CentreForum’s Round-table on School Choice and Accountability : 4th July 2011

4 July 2011

Speakers:

Chair:

CentreForum’s roundtable on school choice and accountability, including presentation of a new school-ranking website, was chaired by Julian Astle, former Director of CentreForum, and the speakers were the Rt Hon Michael Gove, Secretary of State for Education, Mark Dawe, Chief Executive, OCR (part of the Cambridge Assessment Group) and Warwick Mansell, Freelance Education Journalist.

Julian Astle opened the discussion by laying out the need for a system of improved school accountability to go along with the increased liberalisation of the education system in England. He set out the failings of the current system of school rankings; perverse incentives to narrow the curriculum, teach to the test and prioritisation of particular pupils and subjects over others. The report’s aim was to create a source of good, reliable data to parents that was not susceptible to ‘gaming’.

Charles Hotham, report co-author, then demonstrated the prototype ‘one stop shop’ website with its 19 broad measures and single out-of-100 composite rating. Parents could rank schools by any of the individual measures or the composite figure, and could filter the results based on geographical location and various school types or characteristics. For individual schools, users could also view a school statement, Ofsted report and parental comments about the school.

Mark Dawe then spoke, saying it was the system that was responsible for distortions in teaching. He disagreed with a focus on ensuring the equivalence of qualifications, saying they should stand on their own merit. He suggested more weighting on maths and English and also the curriculum range. He also expressed concern over the parental comments section, saying it would be the extremes of views that were reported.

Warwick Mansell began by saying that the idea that accountability should be between parents pupils and schools was ‘very persuasive’, with politicians in the background. The multi-dimensional measure would allow parents to get more rounded view of schools. But he expressed concern about the focus on statistical ratings systems of all types saying that informal discussions were still important. Whilst statistics contain useful information, he said, parents make choices for reasons not related to school quality. Coupled with restrictions on how much popular schools could expand, he therefore thought we should not be too ambitious as to the effects of this change. He feared there was too much focus on statistics and a misunderstanding that they represented individual children’s chances of succeeding; instead, he would prefer inspection reports to be main form of accountability. He was similarly sceptical about 60% of the weighting being placed on exam results. The statistical system assumed professionals were not already minded to do what was best and needed incentives - something he disagreed with, asserting that a similar statistical accountability system would not have made him a better journalist. Similarly, we should not write off the notion of public service in education.

Michael Gove said he was ‘overjoyed by the publication of this paper’, with it vindicating his Department’s realization that the over-reliance on any one indicator would lead to gaming. The more nuanced the system, in his view, the better. He had expected people to take advantage of the publication of data to find new ways to compare schools and was pleased to find that this was the result. He said that if people were unhappy with the government’s indicator, it was now possible to construct alternative measures; that’s what CentreForum have done here. He suggested that if that ranking table has public confidence and the factors within it are well chosen, it could supersede the government’s own measure. He then went on to defend the introduction of the English Baccalaureate as a government measure of performance, saying that the chosen subjects are precisely the ones which the public think are good indicators and the Russell Group value as qualifications. He agreed that a combination of maths, English and progression made a good measure. He accepted that if you make a measurement into a target, you distort the measurement; however, he argued, talk about gaming and teaching to the test can be misplaced. After all, he asked, do we object to teaching to the driving test? No, as these are tests of competence that we value being able to pass. He also said that the best way of having pupils pass a test will be to ensure they are competent overall – schools that have done well under the existing system tend to be good schools in every other regard. Schools in deprived areas that are actually good (e.g. Mossbourne Academy) would previously have been thought bad, but the publication of data proves otherwise. He agreed with Warwick that Ofsted was still needed in order to make human judgements and provide insightful analysis. He concluded, however, by saying that just as in intelligence we need a balance between human reporting and satellite and other data, so in education we need both data and human judgements to reach accurate and insightful conclusions.

Questions were then taken.

Amanda Spielman from ARK Schools expressed concerns about the aggregation into a single number. She said that evidence suggested that once a single number was presented parents would not look beyond it, an annual media feeding frenzy would result, schools would find the easiest places to push up the number ignoring harder ones and a risk of dissonance with Ofsted ratings would arise. David Laws, whilst satisfied that gaming would be hard under the new system, asked how great a risk there would be that some of the factors would still be significantly influenced by social background or catchment area. Tim Leunig, Chief Economist at CentreForum, pointed out that, as a parent, he was most interested in an assessment of a school’s quality specifically for his children and asked how this could be done.

Julian Astle replied that the comparison should be with the present system with its crude league tables as opposed to a perfect alternative and that the left hand side of the website did offer some room for specialisation based on particular parental preferences for their children. Gove and Dawe both said that a good school would likely be good for everyone.

Brian Lightman from ASCL worried that some indicators might be misleading; for example, the exclusions figure might be a sign of a school doing well in removing bad pupils or doing badly in not being in control. He suggested that the report focused too much on imposing systems on schools when it should be school leaders that are the instruments of accountability. Celia Dignan from the NUT, said that teachers feel immense accountability already, not least to their pupils, and that all league tables skew that relationship, putting teachers under pressure. She argued that it is the quality of teaching, not the measurement system, that raises standards. She advocated the need to raise morale in the teaching profession, suggesting better teacher training and more investment in Continuing Professional Development including the possibly of sabbaticals as ways to do so.

Charles Hotham reiterated the point that league tables already existed and that this was designed as an improvement to them. He said that the only real way for any website or measure created to gain public confidence, however, was through government endorsement. Warwick Mansell objected generally to claims that league table systems raised performance, saying that the evidence was at best equivocal. In Australia , for instance, a great deal of data is published but authorities deliberately avoid the ranking of schools based on it.

Darren Northcott, NASUWT, commented on the dual pulls of accountability upwards to the state and downwards to parents. The report was focused on removing gaming in the downwards accountability, he said, but in fact it is usually considered that most gaming occurs as the result of pressure from above. Julian Astle hoped that the government’s targets for schools might be changed as a result of these improved measures being created. In addition he suggested that a point score out of 100 might have a welcome simplifying effect, given the opaque nature of current measures such as Key Stages and other government targets.

Professor Meg Maguire, KCL, made the point that a great deal of evidence suggests parents make choices based on on-the-ground information and quite complicated reasons such as along class lines that are not always rational and certainly not always based on league tables. Also, she pointed out that some locations clearly have much more choice on offer than other (e.g. rural) ones. Michael Gove replied that not everyone needed to exercise choice rationally in order for us to benefit from those who do.

Samina Khan from Tribal Education and a school governor pointed out she sees more staff turnover and exclusions at her school than nearby grammar schools, but they also prevent more exclusions and she would like to see the data account for this. As with the English Baccalaureate introduction, whilst not necessarily a bad thing in the long term, she expressed concern that the introduction of new measures in this way was often unsettling for teachers. Mark Dawe continued the point by saying that increased staff turnover might not be a bad thing if it was needed to turn the school around. Julian Astle and Lim Leunig replied that the balance of different measures should ensure that such necessary action could be taken without it impacting significantly on any school’s aggregate score.

Julian Astle then concluded the discussion by thanking the speakers and Cambridge Assessment and the Bowland Trust for their sponsorship.

• < Prev
• Next >