Transformational Edtech

Transformational Edtech

Transformational Edtech

The edtech industry is expanding rapidly. But which kinds of edtech are going to transform outcomes for students? By “transform” I mean change the way that students learn to make a substantial, positive effect in learning outcomes.

Types of Edtech

There are many types of edtech software, and edtech platforms, but not all of these have a direct effect on how the student learns. A lot of edtech is organisational, for example Learning Management Systems (LMS) and Virtual Learning Environments (VLE). These may improve organisational effectiveness and costs, but they don’t affect the learning process directly. Whether a student hands in an essay on paper or via the VLE doesn’t affect the quality of the essay.

It is clear that the pandemic has boosted the edtech market (Credit Suisse p16). The big winners have been companies supplying online tutoring and MOOCs (Massive Open Online Courses) to students missing out on school and university classes. These companies dominate the lists of edtech companies by market value (Yahoo, Holon). In addition, educational institutions have had to get to grips with distance learning and many have become subscribers of products by tech giants, such as Microsoft Teams and Google Classroom.

What these technologies have in common is that they connect students remotely to teachers and course content. Connectivity can increase access to education: for example, a course on a platform such as Coursera or Emeritus might be a cost-effective alternative to a university course. But connectivity alone doesn’t normally change how students learn; what matters most is what is shared via the tutorial, the MOOC or the video conference. From all the conversations I have had with educators during this time, it is clear that this didn’t change much. Word/Google documents, Powerpoint presentations, pdf files with learning materials and instructional videos are the backbone of online education. Of course, teachers have experimented with novel technology. A good example is Kahoot!, which allows teachers to easily create quizzes for students, and has become very popular – their sales grew 10x during the pandemic. But this is more the exception that the rule. In general, the first wave of the edtech revolution, boosted by the pandemic, has focussed on connectivity and is not transformational in the sense considered here.

What Will Bring About Transformation?

Happily, there are developments in the edtech market that are already bringing transformational changes to education. I want to outline two of them.

Making Use of Cognitive Science: A lot is known about how people learn. In particular, learning occurs by the acquisition of mental schemas – complex structures that link knowledge and create meaning. Knowledge of how to improve this process can be summarised as design practices that lead to good edtech design, for example: dual coding to lower cognitive load; spaced repetition and retrieval practice. Ofsted, the UK education regulator, has written about these in detail in their research findings.

Using Analytics and Machine Learning: Educational software can record every click or mouse movement that a student makes, what type of question they are having difficulty with and so on. There are even projects to measure eye tracking and facial expressions! This creates a vast pool of analytics data to improve learning outcomes. Machine learning can create AI models that automate testing, even for complex questions – allowing fast accurate feedback to the student and the teacher. This will bring a true revolution in how we learn.

Uneven Progress: STEM vs. Humanities

It has often seemed to me that this progress is leaping ahead in edtech for STEM (science, technology, engineering & maths) subjects, but not so for humanities. As both an engineer and a philosopher, I find this both intriguing and disappointing. Perhaps it is to be expected that technology can be applied to technological subjects. It is easy to see how educational robots and lab simulations fit into technology and chemistry classes – and hard to think of an analogue for humanities subjects. But I think the problem is more fundamental.

It is instructive to look at courses on the superb Khan Academy platform. Take a course in coding and you will get step-by-step instructions how to use one function, and then you will move on to another function. The feedback is automated and immediate. If you make a mistake it tells you what you did wrong – immediately. The same is true for maths. This method is really simple to follow and an effective use of analytics. Now look at a course in history. You get a series of video lectures to watch and texts to read. No interaction, no feedback. At the end there is a multiple-choice test of your knowledge. As any humanities teacher will tell you – this isn’t enough. You need to grapple with the subject to really learn – normally by considering the arguments, writing essays to set out your point of view and getting precise feedback from a teacher.

The problem is that learning in STEM subjects can be broken into tiny pieces and repetitively practiced, but learning in the humanities is far more ‘lumpy’. And the same is true for arts and social science subjects, to a greater or lesser extent. In the end, it comes down to acquiring the schemas. I acquire the schema for solving quadratic equations by having the method explained to me and then practicing on multiple examples with fast, clear feedback. Solving x^2 +3x +2 = 0 will help me solve x^2 + 5x + 6 = 0. And the schemas I acquire in STEM subjects build on each other in a straightforward way, so I end up using ideas from the maths classroom in my coding or engineering classes. But what is the schema for Napoleon’s retreat from Moscow? If it is anything more than a jumble of facts, it is certainly much more complex than the schema for quadratic equations. And how does learning about Napoleon help me with my essay on Hitler? Maybe there are some parallels in their histories. But the process of piecing these schemas together is much less straightforward than for STEM subjects since historical events remain largely sui generis.

The result is that Khan Academy is transformational for STEM, but not for humanities. Further examples are the coursework sites Uplearn and Hegarty Maths. They use sound cognitive science principles such as spaced repetition and corrective feedback, also using the analytics to learn what the student is doing well or badly and adjusting the course as necessary. But they don’t cover humanities subjects. Lectures and essays are too big ‘lumps’ to repeat or to gain a lot of analytics data from. And the same goes for all the big learning platforms such as Coursera. Look up humanities courses and you will find recorded lectures and texts to read, and maybe some quizzes. I have been there myself, doing a philosophy degree by distance learning. I needed to write essays and find a tutor to mark them to make any progress. To achieve a transformational change in learning, edtech needs to break humanities subjects into pieces, apply cognitive science principles and measure analytics to give actionable feedback.

Transformational Edtech For A Brighter Future

Edtech in STEM subjects will undoubtedly continue to forge ahead. But there are strong signs of interesting tech applications being introduced into humanities subjects. Data science is used to study authors’ writing styles, for example. And I am particularly interested by the burgeoning use of VR/AR tech in subjects such as history and archaeology. I expect this trend to accelerate.

Endoxa Learning is very much based on a vision of being transformational for learning and practicing critical thinking skills in subjects that require them – especially humanities. We visualise arguments, applying the cognitive science principle of reducing cognitive load by dual coding (see our article on this here). We break arguments into small steps, allowing learning and testing to be more granular and generating useful analytics. And we are starting to use machine learning to expand what is possible in terms of automated feedback for the student and teacher, to make it like a Khan Academy coding class. This is a road which leads to better outcomes for students of all abilities.

I will conclude by noting that the adoption of edtech must be accompanied by a strong ethical framework and the right perspective on its proper role. For example, data privacy rules have to be respected and this may restrict more invasive techniques, such as facial recognition technology. And in my opinion, the role of edtech is to help the teacher, for example by automating marking or pinpointing problems that individual students have, rather than replacing them. Let the computer do the routine work while the teacher coaches students exactly where they need it.


Julian Plumley, CEO Endoxa Learning Ltd.