Among the promises from Education Technology resounding throughout the field these years, is the promise of adaptive, intelligent technologies helping teachers do their job better.
Adaptiveness is not magic. It is built into technologies step by step and “someone” decides, what counts as input in the system and what doesn’t. As an example, your rating of the movies that you liked on Netflix helps adapt the movies, which you will subsequently be suggested to watch. It also matters if you actually decide to watch something (an indirect “like”) and what other people, sharing your taste in movies, liked or decided to watch. Netflix could have chosen different parameters (and, I am sure, have other, less obvious ones at play already). But, the key word here being “taste”; no one is a better judge of your personal taste than yourself. The adaptiveness is driven by subjective markers and rightly so.
To build adaptive technologies in education is quite different from this. Unlike in the entertainment industry, teaching and education aren’t primarily reliant on “subjective taste” but on both objective and intersubjective markers. So, to create technologies that adapt to the student the same way a good teacher would, we can’t rely on what the students “like”. Teaching is about shaping the individual student in ways dictated by objectivity and intersubjectivity, and therefore each student aren’t themselves the best judges of, what they are learning or need to learn – as would be the case if it was just a matter of taste and entertainment.
Adaptive technologies in education need to interact with the student via different techniques, the most common of them being questions, that the students need to answer to demonstrate understanding or knowledge. Asking the right questions can help shape an accurate profile of the student, which the computer can decode and respond to. But there are other ways to interact with the student, that can help the computer get to know the student, such as;
- semi-scripted science experiments
- math problems build with more complexity than right/wrong outcomes and monitoring how the student is solving the problems
- error type analyses performed on some piece of text written by the student
After a number of interactions, the student’s profile is beginning to take form and the technology can suggest appropriate activities for the student or collect the information in small data packages, which will help the teacher gauge the learning, that is or isn’t happening for each individual student or for groups of students. That way the technology adapts to the student or it helps the teacher adapt the instruction.
So – since it isn’t magic and since the artificial intelligence driving the adaptivity of any adaptive technology isn't wise in the sense real teachers can be, it is of great importance that any tech company claiming to be adaptive ensure transparency for teachers, parents, and other real human authorities, trusted with the important task of educating the next generation.
At WriteReader, we have started building new layers in our technology that will support teachers in their ability to adjust the teaching, and you can follow some of the steps we take to do this here on this blog.
This post is a brief summary of a talk about intelligent, adaptive technologies given by Tashia Dam who is the CPO at WriteReader held in Copenhagen Tuesday the 23rd of August at a conference held by Confederation of Danish Enterprise about digitization in the educational field.