Assessment: levelling off?
By Blog Editor, IOE Digital, on 17 June 2013
On Thursday, OFSTED reported that a quarter of those who secured Level 5 at the age of 11 did not get top grades at GCSE; I blogged on it here. On Friday, government solved the problem: it abolished National Curriculum levels. If only all educational challenges could be solved so easily. I picked up the news on a train returning from a day working with headteachers in Somerset and we were powering across the Somerset Levels, which felt somehow appropriate.
The eight national curriculum levels in use now derive from Paul Black’s Task Group on Assessment and Testing (TGAT) Report of 1988. It is a sophisticated report, still worth reading, which tried to devise a national assessment system precise enough to give information about pupil progress but flexible enough to recognise that pupil learning is not simply cumulative. As Black observed, “no system has yet been constructed that meets all the criteria of progression, moderation, formative and criterion-referenced assessment” (para 5). TGAT’s 10-level scheme, drawing together test results, teacher assessment and teacher moderation was intended to secure “ information about … progress … in such a way that the individual child’s progress can be related to national performance” (para 126).
The implementation of TGAT was – as is always the case – partial. The recommendation that “wherever schools use national assessment results in reports for evaluative purposes, they should report the distribution of pupil achievements” (para 84) was lost as progression was mapped through a level-based system . Assessment at 14 was never implemented as intended. Working groups each generated statements of attainment at each level in each subject – Science alone had seventeen attainment targets, so 170 statements of attainment. TGAT’s proposals were, as reform proposals always are, sharply criticised. Some who wanted “simple” tests disliked the attempt to draw external and teacher assessment into a single 10-level structure. Others thought it would lead teachers to think of pupils as numbers. But level-based assessment was rapidly adopted, not least because in areas of the school system it was working with the grain of thinking. Substantial work had been done on graded assessment in languages, GCSE operated levels of response mark schemes, and primary mathematics schemes often had a level-based structure. Lord Dearing’s review of the national curriculum in 1994 moved from statements of attainment to more generic level descriptors, and lopped levels 9 and 10 off TGAT’s structure.
Since 1994, the use of levels has become embedded. Some teachers – wrongly – assign levels to every piece of work, which misses the point about generic descriptors. Better practice happens where teachers have used levels to inform assessment and planning and sharpen the relationship between, say, level 5 writing and level 5 speaking in English, or progression in History between levels 5 and 6. Precisely because levels are not disaggregated, they have informed thinking about what performance looks like across different cognitive domains and how domains relate to each other. Primary and secondary classrooms often feature posters using pupil-friendly language to help pupils identify how to consolidate learning and move on. The application of level descriptors itself involves professional judgement. Black’s aspirations for the way the TGAT structure could shape progression and formative assessment have been substantially achieved.
Ironically, it is what TGAT envisaged as criterion-reference assessments related to national performance that have caused more difficulty. Such has been the focus on key hinge points of the structure – the importance attached to the proportion of pupils securing a Level 4 at Key Stage 2, for example – that schools have engaged in frenetic activity to push pupils over the ‘next’ threshold for key assessments, so that a higher level is achieved, but not always securely. ‘Sub-levels’ –disaggregated progression within levels – have generated data of dubious reliability. Partly for this reason, many secondary schools re-assess pupils soon after they arrive, and, perhaps inevitably, decide to place greater emphasis on these tests than on national curriculum levels.
The government’s announcement that levels are being abandoned suggested that “outstanding schools and teaching schools” would lead the way in devising alternative ways of tracking progress. More likely will be an expansion of external tests, like those from NFER, CEM or from commercial test and publishing companies – News Corporation has been buying into the market in the USA. National benchmarking will still exist through end of Key Stage tests; without levels, these will presumably generate a numerical score, though TGAT warned (para 10) that where external tests are detached from school assessment practice they tend to be “narrow in scope and employ a limited set of techniques….[and] rarely assist normal teaching work”.
Profusion of local and commercial assessment puts at risk the real long-term achievement of TGAT: its contribution to raising standards across the school system. It made expectations are clearer – from Carlisle to Canterbury, from Newquay to Newcastle. Some won’t mourn levels, especially those who have seen them misused. But the level structure focused attention on what pupils need to learn, learning outcomes, and on progression towards those outcomes.
TGAT recognised that “many schools have established common systems for use throughout the school”, but teachers’ use of assessment was “limited by the range of instruments available, by the lack of time and resources required to develop new instruments, and by the lack of external reference and a basis for comparison”, so the “needs of the less able, or the competence of the most able, have hitherto been under-estimated” (paras 7-8). And that, more or less, brings us back to OFSTED’s concerns.