This is the text of a speech I gave to the joint NAHT/NUT conference on the future of assessment, at the NUT’s London headquarters, on Wednesday, February 11th, 2009.
“We get taught things in lessons to prepare us for things like the Sats tests.” 9-year-old boy.
I thought I’d start with this quotation, as I read it only last week and it struck me as shocking and powerful, even though I’ve been investigating the impact of high-stakes testing in schools for six years now and thought I was fairly unshockable.
It was the response of a nine-year-old boy, quoted in last week’s Good Childhood Inquiry by the Children’s Society. He had been asked why he agreed with the statement “my school is helping prepare me for life”.
What struck me most was the boy’s age. At nine, the key stage 2 tests were two years away, and he would have had his key stage stage one assessments two years previously. Yet still “being prepared for life” seemed to equate in his mind with “being prepared for Sats”. The influence of these tests, which actually are relatively unimportant to children’s futures in their own right, is huge. But are they serving children’s needs well?
That is the central question which has pre-occupied me, as essentially an outsider to our schools system. I am not a teacher, and I do not have children going through the system. But I was driven to write a book on this subject after finding, in my former life as a reporter with the Times Educational Supplement, evidence coming my way every week, if not more frequently than that, which appeared to be pointing to downsides with the current system. And I say downsides from the point of view not just of teachers, but, crucially, I believe, for many children. I think this evidence of unintended consequences has to be put in the public domain and debated properly in any evaluation of whether the current regime is fit for purpose. This speech is largely an argument for the Government to introduce a proper and independent-minded review of the assessment system.
I should say here, that I am not against testing per se. The question is whether the current assessment regime, and crucially, the accountability system of league tables, targets, inspections and the rest which is built upon it, helps or hinders education. Neither do I disagree with the notion that schools need to be held accountable in some way for the public money which is spent on them. I merely question whether the current form of test-based accountability is working effectively for children’s benefit, and for that of the nation as a whole in having a generation of young people educated in a rounded sense.
My book is an attempt to catalogue some of the downsides. They range from the months primary schools spend teaching to the test to the focus on particular groups of pupils which are important to schools statistically, the advent of exam-driven schooling in the latter years of secondary education and the use of textbooks tailored very precisely to the requirements of particular GCSEs and A-levels.
The system, I believe having considered in detail the evidence on this subject since 2004, is having powerful negative effects on many pupils. These effects are not always simple. It is often contended, for example, that the testing system is stressful for children. I agree that it is so for many pupils, sometimes horrendously so. Other children will find the tests relatively straightforward and breeze through them. But for many, I just think that months of test preparation are boring and that there are better uses of their time, which schools would go in for, were not so much hanging on the test results, for the adults in this system. I also have serious concerns about the fact that the current regime makes it very clear when a child is “underperforming” or at the bottom of the class, labelling them as such often on the basis of a relatively narrow, if important, range of skills, focusing on the ability to perform well in time-limited English, maths and science tests. Yet the results themselves are not always reliable. Also, the spoonfeeding that the test and exam regime encourages – results are everything, remember, no matter how they are obtained – is being picked up by universities as helping create a generation of passive learners, expecting to be told the answers.
Are any of these effects, about which there is now a lot of evidence, being taken seriously by the Government?
Consider two more quotations.
“It is very unusual to go to schools where everything has been turned over for a large amount of time to focus just on the tests.” David Bell, permanent secretary at the Department for Children, Schools and Families, January 2008.
This was the response of David Bell, now the most senior civil servant at the Department for Children, Schools and Families, when questioned by members of the Commons Children’s, Schools and Families committee on the issue of teaching to the test.
“Many teachers spend too much time preparing pupils for the tests; in most schools, the whole of the spring term, and often time before and after, is devoted to explicit test preparation.” David Bell, annual report as Her Majesty’s Chief Inspector of Schools, 2005.
Yet this was the same Mr Bell, barely three years earlier. You can have different views on what might have prompted his change of opinion on this topic. I personally find that his statement to the MPs falls short of the standards of honesty one would expect in a senior public servant.
But wherever one stands on that, the deeper issue, of taking seriously the evidence of what is going on in schools, remains. The Government has to take on board the effects, considered in the round, of its testing policy in schools. And – despite the wholly welcome move to abolish the key stage three tests – it still is not doing so in any fundamental way. I find it staggering, for example, that while the expert group of which Sir Tim Brighouse is a member has been asked to look at the issue of teaching to the test at key stage 2, its brief appears to extend only to producing new guidance effectively telling schools not to do it. Yet, of course, teachers know that the accountability measures by which the outside world judges them, from league tables to Ofsted inspections, encourage exactly this tendency. Indeed, primary schools under a certain level of test performance for a number of years are now supposed to be considered by their local authorities for possible closure. [See slide, below]
“Where it is clear that a school will not be able to improve their results enough to move above the floor target in 2010, even with additional support, a more radical solution such as closure or use of the LA’s statutory intervention powers will need to be considered.” Government target-setting guidance for primaries, sent to local authorities in October 2008.
It is enough to make the average head of a primary struggling at the bottom of league tables very relaxed about test performance, I can tell. You can imagine the local authority saying to the head, ‘we might close you in a year’s time because of poor test performance, but don’t worry too much about teaching to the test’.
The National Strategies, who are themselves charged with raising test results, have encouraged schools to start detailed test preparation four months in advance of the key stage 2 Sats.
Without an acknowledgement of the reality that teaching to the test is encouraged by the huge pressures of external accountability on teachers, and no ability to scrutinise these effects, the remit given to the expert group just looks naive, and out of touch with the reality in classrooms.
And amazingly, to my knowledge the Government is not even trying to build up its own evidence on the extent of teaching to the test. This is, of course, not the first time ministers have ducked the chance to have a proper look at the effects of what I have called hyper-accountability. Christine Gilbert’s inquiry into personalised learning in 2006 proposed a wide-ranging review of assessment, which was ignored. Sir Jim Rose’s recent primary review was barred from looking at the impact of testing, even though he viewed it as the “elephant in the room” in this particular debate.
In fact, successive inquiries have raised it as an issue. The Good Childhood Inquiry followed those by the Children, Schools and Families Select Committee, the especially exhaustive Cambridge University-based Primary Review, the Nuffield Review of 14-to-19 education, the Royal Society and a host of other scientific organisations, the Government-funded Advisory Committee on Mathematics Education and the teacher unions among others in asking serious questions about test-driven schooling of the current kind.
It is often contended by the Government that parents like the current testing regime. But I would like to see more evidence on this, particularly after the NAHT survey of last week, which pointed in a different direction.
Other evidence continues to pile up on this subject. Only last week, an Ofsted subject report found that some primary schools were so anxious to improve their test results, they were denying pupils any music tuition in year six. If that is happening, it is an outrage. What is a school doing, not teaching its pupils music for a year because it wants to improve test results in English, maths and science? That sounds like I’m blaming the school, but that is not the point of this argument. Structural pressures, over which the Government has control, are what is driving this behaviour and they need to be addressed.
This is slightly flippant and only works if you follow football, but I sometimes feel, instead, that ministers and civil servants have graduated from the Arsene Wenger school of myopia, when it comes to closing their eyes to the unintended consequences of their testing and exams mechanism. Like the Arsenal manager who is forever unsighted when one of his players commits an indiscretion on the field, the likes of Mr Bell and departmental spokespeople profess “not to have seen the incident” when something untoward happens as a result of their accountability system.
The Government might object, here, that it is investigating reform to testing, through the introduction of single level tests. But these do not offer any answer to the problem of months of teaching to the test. In fact, they surely run the risk of making it worse, with pupils facing tests more frequently. The Government counters that this has not happened with the current trial of these new assessments. But that is not surprising, since the pilot has been designed so that the tests which are under investigation are not “high stakes” for the schools involved. So the trial does not reflect the reality of how the tests will work if they do succeed the Sats. Instead, ministers stick to the belief that using the same assessments as a way both of helping children to improve and of holding their teachers to account is unproblematic.
Today’s conference is obviously about the future of Sats. But I think there are bigger issues even than the tests to think about. There needs to be a proper debate not just about test-driven schooling, but about the use of exam statistics, including those related to GCSEs and A-levels, to influence what goes on in classrooms.
While GCSE and A-level results are clearly important to pupils, are they really all that matters in terms of receiving a good education? I would say not, and I guess many people would agree. But from the way the Ofsted inspection system works, for example, it would appear that the idea of what constitutes a good education has been quietly redefined to mean the securing of a good set of qualifications. (See slide, below).
“No school can be judged to be good unless learners are judged to make good progress”
[As measured by test and exam results].
Ofsted guidance to inspectors, February 2009.
The interpretation in the brackets is my own, but it is clear that this is the way Ofsted inspections work. No school without good exam results, and these can be measured in a number of ways, can be adjudged to be good. So, in the inspection system, good exam results define entirely a good education.
The School Teachers’ Review Body, in a report last year, also concluded that “outcome indicators”, the most obvious of which are test and exam scores, are how teachers should be judged. (See slide, below).
“Our strongly-held view is that teachers are accountable for outcomes,
not inputs or activities.” School Teachers’ Review Body, April 2008.
Again, is everything a school does to be valued only in terms of results, or outcomes? By that criterion, good teaching has no intrinsic value whatsoever: it only matters if it results in good measurable outcomes for the pupil. These formulations may make for a relatively easy system of “indicators” or benchmarks of supposed school quality for civil servants to pore over. But, again, is it really all that we want from schooling? I actually was taken, again, by a conclusion from last week’s Good Childhood Inquiry, which defined a good education like this: (see slide, below).
“We believe that exam grades and qualifications must not be seen as the primary objective of children’s education, rather as one of the markers of children’s growth, learning and achievements among many others.”
The Children’s Society’s Good Childhood Inquiry, February 2009.
The new “report card” system being proposed by the Government, which would rank schools according to a wider range of measures than exam results, does offer some concession to the point that there is more to school life than simple exam results statistics. In that sense, it might be more “balanced” than the current rankings. But it really is just another form of league table, which does nothing to address some of the underlying and fundamental problems with this approach. It seems to me to be just another expression of the belief by the politicians, and their advisers, that the way to improve something is to publish a new set of indicators and then cajole the profession with a mixture of carrot and large stick into trying to improve on these measures. There are deep problems with this approach.
Education should be a humanising process, but in some senses I believe pupils are dehumanised by this system: they become someone else’s statistics, or just another production target. Whether the figures being generated relate to pupil wellbeing – and the idea of a league table in children’s happiness, with, presumably, penalties for low-performing institutions, makes the mind boggle – or on exam results, the effect is the same.
And this is not just my own view: on a visit to a school last year I found a group of teenagers complaining that they were fed up with being treated as if they were just a set of numbers, or “walking around with statistics posted on my forehead,” as one pupil told me. Pupils pick up on the fact that, often, the tests are as much about testing the school as about testing them. It is not surprising that they sometimes come to believe the school’s drive to improve its data “or to make the school look good in the local paper”, as one student put it recently on a discussion board, takes precedence over an objective consideration of their long-term learning needs.
And I think one of New Labour’s biggest failings in education – and there have been successes, away from test-based accountability – has been to dismiss and ultimately write off the public service ethos that is surely present in many teachers. I should qualify this: I do not think that many ministers and civil servants would disagree with the notion that many teachers have a strong commitment to their pupils. But the system they have set up does not function with this assumption, based as it is on mainstream economic theory, which views people as basically self-interested individuals. Because the Government sees itself as unable to take the risk of what happens, for pupils, when some teachers are found to lack that sense of public service, or simply are not be up to the job, it has created this hugely intrusive system of hyper-accountability which, I believe, is getting in the way of many good teachers doing their jobs as effectively as they could. The prize, of course, would be a system which offered safeguards for pupils if their teacher were not effective (and by not effective, I do not mean simply that his or her test results were poor) while freeing up the best staff. Such as system is a long way off and should not rest on exams data as an indicator of teacher quality. But it surely is not unimaginable.
This lack of trust, analysed by Baroness Onora O’Neill in her BBC Reith lectures of 2002, explains why the word “tyranny” fits the current system. It is an authoritarian regime, which holds that the way to improve education is by setting relatively narrow, or “focused” statistical objectives and then monitoring obsessively performance from the centre. League tables, which are embraced enthusiastically in Sir Michael Barber’s recent book on public sector reform, are one tool. This structure treats pupils, essentially, as passive agents, whose own motivation to learn is played down in a system where results must go up whether or not the individuals being tested want to work hard. It rejoices in incremental percentage point improvements in test statistics without checking whether those gains actually represent anything significant. And the accountability system increasingly, to me, seems to acquire a mad logic of its own, which is often out of kilter with what the public might consider as important in schooling. What is the definition of a good school?, I find myself asking when I look at how the current Ofsted regime operates. One that analyses data effectively, seems to be a big part of the answer coming back. As someone who had an excellent education with many superb teachers and was, to my knowledge, never set a single statistical target, I find this difficult to understand. This does not mean that statistical analyses of all kinds should be rejected. But they should be kept in their proper perspective: good teaching which builds pupils’ understanding should surely be the ultimate goal, however it is reached. Managerial tools, such as data analyses, are, surely, of secondary importance.
It might seem that I am criticising the Government too much. There have been positive moves on testing in the past couple of years, the decision on the key stage 3 tests being the most visible. But if sometimes the Government’s intentions on some of this agenda appear to be right, it is clearly not going far enough.
If I would say anything to this campaign, it would be this: make the argument that there needs to be an open and honest look at the effects of what could be called in shorthand high-stakes testing. This would take place with the understanding both that school accountability is important, and that some form of assessment is essential. But if the current regime is not working to pupils’ benefits, it must change. Not to have had any serious investigation carried out by ministers, certainly since the Dearing review of 1993 and probably since the introduction of the national curriculum in 1988, is… strange.
I do not think that changing the system in such a way as to retain school-by-school accountability, but which is less harmful to pupils and to the overall quality of national education, is impossible to imagine. There are interesting ideas out there, not least one floated by Cambridge Assessment and the Institute for Public Policy Research which would see teacher assessment becoming dominant, with pupils sitting a small sample of tests to help moderate the teacher assessment judgements. But any move towards reform has to start from a proper consideration of the needs of all pupils, their parents, and of the nation of having a generation of young people equipped to think for themselves and well-educated in the roundest sense. It should not begin from other considerations, political ideology and a theory which fails to face up to the real-life consequences of its own assumptions being the prime potential distractor.
As that nine-year-old will one day hopefully learn, there is and should be life beyond Sats tests.
Oh, and my final slide is a plug for my website, where I have been trying, slightly intermittently I must admit, to build up a register of as much of the evidence on the effects of high-stakes testing in England as I can. It is at www.educationbynumbers.org.uk