than a technological issue. Like anything new, it can seem threatening to trainers and learners alike, who fear the removal of experts from the classroom. But contrary to those fears, adaptive learning enables an individualized, contextual approach that focuses on what learners need and directs instructors to where they are most needed. It also provides a new data stream that Learning and Development (L&D) professionals can analyze to improve their own courses.
However, before adaptive learning can be deployed effectively, three myths must be overcome.
Myth 1: Adaptive learning is always computer-based
Adaptive training at Parexel, a global contract research organization, relies on close interactions with L&D and with learners’ supervisors. Currently, Parexel only uses adaptive learning to train the clinical research associates (CRAs) who monitor clinical trials to ensure that trial protocols are followed and that documentation is appropriate and up to date. Approximately 60 percent of CRA training is online and 40 percent is face to face. Eventually, it will be rolled out to other learners.
“The CRAs usually have scientific backgrounds, but we must provide training regarding regulatory requirements that vary country by country,” as well as trial protocols that vary according to the pharmaceutical product and the trial sponsor, says Albert Siu, Ph.D., corporate VP, L&D, Parexel. Some conflict resolutions skills are included, he says, to help CRAs hone the skills needed to contradict physicians. “For us, time is of the essence, so adaptive learning is very important.”
Parexel first identifies CRAs’ strengths and weaknesses through simulations, on-the-job observations, analysis of their reports, and online and face-to-face training. “These methods triangulate a certain level of competency, which we correlate to a set of training requirements,” Dr. Siu says.
Myth 2: Adaptive learning replaces instructors
At the University of Wisconsin – Extension, reassuring professors that they aren’t being replaced by artificial intelligence (AI) is among the first steps in online curriculum development. “Online, adaptive learning is hard to sell to faculty accustomed to being in the classroom, as well as to students,” admits Ryan Anderson, director of Instructional Design and Development.
Anderson’s team counters their concerns, explaining, “we’re actually helping professors have more sophisticated interactions with students by automating the basic ones.” In his experience, that means designing extension courses in ways that provide extra materials, examples, and problem sets to students who need them. “Students still have access to an instructor,” Anderson emphasizes, “and the instructors have access to students’ tests, so they can identify where individuals are experiencing problems and work with them to overcome those problems.”
Myth 3: AI eliminates training
Computer-based adaptive learning that uses artificial intelligence algorithms doesn’t eliminate the need for training, but it can eliminate the need to train employees to use certain applications by providing just-in-time help.
“Training isn’t about teaching employees to use applications, but about making them productive,” says Amir Farhi, VP of Strategic Development at WalkMe. Therefore, WalkMe developed a transparent layer that sits atop any underlying digital application like ADP or Workday to deliver just-in-time information. The application uses machine learning to deliver help based on users’ history with the application and the content on their screen. Consequently, experienced users don’t see the basic information that new users see and application training classes are minimized.
“The application constantly measures what happens to the user—for example, how long they take to access certain procedures, where they get stuck, how they get help, etc.,” he adds, so L&D professionals can adjust resources accordingly.
Streamline Existing Processes
“Adaptive learning uses a broad array of teaching and learning techniques and technologies. Implementing it is far more subtle and challenging than simply deploying tools or techniques,” Dr. Siu says. Before integrating this approach into a course, Dr. Siu recommends streamlining the processes that are associated with the learnings. By tagging, organizing, delivering, and measuring learning and levels of competencies relative to a set of requirements, he says, “you’ll be better off afterward because you’ll be more informed about what you need.”
Think Holistically
Expanding one’s frame of reference is integral to successful adaptive learning curricula. “What makes adaptive learning work isn’t technique, but mindset,” Dr. Siu says. “For example, people often look at challenges as technical problems, but, in reality, they’re often adaptive problems. Seeing the differences requires the ability to look at the whole system and understand the issues.” Then, with this holistic understanding, L&D professionals can consider effective technologies to address those issues.
“Think big, but act small,” he advises, by breaking big challenges into small steps. At Parexel, he designs the overarching plan and determines how changes will be made within courses. His local L&D team spots trends and determines which adjustments are needed, while the actual adjustments are offshored.
Integrating adaptive learning into the development program requires collaboration from many constituencies, Dr. Siu stresses. For example, business leaders must be engaged early and must see the benefits. “Therefore, form a close partnership with the business units,” he advises. “Adaptive learning can’t be developed alone.”
“You also need strong governance around how you engage others to make collaborative decisions and position the learning agenda for the company,” Dr. Siu says. He’s worked with five companies during his career and is convinced that a centralized L&D organization provides the most effective governance. “Decentralized learning only develops certain pockets of the company, and hybrid L&D organizations become political.”
Use Your LMS
At the University of Wisconsin – Extension, Anderson uses assistive technology and intelligent agents to deliver an adaptive experience. “We use a learning management system (LMS) called Desire to Learn (D2L). Our courses are based on a student’s performance, so instructors may choose to release or withhold information to facilitate success.”
Intelligent agents help professors have closer contact with students by setting up the LMS for alerts. “If a student hasn’t logged into the system for a certain number of days, or isn’t progressing, you can use an LMS alert to reach out to him or her,” Anderson elaborates. “It’s a small thing, but it offers a big benefit.”
The challenge, he says, is in asking faculty members who are used to being in charge to trust the system and to spend their efforts determining where and why students fall short of attaining certain scores.
In terms of integrating adaptive learning into the L&D function, “step back,” Anderson says. “You probably can leverage some things you’re doing already. Before buying new software, understand how you will use it on a fundamental level. Identify what you want to achieve before investigating technology. The LMS, for example, has a lot of data on student behavior that can be aggregated.”
That data can inform curriculum design by identifying the zone of proximal development—the peak performance level a student can reach without additional help, which, thus, identifies where extra materials (such as presenting the same material in a different way, showing real-world applications, various contexts, and in different formats) may help him or her master the material.
“Particularly in courses where students progress at their own pace, an adaptive learning approach can be very helpful,” Anderson says. Right now, adaptive learning is geared toward hard skills with definitive answers. In the next few years, he predicts advancements in AI will help adaptive learning to become common for soft skills instruction, too.