Distance or face-to-face learning
Isn't there also a certain unconsciousness in considering that we can transmit skills involving exchanges between human beings and an engagement of the body using digital modalities?
Publish at February 26 2018 Updated May 22 2024
The amount of data generated by learners during online training is immense. What was the trainees' equipment? How long did they stay on a resource? Did they go all the way to the end? Did they click on links? What were their results at the end, etc.? We can know it all!
But what are we going to do with this data? How can we absorb it? And how do we organize power between trainees, trainers, training centers and customers?
Training platforms produce data continuously. Every connection, every click is recorded. Everything pedagogues have ever dreamed of knowing about learning profiles, the rhythms that encourage concentration, the type of assessment adapted to a given learning style, is within reach, provided we take the trouble to analyze this data... . The data could then be used for continuous improvement.
In its document "The 5 C toolkit", Elucidat suggests collecting information before, during and after training. The selected data answers the following questions:
Other platforms also offerlearning analytics, which are often highly relevant too. And face-to-face training courses draw inspiration from them. Learning Catalytics, for example, enables teachers to receive real-time feedback on their teaching, enabling them to adjust their approach.
So we have the means to evaluate every screen page, every quiz question, every serious game action... But what can we do with it? The temptation to always improve or refine what already exists comes up against a lack of time, and the cost of production or improvement.
The idea of offering differentiated courses according to the profiles that emerge or the results obtained is seductive. But unless you're training tens of thousands of people in large groups, the economic viability of the idea is not certain.
On the other hand, the platform's data can tell us whether an event that we're proud of, that required a lot of energy and resources, and that the customer really liked, is really appreciated by the trainee!
Other questions also arise. Let's imagine that on a legal training course, my platform shows me that certain categories of learners spend more time on diagrams and videos than on legal texts or case law. Should I offer them more videos, diagrams and animations? Or, on the contrary, should I encourage them to consult the source documents, even if they are more arid? It's a pedagogical choice. I can also imagine not choosing and leaving several resources available for the same training objective. And in this case, I trust the learner to determine what's best for him or her! In the first case, adaptive learning avoids overtaxing the brain's plasticity. But is it a good thing if all training devices are always adaptive?
Although they don't lend themselves well to statistical processing, open-ended questions, discussion forums and even learner observation are essential for understanding what the trainee perceives and feels. This is an essential complement to the more or less automated processing of the flood of data that e-learning platforms can produce.

There is also the danger of producing recipes that would standardize training courses. We're thinking here of film script design. Statistical studies have shown when it's best to build suspense or when to leave room for lightness. However, if these recipes are applied, if there's a lack of risk, a touch of madness, a little offbeatness, the film just doesn't work...
Making training and its mechanisms transparent, obtaining data on the slightest behavior of trainees, seems to serve a common interest that nobody can dispute. In his article "Learning analytics and the Phantom Menace", Andrew Hope tells us that we hardly dare question a technology that aims for more training and engagement.
And yet Andrew Hope shows us that continuous assessment transforms the way teachers and students see themselves, and the power relations between them. He draws on Michel Foucault, the French philosopher and historian who died in 1984 and wrote Surveiller et punir in 1975. He uses the image of the panopticon, Bentham's 19th-century architectural invention. Bentham imagined a tower at the center of the prison, giving a view inside each cell. The person at the center can call in the guards, provoking a reaction if an inmate behaves inappropriately. As the guard cannot be seen, no-one knows at any given moment whether they are being watched or not. So everyone ends up behaving as if they're being watched at all times, and every inmate is quick to explain the rules to new arrivals. The rules of power end up being internalized.
The transparency provided by e-learning platforms on the activities of learners, trainers and training centers is not neutral. It enables precise control, and therefore necessarily has an impact on behavior. Andrew Hope reminds us that Michel Foucault always associated the notion of power with that of knowledge.

Gradually, for reasons of statistical objectivity, trainers are running the risk of surrounding themselves with numerous dashboards and data of all kinds. Even if the efforts made to simplify these data, to make them easier to read and to bring out the salient facts are considerable, it all takes time.
But this analysis work should not replace exchanges with learners. On the contrary, it can be very useful in preparing a teaching interview. Asking a learner about his or her strategy, learning environment, difficulties or preferences remains an essential step.
The trainer who takes a certain pleasure and, let's face it, a feeling of mastery in collecting data on his trainees or students will soon discover that his line manager or client also has his own dashboard, and is also monitoring his activity. In a panoptic system, the guard has considerable power over hundreds of inmates... But he too is controlled by the interventions he provokes. An activity that produces data seems more objective and reassuring than one based on intuition and improvisation. But when dashboards multiply, they freeze methods and limit the scope for innovation.

The promises of learning analytics are seductive, and are bound to change the way trainers and teachers work. But the sense of control they bring must not push the pedagogical relationship to one side. Rather than making the trainer an all-powerful figure who knows everything thanks to his dashboards, this data should help the trainee to situate himself in his progress and better understand his learning preferences.
They will undoubtedly serve as the basis for studies that will help to distinguish between teaching methods, and will therefore be very useful to trainers. Let's bet, however, that they'll keep a healthy distance when it comes to proposing recipes, and that they'll continue to invent and innovate!
Illustration: Frédéric Duriez
Resources :
Surveiller et punir. Naissance de la prison, Paris, Gallimard, 1975, 328 p.
Andrew Hope Learning analytics: Foucault and the phantom menace - accessed February 25, 2018
http://newmediaresearch.educ.monash.edu.au/lnm/learning-analytics-foucault-and-phantom-menace/