Why the number of failing schools can be a good thing

I've written a column for The Times on the primary school test results.

Yesterday’s league tables brought mixed news for England’s primary schools. Overall, standards are rising, particularly in literacy, with more children than ever achieving good results in the 3Rs. But nearly 800 schools — one in 20 — failed to achieve the target of 65 per cent of 11-year-old pupils reaching the expected standard in reading, writing and mathematics.

The better results are evidence that Michael Gove was right to set higher expectations for primary schools — last year the threshold was just 60 per cent.

It is also heartening that the gap in attainment between rich and poor pupils has narrowed further this year, reflecting a strong focus on disadvantaged pupils and increasing use of evidence of what works from bodies such as the Education Endowment Foundation (EEF).

But there is still a big challenge. And it is only as a result of successive governments being prepared to open what was once a secret garden in education that we know its true extent. Kenneth Baker’s 1980s reforms and David Blunkett’s tougher targets in the 1990s helped to make possible the challenges set by the coalition. Don’t forget that fewer than half of pupils reached the expected literacy standard in 1996.

Read the full piece here

Carol's GCSE maths split could add up to a solution

Carol Vorderman has come up with one very good idea about what to do on Maths, but some of her other prescriptions need more careful thought. Her report for the Conservative Party has attracted standard diplomatic language from Education Secretary Michael Gove, suggesting no great enthusiasm for her findings, but one of her ideas really should be taken forward. Vorderman is absolutely right to argue that GCSE Maths should be split between a functional exam and an exam focused on algebra and trigonometry. It is vital that everyone understands how to calculate, use fractions and percentages; not everyone will master calculus.But it is also important that any functional test is developed with business, if it is genuinely to achieve its aims. Otherwise it could act as a hindrance to achieving wider qualifications - as the Diploma has shown.


However, if a young person gets a good GCSE in functional maths, it is less obvious that they need to continue studying to the same standard to18, as Vorderman recommends: of course maths should be integrated into other courses as appropiate, and those who don't get a C grade at 16 should at least continue to aim for this grade in functional maths. Unless we move towards an International Baccalaureate style exam at 18, it is harder to see how compulsory maths could fit in with our current A levels.
 
Vorderman's attack on national testing in primary school seems more perverse: teaching to the test has always occurred, and happens at GCSE and A level. The issue is what they are being taught and whether being tested helps youngsters to retain what they have learned. If testing currently doesn't do that enough, improve the tests but don't abandon something that helped improve maths teaching in primary schools (along with Labour's numeracy strategy, which Vorderman launched) considerably. There may need to be a renewal of the energy that the numeracy strategy brought, but it would be wrong to replace testing with assessment at 11; doing so would be a huge setback for primary maths.

Passing the test

Lord Bew's review of Key Stage 2 testing seems to have some sensible proposals today. Capturing a school's achievement over three years was used for Labour's school achievement awards, and is probably a fairer way to measure a primary school's performance: indeed it could allow all primary schools to have their results published, including smaller village schools. Schools can suffer with a poor intake one year, and the smaller the cohort, the greater the danger.

His suggestion that composition tests should be marked by teachers is clearly intended to appease the anti-tests lobby. And if any aspect of the tests were to be passed to teachers to mark, this should be it. With other elements of English still externally tested, including reading, speaking, listening, grammar, punctuation and vocabulary, internal assessment of creative writing is not unreasonable. However, there should be strong external moderation if the results are to count in a school's league table score.

What Lord Bew seems not to have done, for which we should be grateful, is to suggest the abandonment of testing at the end of primary school. This is his most important finding: testing has opened the secret garden where parents were kept in the dark about how schools are doing, it is significantly improved standards in the 3Rs and it has helped to guard against low expectations, with improvements greatest in many inner city areas like Tower Hamlets. He also seems to have dropped an idea floated by Michael Gove before the election that testing take place at the start of secondary school, rather than the end of primary. While many secondary schools do re-test pupils at the start of Year 7,  it would have been unfair to judge primary schools on the basis of such tests, particularly as secondary schools have an incentive in value added measures to lower the starting point of their intakes.

The teaching unions may not change their opposition to primary school testing. But Lord Bew - whose expertise has previously been as one of the most incisive historians in Northern Ireland - has produced judicious proposals that the Government ought to accept. Testing is rightly here to stay.

Boycotts, testing, targets and cuts

The publication of today's national primary school test results - the last taken under the Labour government - show both the limitations of the NUT/NAHT test boycott and the importance of externally validated national testing. The improvements since 1995 when the tests began have been remarkable, and this year's Maths results are the best ever. When we took the decision to make Key Stage 4 the expected rather than the average standard (ahead of the 1997 election) we faced a lot of criticism then from those same unions that felt a boycott somehow benefited students this year: they said we were being too ambitious. But using targets, particularly floor targets, combined with the literacy and numeracy hours in the late 90s allowed a big change in results. The fact that this is now a standard that can be achieved by 4 in 5 eleven year-olds vindicates that ambition. Over 100,000 more pupils each year make the grade. The schools minister Nick Gibb is right to argue that even more should reach this standard.

But he and his colleagues must be careful in what they are cutting. If Gibb wishes to strengthen the quality of teaching in primary English and Maths, he needs to find ways to incentivise schools to do so. Simply providing a cash premium on its own for poorer pupils without such incentivisation will make little difference, especially if it removes money from programmes directed at the 3Rs and improved in-service professional development. The majority of pupils and teachers defied the unions, enabling today's results to be published. But it is important that the coalition balances its welcome commitments to strong accountability and a pupil premium with the sort of floor targets and agreed ambitions that have also played such an important part in the improvements of recent years.

Testing times

I've written this post on the NUT/NAHT test boycott for the Public Finance blog:

News that the National Association of Head Teachers and the National Union of Teachers plan a boycott of the tests for 11 year-olds in English and Maths may be cause for immediate celebration on the part of some pupils and teachers. But it should worry the rest of us – and any teachers genuinely concerned with improving standards.

The unions say the tests require excessive workload and force teachers to drill pupils in English and Maths when they could be doing other things. Yet, independent experts found the tests to be ‘educationally beneficial’ while reminding teachers that they could teach the basics more effectively through good teaching rather than drilling for tests. Since there is still a 22 point gap between the achievements of poor and better off children, the boycott will have a disproportionate effect on the disadvantaged, those who were failed most by schools before the increase in accountability in the nineties that included testing and performance tables. National test data are vital in raising teacher expectations as well as revealing where primary schools need to improve.

It was interesting to note the lack of enthusiasm even among the unions’ own members for this boycott (which is opposed by other unions). Most members of both unions chose not to vote in the ballot, which meant that only a third of NAHT members and a quarter of NUT members actually voted in favour of this action. This may make it a lot harder to enforce the boycott.

Of course, there is room to reform the tests. But it is absurd to suggest that a single set of tests in English and Maths at the end of primary school – the only independent measure of primary schools – is excessive. Equally, at a time when we have seen the adverse effects of self-regulation in parliament and banking, it is untenable to suggest that teachers should mark themselves. Whatever happens, it is vital that whoever becomes education secretary after the election makes clear that testing is here to stay.

Gove's welcome willingness to perform U-turns

Michael Gove was complaining during his many valiant - but unsuccessful - attempts to defend Lord Ashcroft on his taxes last night that the Tories were somehow being held to a higher standard than Labour over non-Doms. Michael White explains why this is not so with Ashcroft. But there is another area where the Tories are being allowed to get away with murder: their ability to perform U-turns unnoticed (unless they break their sacred marriage vows with the Daily Mail).

The latest came yesterday, barely noted in the Independent and ignored elsewhere. Alongside an announcement on Ofsted that I have detailed earlier, Gove declared that the story that dominated the news a few months ago - that he would have secondary teachers marking primary pupils' test results at the start of Year 7 - was just an idea he was floating. All those headlines may have come in handy highlighting the party's flexibility to grumpy teachers and giving the impression of new policy in a generally policy-lite party. But now it had served its purpose, it could go. According to the excellent Richard Garner, the last remaining truly experienced education correspondent in the national press:

he appeared to back-track from a plan to transfer tests for 11-year-olds to the first term of secondary schooling – a move designed to reduce teaching to the test in the last year of primary schooling....He said he had only “floated” the idea and could see arguments both for and against it. “We do need to maintain objective testing to discover the level at which children are working,” he added. “There is no way that we will get rid of end of primary testing..."

This is a welcome U-turn, if that is what it is. And I reported in Public Finance last summer that Gove was already having his doubts about letting secondary teachers mark primary tests. But surely it deserves the same celebratory scrutiny that normally accompanies major political U-turns - at least when they come from Labour ministers?

Why a testing boycott would hurt poorer pupils most

I have a column in today's Independent arguing that the unions' wish to scrap testing would damage poorer pupils most.

Last week the National Union of Teachers and the National Association of Head Teachers started to ballot their members on a planned boycott of this year's national English and Maths Key Stage 2 tests for 11-year-olds. At the same time, the National Equality Panel laid bare the extent of economic and educational inequality that still exists in Britain today.

The NUT and NAHT argue that the tests impose an excessive workload on their members, and force teachers to drill pupils in English and Maths when they could be doing other things. Yet the independent expert group on assessment, comprising five
experienced education figures, found these tests to be "educationally beneficial" and pointed out that the best way to prepare for Key Stage 2 tests is "through a varied programme of high-quality teaching throughout the year, not through repeatedly sitting practice test papers".

Aside from the revelation in the equality panel's report that there is still a 26 percentage point gap between the achievements of poorer children and their better-off peers at the age of 11, it is clear that the planned boycott by the two unions is wrong-headed. National test data allows teachers to compare the results of different schools, and of schools in similar circumstances, helping to drive faster improvements in poorer areas such as east London. It is crucial to revealing hidden weaknesses so that they can be properly addressed.

But the achievement gap that remains – and the fact that one in four pupils fails to reach the expected standard in both subjects – shows why it is so important to retain the accountability and openness that has helped to drive improvement in so many schools. If testing is abandoned we will lose such vital information in the future. Yet both major parties have been trying to appease the unions with proposals that – despite their protestations to the contrary – could see the demise of this key indicator of primary school performance.

Ed Balls, the Secretary of State for Children, Schools and Families, said before Christmas that he was "not closing the door" on scrapping the tests at 11 in 2012, if Labour is re-elected, provided he was confident that teacher assessments could provide a reasonable alternative. And his Conservative shadow had earlier proposed moving the tests from the end of primary school to the start of secondary school, with secondary teachers doing the marking, and those marks used to create primary performance tables. Both proposals would reduce independent scrutiny. Teachers already do plenty of their own assessments, but theirs is not an independent judgement. If the results are published and become the main indicator of primary school success, the incentives for cheating will be greater.

The Tories' ideas are equally problematic. Secondary teachers have a different vested interest. They are judged in part on the progress they make with each pupil from the start of school to their GCSEs. If they mark pupils down in Year 7 it will look as if they have made greater progress by Year 11. And the moment such results are used in primary league tables, there will be justifiable uproar in primary schools.

Despite suggestions to the contrary, parents strongly support the tests. An Ipsos Mori poll for the Government in 2008 showed that 75 per cent of parents think information on the performance of primary schools should be made public, and 70 per cent of parents place value on the tests in providing information about how their child's school is performing. The unions say that teachers can be trusted to perform the assessments honestly. And while most teachers undoubtedly can be, the unions hardly inspire confidence when they counter scientific and representative polls with unscientific self-selecting ballots of parents collected by their members. At a time when we are regretting allowing self-policing by MPs of their expenses and the lack of regulation for financiers, it would be extraordinary to contemplate removing the most important piece of independent scrutiny that we have in our primary schools.

Despite their flaws, tests remain our best way of providing independent information to parents and taxpayers about how individual schools are doing and giving teachers
comparative data that they can use to improve. Instead of dreaming up alternatives, politicians should be vigorously defending the tests. Abandoning them will limit the chances of many of our poorer pupils getting the education they deserve. They will be the real losers.

Just 18% of NUT members back their union's planned test boycott

The Guardian reports the humiliating lack of interest among NUT members in the SATs boycott that has had Ed Balls and Michael Gove scrabbling to appease the teaching unions. Apparently, only 25% of NUT members bothered to vote, and 75% of them wanted a boycott. In other words, just 18.5% of NUT members felt strongly enough about the issue (that teaching union leaders say is top of teachers' agendas) to support the boycott which would damage children's education and futures. If the NUT doesn't now abandon its efforts to obscure schools that fail their pupils, ministers - and their Tory shadows - must stand up to their silliness and robustly defend the only externally marked tests that pupils are required to sit in primary school as a vital measure of accountability. We should hear no more about replacing tests with teacher-marked assessments whether at the end of primary or the start of secondary school.

Balls should put children's futures before the teaching unions

Until yesterday, Ed Balls had commendably resisted attempts by the teaching unions to scrap accountability, by refusing to abolish testing and league tables. Yesterday, he announced pefectly reasonably that teacher assessment results would be published in league tables alongside test scores. What was unreasonable was his suggestion that Key Stage 2 tests could go, in a shameful attempt to curry favour with the teaching unions.

Primary schools face one set of external tests. They are vital for parental confidence, Ofsted inspections and public accountability. Doubtless many of the assessments will match the test scores, which is as it should be. But if the system becomes entirely self-policing, the dynamics would change considerably. There would be no real pressure to achieve in the basics, and this will particularly hurt boys' achievement in English. It will also hurt standards in primary education, where accountability has been important in the improvements over recent years.

I have been critical of Michael Gove's proposals to shift the tests to the start of secondary school, partly because it too was an attempt to appease the unions, but also because it would delegitimise the results as secondary teachers have an interest in marking down pupils at the start of secondary (Gove has since suggested that he would probably go for external marking). Both Balls and Gove should make clear their unequivocal support for externally set and marked national tests, and stop putting the interests of the teaching unions ahead of children's futures and school standards.

This post was quoted in the TES.

Too much, too young?

It is a bit rich for Prof Robin Alexander and his cohorts on the self-styled 'biggest review of primary education in 40 years' to accuse politicians of ideological interference in schools when they have just presided over a review which has always has at its core an ideological opposition to national testing. So, nobody can be surprised at its best publicised conclusions, that national tests should be scrapped and 'formal learning' should start later.

Yet for a study apperently groaning under the weight of '4,000 pieces of evidence', it is fascinating that it takes a practitioner to explain what nonsense underlies the suggestion that five year-olds are being subjected to rote learning of the sort that a Victorian would recognise - a claim inevitably treated with reverence on the BBC this morning.

In fact, there is not a great deal of difference between what is taught in pre-school in most other countries with a later starting age and what actually happens here. Nor is there a lot of difference between the extent to which some youngsters are taught well at home and others need schools to make up for a lack of home support (as a conversation with any Helsinki primary teacher will attest).

As Liz Steele, a not untypical English primary headteacher writes in the Times today:

Children starting reception class at 4 or 5 are not made to sit in serried ranks and chant times tables. They play, run around outside and often barely notice the transition from pre-school — except that they stay all day instead of just for the morning. While they are gardening, cooking and role-playing they learn about sounds and how to recognise letters.

But it is not a rigid curriculum, it’s learning through experience. For some pupils — who would otherwise not get these opportunities — the chance to interact is crucial to their development. They thrive and grow by being around adults who are trained in how to engage them and by playing alongside other children. Their speech improves and they are supported in a community rather than being taken to towns and clinics where their background and family are unknown.

If the environment is right, with learning experiences and good role models in the classroom, then it is best for some children to start school as soon as possible. Their breadth of knowledge increases and their basic skills improve. They are initially quite ired but we don’t see children falling asleep in corners and they soon become used to it and are quite lively.

It is not the middle classes who would lose out by the change proposed by the Alexander review, but those from backgrounds where there is too little support and stimulation at home. They are starting to fall behind by age three and will fall much further behind if they are denied the sort of purposefulness that is offered in thousands of primaries like that led by Ms Steele. Indeed the Alexander notion that because some parents won't read with their children at home - the main 'homework' expected in primaries - schools should not expect it to happen is defeatism of the highest order.

"Give us back our schools," is the TES headline to its story about the report. And that just about sums up the reasoning behind the desire to scrap tests. For those who object most to national testing are not a generation of youngsters obessesed by competitive computer games but some of their teachers and rather more of those running the teaching unions, who still haven't got used to the idea that schools are not run for the comfort of the staff but in the interests of the pupils; and as taxpayers we have a right to look occasionally at how they perform. Those who regard this as "Stalinism" clearly forget that when the first tests were taken in 1995, less than half of eleven year olds reached a standard that 70-80% reach today.

The primary purpose of primary schools must be to develop literacy and numeracy to a reasonable level in all their pupils, as well as providing a broad curriculum. I would have no argument with stripping back the compulsory curriculum to enable this to happen, and to allow 50% of the day to be decided at school level provided that children were taught to read, write and do arithmetic. I also happen to agree with the suggestion that there should be more specialist teachers in primary school. One of the benefits of the numeracy strategy was that it provided teachers with an understanding of a subject where many lacked confidence, and for which initial teacher education had left them ill-prepared.

But to imagine that children will be better off with less focus on English and Maths, and no proper independent school-level accountability is bizarre. We tried that before the mid-1990s and it didn't work. The only surprise about the Alexander report is that it took so long for them to reach the conclusions they had intended to reach from the start.

I've also blogged on this subject at Public Finance and John Rentoul has picked up this posting in a strong piece for his blog.

Don't lose a primary focus on the basics

I have a column in today's Independent, reiterating the case for testing after last week's disappointing Key Stage 2 results. You can read it here. Here's an extract

The most important feature of the years between 1995 and 2000, when results rose rapidly, was single-minded momentum. Schools were in no doubt that their top priority was the 3Rs. That momentum disappeared as ministers merged the literacy and numeracy strategies into a catch-all primary strategy, and began to give equal emphasis to welfare issues. Of course, primary schools should offer a broad curriculum and care for their pupils' welfare. And they should ensure that pupils who can read get access to plenty of good books.

But the focus of government needs to be clear. After all, those pupils who don't reach level 4 are unlikely to gain five good GCSEs. And without them, they will find it much harder to get a good job or progress to university...... let's not lose sight of why testing and accountability were introduced – too many primary schools assumed that children would acquire the 3Rs through osmosis or their parents. The losers in this lottery were poorer pupils without supportive or fluent parents.

And, given the evidence that rigorous phonics, a sound grasp of grammar and punctuation, and a facility with mental arithmetic are essential building blocks for the 3Rs, governments should not shy away from expecting schools to deliver them. But if government wants schools to do better, it must avoid mixed messages.

Ministers made a big error allowing the abandonment of Labour's literacy and numeracy strategies to be seen as an admission of their failure. And the Tories played a foolish game when they suggested that the Key Stage 2 tests could be replaced by teacher-marked tests in secondary school (they have since indicated that external marking is more likely).

The main reason there has been so little recent progress has been that – with the exception of a renewed focus on phonics – there has been too much muddle over what matters in primary schools. The latest thematic review of the curriculum is in danger of adding to the confusion. But the strategies should not be revived: schools must be held accountable instead for the extent to which they successfully teach the 3Rs, and encouraged to use the best of the available commercial teaching programmes. Ofsted inspections should ensure that children are taught the basics properly, and show how some schools succeed against the odds.

Yet, in the end, the only way that we can know whether individual primary schools are doing their job is through independently set and marked tests. To imagine that our children would be better taught without them is to ignore the lessons of history.

What do the test results tell us today?

Today's Key Stage 2 test results may not tell us an awful lot. Despite the headlines, the fall in English - one point - is neither here nor there, and could disappear after the final results are published, though it may signal lost momentum, as the pressure has eased on primary schools.

Also, as Sarah Ebner points out, it is worth recognising that a level 3 is not the equivalent of illiteracy; when David Blunkett decided that level 4 should be the expected standard, he did so in the knowledge that it was conceived as the average. The truth is the tests show no significant change from last year; that is in itself of concern.

But there are four things we should take from the tests. The first is how important it is to have independent evidence of how pupils are doing in each school in the 3Rs, and how far pupils have come since 1995 with the tests. The second is that writing remains weaker than reading probably because schools don't do enough to encourage creativity alongside the grammar (though do remember that grammar and punctuation had disappeared for too long), and haven't yet found a good way to enthuse boys.

And the third is that there are plenty of schools in disadvantaged areas that get better than average results. Their secrets should be kitemarked and shared in place of the now defunct national strategies. Their heads should also do more mentoring of their weaker colleagues. Finally, the weakening of level 5 results since 2007 suggests that not enough is being done to stretch bright pupils. In a significant minority of schools, this should be the expected result for most pupils rather than level 4.

But before we indulge in a bout of collective government-bashing and educational dystopianism, let us not forget that in the first year of the tests in 1995, just 49% of pupils gained a level 4 in English and 45% did so in Maths (See Table 1 and Chart A). 80% did so in English and 79% in Maths this year. Had the 1995 results been replicated today, 178,870 fewer pupils would have gained a level 4 in English and 196,180 fewer pupils would have done so in Maths.

Why are the Tories abandoning primary school accountability?

The Conservatives' plan to drop externally marked national tests for 11 year-olds from the end of primary school in favour of teacher assessment at the start of secondary school may reduce duplication given that many secondaries use CAT tests, but it could seriously reduce the accountability for primary schools that is so important in improving standards in the basics of English and Maths.

With their policy of greater autonomy for primary schools, through more independent state-funded primaries, it is more important than ever that there should be independent and reliable external scrutiny and accountability.

Shadow schools secretary Michael Gove says that they would aggregate the results of the tests to publish data on primary school performance. In that case, given the importance of the results to analysis of primary school success and parental choices, it is hard to see how external validation and appeals could be avoided.

After all, there is a real incentive for secondary schools to mark the results down so that they can claim to be making more improvement than they have actually made. It will be primary schools that will be putting on the pressure for such validation.

More to the point, the claim that this measure would 'free up' primary schools in their final year seems misplaced. Good schools already strike a sensible balance, but see the tests as an opportunity for valuable pre-secondary school revision in the basics. This would probably not change.

But those schools that over-prepare for the tests are hardly likely to reduce this just because tests that inform league tables are to take place in September rather than May or June. More likely, they will extend the preparation period up to mid-July and organise summer schools to support catch-up. And if the more pushy parents realise the tests are to be used for setting and streaming, those who can do so will step up their use of tutors. The transition period between primary and secondary - already a stressful time for pupils - could become a lot more so.

So I wonder whether Michael Gove has really thought through the consequences of his proposal. It would be shameful if he had reduced effective primary school accountability in favour of a cheap headline and the transient praise of the teaching unions.

This posting has been picked up by the Bickerstaffe Record blog.

Keep tests and tables

I have a column in today's Independent arguing the case for keeping primary school tests and the importance of those results continuing to be published nationally. You can read it here.

A reasonable way forward on testing?

Reports about the likely recommendations of the expert group on testing have focused on the plans to replace the Key Stage 2 science tests with assessment. Since this blog recommended this as a way forward - and it has been welcomed by the science teachers' association - this is a sensible way of reducing the amount of testing whilst focusing on the basics.

Of course, it is the primary responsibility of primary schools to ensure that children know how to read, write and add up properly by the age of eleven. And the state has a duty to parents and taxpayers to make sure that they are doing so. Of course, there may be different ways of doing this - and the expert group seems wisely to be urging more piloting of the so-called progress tests that pupils would take as they are ready - but such tests must be externally set and marked. And they should be taken at least once by every pupil in every primary school.

Of course, the NAHT and NUT don't like this. But the NAHT general secretary - fresh from trumpeting his dodgy 'poll of parents' - told Today this morning that the scrapping of science 'narrows the curriculum too much'. So, he is no longer concerned to reduce the 'burden of testing'? Logically, then children should be tested in every subject, not none, as Mr Brookes admits that the presence of tests concentrates teachers' minds on getting children to pass them.

In truth, the revision that pupils do for tests in English and Maths can and should be a time to ensure that they have learnt how to read and write, and that they know their grammar, spelling, punctuation and mathematical rules. How does Mr Brookes imagine these children will get on in secondary school if they can't?If there is too much pressure on pupils, that is the fault of schools and parents, and not the tests.

But there is one respect where the expert group and those teachers' leaders who oppose a boycott may be being disingenuous, if advance reports are right: with respect to the publication of results. As a general principle in a society where freedom of information is the norm, it would be intolerable not to publish the test results. So, the results would appear in school report cards. Fine.

But there is a suggestion that the government might stop publishing the results on the same day, to avoid newspapers compiling league tables. We did, in fact, leave publication to individual local authorities in 1998 precisely for this reason, and apart from inconveniencing the education correspondents, it made no difference. The Press Association had to work a bit harder. There was less celebration of good and improving schools. So we returned to national publication.

With report cards, the newspapers will still publish the results. Surely, the DCSF must continue to make it easy for parents and others to search for any school's results on its website?

Of course, the real reason why these unions don't want the tests is not that they really think they 'damage' children. It is that they don't like the scrutiny that their publication brings. But such scrutiny has spurred so many of them to improve in recent years. The tests should stay; and their results should remain easily accessible to all.

The NAHT 'poll' that shows why we need independent testing in primary schools

I listened to NAHT general secretary Mick Brookes telling us this morning that we could trust his members to do their own assessments of how good their own schools were, without any need for troublesome tests or pesky performance tables.

In the midst of the discussion, Mr Brookes informed us that his stance enjoyed the overwhelming backing of parents, according to an opinion poll that his union had done. Given that an independent Ipsos MORI survey for the DCSF showed a rather different result - and it was conducted using a representative sample of parents by a reputable polling company - I thought this a bit fishy.

So I checked the NAHT's website for more details of this 'poll.' In fact, there is no independent opinion poll. Instead, the NAHT has pulled together a self-selecting 'sample' of parents' views collected by NAHT members and collated with no attempt at weighting.

The press release issued in February even suggests that embarrassed NAHT press officers know the whole thing is a crock. "Whilst we recognise that this is not an independent survey, it represents a substantial body of opinion," they plead.

So, let's get this straight. The union which tells us that we can trust its members to tell parents exactly how well their child's school is doing in English and maths without independent national tests or published results is quite happy to defy the rules of statistics to get the data it wants, and to pretend that its data is of some value even though it shows results exactly the opposite of properly conducted independent research.

And we are supposed then to take their word that we don't need any independent look at how primary schools run by its members are performing?

Boycotting tests would be a slap in the face to taxpayers and parents

The fairly predictable decision of the National Association of Head Teachers to ballot its membership on a boycott of national tests for eleven year-olds is nothing to do with the absurd notion that a single set of externally marked and set tests in the first fifteen years of a young person's life is 'damaging' and everything to do with the view of some professionals that it is impertinent of parents and taxpayers to wish to have a degree of accountability that is independent of individual schools. After all, it is up to individual NAHT members to decide how much they want to 'pressure' their pupils, and to pretend this is all the fault of the tests is not only disingenuous but is disproved by the many successful schools that manage to get the balance right.

The issue here is not whether the existing tests might be improved; they can. The Government is already reviewing them, and it is likely that the conclusion will be more tightly focused English and Maths tests. There is also merit in doing much of the testing online. Rather it is the growing view that professionals should be accountable to nobody but themselves, despite the fact that the taxpayer foots the bill (the Tories have already swallowed this view in their plans to leave doctors free of any patient-focused targets).

Having seen what happens with excessively light-touch regulation in the private sector, it is extraordinary that public servants should demand the same for themselves. Politicians of all parties should make clear that anybody boycotting the tests is in breach of their legal obligations and their duty to parents; they should do so quickly and without equivocation, and explain the consequences. Otherwise, all the gains in openness in our schools system over the last two decades will be endangered, and parents will once more find themselves excluded from a secret garden in which their enquiries are unwelcome and there is no external check on the assessments being made by teachers about their child's progress before GCSEs.

Test and publish

When David Blunkett experienced the wrath of militant teachers at the NUT conference in 1995, the NASUWT's then general secretary Nigel de Gruchy ensured that the shadow education secretary had a standing ovation the following week when he turned up at their conference. So it was no surprise to read tongue-in-cheek reports today that NASUWT would strike if the tests for 11 year-olds went, a neat juxtaposition with the NUT's rallying call for a boycott of the tests by its members.

But behind the headlines there are two problems. The first is a headache for Ed Balls. He wants to move to a system of testing when ready, which would probably involve teachers in more work than the current tests, as there would be several testing opportunities each year; though the NASUWT objections are also to the NUT's absurd idea that the tests should be marked by the teachers themselves. There are good arguments for the Balls change, as well as the status quo.

There is a second problem that should worry us all. Keep tests, but scrap tables, say NASUWT. DCSF doesn't itself publish league tables. Its search engine directs you to individual school results. You can also see in an alphabetical list how those results compare with other schools if you search by local authority. Newspapers compile the tables using discs supplied by the DCSF. What NASUWT is demanding is that the government censors the information it supplies newspapers and the public. That is no more tenable than ditching the tests.

Parents back tests - ministers need to do so more clearly too

One of the arguments made by the two teaching unions threatening to boycott national tests next year is that they have the backing of parents.

Well, they may have on the sort of self-selecting surveys that teaching unions seem to specialise in. David Aaronovitch did a good job exposing the latest ATL survey proclaiming the death of school discipline to the sort of basic scrutiny that seems to have eluded most newspapers.

And when the DCSF conducted a proper opinion poll - using Ipsos MORI and based on a scientific sample - it found a rather different picture when it comes to what parents think of tests.

• 75% of parents think information on the performance of primary schools should be available to the general public;
• 70% of parents place value on the tests in providing information about how their child’s school is performing;
• 78% of parents with children who have done the tests think they are an accurate reflection of how their child is doing;
• 69% of parents think test results are useful for teachers when their child progresses to secondary school

Even when given the chance to say whether the tests should stay as they were or be replaced, 44% of parents said tests should stay as they are, with 36% of parents wanting them replaced, with most of the latter group wanting testing to stay. As the DCSF press notice pointed out, this means that 55% of parents expressing a preference think tests should stay at they are.

Despite last year's fiasco, tests still retain majority parental support. So Ed Balls should be wary of sending mixed messages to the teaching unions at their annual whingefests. The main headline on the DCSF press release showing A MAJORITY OF PARENTS BACK TESTS was not that; instead it was KEY STAGE 2 TESTS NOT SET IN STONE, ED BALLS TELLS TEACHERS' UNION CONFERENCE with the survey results given second billing.

The truth is that the Government is looking at a more complex alternative to the current tests, which would see most pupils taking more than one test in each subject in their last two years of primary school, and wants to publish the results on a new report card. Which is fine if they can find a straightforward way to present the results, whilst getting the marking right, crediting high achievers and avoiding the low expectations which CVA value added tables have fostered.

Balls was right yesterday to make clear that some form of national testing would remain and to condemn the threatened boycott. But we need to hear much stronger and clearer support for national testing and published results from ministers, more often. The parents they polled have shown them the way.

How to reform tests

I have a piece in the latest edition of The House Magazine, continuing my arguments in favour of a reformed testing system in schools. Here's an extract:

Testing has – with regular inspections and pupil-level targets – been part of a cultural shift in schools that has brought real improvements. But it is not just through greater accountability that testing has helped improvement. The data from tests – combined with the power of computers – has given teachers access to data that they can use to ensure that they get the most out of individual pupils. By seeing what the best students in similar circumstances can achieve, they can lift expectations in their own classroom.

There are still ways to simplify the tests whilst keeping the level of independent scrutiny that they provide for every primary school. First, teachers could still be encouraged to enter pupils for tests when they are ready. Pupils should not be prevented from scoring higher-than-average grades on the papers, but many might be tested at age 10 rather than 11. In schools that devote months to the preparation for tests to the exclusion of other subjects – and those that do are rarely the best schools – this should allow a better curriculum mix.

Second, marking pressure could be reduced by confining external marking to English and Maths, leaving Science to internal assessment. Those are clearly the most important subjects for any child if they are to access the rest of the curriculum. Third, there may be a place for multiple choice questions in Grammar, Spelling and Mathematics, which can be marked by computer, though there must still be a place for independently marked writing tests.

Whatever happens, we cannot afford to abandon the level of independent scrutiny that testing provides for primary schools. Without it, we will fail future generations of children.