Thursday 19 January 2017

Primary assessment: NAHT call for end of Key Stage 1 statutory assessments and broader judgement of school effectiveness

The National Association of Headteachers (NAHT) yesterday published the Assessment Review Group's report on primary assessment. The Group was set up in May 2016 so this is speedy work that reflects the great concern from parents and teachers about last year's chaos around SATs and the new curriculum, teacher and pupil stress and the narrowing of the primary curriculum through the domination of high stakes testing. The report does not go as far as seeking the abolition of Key Stage 2 SATs, a demand that arose from teachers and parents last year, but it does argue for the end of Key Stage 1 SATs.

The report, LINK claims it is a contribution to the process of seeking a consensus around how to 'redress the balance' and suggests :
Have two statutory assessment points for primary pupils 

Statutory assessment in primary school should be restricted to two points, Reception and Year 6, in order to create the space in between for schools to focus on delivering a broad and balanced curriculum, appropriate to the needs of all children. Throughout the primary phase, schools should be free to determine their own processes and procedures for pupil assessment, informed by widely available evidence of best practice, that allows teachers to maximise pupil learning and progress. 

High stakes testing narrows the focus of the curriculum to that which is tested. The group do not believe statutory testing should be used by the government to influence teaching, learning and pedagogy. The various screening checks deployed by the government, including phonics and the proposed multiplication tests, should instead become part of the national sampling framework.

Introduce a start of primary school statutory assessment 

In order to establish a baseline from which to measure progress, teachers would carry out an observation-based assessment during a child’s first year in primary school. This should take the form of a single, nationally agreed assessment to avoid a repetition of the problems experienced in 2015/16. We anticipate that a moderation process would be necessary to support this. Great care would need to be taken when designing such an assessment, with significant input from Early Years experts. It is important that the results of this assessment should not be used to set targets for individual pupils or as a predictor of their future progress. Instead, the data from this baseline should be used solely as part of a cohort level measure of progress at school, local and national level. 

Whilst it was relatively clear that the end point would be the summer term of year six, agreeing on the best ‘start point’ or baseline proved one of the most challenging issues the group faced. There was general agreement that the initial assessment or ‘start point’ should be as early as possible in a child’s time in school, in order to take full account of the progress they make throughout their primary schooling. There is much to consider regarding any baseline assessment and these issues are outlined later in this report. 

Remove end of Key Stage One statutory assessments 

In the proposed model there would be two statutory assessment points. One at the start of a child’s time in primary school and one at the end. The key measure arising from statutory assessment should be the progress children make between these two points therefore end of Key Stage 1 assessments should be removed as a statutory requirement. 

Streamline and improve Key Stage Two statutory assessments 

At the end of year six, in the medium term, we envisage statutory assessments in reading, maths and writing would continue in some form. Reading and maths would continue to be assessed through a national test, externally set and marked. Writing would remain teacher assessed through an improved system that focuses on the overall quality of a child’s writing rather than the component parts. Early evidence suggests that comparative judgement may provide a workable and valid alternative to current arrangements for teacher assessment of writing. 
 
Make statutory tests accessible and enable pupils to show progress 

Statutory assessments and tests must be designed in such a way that the majority of children are able to access them. At the very least, tests should be structured so that the questions, and where appropriate any texts, appear in order of difficulty. Serious consideration should be given to removing the hard time limits for statutory assessments, particularly in reading, and replacing these with a minimum and maximum time limit so that children can focus on demonstrating what they can do rather than test technique. Inevitably there is likely to be a very small proportion of children with more significant special educational needs who are not able to access the tests. The Rochford Review has offered some interesting and potentially useful recommendations in this specific area which should be considered fully. 

Introduce national sampling and assessment banks 

Within this model, the government would have the option of carrying out national sampling if there were a need to monitor standards in particular subjects or aspects of the curriculum. The data produced through sampling should be used to gain an understanding of national standards. It should not be used to hold individual schools to account but could provide national data against which schools can evaluate themselves. In the long term, there is potential for national sampling to replace the current model where every pupil takes every test at the end of Key Stage 2. 

All schools would be expected to have robust assessment processes in place and to be able to explain how they use these to support pupils’ learning, to identify and intervene where pupils are falling behind, and to report to parents. Schools should be mindful of the recommendations made in the Commission on Assessment Without Levels Final Report when designing such processes (DfE, 2015). To support teachers and schools, a national bank of assessment materials should be made available. Such resources would also help teachers in assessing the progress children are making against national expectations.
Report pupil performance as a score on the national scale 

The terminology used to describe pupils’ attainment in 2016 (working towards the expected standard, working at the expected standard or working at greater depth within the expected standard) was unhelpful, arbitrary and demotivating. Such an approach also fails to recognise and celebrate the progress that a significant group of pupils have made. The group were particularly concerned about the effect on those pupils who, despite making significant progress, could only be judged to be working below expectations at both the end of Key Stage 1 and the end of Key Stage 2. Stopping the use of such terms and simply reporting a child’s scaled score would be a positive step forwards. 

Accept data is only one part of the picture of school effectiveness 

It is important to reiterate that this model should be viewed in light of the overarching recommendation that any data produced from such statutory assessments should be seen as only one element when judging school effectiveness. Schools should not be held to account on the basis of this data alone. It is also important to  recognise that such statutory assessments will never be able to capture all aspects of a child’s progress or all the different ways in which a school contributes to the progress a child makes. 

No one single set of results should lead to negative consequences for the school. All data should be considered over a rolling three year period. There needs to be a recognition that cohorts of pupils vary; a dip in results in one year does not necessarily equate to a decline in school effectiveness. Basing interventions on such a short-term approach is unlikely to be helpful or indeed valid. 

End floor and coasting standards as determinants of intervention 

The use of floor standards and coasting standards to determine intervention in individual schools should be stopped. Instead there should be a greater level of dialogue between schools and those that seek to hold them to account, including RSCs. The starting point should be a discussion around the data to understand the context and story behind it. Any intervention at this point should be supportive, recognise the knowledge and understanding of the professionals working within the school and be based on working with the existing leadership team in the school. 

In an ideal world, data from assessments should be used as part of the inspection process. The results of the inspection may, if appropriate, trigger supportive intervention, and the RSCs (Regional School Commissioners) should base their work on the inspection results rather than independent evaluations. This streamlines the accountability system without reducing rigour, inserts the necessary expert judgement into the process, reduces conflict and duplication, and minimises the level of fear and uncertainty.
The report  makes the case for a separation of the statutory assessment and the school's own internal assessment procedures and calls for an emphasis on assessment to help children progress further - something that SATS do not do.
It is all too easy for statutory assessment to become entangled with in-school assessment - particularly when schools are driven to predict and provide data on future performance in statutory assessments. Under these conditions, in-school assessments inevitably take on the form of statutory assessments, in order to produce compatible data, however inappropriate this form may be to support teaching and learning. We should shift away from predictions of future performance and focus more on capturing accurate pictures of current performance of pupils against expected standards for their age. This has a major impact on what data should and shouldn’t be asked for. 

The core focus of assessment should be on supporting learning, not simply tracking progress. To help maximise the progress children make, we should expect all schools to have highly effective and robust assessment processes in place. These are entirely separate from statutory assessments but should give a clear sense of how children are progressing, and how they can be supported to progress further. Such information should allow teachers and school leaders to identify which children need additional support or challenge and in which specific areas.
The reports quotes rsearch on the impact of children's background on the level of achievement, including level of education and earnings:

Research therefore supports the fact that judgement of a school’s success or failure on the basis of statutory tests is unjust and unreliable. No intervention should be triggered on the basis of test data alone. Rather, the results from statutory assessments should trigger further discussion leading to a qualitative expert judgement. We should also remember that superficially good test results can be achieved at a high price in terms of curriculum breadth, extra-curricular activity, pupil welfare and school sustainability - none of which are evident in the raw data. Over reliance on data is simply naive and in some instances dangerous.
Last year's changes made many pupils feel that they were failures and the report tackles this head on:

A basic expectation of any assessment system is that it should recognise the progress made by all children. The current interim framework and assessment materials fail to do this. Simplistic, overarching labels such as ‘working below the expected standard’ mean that the progress of too many children is ignored and too many children are effectively labelled as failing and the cumbersome bureaucratic language does not conceal this perception from pupils or their families. This is not only unhelpful to the school but it also sends entirely the wrong message to our pupils, potentially having an impact on their future motivation.
The report calls for a recognition of the unfairness of judging schools on attainment data alone: 
Whilst any form of data from statutory assessment alone should not be used to judge school effectiveness, if such data is to be used as part of the wider picture when holding individual schools to account then the fairest way to do this is by focusing on the progress pupils make. Attainment is important and all teachers want as many children as possible to reach the highest standards. However, when it comes to holding schools to account, it would be grossly unfair to base comparisons on attainment when children’s starting points can be so different.
Nick Brook, Deputy Geberal Secretary of the NAHT concludes by focusing on how the wider picture will impact on any reform of assessment :
Firstly, we must look again at how data from statutory assessment is used to hold
schools to account. Over-reliance on statutory assessment data raises the stakes of testing and ultimately distorts curriculum emphasis and outcomes. Unless we address some of the worst aspects of the current accountability system, including acceptance
of the inherent limitations of data, even the most sensible assessment arrangements will become skewed. Floor and coasting standards cast a shadow of fear over many schools and school leaders. Poor test results can trigger an avalanche of interventions, based on
a presumption of school failure, which are distracting at best and career ending at worst. It is easy to understand why schools in this shadow struggle to recruit teachers and leaders. There needs to be better join-up amongst those that hold schools to account and a more constructive approach to intervention. Most importantly, we need to replace the presumption of failure with an expectation of support.

Secondly, better governance of the assessment system is needed, leading to a stable, proportionate cycle of design, evaluation and implementation for every national assessment. Effective national test design is a complex skill which requires careful thought and substantial evaluation. The scale of national assessments in a system the size of England means that effective implementation of change is a major challenge in itself. Frequent reforms and constant tinkering around the edges can therefore have a negative impact on quality.

Thirdly, assessment for learning is not an intuitive skill possessed by all. There needs to be substantial investment in the training and development of staff in schools if this is to be done universally well. Not all schools or academy chains will have in-house expertise to draw upon and external support will come at a cost. We know that school budgets are already at and beyond breaking point, following real-term cuts since 2010. More resources are required. Additionally, the development of national assessment banks will require investment to ensure the highest quality materials are available to schools. These cost pressures should however be offset by savings achieved by reducing the amount of statutory testing required within these proposals.
The new Secretary of State, Justine Greening, has shown a desire to listen and a willingness to set right mistakes of recent years. With political will and genuine engagement with the profession these challenges are far from insurmountable.


No comments: