What ETAG should say

The first meeting of ETAG, attended by Michael Gove and Mat HancockThe Education Technology Action Group, set up by the DfE and BIS and announced by Matt Hancock at BETT, has made a shaky start. This is what I think they should end up saying to Ministers.

In my post “Land ho!” of 16 December, I welcomed the noises being made at that time by Matt Hancock, Minister for Skills & Enterprise at BIS, about the government’s new, more proactive approach to education technology. This led to the announcement at BETT on 23 January of a new advisory group, the Education Technology Action Group, to be chaired by Stephen Heppell. The most that could be said so far is that ETAG has had a slow start.We didn’t hear anything of substance until 23 April, when it published a series of questions that are to form the basis of a consultation, which is to run until 23 June. In my view, the questions are not particularly helpful. Nor have they attracted any significant response in the first couple of weeks, there having been only a couple of dozen substantive tweets using the #etag hashtag. But I am looking forward to engaging in the consultation and, by way of encouraging the debate, publish below my own views on what ETAG should say to Ministers.

ETAG’s remit

Before analysing the ETAG consultation questions, it is worth copying the ETAG remit, which is published on the ETAG home page.

The Education Technology Action Group ETAG will aim to best support the agile evolution of the FE, HE and schools sectors in anticipation of disruptive technology for the benefit of learners, employers & the UK economy.

Setting aside the ugly split infinitive, I think the remit is unhelpful because it suggests a rather passive approach, waiting patiently for the disruptive technology that I suppose we should expect to arrive any time soon. But what happens if Godot isn’t coming? And how much trust can be put on recommendations for how government should react to something that hasn’t happened yet?

Maybe members of the ETAG group had the same reaction to this remit as I do because there appears to be an informally agreed codicil, which runs as follows:

…and we have also been asked to identify any barriers to the growth of innovative learning technology that have been put in place (inadvertently or otherwise) by the Governments, as well as thinking about ways that these barriers can be broken down.

This is the interesting bit of the remit—and the one that I assume everyone will be focusing on.

The ETAG consultation questions

Before I publish my own policy proposals, let me comment on the ETAG questions, which I paraphrase in my comments below. I also include comments that I tweeted on 8 May.

Learning will be significantly more global – so how do we get there?

ETAG TweetA key problem with the tone of most of the questions is that they assume that everything is going along just fine and we are all headed for what members of the group may regard as ed-tech nirvana. This is an aspect of Stephen Heppell’s undoubted appeal as a speaker: he exudes optimism. But the assumption that lies behind the remit is that this is not going to happen without some sort of government intervention. If you accept the assumption of the ETAG remit, learning is not (inevitably) going to be more global; if you accept the assumption of the question, then ETAG is redundant.

I would dispute that the sort of globalisation that Stephen Heppell advocates, in which students network with their peers in other countries and become good international citizens, is going to have a significant impact on pedagogy. It may be a desirable end of education—but that it is a curriculum issue, not a pedagogical one, and therefore out of scope. Is it really such a great help to chat about your six times table with a Korean friend on Facebook? What are recommended by ETAG as “simple pragmatic ideas like this Primary School YouTube clip” appear to me to have no pedagogical merit at all but merely to represent social media trivia.

What may indeed be global is the market for education technology products—but that is something different.

Servers and services will be cloud based…


I think that this is an important topic for discussion, though still somewhat unfocussed. It is certainly an uncontroversial observation about the direction of generic IT markets—but it is difficult to anticipate what effect this will have on education technology. In my view, it will make education technology more viable by consumerizing the technology and making its acquisition simpler for schools with limited technical resources. It is a trend that removes the amount of expertise that is needed locally to install and administer complex systems. It will also allow (subject to privacy concerns, discussed below) the collection of the significant quantities of AI training data by commercial providers that will enable advanced software functionality.

None of these issues represent barriers.The particular problem which is brought up in my mind by the ETAG blurb concerns the “private/public mix”—raising the spectre of a corporate takeover of education—and it is a spectre that may well have some substance. The rhetoric about learner-centred education, multiple identities and learner ownership of data is all somewhat naive. Learner-centricity on the web is often a front for unacceptable forms of corporate control: Facebook is not run by teenagers (OK, it’s run by twenty-somethings but they are wealthy, corporate twenty-somethings). While I am the first to encourage the greater involvement of industry in education, the role of industry should be about providing the means of educating, not taking over control of the ends. Proprietary lock-in, a massive problem in our current ecosystem, is a major threat to the governance of education in the future. Interoperable learning management systems are therefore essential for reasons of governance and control, as well as of pedagogy.

On-line learning for all – as an entitlement


“Entitlement” is a euphemism for “compulsion”. It is the main reason why so many ed-tech initiatives have turned out so catastrophically over the last twenty years that poorly conceived and executed initiatives have been forced on the system by well-meaning governments. But I worry that there is a preponderance of people on the ETAG group who rely on government funding for their livelihoods and are therefore predisposed towards top-down, government-driven schemes.

I am not worried that ETAG will actually spawn any such prescriptive schemes—not this side of the next election, anyway. The approach has been so comprehensively discredited by previous experience and is so alien to the views of current Ministers, that If ETAG makes any such recommendations, they will surely be given short shrift. But that would be a terrible waste of a good opportunity.

To be nit-picking, it is also worth commenting on the talk about “online learning”. This has been around for many years in digital forms, and for many decades before that as non-digital distance learning. It addresses problems of learner isolation but does not address problems of scale and effectiveness in mainstream education. We should use the term “ed-tech” instead of “e-learning” or “online learning” for these issues, which are the ones that really matter.

Student ownership of their own “big” data


etag_tweet_5What are commonly called “privacy” issues will become urgent once we start making headway with data—and this is certainly something that the ETAG group needs to address. But it is not yet an issue because we do not yet have any data in schools: I recommend Donald Clarke’s paper on this point at .

When we do come to address the issue, concepts of “ownership” of personal data will not be helpful. Does this concept mean, for example, that if I were a paedophile, other people could not warn each other about me? My sexual predilections count as personal data, after all. Or, if you think that the legality of my behaviour is the issue, could a prospective employer not insist on seeing a reference from my current employer or school before taking me?

Then there is a point about pragmatism. As ex-Sun CEO Scott McNealy said in 1999, when it comes to the internet “you have zero privacy anyway—get over it”.

This is why I think that the concept of “trust” is more useful than “privacy” or “ownership” of data. Data collected for the purpose of education, in the context of a particular fiduciary relationship, should be used only for education in that context. But this restriction has what some people might view as two loopholes:

  • that education includes assessment and qualification, not just learning;
  • that education is a collaborative effort involving many providers, both within and without the institution that is primarily responsible for the learner—and rules that restrict the sharing of data between those partners are likely severely to impede the effectiveness of new forms education technology.

For these reasons, I suggest we need a machine readable, data management description language which can specify the data handling protocols that will be followed by particular educational providers, their suppliers, sub-contractors and associated institutions. These machine-readable protocols can then be expressed in human-readable consent forms, government regulations and international standards, as well as (critically) being automatically enforced by education technology software systems. This will enable the free exchange of data for legitimate purposes, while drawing clear lines that limit the dissemination of data beyond a trusted boundary, and protecting practitioners from tiresome regulations and serious liabilities.

Creating that data management description language will not be a trivial task (there are particularly complex issues regarding the reliable anonymisation of data, for example)—but in my view it is doable, and what is more, worth doing. And the reason that it is worth it is not because I favour the imposition of restrictive rules but, to the contrary, because I want to enable the sort of data-sharing that can transform education. This sharing will only happen if people are given effective reassurances that data is not being misused. The recent political furore about the sharing of anonymized medical data for research purposes, or the failure of police and social services to share data about the Birmingham grooming gangs are two examples of what happens when clear, enabling rules are not put in place.

Technology will be even more personal

In response to the ETAG statement that

With wearable and personal [sic], “powering down” or “we don’t use them here” will not be an option (“take off your smart glasses!?”) so the challenge for education is to take best advantage of the opportunity. We can’t “lock and block”, we can’t “ban and wait”, we can’t appropriate each technology with a “standard” acceptable version.

I tweeted:



I am not against exploiting the technology that learners bring to the classroom, but only if it can be used productively in the classroom context. The BYOD lobby has, on the whole, resolutely ignored the problems of distraction, the problem of destructive addiction to gaming or social networking, and the problem of equity in the classroom created by BYOD. I do not say that any of these problems are insuperable—I just do not accept that a consultation like ETAG should hard-wire in assumptions that say “we can’t ban x”. We can do what we like.

Part of the problem with the BYOD argument is an unhelpful underlying view that teenage street culture is somehow superior to the dusty knowledge to be found in the school library. The idea that all our learning objectives must change now that we are in the twenty-first century, or Marc Prensky’s discredited idea of the “digital natives”, or that textese is a satisfactory replacement for traditional grammar, all tend to this position, which is essentially subversive of formal education.

My guess is that hardware providing satisfactory levels of performance will become so cheap that providing standard classroom response systems will often be a preferable solution to BYOD. Ultimately, its not the hardware that’s missing. On the whole, hardware is cheap and plentiful. It is the education-specific software, which in most cases does not exist at all.

We will know a LOT more about how we learn

This may be true, but only after we have successfully implemented education technology in our schools, which is the problem that the ETAG group is meant to be addressing.


Better measures of performance



In my view, this is a critical and difficult question—difficult partly because so many educationalists appear to be hostile of the idea of measuring learning at all. But beyond the reaction encapsulated in my tweets above, I will leave a full discussion of what is a complex issue to another occasion.

New emerging teaching and learning

I didn’t tweet a response to this question—but it goes to the heart of the problem with ed-tech. The assumption that teachers are going to lead innovative new forms of teaching and learning, without access to the “tools of the trade” that only industry can supply, is the key misapprehension that has dogged us over the last twenty years. The evidence on this approach is already in—and we know it doesn’t work.

Which queues my own proposals for an industry-led approach.

My policy recommendations

The following recommendations can be downloaded as a PDF at Stimulating innovation in education technology. If you find the writing a little dense: that is because I have tried to fit it all onto three sides of A4:

  • page 1 provides background;
  • page 2 covers my three policy recommendations;
  • page 3 anticipates the outcomes of these recommendations.



Barriers to the development of successful education technology

Expectations that digital technology will significantly improve educational provision have not yet been realised. Progress has been inhibited by governments that have tried to lead innovation themselves; that have looked for leadership to non-technical teachers; and that have failed to perceive the need for education-specific, technical innovation of a kind that can only be led by industry.

The role of government

The UK government is currently reviewing its policy in an attempt to find a middle way between government prescription and complete laissez faire. This paper responds to that review by proposing policies to promote a competitive and innovative market for education technology, without attempting to pre-judge what such a market will produce.

The importance of interoperability

Many educationalists fail to understand the critical importance to technical markets of interoperability, highlighted by BESA’s 2008 Policy Commission as a key lever for change. Without the three-pin electrical plug, the market for electrical appliances would not exist.

The iterative, dialogic and adaptive nature of teaching means that education technology is almost wholly dependent on good data interoperability. Neither instructional software nor learning management software can work effectively in isolation: the former requires automatic access to data about students, their past performance, their current proficiency and their future learning objectives; the latter requires access to data describing the outcomes of previous learning activity.

The technical, education-specific standards required for such data interoperability do not yet exist. In these circumstances, there is an increasing realisation that MOOCs have over-promised their ability to deliver adaptive learning systems and useful learning analytics. Meanwhile, uncompetitive markets based on previous bureaucratic procurements perpetuate themselves as key suppliers leverage their dominant positions.

Many attempts have been made to address this problem: all have failed. There is a common misconception that technical standards for interoperability can either be devised and mandated by government or developed through consensual agreement. Such approaches fail because interoperability specifications have a close relationship with product innovation, driven by market competition.

It follows that government should refrain from leading work on technical standards. It should instead reward interoperable outcomes and support processes that decouple the development of initial technical specifications from later standardisation.

Recommended government interventions

Establish a technical specifications incubator

The government should provide seed-funding for a technical specifications incubator. The incubator would provide a governance framework under which self-selecting groups of suppliers could publish open specifications to improve the interoperability of software. After the technical efficacy of specifications had been proved by working demonstrations, the incubator would award badges to compliant products, informing consumers of the benefits that they offered and guaranteeing the reliability of the proposed solutions.

Only when a particular specification had achieved recognition in the market could it proceed to formal standardisation through BSI. By driving international standards, UK industry will gain a head start in an important, emerging, global market.

Establish an online catalogue of education technology products

The Department for Education should provide seed funding for an online catalogue of education technology products. By identifying and describing products authoritatively and transparently, the catalogue would support the provision of associated services, such as:

  • certification, as provided by the incubator and other standards organisations;
  • online marketplaces and other e-commerce platforms;
  • open source communities;
  • product review sites.

The catalogue would support a heterogeneous solution to the problem identified by Sebastian James, whose Review of Education Capital called on government to fund “an online price comparison catalogue” for education technology products.

The online catalogue should present a low threshold of entry to new products, allowing reputation rather than regulation to determine their ultimate success.

Stimulate the development of a professional press

Healthy markets require informed consumers. Education currently lacks a professional press for education technology of a type that would:

  • provide comparative product reviews;
  • discuss the appropriate use of product types in the classroom;
  • examine the evidence for the efficacy of product types and their associated pedagogies, linking to academic research where available;
  • encourage professional debate about what works best.

The Department for Education should open a conversation with interested media groups to establish what further measures can be taken to stimulate the development of such a press.

Anticipated outcomes

The purpose of the following horizon-scan is not to specify but rather to anticipate the outcomes that will follow from the enabling actions recommended above.

New types of education-specific technology

Current methods of education fail to provide feedback to students at sufficient quality and scale. This weakness will be addressed by new forms of instructional software, that will implement at least three fundamental paradigms to offer feedback:

  • by machine feedback (automatic essay marking, quizzes, serious games);
  • by exploratory environments (digital simulations, creative tools and “micro-worlds”);
  • by digitally mediated and structured peer instruction.

These instructional strategies will be complemented by learning management software that will manage assignment, activity sequencing, learning analytics, and e-portfolios.

New standards of data interoperability

The pivotal requirement for data interoperability is to enable learning platforms to launch individual learning activities and harvest learning outcome data (interactions data, marks and grades, competency statements and artefacts created by the student).

Data will also be shared between different components of the learning management system and between different activities (e.g. a spelling quiz might prioritise words that a student had been getting wrong in recent written work). Adaptive sequences of learning activities will capture proven learning designs in reusable form; and education-specific hardware (such as classroom response systems) will be able to interface to any third-party software, encouraging innovation in learning content.

Anticipated benefits

By delegating the routine aspects of teaching to machines, more targeted use can be made of highly-qualified graduate teachers, whose jobs will become more interesting, more productive and better paid.

The automatic and timely sharing of information about learning will support greater collaboration between members of the teaching team, more specialisation in roles, better supervision and mentoring; as well as encouraging the involvement of parents.

Better monitoring of student progress will deliver more consistent outcomes, while personalised learning pathways will combine academic rigour with the diversity of provision that is appropriate to a technically advanced, liberal society.

The encapsulation of pedagogies in digital form will help disseminate best practice through the profession.

The collection of data to support teaching will also support monitoring and research, providing empirical evidence for what works best, supporting continuous optimisation of pedagogy, and maintaining minimum standards.

Learning analytics systems will compare the predictive reliability of different qualifications, helping to pioneer new more authoritative approaches to assessment that offer greater continuity between formative and summative modes.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s