Does EdTech Disappoint, or Do Our Expectations?

Different philosophies lead to different diagnoses of EdTech failure

Educational technology carries an old promise in new clothes. Each generation arrives with its own emblem of transformation: radio, television, computers, tablets, apps, platforms, artificial intelligence. Each is introduced as if education were waiting for the right device, the right interface, the right technical solution to unlock its unrealised potential. Yet, again and again, the pattern is familiar: enthusiasm, investment, adoption, and then a quieter reckoning with limited or uneven effects on teaching and learning.

This disappointment is particularly striking because the technologies in question are not merely administrative systems sitting at the edge of schooling. They are tools explicitly designed for learning: digital books, educational apps, assessment systems, learning platforms, or, more recently, AI tutors. These tools enter the classroom with pedagogical intent. The providers of these tools claim that they will support educational activities, personalised feedback or assessment. If such tools do not deliver the promised gains, the natural question is to ask why. Why is impact difficult to realise with EdTech? Here are couple of possible answers to that question from some key EdTech researchers.

Distance between product and user

Many technologies are designed far from the lives of those expected to use them. The imagined learner is often cleaner, simpler, and more compliant than the real child. The imagined classroom is calmer than the real one. The imagined teacher has more time, and more training, more support from the leadership team and fewer competing demands. In this distance between production and use, the technology acquires its first weakness: it is built for an abstraction, not for a situated human practice.

Historical amnesia

EdTech is developed as part of innovation and the innovation language follows the narrative of novelty and transformation. Audrey Watters in her book Teaching Machines has described the “teleology” of educational technology as the persistent belief that technology represents an inevitable march toward progress, even though history offers a more hesitant lesson. Although the technologies change, the rhetoric remains, Watters writes. Transformation is promised, adoption follows, and then schooling absorbs the innovation without being fundamentally remade.

The Add-on Problem

Larry Cuban’s account of computers in the classroom being “oversold and underused” remains powerful because it names a recurring pattern: technologies may enter classrooms, but often as add-ons: drill practice, typing, presentation, supplementary activities. The device is present, yet the pedagogy remains intact. The classroom continues according to its older rhythms. Technology, in such cases, is unsurprisingly, not a catalyst for new practices and educational impact is out of reach.

Context shapes impact

No educational tool operates in a vacuum. Schools are social, institutional, political, and cultural spaces shaped by curricula, accountability systems, professional norms, relationships, resources, leadership, and so many other factors. Neil Selwyn’s critical work reminds us that technology is always entangled with such power and social conditions. Sara Hennessy’s work similarly foregrounds teachers’ professional learning, classroom practice, and relational conditions in influencing EdTech’s impact. A tool may be well designed and still fail if the ecology around it cannot support meaningful use.

Evidence of Impact as RCT or co-design

Perhaps EdTech disappoints partly because it is judged using methods that are ill-suited to educational technology. Randomised controlled trials are often presented as the gold standard, yet few EdTech interventions have demonstrated positive effects when evaluated in this way. Borrowed from the medical sciences, this approach has been questioned by educational researchers, who argue that classrooms are not clinical settings, teachers are not neutral delivery mechanisms, and learners do not receive interventions in fixed doses under controlled conditions. If technologies are co-designed with teachers and children, and if evidence of impact includes implementation processes and teachers’ experiences, then educational impact rests on a different kind of evidence base. Debates about what counts as evidence have persisted for years and continue to divide the EdTech field.

Many misalignments, many perspectives

These are just some examples but they show that the “disappointment of EdTech”, then, may be the result of many misalignments: between design and use, promise and history, technology and pedagogy, tools and contexts, methods and their meanings.

Different epistemological positions lead people to ask different questions, use different methods, and therefore reach different conclusions. The same is true in the EdTech case - our conclusions about its value depend on the assumptions we bring to the question.

If we see technology as a driver of inevitable progress, then disappointment appears as a failure of implementation. If we focus on design, we notice the distance between those who build technologies and the teachers and children expected to use them. If we focus on context, we see that schools are social, cultural, institutional, and political environments that no tool can bypass. And if we focus on evidence, we must ask whether the methods used to judge EdTech are adequate to the complexity of learning.

The question of whether EdTech works should thus always be preceded by an explanation of the underlying philosophy and epistemology of what counts as evidence and what counts as learning. This lecture, delivered as part of the AIEOU Webinar Series at Oxford University, explores these questions in more depth. Watch the video here or read the transcript here.