ATL

Assessing to learn

Informed by research evidence and member views, this position statement contributes to the assessment debate and provides a framework for future work by ATL.

ATL's vision of education for children and young people is founded on their development as physical, moral, social and intellectual beings. Central to this belief is the revitalisation of teachers' roles as professionals, equipped and empowered to lead a professional debate on learning and teaching. ATL's curriculum policy, Subject to change, depicts a role for teachers where, within a local context, they have input into curriculum content development, through which the skills defined in a light national curriculum framework could be developed.

ATL's skills-based curriculum model requires a radical re-think of the current knowledge-based assessment system. Indeed, the impetus for a change to the assessment system has come from many quarters and not just in terms of curriculum reform. There has been a growing consensus amongst education professionals and researchers that the current assessment system is flawed due to its impact on learning and the reliability of the inferences made from current outcomes.

ATL believes that teachers are at the heart of developing a better alternative. There is little doubt that teachers have a positive role to play in formatively assessing their pupils, particularly through strategies such as Assessment for Learning (AfL). However, from foundation stage onwards, teachers also have an equally strong role to play in summative assessment and ATL recognises that teacher assessment provides rich, reliable and valid data on pupil achievement at local and national levels.

Informed by research evidence and member views, this position statement contributes to the assessment debate and provides a framework for future work by ATL. The general principles in this position statement apply across the UK; whilst most of the proposals relate specifically to England some of the proposals in this document are already working well in other parts of the UK.

The current assessment picture

Our current pupil cohorts face a barrage of national tests and exams throughout their school careers, from Key Stage tests to GCSEs, 'AS' and 'A' levels. Our children are widely understood to be amongst the most tested in Europe, yet pupil performance continues to slip down the international league tables. In addition, research and member evidence has demonstrated that the tests and exams, within the current high-stakes national assessment system, have a negative impact in that they:

• narrow the curriculum and reduce flexibility in curriculum coverage

• undermine the Every Child Matters agenda and have a negative impact on pupil attitude and motivation

• depress staff morale and lead to 'teaching to the test'

• stifle creativity in teaching and learning

• result in lack of engagement with subjects deemed to be more difficult.

Employers, colleges and universities express strong concerns regarding the impact of the current English assessment system. Colleges are required to provide remedial classes for students, employers find that their recruits are unable to practically apply knowledge and universities find that students have not been given the opportunity to develop the deeper thinking and analytical skills needed for higher education. Tests provide a large quantity of data. This data is used for too many different purposes and the inferences drawn are often inaccurate. The data does not provide meaningful information about their child's learning for parents.

In this current milieu, teacher assessment is seen as subordinate to external testing. ATL believes that there needs to be a fundamental re-balance in the relationship between formal external assessment and teacher assessment. ATL proposes, as an alternative, an assessment system based on teacher formative and summative assessment, which would meet current system needs and be flexible enough to support curriculum reform.

Teacher assessment as the reliable and valid option

Currently, tests are deemed to be more reliable than teacher assessment by the public. However, research indicates that up to 30 percent of candidates in any school public examination or test in the UK will receive the wrong level or grade. ATL recognises that this lack of reliability is systemic and inevitable given the nature of the testing system and its intrinsic sampling errors; short writing tests do not provide a good indicator of pupil ability.

Teacher assessment, however, through regular and comprehensive assessment of each pupil and using evidence drawn from the full range of classroom activity, avoids the narrow sampling error which makes tests unreliable in measuring actual pupil achievement and skill.

National tests can also be subject to bias; research has found bias in terms of gender, first language and SEN in the results of standard tests and tasks, as well as the bias of unfamiliar situations. ATL recognises concerns that teacher assessment may be biased. We believe that this risk can be mitigated by effective training and moderation. Non-relevant assessment factors, such as pupil behaviour and gender, can be recognised as potential sources of bias and guarded against by teachers and moderators.

What is often not considered in the argument is the validity of test results; that they convey the achievement of specified skills and knowledge to pupils and other stakeholders. A narrow paper-based test reliant on memory recall reveals little beyond the ability of pupils to take tests, for which many are now coached. They report little on broader abilities and skills and therefore fail in this respect. Teacher assessment is likely to be more valid, by varying the tools of assessment and measurement and by judging abilities and skills in more realistic and varying contexts.

ATL calls for a comprehensive review of the current national assessment system that publicly acknowledges recent research, including findings related to unreliability, bias and lack of validity.

Features of an effective teacher assessment system

ATL's vision of a teacher assessment system is one that builds on current expertise in the profession, the development of school networks to support good practice and a rigorous moderation system. Currently, every teacher assesses their pupils' learning; ATL's vision is one where this assessment carries weight and is not undermined by a public emphasis on test results as the only reliable indicator. However, our members recognise that many teachers no longer feel confident in taking a greater role in assessment due to the de-skilling effect of a reduced role over a period of several years.

ATL envisages an effective teacher assessment system that possesses the following features:

• teachers as participants in the development of assessment criteria, sharing them with learners

• expert assessors (like the currently under-development Chartered Educational Assessor role) supporting the teacher assessment process locally, with a long-term aim of at least one assessor in every school

• assessment tools such as AfL strategies and banks of tests, created and used by teachers

• professional networks within and between local schools, supporting assessment and its moderation and building on already formalised networks, eg diploma-delivery networks in the secondary sector

• moderation options such as the pairing of schools nationally to moderate each other, or locally-based moderation within and between neighbouring schools

• online resources (such as e-portfolios) where appropriate and according to the needs of teachers' subjects, pupils, availability of resources and the skills to be validated.

This teacher assessment system cannot thrive in tandem with national external tests. ATL believes that there should be no national testing system prior to the terminal stage of compulsory schooling and that teacher assessment is key to that final assessment.

International evidence clearly links high pupil achievement with systems that postpone national assessment and selection. Bowing to political imperative and allied parental fears, there may still be a need for national tests in key skills, such as English, Maths and IT competency, which can be offered through a bank of tests on a when-ready basis, but only at the level of functional competence.

In terms of meeting the needs of national monitoring of educational standards, ATL proposes a system of national cohort sampling, thus reducing the overall test burden whilst increasing the relevance and breadth of learner evidence. This would entail regular across-the-curriculum surveys of small random samples of pupils and is a system adopted by other countries, eg the Scottish survey of achievement, the National assessment of educational progress in the US and the National education monitoring project in New Zealand.

A new professionalism

The implications of teacher assessment for initial teacher training and early and continuing professional development are profound. An increased proportion of initial training time on assessment and a corresponding emphasis in the professional development opportunities available to new and established teachers would be required. Investment of money and time would be critical to the achievement of a profession re-vitalised in assessment.

This investment must be measured against the current financial cost of the testing and examination system, which has risen dramatically in the last few years with the proliferation of tests. The current cost has been estimated by PricewaterhouseCoopers as £610 million, a figure judged to be conservative. ATL's vision sees the incorporation of teacher assessment as a central part of teaching, not an add-on, and moderation becoming a vital part of teachers' continuing professional development.

ATL has identified the key areas of professional development support required for teachers:

• development, understanding and application of assessment criteria

• the development of assessment/test models

• moderation

• various assessment approaches and tools, particularly AfL.

ATL recognises that the teaching profession has faced the challenge of continual initiatives and change, driven by political rather than educational imperatives. We must be clear that it is education that drives this proposed change and that a new assessment system, such as outlined here, whilst challenging in its inception, offers teachers the opportunity for a more appropriate and effective use of their professional skills.

Key challenges

ATL's recent member survey indicates that time and workload are key issues for teachers in addressing teacher assessment. Preparation for tests currently takes up a disproportionate amount of teaching time. In 80 per cent of primary schools, the amount of time spent on Key Stage 2 test preparation has increased in the past 10 years, with a similar pattern of responses from secondary schools.

Comparisons with Scotland, which operates a system far closer to teacher assessment, demonstrates that reduction of national testing has a significant impact on teaching time. Equally important for teachers is the amount of time that a system based on teacher assessment would take outside the classroom, in developing or choosing suitable assessments, in marking assessed work and in moderating and recording judgements, for example.

Teachers already carry out this work, in support of the national testing system, in non-teaching time. ATL will continue to gather member evidence on teachers' working hours and campaign to ensure that it is manageable and meaningful.

Previous teacher summative assessment experience has existed as part of the overall high-stakes national testing and examination system and ATL acknowledges the negative impact that this system had on teachers' experience. For example, concerns with plagiarism and constant re-working of submissions made coursework an onerous burden on teachers.

Teachers know that it is the high-stakes nature of the current testing system that provides much of the motivation for poor practice and for plagiarism. ATL believes that the local model of curriculum and assessment that we advocate would remove much of the opportunity for plagiarism by increasing the variety, in terms of task and context, of assessment tasks such as coursework. A low-stakes assessment system would remove the perceived necessity for the frequent re-working of coursework with the attendant frustrations for teachers and pupils alike.

The bedding down of a teacher assessment system will demand time, work and professional commitment. However, ATL's vision of a teacher assessment system will not only remove much of current teacher workload around practice tests, re-submission of coursework and the re-sitting of modular tests but will replace it with an assessment system that directly supports their teaching, engages pupils and allows teachers more creativity and depth in their teaching and their pupils' learning.

Proposals for change

ATL calls for:

• An urgent government review of the current national assessment system which publicly acknowledges recent research, including findings related to unreliability and broader concerns regarding impact on curriculum, learning, pupil and teacher attitude and motivation and validity.

• Investment in assessment-based professional development, from specific training on assessment to the formation of local networks of assessment expertise and experience. A greater emphasis on teacher assessment in the professional standards framework, including QTS, induction and leadership.

• The development of published exemplification materials; with input from teachers, a bank of resources which is constantly reviewed and updated and is owned by teachers.

• Rigorous piloting and evaluation. Pilots that combine assessment for learning approaches alongside teacher summative assessment but in schools exempt from national testing so that valid inferences can be drawn from the pilot data.

• Balanced accountability, including: an exploration of external monitoring alternatives (cohort sampling); a review of Ofsted's role, exploring alternative methods of accountability that emphasise support for improvement and re-focus priorities on learning and the assessment that supports it rather than narrow grades; the removal of league tables. Teacher assessment relies on school collaboration and support, something which league tables and the resulting culture of competition undermines; an exploration of more meaningful ways of being accountable to parents.

ATL recognises teachers' weariness with endless politically-driven change. We therefore propose an incremental process of change towards a teacher-led assessment system with no national summative assessment until terminal stage, the latter echoing the vision of the Task Group on Assessment and Testing and Tomlinson.

ATL believes that only a highly professionalised and assessment-expert workforce, with teachers as key players in curriculum and assessment development, can ensure that our children and young people have access to an education that engages them and which, ultimately, nurtures and testifies to a fuller range of their abilities and skills.

If you would like further information or to comment on this briefing paper please do so by contacting ATL.