The other day a very talented and committed member of my team suddenly looked very worried and asked me: "What if we're wrong?"

"What do you mean?" I replied.

"What if we're wrong about digital technology and education? What if it doesn't make any difference? There's no evidence from previous trials that technology makes any difference to attainment. What if we're wasting money?"

I wish I could say I made a rousing speech that convinced her and everyone around me that we were on a noble mission that could, and inevitably would, result in a profusion of digital education experiences that would achieve greater impact (at, of course, a lower cost) than the world has ever seen before. 

Instead I chuckled nervously, said, “Yes, we ought to have a strong response to that”, and rushed off to pick my daughter up from childcare. 

But I've been thinking about it a lot. The first conclusion I came to on my way to the tube (call it l'esprit d'escalator) was that, right or wrong, we have a responsibility to look at digital technology in the context of education. Connection to the internet is the most powerful mass communication tool since the invention of the book (I'm willing to have that argument if anyone wants it). It has affected how we work, talk and think - not to explore education in this context seems wilfully short-sighted.

The lack of evidence for the success of technology in the classroom is, by and large, indicative of a lack of evidence, full stop. Instead, we have anecdotes, individual case studies, expensive and irreplicable islands of excellence and futurology (guesswork). 

One thing we're pretty sure of is that unless technology and teaching change together, the impact is always going to be minimal. A digital replication of analogue practice is never going to change much for learners (I'm looking at you smartboards and ebooks). Technology and pedagogy must develop in partnership in order to, as Michael Fullan puts it, “lift the lid off learning”.

This means there's a bit of a journey to go through before we can even get to rigorous evidence. First, we need promising technologies (more on that later) that suggest new models of teaching - to achieve this requires technologists and teachers to work together to define both the product and its use. 

Once a replicable model of technology-enhanced pedagogy has been defined in the classroom, if it shows promise, it must be taken to a statistically relevant scale. This has all the usual problems of fighting firewalls, connection speeds, access to hardware, etc. in at least 30 schools. 

Only at that point can we conduct an RCT or other rigorous evaluation.

And this isn't a once-off process. Digital technology moves so fast and has so many variants that this process will be required again and again as new products, that inspire thinking, come to market.

At Nesta, what we hope to do (and need partnerships to achieve) is create a system where academics, technologists and teachers can work together to identify areas of promise for technology, articulate learning challenges where digital technology might be useful, and then collaborate to create useful educational models where teaching and tech work together.

The first step we've taken is to commission a report that will be published later this month. It takes a very wide look at the existing landscape of educational technology. It maps these, not by product type or hardware required, but against learning acts that we know have an impact - for example, effective feedback or collaborative learning. 

The results are intriguing, not least when it reveals that the narrow range of learning acts that commercial products support. A picture begins to emerge where there are big areas of promise for the development of really useful new digital technologies to support innovative and effective pedagogy. 

The next step is to use the report to galvanise tech companies, teachers and academics to work together, and to create new models for tech-supported pedagogy that are ready for evaluation. Only then can we find out if we're right or wrong.

And what if we're wrong? What a wonderful piece of progress, to know for sure that an avenue is no longer worth exploring. 

But what if we're right...?