The emphasis will no longer be on teachers’ ability to break down individual pieces of children’s work into their constituent parts, but to judge the overall quality of a child’s work in terms of “best fit”. This will require a fundamental change in some teachers’ thinking: from thinking quantitatively (ticking-off and adding up individual features) to thinking qualitatively (developing judgments of overall quality).

However, both approaches require a clear view of what counts as quality. This means understanding what standards of performance are expected or accepted at each level. The level descriptions cannot supply this kind of information by themselves because they are generalised and require interpretation in relation to concrete examples.

Over the past four years, teachers have developed their understanding of standards through local authority-run activities, particularly meetings to which teachers have brought samples of children’s work to discuss and agree the attainments and levels demonstrated.

LEAs no longer have Grants for Education Support and Training (GEST) funds to help finance such assessment training but teachers will still need support. Professional judgment does not just happen, as some of the teachers’ unions seem to imply; it needs to be fostered, particularly if there is an expectation that one teacher’s professional judgment should be similar to that of others. So where is this support going to come from?

In what remains of their 199495 GEST allocation, some LEAs are doing what they can to prepare for level descriptions by compiling samples of children’s work to illustrate interpretations of levels, and others to illustrate “best fit” judgments - although schools will not be obliged to use them.

In July, the School Curriculum and Assessment Authority and Awdurdod Cwricwlwm ac Asesu Cymru (the curriculum and assessment authority for Wales) plan to publish guidance, along the lines of the old School Examinations and Assessment Council’s well-regarded “Children’s Work Assessed”, on establishing a basis for “best fit” judgments. Further advice is likely to focus on the need to secure common interpretations of standards across the school. These will need to relate to national standards as illustrated in the new version of “Children’s Work Assessed” and in the national tests.

This much SCAA and ACAC can insist upon, because OFSTED will look for evidence of such procedures in its four-yearly inspections of schools. However, in two other vitally important areas they are likely to remain silent, because of resource implications andor the anticipated reactions of the teachers’ unions. First, there is unlikely to be any strong advice about procedures for developing common standards across schools because LEAs will no longer have GEST funds for training. Anything that is done will be at the discretion of headteachers, who will need to commit their own funds to joint activities. Secondly, there is unlikely to be advice on on-going “formative assessment” - assessment for learning - because, in accepting Dearing’s advice that this should be the professional responsibility of teachers, government agencies can only properly concern themselves with statutory end-of-key-stage teacher assessment (TA).

All of this leaves a great deal in the hands of headteachers who will need to think about the in-service opportunities teachers need if they are to develop sound professional judgments. Teachers need not only to learn how to use level descriptions, but how these understandings can be used to help children learn.

Government interest is now clearly focused on assessment for accountability. It will therefore be up to schools and teachers to rescue the potential of assessment for learning.

So what might primary headteachers do to promote effective assessment in their schools, that meets statutory requirements but also promotes the wider goals of children’s learning? Here are some suggestions: * Give teachers access to training in assessment skills and techniques that will enable them to analyse what children know, understand and can do but also gives them information about children’s misunderstandings and difficulties. This is likely to include techniques of observing and questioning as well as “marking” class work.

* Provide opportunities for teachers to discuss children’s learning using both tangible and ephemeral evidence. The purpose will be to agree and check standards with reference to external materials, but also to discuss strategies for teaching and learning and to engage in joint curriculum planning to promote continuity and progression. (Building in assessment opportunities should be an explicit part of curriculum planning.) * On the basis of these discussions, develop a school portfolio of assessed samples of children’s work (some samples can be in the form of teachers’ notes of observations) agreed at the various levels. The portfolio could have two sections: one providing interpretations of the levels drawn from the work of different children; the other giving “best fit” judgments in relation to a range of work from individual children. This would become reference material for teachers, governors, inspectors and parents.

* Commit some resources to allow teachers to meet with colleagues in other schools to refine their understandings of common standards by discussing children’s work from different schools. The school portfolio of evidence could be used in this context in conjunction with the materials produced by SCAA and ACAC.

A programme such as this will take time, money and commitment and some schools will be tempted to follow an easier alternative, such as using past test papers to determine a TA level that can be statutorily reported to parents while allowing teachers to “do their own thing” in day-to-day assessment. This would be to miss an important opportunity to make assessment work for real educational benefits in terms of children’s learning and teacher development.

* Dr Mary James is a lecturer at the University of Cambridge Institute of Education. With Colin Conner she has evaluated the moderation of assessment at KS1, from 1991 to 1994, on behalf of six East Anglian LEAs.