Adaptive learning uses artificial intelligence to offer personalised training to learners. This approach is a powerful lever for accelerating an individual’s skills development by tailoring training more precisely to their needs.
What exactly is adaptive learning?
We all learn differently. It is not physically possible to have a trainer support every single learner, but adaptive learning – which was introduced into American universities some fifteen years ago – manages to achieve something close to it.
Adaptive learning is a pedagogical approach that provides an individual with an evolving learning experience, thanks to the support of new technologies and our knowledge of cognitive sciences. It uses artificial intelligence to offer learners training programmes that are adapted to their profile or career path and can be followed in a personalised way.
Personalisation involves:
- Training paths (i.e., the choice of modules and their order)
- Content (particularly the level of difficulty)
- Teaching methods
For example, if a learner scores poorly on a mid-course assessment, the adaptive learning platform will suggest a module to help the learner revise the subject in question.
How does adaptive learning work?
Adaptive learning relies on algorithms capable of analysing large amounts of data about a learner’s performance in real time. This data (from the Learning Management System) is collected beforehand and as the learner progresses throughout the course. Information analysed by the algorithms includes training courses they have previously taken, the time spent on each training module, their level of subject mastery, their level of completion of each module, their behaviour during the training course, and more.
By cross-referencing all this data, artificial intelligence defines unique learning paths for each learner to help them improve their skills. More generally, artificial intelligence is also capable of learning from users, conducting what we call ‘machine learning’. If many learners fail a single exercise, for example, AI will lower the level of difficulty. Thanks to this approach, adaptive learning software can adjust training courses virtually in real time and in an automated manner.
Micro adaptive and macro adaptive: what are the differences?
Adaptive learning works on two levels: macro and micro adaptive learning. The first applies to the whole program, while the second is concerned with each module within the program. In the micro adaptive framework, a quiz, for example, offers questions to the learner based on his or her answers to previous questions. The content therefore adapts to the learner's interactions.
When adaptive learning works on a macro scale, a course is adapted to the learner's exact needs in terms of skills. The further they progress in their training, the more the content will be refined according to the knowledge they have acquired. In this case, a self-assessment, which may take the form of a quiz or an interactive game, is offered to the learner to find out where he or she stands on the learning curve for the target skills. This level of mastery will then impact the granularity of the training. Artificial intelligence will only offer the modules that are truly useful to the learner.
What are the benefits of adaptive learning?
The first benefit is efficiency. Personalised, tailor-made training inevitably gives better results than training that is standardised for all learners, who have neither the same expectations nor the same level of knowledge on a subject.
Moreover, there is less ‘waste’ in adaptive learning, since learners only learn what they need to. The learning experience is also improved. In traditional linear courses, completion rates are not maximised, as the training content is sometimes too far removed from real needs and not sufficiently adjusted to what will genuinely help learners progress. With adaptive learning, learners are more engaged in their training because it is relevant to them. Their motivation is therefore increased tenfold, as is their level of commitment.
Once the course is over, learners retain the knowledge they have learned more effectively and improve their skills more quickly, even when the content is dense and the subject matter complex (for example, with hard skills training). In this respect, adaptive learning responds to the challenge of anchoring knowledge.
Case Study: IFCAM experiments with adaptive learning
In France, several experiments in adaptive learning have been conducted. In 2019, IFCAM, the Crédit Agricole Group's university, deployed an adaptive learning program. This project from a major bank involved a Bachelor's degree course, made up of 7 teaching blocks. Each block was made up of several MOOCs that employees followed using the Moodle platform.
Within one of these MOOCs, employees could take ungraded quizzes to assess their mastery of the content. Adaptive learning made it possible to offer each learner questions based on their strengths and areas for improvement. The implementation of this approach finally led to a 4-point increase in final exam results – from a 91% pass rate in 2018 to 95% in 2019.
Cegos initiated an adaptive learning approach as part of a project aimed at supporting HR managers new to their posts. A survey of mastered knowledge was created to evaluate the skill level of the trained populations, and the course was adjusted to align closely with each learner’s expectations.
Challenges for training professionals
Training professionals wishing to capitalise on the adaptive learning approach face several challenges. The first is the need to access huge amounts of quality data and the ability to take advantage of it. Secondly, implementing adaptive learning means being able to transform your training content so that it is as granular as possible. Finally, it requires having training content in different formats (videos, checklists, e-learning activity modules, etc.)
Adaptive learning actually helps trainers to guide their support, by tracking the progress of participants and the levels of learning achieved. They will thus be able to better personalise their contributions during live sessions.
Professionals are certainly ready to take up all these challenges. According to the Transformations, Skills and Learning Barometer conducted by Cegos in July 2022, 53% of French HR professionals say they want to personalise their training more, compared with 43% in 2021.
It appears adaptive learning has emerged as one of the major training methods. The following decade could therefore see AI applied more comprehensively throughout the training industry.
This article was written together with Aurélie Tachot, a journalist and employment specialist.
