Thus he restored the status of assessment as a continuous process, with pupils’ progress being monitored throughout the key stage, and culminating in their level of attainment being reported at the end of the key stage, which will become statutory requirement next summer.

Most day-to-day language classroom activities can be given an assessment focus, and can provide information about pupils’ performance. The School Curriculum and Assessment Authority and its Welsh counterpart, ACAC, have jointly published a collection of pupils’ work - Consistency in Teacher Assessment: Exemplification of Standards, Key Stage 3 Modern Foreign Languages - which demonstrates how work done in each of the four attainment targets can be assessed against the national curriculum level descriptions, and how this can be used diagnostically to help pupils progress further.

Last month saw the publication of further help for teachers - a pack of optional tests and tasks, designed to help teachers track their pupils’ progress during the key stage, and also make a summative judgement at the end of key stage 3. These teststasks act as externally-provided benchmarks to stand alongside the range of everyday sources of information about pupils’ achievement - they supplement but in no way supplant teachers’ professional judgement.

Initially a few test units in French and German will be available to schools; in spring 1997, a further set of French and German units will follow, along with Spanish and Urdu test units. The materials have been developed at the University of York on behalf of SCAA. The project team included practising teachers and the format of the materials has been trialled extensively in schools.

The test units, and the activities within them, are linked thematically and cover the range of topics suggested for key stage 3 areas of experience (everyday activities; personal and social life; and the world around us). They also cover as much of the programme of study 1 as is possible within the structure of the tests.

Since the tests are not formally taken or externally marked, this has created an opportunity for the tasks within them to be based on activities familiar to the languages classroom, but not feasible in the formal set-up of the GCSE. Thus there are examples of pair-work and survey tasks (attainment target 2), poster creation and word-puzzles (attainment targets 34), and - in the second batch of materials due to be published in spring next year - examples of extended reading as a means of assessment.

However, there are also aspects of the teststasks that look ahead to the new GCSE format. All the activities are presented to the pupils in the target language, though the accompanying teacher materials are in English (or Welsh). Scene-setting and rubrics are generally concise, and there are worked examples, symbols and pictures to help pupils’ understanding and to support them in carrying out the tasks. Across the four languages, there will be more than 200 tasks for pupils. Many involve mixed-skill activities, but tend mostly to focus on a single attainment target for assessment purposes. Each task is accompanied by: * teacher’s notes, which explain what the task entails and how it relates to specific aspects of the curriculum, in particular elements of the level descriptions; tape script, where appropriate, and answers to listening and reading tests. The notes sometimes suggest how tasks could be adapted or extended, or given a different focus, and also how IT elements could be incorporated * an assessment scheme which shows the level or levels that the task is pitched at, and explains why, via a summary of the characteristics of the level(s) which are appropriate to the task.

The activities are differentiated to allow pupils to show evidence of achievement throughout the whole range of levels 1 to 8 where this is appropriate. Tasks that principally assess the more active skills - speaking and writing - sometimes offer scope for a range of achievement according to outcome.

Those that assess the more receptive skills - listening and reading - where the difficulty of the listening or reading stimulus input is important, are usually based on criteria relating to one level description. Each activity has been assigned a level or band of levels as a guide to teachers, although the accompanying handbook stresses that no single piece of work can ever guarantee “a level achieved”.

This “best-fit” message is reinforced in the Exemplification book - “You will arrive at judgments by taking into account strengths and weaknesses in performance across a range of contexts and over a period of time, not by focusing on a single piece of work”. Most of the materials are aimed at levels 2 to 6, since the national curriculum states that by the end of key stage 3, the performance of most pupils should be within that level range.

However, a few tasks focus on achievement at level 1, and at level 7 or 8. It was heartening, during the trials of the materials, to note that teachers felt that the levels achieved in the tests by their pupils matched their expectations.

This use of the level descriptions across the four attainment targets to assess pupils’ work in terms of national curriculum progression should be more informative than traditional marks and grades, and help answer one of the Office for Standards in Education’s criticisms, namely that reporting to parents tends to comment on a pupil’s attitude and effort at the expense of saying what the pupil can do.

These teststasks are not designed to be used as one-off, end of key stage exams, and in the absence of such a mechanism, teachers are placed at the centre of the assessment process.

The optional tests and tasks aim to fulfil the two main functions of assessment: telling the learner, the teacher and the “system” (such as parents and school management) how well a pupil is doing, and what needs to be done next.

No assessment materials can be ideal for every situation, but as Ted Wragg, professor of education at Exeter University, has pointed out, one of the problems with pupil testing is that it is assumed that the perfect calibration lies just around the corner, if only someone can find it.

David Crossland, director, and Phil Horsfall, deputy diretor, led the University of York project on behalf of SCAA to develop optional tests and tasks in French, German, Spanish and Urdu.