Writing for LSN in 2014, the late Maurice Holt noted the ‘transatlantic bridge tendency’ in English education policy – which, for more than three decades now, has ‘taken its cue from developments in the US’. Since the 1980s, England and the US have pushed the market-based reform of state education further than any other countries. This process – the marketisation of public education, and its partial privatisation, by means of academies here in England, and charter schools in the US – has been driven forward by systems of ‘accountability’ which involve frequent mass testing of students.
Recently, however, there have been some puzzling developments. Here in England, Ofsted has just announced that it will no longer grade schools on ‘outcomes’ -- i.e. test and exam results -- but rather on ‘quality of education’. Nonetheless, the government continues to roll out new standardised tests: a times tables test in Year 4, a new baseline assessment in Reception (which one test provider has described as ‘verging on the immoral’).
Significantly, both of the new tests will be computer-based. A feature of the schools reform movement on both sides of the Atlantic has been the quiet but very determined promotion of education technology, or ‘ed tech’. England is a few steps behind the US in this respect. But we have already seen the Coalition’s ‘Tablets for Schools’ initiative; Gove’s frequent meetings with Rupert Murdoch and other News Corp execs, while a Murdoch company was busy developing a ‘tablet-based curriculum’ (here); DfID’s support for tech-driven ‘low-fee’ private schools in Africa (here); and the launch of lobby groups like Edtech UK. At the same time, the bigger academy chains are already experimenting with various forms of computer-based instruction: examples include Ark Schools’ ‘blended learning academy’, United Learning’s ‘United Classroom’ and DRET’s ‘e-learning’ initiative. As Tamasin Cave has argued, privatisation is turning English state schools into ‘a magnet for tech interests’ (see here and here).
The new ed tech is made possible by networked computers -- more specifically, cheap, lightweight tablets and laptops -- and the internet. It gives schools, and other organisations, the capacity to collect, store and manipulate vast amounts of data: students’ test scores above all, but masses of other data too. We are seeing the emergence of what might be called a platform model of education, based on online ‘learning management systems’. But few people are aware of how rapidly this particular model of tech-based, ‘data-driven’ schooling is spreading through our education system – or how hard this model is being pushed by powerful actors in the political, business and financial worlds.
And the tempo is picking up. Damian Hinds appeals to the tech industry – ‘both the UK’s burgeoning tech sector and Silicon Valley giants like Apple and Microsoft’ – to ‘launch an education revolution’. The DfE’s working group on data management advises stressed-out teachers that ‘electronic systems offer the prospect of quicker and simpler [data] collection and almost real-time analysis’ -- and that schools’ use of these systems will ‘support the MIS [management information system] market to develop and diversify’. The very latest data source, the new on-screen maths test for eight-year-olds, is pitched by the DfE as an ‘innovative use of technology in testing’. And so on.
Meanwhile, Amanda Spielman bemoans the ‘tyranny of metrics’ and lectures audiences on Campbell’s Law (for the latest example, see here).
For anyone trying to make sense of this situation, the work of the American blogger Alison McDowell (aka ‘Wrench in the Gears’) is required reading. A parent activist who was part of the national ‘Opt Out’ campaign against standardised testing, Alison was one of the first people to see that the education reform movement is entering a new phase -- which she calls ‘Ed Reform 2.0’. Her work is as relevant for teachers, parents and campaigners here in England as it is for those in the US, and she kindly agreed to be interviewed for the Local Schools Network and Privatising Schools.
MB: Your research is closely tied to your experiences as a parent with a child in the public school system in Philadelphia. Could you say more about this? Is there something about the Philadelphia school district which brought the issues into focus for you?
AM: Philadelphia has a large, under-funded, urban school district. A significant percentage of our families live in poverty. For the past sixteen years, the district has been run by a School Reform Commission, which was effectively controlled by the state government. In other words, Philadelphians had no say over what happened in our schools.
Throughout my daughter’s elementary years I was active at the school level, volunteering and fundraising. I occasionally testified at public meetings, but mostly stayed in my lane. This changed in 2013, however, when the School Reform Commission decided to close 23 schools and lay off over 3000 teachers. Scores from end-of-year standardised tests were used to justify closures and turnover of operations at ‘underperforming’ schools. At the same time, it was decided to implement a punitive school report card programme, an initiative underwritten by the Michael and Susan Dell Foundation (Dell Inc.’s philanthropic arm).
A number of parents and teachers banded together to launch a campaign around opting out of standardised state tests. We thought that if we could encourage enough parents to boycott the tests, it would make it harder for officials to use test scores as weapons against schools and teachers.
There was a national ‘Opt Out’ movement for several years. I was part of a network of loosely affiliated chapters under the umbrella of United Opt Out. In Pennsylvania, we have a legal right to opt out of the tests, on religious grounds. Nevertheless, it surprised me when, in 2015, officials agreed to my request to translate state-provided literature about opting out, so that all of the families in the district would know their rights. District leadership had a reputation for putting up roadblocks in the face of even basic requests. So why would they readily co-operate with an action that would weaken their position with regards to testing?
In October 2015, the Obama White House put out a Testing Action Plan. This drew attention to the ‘undue stress’ caused by ‘unnecessary testing’, and called for an end to the ‘“drill and kill” test prep that is a poor use of students’ and educators’ classroom time’. Opt Out campaigners were excited. This seemed to be the breakthrough that we’d been waiting for.
In fact, it was a good, old-fashioned ‘bait and switch’.
MB: Can you say more about that? Amanda Spielman, a former financier, last year became the head of Ofsted, England’s school inspection agency. Since then, she has regularly attacked test-based accountability, describing schools as ‘exam factories’ where children are simply ‘jumping through accountability hoops’. We’re now being told test and exam results will no longer be the main focus of school inspections. Which is all very strange, given that Ofsted is, and has always been, a key driver of the system.
I think we are seeing the same agenda playing out in both countries. Both here in the US and in England, the education reform movement began with punitive measures which put public education systems under enormous stress. We experienced this under George W. Bush’s No Child Left Behind Act of 2002 (NCLB), which put the full power of the federal government behind test-based accountability. NCLB caused grave harm, driving large numbers of teachers out of the profession, allowing charter school chains to expand rapidly, and giving a huge boost to the testing and test prep industry. And, of course, NCLB normalised the idea that the education of human beings is something that involves relentless testing, and the collection of huge amounts of data on children and young people.
Obama’s Testing Action Plan seemed to mark a change of direction. The Plan was followed by the Every Student Succeeds Act (ESSA), which became law in December 2015. Many hoped ESSA would put an end to over a decade of ‘test and punish’. But, in fact, it simply changed the nature of the testing by emphasising ‘formative assessments’ taken multiple times a year.
ESSA broadened accountability measures beyond scores in reading and maths tests. Now schools are expected to track student ‘growth’, graduation rates, proficiency for English language learners, and a fifth measure of school quality – such as social-emotional learning, school climate, college or workforce preparedness, or parent engagement.
The shift to ‘multiple measures’ means that the amount of data being gathered and processed by public schools has increased exponentially. And this is where the tech comes in. If you look at the provisions embedded within ESSA, you will see that the ‘high quality’, ‘smart’ or ‘innovative’ assessments mentioned in the Testing Action Plan -- supposedly the humane alternative to the big, end-of-year, standardised state tests -- are essentially ‘technology-based’, to use the language of the Act. There are explicit references to tech-driven school models like ‘blended learning’. With ESSA, federal education policy enters the age of the online learning platform and ‘data warehousing’.
Back in 2015, my colleague Emily Talmage, an elementary school teacher and Opt Out campaigner, warned us that the Testing Action Plan was essentially a Trojan Horse for ed tech. She was right.
MB: For you, then, the Every Student Succeeds Act marks the real beginning of ‘Ed Reform 2.0’.
AM: Yes. To summarise very briefly, Ed Reform 1.0 was characterised by high-stakes end-of-year testing, school closures and the expansion of charter schools, the deprofessionalisation of teachers via alternative training programmes, and a push towards voucher systems, allowing public funds to be redirected to private schools.
Ed Reform 2.0 is all about tech. Online curriculum and continuous testing – or ‘formative assessment’ – rolled out out via ‘personalised learning’ platforms. Data collected across all subject areas, rather than just literacy and maths. Letter grades replaced by ‘mastery’ rubrics. Data on non-cognitive ‘skills’ gathered via computerised behaviour management systems, like ClassDojo. The data-driven mentality now extends into every aspect of school life. Some schools are even adopting rubrics for recess and play! Accountability has not lessened at all. We are literally testing all the time.
MB: What has been the impact of ed tech in the Philadelphia school district?
AM: In some ways, we’re not nearly as far down the road with ed tech adoption as other districts. We don’t have a district-wide ‘1:1 device programme’, where all students are issued with Chromebooks, via which most of their lessons are then delivered. But our district has opted to put tens of millions of dollars in recent years into online education and data management.
This is being done despite the preference of parents and teachers to direct resources into reducing class sizes, restoring electives and extracurricular activities, ensuring all schools have a proper library with a certified librarian, and addressing unsafe building conditions.
MB: Can you say a bit more about these 1:1 device programmes? This is something that is being implemented -- perhaps more gradually -- in English schools too, especially the big academy chains (see here for one example).
AM: As the price point for tablets and laptops has dropped, it has become increasingly common for students, even in low-income school districts, to be issued with their own devices through 1:1 initiatives. By 2014, about a third of US K-12 students were using school-issued devices (here). And since Google’s Chromebooks came onto the market in 2011, their presence in our schools has grown explosively. In 2016, nearly 60 percent of the mobile devices sold to US schools were Chromebooks, as compared to less than 1 per cent in 2012 (here).
Parents are finding that schools no longer provide print textbooks. Instead, students must access content via screens, which has spurred growing concern around vision problems, and reduced reading comprehension and retention. The state of Maryland recently passed a law requiring research into best practices for screen time for children, including device use in school settings.
MB: Ark Schools, an academy chain set up by a group of hedge fund managers, is planning a ‘blended learning academy’, on the model of the California-based Rocketship charter schools (here). It’s perhaps worth noting that Amanda Spielman is closely associated with Ark, having been part of the chain’s original management team.
AM: To understand blended learning, it’s easiest to start with the phenomenon of wholly online ‘virtual’ schools. These have become a feature of the US education landscape in the past decade. They are mainly operated by private companies, or charter management organisations (CMOs) – the equivalent, I suppose, of your academy chains, except that some CMOs are for-profit businesses.
During the 2013-14 school year, over 1.7 million students in the state of Pennsylvania attended one of 16 online virtual schools. The shift of public funds into virtual charter schools, or ‘cyber-charters’, has obviously been financially destabilising to school districts across our state. There have been a number of cases of financial mismanagement and fraud by the operators of online schools -- not just in Pennsylvania, but other states too.
And studies continue to show that most students are not well served by online instruction (see here).
MB: That makes me think of the OECD study of computer-based instruction, which found that it has no effect whatever on students’ performance (here). This report got very little attention, despite all the talk about ‘evidence-based policy’ and ‘what works’. But let’s get back to blended learning...
AM: Well, besides charter school operators offering home-based, 100 per cent online instruction, a growing number of traditional classrooms in the US are experiencing a shift to digital education via online learning platforms. Though it can take a number of forms, blended learning is basically a ‘hybrid’ model which involves supervised instruction away from home, in a ‘bricks and mortar’ school. Instruction is partly online, with students supposedly having control over the ‘pace’ of their learning. ‘Self-paced’, ‘self-directed’, ‘student-centred’ -- and ‘personalised’, of course -- are all buzzwords in the marketing of blended learning, which you may be starting to hear in England.
Many blended learning schools operate a ‘rotational model’, in which the school day is divided into thirds. Students spend part of their day pursuing online learning individually, part of their day collaborating with peers via computers, and part of the day in small group instruction with a human teacher.
Although it is branded as ‘personalised’, the reality is that online instruction basically consists of collections of digital worksheets or playlists of videos – sometimes called open education resources (OER) – accompanied by quizzes and multiple-choice tests. Lots of tests. The student watches a video or reads a bit of text, completes a worksheet, and takes a test. If the score is high enough, the software serves up the next worksheet or video. This is what they call ‘formative assessment’, which is ‘integrated’ or ‘embedded’ in the ‘curriculum content’. It’s basically a system of continuous testing -- or, put another way, of ‘real-time’ data collection.
One of the more successful purveyors of online videos is Salman Khan, a former hedge fund analyst. The Khan Academy has put out a series of videos pushing blended learning, which can be found here.
MB: Who are the people promoting ‘personalised learning’ in the US?
AM: The big charter school management organisations, like KIPP and Aspire. Philanthropic foundations -- including the Bill and Melinda Gates Foundation, of course -- which fund think tanks and ‘advocacy’ groups, as well as specific initiatives. Silicon Valley moguls like Mark Zuckerberg. Facebook assigned a team of engineers to help a CMO based in California, Summit Public Schools, to develop a ‘playlist’-based ‘personalised learning platform’, which they call Summit Basecamp. And a big chunk of Zuckerberg’s personal investment fund, the Chan-Zuckerberg Initiative, is dedicated to ed tech, especially personalised learning.
Reed Hastings, the billionaire CEO of Netflix, is another charter school fan and a promoter of tech-driven schooling. He was an early supporter of the Rocketship schools, as well as an investor in DreamBox, a company which developed the maths software used by the chain (see here).
I would also mention the Christensen Institute in San Francisco. Clayton Christensen is a professor of business management at Harvard, known for his work on ‘disruptive innovation’ in the fields of education and healthcare. His Institute has done a lot to create the language and ‘messaging’ needed to sell blended learning to the folks on the ground. It was his team, for example, who came up with the notion of the ‘rotational model’ (here). None of them are teachers, obviously.
MB: I remember, a few years ago, reading the job ad for the founding principal of Ark’s Blended Learning Academy, and realising that large parts of it were literally cut and pasted from a paper put out by the Christensen Institute!
Nevertheless, Ark’s ‘comms’ people have had difficulty selling the model, and there are no references to tech in recent marketing for the school, now called the Pioneer Academy. Back in 2014, they were telling us that blended learning would ‘improve cost efficiency through staffing and school design efficiencies’, and allow for ‘revised teacher roles’ (here). That sounds ominous.
AM: In 2008, before the Great Crash and the onset of permanent austerity, Christensen was already hailing ed tech as the key to the ‘efficiencies’ demanded by the long-term underfunding of public education. And I think I agree with him here. Personalised learning is essentially a means by which schools can push up class sizes and reduce staffing costs, hiring non-certified assistants to monitor students while they are online. At one stage, the Rocketship chain was putting 100 students -- some as young as five! -- into a single room full of computers. This was the so-called ‘flex model’, which admittedly was quite short-lived (see this report from 2014).
As well as a shift to supervision of students on devices by support staff rather than certified teachers, there’s also a growing emphasis on ‘out-of-school time’, so-called ‘project-based learning’ with community and workplace ‘partners’. I’ve pulled together resources on this new phase here.
MB: We haven’t yet seen anything as extreme as the ‘flex model’ here, although one academy chain has piloted classes of 60, with only one qualified teacher in the room (here). But the pattern is the same. By starving public schools of funds, austerity policies have opened the door to ‘innovations’ and ‘solutions’ -- often tech-driven -- which are at best educationally unsound, and at worst abusive.
AM: To give you another example, maybe a bit less dramatic, but still disturbing. These online learning systems rely on unique log-ins, and the youngest children – five-year-old kindergartners, or even four-year-old pre-kindergartners – cannot remember their usernames and passwords. So they are issued QR code badges that are scanned with the device camera. These ‘Clever’ badges are widely used in the Rocketship schools.
And now we are seeing online learning platforms targeted at pre-schoolers -- like UPSTART, a ‘kindergarten readiness programme’.
MB: Ben Williamson, an academic here in the UK and the author of Big Data in Education, argues that ed tech extends the ‘data-mining’ techniques of the giant tech firms into the classroom (his blog can be found here).
AM: And it’s not just about collecting and analysing test score data, or even monitoring students’ interaction with their laptops or tablets. A whole range of data-gathering devices and sensors are gradually being introduced into schools. As long ago as 2012, the Gates Foundation was funding research on biometric bracelets for students, which were supposed to measure their levels of engagement during lessons (here). The bracelets, known as Q Sensors, are sold by a US firm specialising in ‘emotion measurement technology’, Affectiva Inc.
Another sign of things to come is AltSchool, a group of ‘mini-schools’ set up by Max Ventilla, formerly a senior product manager at Google. These are boutique private schools, and the classrooms look rather different from those at Rocketship or Summit. But the level of surveillance at AltSchool is extraordinary: as well as the streams of data from the students’ devices, every classroom is packed full of microphones and cameras. This is sold as liberation from the ‘incredibly inaccurate, invasive standardised-testing regimen’ (here). All the data fed the development of a new online learning platform, which is now being rolled out into the public system, while Ventilla shuts down his micro-schools (here).
MB: Is this where virtual reality systems fit in too? Our education secretary, Damian Hinds, enthuses about the pedagogic possibilities of VR, which, he says, is ‘helping children take virtual trips through the Amazon’ (I assume he’s talking about the river rather than the warehouses). And a UK company is already promoting the classroom use of VR headsets (here).
AM: Adoption of virtual ‘field trips’ as a mode of instruction is one more example of how tele-presence is being normalised, despite serious health concerns over children’s use of VR. And, yes, VR systems capture vast amounts of biometric data. Most people do not realise when they put on a headset and hand pieces, they create as much data as they consume. In addition to eye tracking and body positions, the systems can also capture heart and respiration rates, blood pressure, and emotional states (see more information here, here, and here). Naturally, Google is pushing the use of VR in schools, with the Expeditions app; the company is currently running a Pioneer Programme in the UK.
MB: Let’s talk about Google. They’ve been getting into schools in a big way – not only because of the success of the Chromebook, but also via the free apps known as the G Suite for Education. English schools, like US ones, are rapidly adopting both Chromebooks and the G Suite. Some of these schools are now ‘Google Reference Schools’. And, as in the US, teachers can become ‘Google Certified Educators’ (see here).
As privacy campaigners in the US have argued, Google’s classroom apps are collecting massive quantities of data from schools. To be fair to the company, however, I don’t think their aim is simply to get loads of new data to sell to advertisers. Patricia Burch, an expert on ‘the new education privatisation’, has said that what Google is doing in US and English schools is ‘very creatively using public resources’ to build new markets – and to carry out R & D at public expense. Would you agree?
AM: Sure. The vast amounts of data harvested via the G Suite for Education are being used to develop new products, and to build monopoly control of the ‘education space’. In a way, Google is turning publicly-funded schools into a company lab. This is where ‘personalisation’ comes in. The other side of big data is machine learning. As you know, the use of algorithms to process vast amounts of data, allowing the ‘micro-targeting’ or ‘individualisation’ of content and advertising, is a key part of the operating models of Facebook, Amazon, and Google, as well as other big tech firms.
People like Zuckerberg and Reed Hastings – and, I guess, the engineers behind Google for Education – think that the same techniques can be extended to education, making possible ‘adaptive’ or ‘intelligent’ learning systems, which would function a bit like the ‘recommendation engines’ used by the giant platforms. In other words, the automation of teaching. But to develop and ‘train’ those adaptive learning systems, they need data, and lots of it.
MB: Anthony Seldon, an influential voice in English education debate, has just published a book called The Fourth Education Revolution. He claims that robots – in other words, ‘intelligent’ learning systems or ‘AI personal tutors’ – will replace teachers within 10 years.
AM: Some reformers do indeed foresee a time, within the next two decades, when artificial intelligence, or AI, largely supplants human teachers. These futurists imagine AI ‘tutors’ or ‘learning assistants’ delivering ‘just-in-time’ content for in-demand workforce roles.
Personally, I find this vision rather dystopian. It would involve students being fitted with wearable biosensors that extract ‘real-time’ data to guide the delivery of online content. Here in the US, a firm called BrainCo has already developed a brainwave monitoring device for classroom use and is now selling it to Chinese markets.
MB: You’ve warned that the teachers who are leaning heavily on Google’s platform – all those Google Certified Educators – are working, for free, to develop the AI systems which will eventually replace them.
AM: Yes. What Google calls the ‘paperless classroom’ will, sooner or later, be a teacherless classroom. And, in the end, there may well be no classroom there at all.
But which is ultimately better for society: an algorithm that ‘learns’ every student in a class and delivers, let’s say, a pre-determined reading selection to each individual, which they are then tested on -- or a human teacher who selects an all-class reading which leads to a lively debate? The first scenario isolates the student, and basically forecloses creative thought in the service of data production. The second opens up chances for students to gain new insights through interaction and co-operation, while limiting the opportunities for digital surveillance.
MB: You’ve always been very clear that the problems with ‘online learning systems’ go beyond issues of privacy and data security. Even if we could be confident that all the data collected by Google (for example) will not be used for commercial purposes, and are beyond the reach of hackers – even then, there would still be reasons to be concerned about the spread of tech-driven schooling.
AM: While student data privacy is important, I’d caution activists that we don’t want to ‘win’ on privacy but end up with secure algorithmic learning. The reformers are keen to talk about privacy and appropriate use of data. Let’s not allow them frame the discussion in this way. We must ground ourselves in the importance of good pedagogy, which respects the humanity and personal agency of both student and teacher. Privacy concerns can play a role in educating the public, but to win this war we have to ground our strategy in the rights of parents and teachers to unplug students from online learning systems altogether, not simply to secure the data in those systems.
MB: It’s often noted that the Silicon Valley elite prefer to send their children to tech-free schools, like the Waldorf schools (here).
AM: Yes, because the teachers in those schools understand -- I would hope -- that learning is not a linear process. It’s an organic one, with occasional doldrums sometimes followed by great leaps of understanding. Human teachers do not view their students as data points subject to precision engineering. They sees them as contributing members of classroom communities, each with unique talents, strengths, and weaknesses.
Screen-based isolation, and an emphasis on targets and tests, gradually erode students’ capacity for creative co-operation. It saddens me deeply that children in public schools with low test scores, schools which have been subjected to a ‘turnaround’ process, are often compelled to use online learning systems for in-school and at-home instruction. The children who most need meaningful personal connections with their teachers and fellow students are denied that right; instead data-driven ‘fixes’ are substituted for human care.
MB: What questions should parents be asking about the use of ed tech in their children’s schools?
AM: I think every parent needs to ask: does the technology used in your child’s school empower students to share their own insights and creativity with the larger world? Or does it transform them into consumers of pre-packaged (and usually proprietary) content, while monitoring their behaviour, gathering data, and using that data to generate a ‘profile’? If a computer program requires a login from a child, and cannot function without having access to their previous interactions with said program, then I would have serious reservations about it.
Because, in this model, the algorithms become the de facto gatekeepers to knowledge. The information fed to students is determined by their profiles, reinforcing the position each child is expected to occupy in society. The data says you have no talent for languages. Your profile says you’ll never need physics. Our ‘dashboard’ says you’re behaviourally non-compliant, which is too bad given your high intelligence. Education systems have always profiled students; but developments in Big Data, machine learning, and predictive analytics have the potential to make existing systems considerably more oppressive.
MB: Back in 2016, shortly after ESSA became law, you wrote that Opt Out families had become ‘pawns in the game of a fake “assessment reform” movement’. What would be your key message for campaigners against test-based accountability and privatisation here in England?
We thought ‘Opt Out’ would be an effective tool against privatisation, but instead it ended up reinforcing reform arguments for ‘innovative assessments’ -- which means tech, which means online learning platforms, which means ‘real-time’ testing and data capture, the tracking and profiling of students. We didn’t realise this until it was too late. Just because the powers that be seemed to be listening to us, didn’t mean that we had won. The battle had just moved into a new phase. This was very hard for people to hear, because those engaged in the Opt Out movement did so often at great emotional cost. Everyone wanted to think we’d won; but we hadn’t won. The terms of engagement had simply changed.
I would say: learn from our experience. Be very wary of this talk about scaling back the accountability system, about the need for more and better measures of students’ ‘progress’. All this hand-wringing over the corrupting effects of test-based accountability -- it’s a bit of a con trick, as I said before. To sell tech-based, data-driven schooling to teachers and parents here in the US, we’ve seen a wholesale hijacking of the language and concepts of progressive education: the need to educate the ‘whole child’, to cultivate ‘critical thinking skills’ and ‘creativity’, to give students opportunities for ‘project-based learning’, and so on. It’s just a marketing pitch.
Public education is in the process of being remade as a profit centre for tech and financial interests. Policymakers have no intention of stopping the data collection game -- quite the reverse. Data is what fuels the machine.