This memorandum is AQA's response to the invitation from the Select Committee to submit evidence in connection with the Select Committee's inquiry into Testing and Assessment. The memorandum offers background information on AQA and then focuses upon those of the questions posed by the Select Committee in its published terms of reference on which AQA is particularly well-placed to express a view. In doing this, issues raised by some of the other questions posed by the Committee are also addressed.

The essence of this submission is an argument for choice in Key Stage 4 and post-16 qualifications offered by competing Awarding Bodies. There should be choice between modular and non-modular examinations; choice as to the use of properly controlled teacher-assessed coursework; choice between GCSEs, A-levels, Diplomas and Baccalaureates. Effective regulation is required in such a system to maintain public confidence and ensure that young people are treated fairly. However, there are no insoluble technical problems to achieving this aim.

As in other areas, competition between Awarding Bodies drives both technical and educational innovation and helps to reduce the burdens and costs of the assessment system as a whole. Again, there are no insoluble technical problems about ensuring comparability of standards in such a system and no evidence of any consistent problem in this respect at present.

Most importantly, choice in qualifications is consistent with the widespread desire to move learning in a more personalised direction and what we know about the varying ways in which different young people learn and are best motivated.


The Assessment and Qualifications Alliance

1. AQA is the UK's leading Awarding Body and, as a long-standing provider of high quality general qualifications at GCSE and A level, the awarding body of choice for schools. We are a non-profit making educational charity so all our income from examination fees goes into running and developing our examinations and other services to schools and colleges. We place great emphasis on engagement with our stakeholders in educational centres to ensure we are fully meeting their needs. As the UK's main Awarding Body for general qualifications, one of our primary roles is to engage with our regulators and policymakers on issues of curriculum design and wider educational and assessment policy, utilising our educational research department which has a considerable international reputation. One of our priorities is the effective use of innovative technology to facilitate and modernise assessment techniques. AQA is pioneering the introduction of new methods of electronic assessment and marking that increase accuracy and reliability while maintaining and enhancing the integrity of the examination system.



Why do we have a centrally run system of testing and assessment?

2. The National Curriculum tests at Key Stages 2 and 3 are centrally run by QCA but at Key Stage 4, and thereafter, competing Awarding Bodies provide a choice of assessments and qualifications. Including AQA, there are three General Qualification Awarding Bodies in England, one in Wales and one in Northern Ireland. All five of them are able to offer qualifications in England, Wales or Northern Ireland.

3. In the broader sense of assessment in support of the teaching and learning process, we do not have a centrally run system, of course, and there is much good practice carried out by teachers in their own classrooms, some involving material or tests purchased from publishers and other suppliers, some constructed by themselves. Assessment by teachers is the essence of good teaching which requires constant evaluation of the specific strengths and weaknesses of each pupil and planning of the best way to develop their knowledge and understanding further. More formal assessment for learning has a significant part to play and can build upon and illuminate the teacher's informal knowledge of their pupils.

4. Testing can act as a motivator for students and teachers but, in general, it can only improve attainment directly if the outcomes of the tests are used to guide subsequent teaching and learning. To be effective in this regard, assessment instruments need to be designed primarily with this purpose in mind and to provide more than a general snapshot of achievement. Although it is possible to extract some formative information from summative assessments like the National Curriculum tests by looking in detail at the pattern of an individual pupil's performance on the different questions in the test, their use for this purpose is inevitably limited.

5. However, the primary use of the National Curriculum tests is accountability for schools and monitoring of the performance of the education system as a whole. They have a second main purpose of providing standardised information for parents and others about the progress of individual pupils. It is sensible for them to be centrally run for both these purposes. Although an arrangement in which there is local control of progress assessment (say, at Local Authority level) with some system for ensuring that the results can be aggregated to form a coherent national picture for accountability purposes are possible, they are likely to be more expensive and burdensome than a single national system because of the need to establish comparability of standards across many different local assessments.

6. At Key Stage 4 and beyond, the need for accountability measures is met by collating outcomes from the qualifications which the Awarding Bodies already provide for individual students. This ensures that no additional assessment burden is placed upon schools and students for accountability reasons. The primary purpose of the work of the Awarding Bodies is certification of individual achievement. The credibility of the qualifications is enhanced because much of the assessment is external to the school or college and centrally regulated, albeit not centrally administered.

7. The Awarding Bodies operate in competition with each other and, as in other industries, competition encourages them to be innovative and to operate to high standards of service and offer high quality examinations. Modernization and innovation in assessment is one of the benefits of competition. For example, in large scale GCSE and A-level examinations, AQA is leading the way in the use of electronic assessment, where candidates answer on computers rather than on paper. A great deal of development work is also underway to improve the quality, security and control of the marking process by the use of modern technology. Quality of marking is clearly an issue of great concern to the schools which are our customers and it is unlikely to be a coincidence that these important developments are progressing much faster for qualifications where there are competing Awarding Bodies than for National Curriculum Tests where there is no competitive pressure.

8. Even in an era when the content of many examination requirements is largely determined by the regulator QCA, there is scope for significant curriculum innovation which is encouraged by competition between the Awarding Bodies. For example AQA's recently developed Foundation Certificate in Secondary Education in Modern Languages was welcomed by Lord Dearing in his recent report on Language teaching and is intended to help reverse the decline in the take up of Modern Languages in secondary schools. We are also developing an AQA baccalaureate qualification which will enable A-level students to have the breadth of their educational experience, including enrichment activities such as work experience or significant contributions in the community or to their school or college, to be formally certificated alongside their A-level results.

9. There is significant scope for competition in the level of service and support for teaching and learning that an Awarding Body provides. AQA offers extensive programmes of teacher support at the beginning of a new syllabus and regular review opportunities as the syllabus becomes operational. High quality and rapid feedback comes from examiner reports after each examination and we ensure that comprehensive, innovative and motivating teaching and learning materials are ready for use as soon as a school has selected AQA as its examination provider. The provision of teaching and learning support is a major, and growing, part of the work which, as an educational charity, AQA does and is one of the things which differentiates us from competitors.

10. Some critics believe that the existence of several Awarding Bodies is bound to lower standards because they will compete by awarding more high grades. The evidence does not support this. There is little correlation between market share and pass rates and thus little incentive to compete on standards but, in any case, the General Qualification Awarding Bodies all see their role as providing educational services for the benefit of young people. This, and the credibility of the sector as a whole, clearly depends upon comparability of standards between the Awarding Bodies and we therefore work together on research to ensure that our standards remain comparable, publishing the reports of that work for public scrutiny. As our regulator, QCA also does research in this area and has found no evidence of significant and consistent differences of standard between the GCSE and A-level awarding bodies.

Does a focus on national testing and assessment reduce the scope of creativity in the curriculum?

11. The crucial issue here is the status given to the results of testing and assessment and the extent to which they become the primary focus of teaching and learning. There is a clear tension between the need to ensure that schools are properly accountable on the one hand and the need to allow room for curriculum innovation and inspired teaching on the other.

12. There is no reason why a suitably structured and restricted range of national testing and assessment should swamp the curriculum and constrain creativity. Clarity about the purpose of national testing and assessment is crucial to its design. If the intention was only to measure the performance of schools, then there would be no need to assess every pupil across the entire curriculum at each Key Stage - a sampling approach within each school would be adequate. In the same way, if the intention was only to monitor the system as a whole then sampling of schools themselves would be possible within a system which would still provide effective national accountability. If, on the other hand, the intention is to provide information about the progress of individual pupils, then a more comprehensive assessment regime, like the present one, is necessary.

13. In this connection, we welcome the work which QCA is proposing on the construction of diagnostic National Curriculum assessment material for use by teachers wishing to assess their pupil's progress when they are ready for it. Once established, these assessments could replace the use of Key Stage tests for individual purposes, allowing those tests to be scaled down, reducing the burden of testing for accountability purposes upon pupils and improving the cost-effectiveness of the accountability system.

14. Underpinning the testing and assessment regime is, of course, the National Curriculum itself. The latest revisions to the programmes of study (in Science, for example) appear to be taking an approach based more on key concepts and requirements, rather than specifying all that must be taught. This is to be welcomed as widening the scope for creativity in the classroom.

Who is the QCA accountable to and is this accountability effective?

15. QCA operates to remits from the Secretary of State in relation to testing, assessment, qualifications development and operation. From its behaviours, it appears that QCA sees its main accountability as lying there.

16. There is therefore considerable potential for conflict of interest between QCA's twin roles of developer and regulator. The clearest example is in National Curriculum Tests which QCA simultaneously operates (through its NAA division) and regulates. However, there are many other instances. For example, QCA produces the current Basic and Key Skills tests and requires the Awarding Bodies to use them, rather than develop their own. As a result, regulation is weak and there are serious problems about accountability for the quality of the test material and assessment processes used.

17. However, the most important problem arising from the lack of clarity about QCA's accountability is encapsulated by its recent decision to regulate the entry fees for new A-levels and Diplomas. This is within QCA's statutory powers and entry fees are clearly a legitimate area of engagement for a qualifications regulator. But there was no visible attention paid to operational efficiency or cost-effectiveness during QCA's work on developing the structures and operating rules for these qualifications, and nor is there reference to these matters in the criteria for their accreditation. There is a need for clear separation between the qualifications regulator - with real independence from Government - and the agency working to Government remits to develop new families of qualifications in furtherance of policy. Such a separation could make a major contribution to the proper consideration of the impacts, both educational and financial, of new qualifications that are being considered and developed as part of national policy. At present these important considerations, which have major implications for the administrative burdens on schools and significant impacts upon school budgets, are given little or no priority until the design phase for new qualifications is over - a highly unsatisfactory, if long-standing, state of affairs.

What role should exam boards have in testing and assessment?

18. Exam boards (Awarding Bodies) are concerned primarily with the provision of qualifications to enable individual students to progress to subsequent stages of education or employment. The use of qualifications for this purpose has the effect of ensuring that, as far as possible, those who engage in further study or particular jobs are equipped to do so. In this way, qualifications contribute to greater efficiency in the employment of the nation's human capital than the alternatives (essentially influence, patronage and random selection). Equally important, qualifications contribute to social cohesion by providing a widely accepted and essentially meritocratic basis for sharing educational and vocational resources.

19. As a result, the Awarding Bodies are the organisations in the UK with the greatest knowledge and most practical experience of educational assessment. AQA, for example, prepares assessment materials and organises examinations on a national scale every year (we set, mark and process approximately 13 million individual student assessments every summer). AQA also has the largest research facility of any of the exam boards in the UK, making our Research Division the largest group of professional researchers anywhere in the UK who are working solely on educational assessment.

20. With this background, it would be sensible for Awarding Bodies like AQA to play a central role in National Curriculum testing and assessment within the UK. However, there is little incentive for an organisation like AQA to involve itself in current National Curriculum testing because the educational work is inevitably tightly specified, leaving no scope for curriculum development and, historically, development of the assessment processes is marked by an aversion to innovation which makes the use of modern technology and the delivery of cost-efficiencies difficult. The result is that contracts to operate National Curriculum tests are essentially about delivery of routine administration on behalf of QCA, rather than educational or assessment development work.

21. This may be about to change to some extent with the proposed development by QCA of National Curriculum assessments which pupils will take when they are ready, with the intention of supporting personalised learning backed by diagnostic assessment in classrooms. It will be crucial for this work to be approached as an opportunity to develop National Curriculum assessments which are innovative in terms of assessment processes if the expertise of the Awarding Bodies is to be engaged, as it should be, in these important national developments.



Is the testing and assessment in "summative" tests (for example, GCSE, AS, A2) fit for purpose?

22. The purpose of GCSEs and A-levels is to identify within reasonable parameters of confidence, and certificate, a student's performance in a particular subject at a particular time. The assessment arrangements are fit for this purpose and based upon educationally valuable specifications which are only accredited following rigorous consideration by QCA. Specifications have to meet national qualification and subject criteria and assessment processes have to comply with a statutory Code of Practice. As a result, there is no evidence that standards of demand and reward are not broadly consistent across Awarding Bodies and over time so the confidence that end users of the qualifications have in them is not misplaced.

23. The standards of attainment which GCSE and A-levels represent are widely recognised, understood and valued by Further and Higher Education and employers. The qualifications are highly valued by young people and their parents.

Are the changes to GCSE coursework due to come into effect in 2009 reasonable? What alternative forms of assessment might be used?

24. In 2006, Examination coursework in GCSE and A-level examinations was the subject of an important report from QCA. The report confirmed the value of coursework in many subjects but the issues of candidates receiving assistance with their coursework and of plagiarism, especially involving the use of the internet, were highlighted. QCA outlined a number of initiatives to address these issues, such as improving the understanding of what is acceptable in terms of assistance. AQA is fully committed to these initiatives and has been actively working on them with QCA and other awarding bodies.

25. QCA has also reduced considerably the amount of coursework in examination requirements, particularly for the new GCSEs. There has also been a reduction in the amount of coursework involved in A-level for courses starting in 2008, as part of the change in the number of assessment units in most subjects from 6 to 4. Is the reduction in the amount of coursework in GCSE and A-level examinations the best policy to pursue? To answer this question, we need to understand how we have arrived at the current position.

26. Although coursework is the term that everybody uses, teacher-assessment might be a better one. Most GCSE and A-level coursework components consist of a well-defined piece (or pieces) of work which students complete during their course and which is then marked by their own teacher. Rigorous external moderation procedures are applied by the Awarding Bodies after the assessments have been made, in which external examiners re-mark a sample of work from each school to ensure that every teacher's marking is done to the same standard.

27. However, examination coursework was originally intended to be work carried out during the course, not an additional examination requirement such as an extended essay or project. For example, the early requirements for examination coursework in GCSE English Literature were for assessments by teachers of pieces of work produced during the course of study across a range of genres, periods, and so on - work which arose naturally as part of the study of literature over two years. This approach leads to a wide range of work being produced and assessed and requires significant professional participation in standardising and moderating the work. Consequently, moderation of this sort of coursework focussed upon training and professional development in meetings of teachers organised by Awarding Bodies, as well as on external checking after the event.

28. Between the 1980s and the present day, coursework changed in its nature and the perception of it changed equally. Concerns expressed, but never justified with substantial evidence, about the extent to which the original approach involved trusting in the professionalism of teachers led policy makers to seek increasing amounts of control over the nature of the work assessed and direct moderation of the marks awarded. The consequence is the situation we now have where a more formulaic and controlled approach leads to less motivation for students and more of a sense of burden for teachers. The very tight definition of the coursework which candidates have to do facilitates plagiarism and other practices which now cause concern.

29. In essence, the historical attempt to reduce risks relating to teacher professionalism by increasing amounts of control, has created different risks relating to the authentication of coursework as the work of the students themselves. The question is whether it is now possible to move back towards a system in which examination coursework arises more naturally as part of the student's learning and is assessed by teachers working under a framework of quality assurance, as well as quality control.

30. Ironically, at the very time when QCA is reducing coursework in many GCSE and A-level examinations, the new Diplomas will involve subjects and units which, by their very nature, require assessment by teachers of vocationally related activities, many of which involve hands-on practical work. The resources required to externally examine these activities would be prohibitive and they do not lend themselves to re-marking after the event by an external moderator. Similarly, Applied GCSEs are already fully operational and feature a pattern of two internally assessed portfolio units and one external assessment. Teaching for Applied GCE A-levels started in September 2005 and this summer (2007) will see students being awarded the first full Applied A-level results. These are a key part of the strategy to broaden the range of learning available to 14 to 19 year olds and, in general, consist of 2/3 coursework and 1/3 external assessment. They involve an approach which puts major emphasis on the accreditation of individual teachers, who retain their accreditation even if they move centre, subject to light moderation.

31. With appropriate structures of support and moderation, teacher assessment can provide valid and reliable results within an environment that is encouraging to students, rather than daunting. The concerns that have been voiced about possible plagiarism through the use of the internet and malpractice through the input of others (fellow students, parents, and so on) are valid, but appropriate action can be taken to ensure fairness and that appropriate grades are awarded. Work which arises naturally during the course is much more varied, enabling teachers more easily to detect plagiarism, especially if the work is done, at least in part, in the classroom under controlled conditions. Discussions over what represent adequately controlled conditions are currently taking place to define criteria for teacher-assessed components and AQA will put those outcomes into action. It is possible to set assessment pieces that are of a task-like nature but which can be completed in the classroom, under direct supervision and within a restricted timeframe. In these sorts of ways, plagiarism can be made much more difficult to do and easier to detect so that, if it does occur, it can be identified and penalised.

32. In the world outside education it is relatively unusual for anyone to sit down and think and write continuously without the opportunity to explore additional resources or check points with colleagues. Such opportunities reinforce the value of collaborative and team building skills. Many modern coursework requirements are artificial to the extent that help from colleagues - in this case teachers and parents - is artificially restricted. If we are preparing our young people for the world beyond school, then activities which require elements of independent study and research but also sharing and cross referencing with others must be appropriate.

33. These arguments show that alternatives to reducing the amount of coursework in examinations are possible and consistent with current policy developments. They also reflect the fact that, on a large scale, many practical skills cannot be assessed externally in a valid and cost-effective way. As is implicitly acknowledged in the new Applied GCSEs, Applied GCEs and Diplomas, some assessment is best done by teachers - a further example would be oral skills. QCA has therefore tried to identify those academic subjects which require coursework assessment and has prohibited its use in others. This is a restrictive and narrow approach which unnecessarily limits the choice available to teachers and students. If teacher assessment can provide good quality assessment when it is essential, there is little logic in prohibiting it as an option in any subject.

What are the benefits of exams and coursework? How should they work together? What should the balance between them be?

34. Balance is everything here. Exams ensure a level playing field, but can distort teaching and learning. Coursework, where it is well designed and well implemented, supports learning but, as noted above, makes it more difficult to ensure a level playing field. Each form of assessment can assess different aspects of knowledge, skills and understanding and they work well together where they do not assess the same things but are used in a complementary way.

35. There is no single answer to the question about balance. One balance is not appropriate for all subjects. A judgement must be made as to which assessment strategies best fit the particular subject and course specification and which strategies will provide the most valid and robust assessment of the student's abilities in each area of learning involved. By this means it is possible to produce an assessment scheme that is supportive of good teaching and facilitates learning. This is the aspiration which should determine the choice of balance between exams and coursework, or any other form of assessment.

Will the ways in which the new 14-19 diplomas are to be assessed impact on other qualifications, such as GCSE?

36. The multi-component nature of the Diplomas will impose pressures on students that are different from those which arise from the separate demands of a number of GCSE or A-level subjects. This will be particularly true at Level 2 where there is strong evidence that the requirement to pass the functional skills at that level before a Diploma can be awarded will be a severe challenge for many students. If a student is finding it difficult to maintain the requisite level of success across the range of demands of the Diploma it is likely that they will seek to reduce the pressures from elsewhere to enable them to focus on their Diploma work. However, this may well be a desirable trade-off if it replaces poor results in a range of GCSEs with success in a coherent Diploma course. The number of GCSEs taken by many young people is, in any case, an issue which is worthy of review and pressures to reduce it somewhat are not self-evidently disadvantageous, given that there is a requirement for the functional skills and other broader activities alongside the vocational cores of the new Diplomas.

37. There is, however, a related matter of major concern to AQA which arises from the proposed relationship between Functional Skills and GCSE qualifications in English Mathematics and ICT. For GCSE courses in these subjects starting in 2010 it is presently policy that students will be required to achieve Level 2 in the relevant Functional Skill to be eligible for the award of a Grade C. Our research into the potential impact of this 'hurdle' suggests a major risk that there will be a significant consequential reduction in the number of students achieving success at Grade C or better in these GCSEs, particularly in English.

38. Of course, it could be that the situation will be different in practice when the new GCSEs in English, Mathematics and ICT are first certificated in 2012. Levels of achievement in the functional skills will, hopefully, have improved significantly as a result of a strong emphasis on their teaching and learning in the intervening years. This is clearly the intention of the policy involved. However, the policy will require significant investment and careful monitoring of its success in order to ensure that GCSE standards can be maintained when the functional skills 'hurdles' begin to operate in 2012. The consequences of a major change in GCSE standards and outcomes in English and Mathematics in that year would be problematic because of the compulsory inclusion of these subjects within school performance measures at Key Stage 4. More important, however, is the potential for very grave injustice to be done to the young people affected, as they compete with those from the year before for the same jobs and places in Further and Higher Education.

Is holding formal summative tests at ages 16, 17 and 18 imposing too great a burden on students? If so, what changes should be made?

39. The burden is considerable at age 16 because of the range of subjects commonly taken. Where students are clearly going on to post-16 study, what purpose do 8-10 GCSE grades serve? Where students are clear which subjects they want to take through to post-16 education, it is entirely possible to have a system where they do not take the 'summative' GCSE examination in those subjects at 16. This was an aspect of the old O-level and A-level examinations which operated successfully in many schools. Clearly, in the current world where there is more mobility at age 16 there is a greater need for certification of achievement at 16 but for many students a record of subjects studied would serve the purpose adequately in subjects which they plan to study further - certainly when that further study is to be at the same school.

40. Similarly, there is little point in AS certification (and thus examination) when a student plans to progress to A2. If AS and A2 were decoupled, so that AS was no longer formally part of A-level examinations, such students could bypass assessment at AS and proceed straight to A-level. For the rest, AS could continue to act as a useful qualification, certificating their work in the first post-16 year and enabling them to choose their full A-levels informed by their progress so far. Such a decoupling arrangement would need the support of Higher Education as it would impact upon admissions arrangements but it would offer the opportunity for the A-level assessment to be more holistic in nature which is something for which many HE institutions have expressed support.

To what extent is frequent, modular assessment altering both the scope of teaching and the style of teaching?

41. Modular assessment arrangements can give rise to a perception that the subject as a whole is not addressed as it would be with a wholly end of course assessment. If the modular assessment structure is accompanied by a modular approach to teaching and learning, then opportunities to explore broader issues arising from the studies are reduced. In most subjects, particularly at A-level, these sorts of issues require a more comprehensive understanding of different aspects of study. It takes time to accumulate the knowledge and skills required to address real issues rather than ones that have been contrived to fit within the confines of the limited areas of study currently in focus. In order to attempt to address such concerns a number of artifices have been added to A-level, such as synoptic assessment and stretch and challenge questions, but with varying degrees of success.

42. On the other hand, modular examinations and courses enable greater flexibility of provision and can improve achievement through better structuring of courses, increased accessibility of course content to students, valuable feedback about progress and continuing motivation. They also help to ensure that students who do not successfully complete entire courses nonetheless receive proper recognition and certification of what they have achieved. Some institutions - particularly in the FE sector - use module results as part of their accountability arrangements for individual teachers.

43. It is for these sorts of reasons that many Higher Education courses of study are now modular in nature, despite precisely the same implications for holistic understanding of the subject being an issue at that level. It is also important to note that unitisation (modules) is a design feature of the new Diplomas and vocational qualifications generally.

44. The right response to this sort of situation in which there are clear strengths and weaknesses in a particular assessment model is to provide choice for schools, colleges and therefore students. One student's motivation from regular module tests is another student's stressful burden and a choice should be provided, rather than a single model to fit all. This was the case prior to the introduction of the Curriculum 2000 A-levels and could be so again if QCA were to develop appropriately flexible criteria for A-levels.

How does the national assessment system interact with university entrance? What does it mean for a national system of testing and assessment that universities are setting entrance tests as individual institutions?

45. University resources are not unlimited so we have to choose which of our young people we are going to make them available to. Putting this selection process on a sound basis is the key purpose of A-level examinations but there is currently a significant problem in using A-levels for selection to some university courses: a minority of University selectors find themselves faced with more applicants than they can accommodate who have straight A grades from their A-levels.

46. Does this mean that A-levels are too easy? The average Grade A pass rate for A-levels in 2006 was about 24% - meaning that almost a quarter of the candidates for a typical A-level got a grade A. There are five grades altogether, so having a quarter of candidates in the top grade means that the grade scale is not providing as much differentiation between the candidates as it could. This is the result of maintaining the standard represented by the grades over a period when young people's achievements improve significantly. Judged educationally, it is cause for celebration. But judged from the standpoint of the effectiveness of A-level examinations for university selection purposes it is a problem.

47. On the other hand, if we look at the complete results of individual candidates, most of whom take several A-levels, we do not see a general problem of differentiation:



4 As

3 As

2 AS

1 A

Cumulative % of candidates





Cumulative % of 18 year-olds





48. Only just over 9% of candidates get 3 or more Grade As - and that is just 3% of all the 18 year-olds in the country in any one year because only about a third of 18 year-olds take any A-levels at all. Only about 11% of our 18 year-olds get one or more Grade As at A-level. From these figures, it is very hard to argue a case for making it harder to get a Grade A at A-level, but there is a case for introducing a Grade A* to differentiate within the top 25% of candidates in each subject and to ensure that A-levels remain fit for their purpose of providing the basis for fair selection for university courses of all types.

49. AQA is therefore strongly supportive of the decision to introduce a new A* grade at A-level. There are to be revised A-level courses starting in 2008 which include some more demanding and open-ended questions and the present policy is to bring in the new A* when the first awards are made for those new A-levels in 2010. However, there is enough information in the marks of candidates taking current A-levels to enable us to differentiate effectively between candidates within the present Grade A and we believe that we could, and should be permitted to, introduce the new A* two years earlier than that, in 2008. We would urge that this be given serious consideration.

50. Of course, there is detail to be worked through about the precise mechanism for awarding Grade A* but AQA has already done a substantial amount of research on this topic. There are other important matters to be decided as well - not least the question of how many UCAS points the new A* grade will be given. However, none of these technical matters is difficult or complex enough to prevent A* grades being awarded and used in university selection from Summer 2008.

51. There are two issues of concern about the introduction of an A* grade. One is the increased pressure which it will exert on young people. Some of those who currently aspire to achieve straight Grade As will now aspire to straight Grade A*s. For some, that will be an appropriate aspiration - and it is not the business of educators to persuade their students out of having high expectations of themselves - but for others it will not be an appropriate target and it will be vital for teachers to provide wise counsel to students about how best to spend their time and energy during their A-level studies. It will be essential for everyone to understand that the standard of Grade A has not changed - that a Grade A at A-level remains an excellent result which only 11% of our 18 year-olds achieve.

52. The other concern which has sometimes been expressed about the introduction of an A* grade at A-level relates to widening participation and, particularly, the issue of inclusiveness. In particular, independent school students are over-represented among those awarded a grade A, given the proportion of all candidates educated in independent schools. Independent school students make up 14% of all A-level candidates but 28% of those with grade A. Will they predominate even more in the new Grade A*? Preliminary research carried out by AQA across a range of subjects, including Art, Sciences, English Literature, Foreign Languages and Social Sciences suggests that the proportion of independent school pupils in Grade A* is likely to be a little higher than it is for Grade A, at around 33%, with 18% coming from FE and 6th form colleges and 47% of candidates with A* grades coming from maintained schools. There is a challenge here for the maintained sector as a whole but these are only average figures. The reality is that there is a range of achievement in all school types, with many maintained schools and colleges more than matching their independent colleagues in terms of examination results. It can only be good if the introduction of a new Grade A* at A-level serves as a spur to further improvement in schools of all types.

53. Certainly, the introduction of specific university entrance tests is a backward step and not conducive to an education system that seeks to give the same opportunities to all. Such tests bring significant risks of curriculum distortion and problems of social inclusion as a result of differential availability and level of preparation for candidates from different backgrounds and in different types of schools. A national assessment system which meets the selection requirements of all universities is a much more equitable approach and is the pattern in use generally around the world.

54. One of the ways of selecting people for University which is sometimes suggested as being better than A-levels, precisely from a social inclusion perspective, is the use of aptitude, or reasoning, tests. Those who propose aptitude - or reasoning - tests for university selection usually make two, related claims:

They claim that assessing aptitude or reasoning ability, rather than class-room learning, removes the effects of schooling. So they claim it is a fairer way of selecting which will help to widen participation and promote social inclusion.

They also often claim that it is better because it provides better predictions of success at university.

55. Neither of these claims is consistent with the evidence. In the USA aptitude tests have been used to select students for university for years, with the consequence that a great deal of time is spent practising for the tests. And in this country when selection at 11 was common primary school children spent many hours practising for their 11 plus. The fact is that scores can be significantly improved on any sort of test by practice and preparation. All the evidence is that teaching improves aptitude test scores - they are not "school proof" and are not, therefore, inherently better than examinations in terms of ensuring that selection is fair or socially inclusive.

56. Nor are aptitude or reasoning tests any better predictors of success in university than examinations. In fact, in this country, studies have shown A-level to be the best single predictor of success at university, albeit not a very good one. Around the world, there are countries which use tests of various kinds, countries which use examinations and plenty of combined approaches. Overwhelmingly, the evidence is that none of these selection methods provides very reliable predictions of students' performance at university. And nor is it surprising that it is difficult to predict the future 3 years hence - a future which involves a quite different approach to education and some of the most significant developments in many students' personal lives.

57. So aptitude tests are neither fairer nor better predictors of success than exams like A-levels. But their downside is that time spent practising and preparing for them could be spent on learning which has real benefit. Preparing for exams like A-levels which are embedded in the curriculum is about learning which has purpose. The use of tests which supposedly assess reasoning ability or aptitude, independent of schooling, can - and usually does - distort the curriculum significantly.

58. But if it is to be examinations, rather than aptitude tests, which we use to select young people for university, would baccalaureate style examinations be better than A-levels? Baccalaureates compel each student to follow a broad range of study. This may mean including mathematics, their mother tongue, a foreign language, perhaps something about the theory of knowledge in their course - as well as the subjects in which they are specialising. Choice for students is therefore constrained. Such constraints are not self-evidently desirable, especially at a time when individualised learning is seen as an important tool in the drive to encourage more young people to stay in education after the age of 16, and to go to University. For example, when there is concern about the decline in the number of people studying physical sciences, we would surely not want to prevent a talented scientist from going to University because they either cannot , or will not, study a modern foreign language.

59. A-levels, unlike baccalaureates, offer a wide choice for students in which broad study or specialisation is possible, depending upon their own interests and enthusiasms. And it is interest and enthusiasm which leads to learning which lasts, rather than learning which is done purely to get a qualification and which is then forgotten afterwards.

60. A-levels are also available in ones and twos. It is often not appreciated that 25% of our 18 year old A-level candidates take only 1 or 2 of them. This means that students who have a particular talent, but who would struggle to succeed in all the elements of a baccalaureate, can obtain valued qualifications which fully recognise their achievement. We should not deny those students that possibility, perhaps forcing them, instead, to take a lower level baccalaureate qualification which does not do justice to their particular talents in Art, Sport, Music or, indeed, in English Literature or Mathematics, if that is where their abilities lie. Such an approach would be inconsistent with the aim of developing each individual to the full extent of their capabilities, both for their own benefit, and that of society at large.

61. Baccalaureates should be offered as a choice for those whose educational needs are best met that way and AQA is currently developing its own baccalaureate qualification which will certificate student's A-level results alongside the completion of a personal in-depth study, learning about thinking or citizenship and their wider activities such as work experience, contributions in the local community or personal development programmes such as the Duke of Edinburgh's Award scheme.



AQA's Director General, Mike Cresswell, would be delighted to give oral evidence to the Select Committee if required and would welcome the opportunity for further discussion and clarification of any of the points raised in this memorandum.

As well as being Director General of AQA, Mike Cresswell is a Visiting Professor at London University's Institute of Education. He has worked on national and international surveys of students' mathematical attainment at the National Foundation for Educational Research and was an active researcher on assessment matters for many years before the formation of AQA for which he was Director of Examinations Administration before taking up his present post in 2003.

He has a national and international reputation as an expert on assessment - particularly on the topic of setting examination standards. He has published many papers, research reports and books on assessment. He served on Tomlinson's Assessment Group and continues to be closely involved in national discussions on the introduction of the new Diplomas, contributing especially to the recent work on methods of awarding candidates' grades.

Mike believes the major challenge of the next few years in the assessment field is the search for ways of bringing technology to bear on assessment to reduce costs and burden on schools and students in a way which both improves assessment quality and retains fitness for educational purpose.