Widely differing approaches to writing assessment could skew comparisons between schools, ministers are warned
Headteachers have called on education secretary Nicky Morgan to stop the publication of this year’s Sats data, warning that difficulties with the tests mean comparisons between schools will be “very risky”.
In an open letter to Ms Morgan, the NAHT headteachers’ union argues that the lack of preparation for schools, inadequate time to implement the new curriculum and variations in approaches to assessments caused by “delayed and obscure guidance” could undermine confidence in the data produced this year.
The letter states: “It is not just that the marks may be lower overall, which could be addressed, but that they will vary in unpredictable ways. We know of widely different approaches to writing assessment across the country, for example.
“And the content and sequencing of the reading test meant that lower attaining pupils had little opportunity to show their progress. This may result in a skewed distribution of marks that simply setting a lower threshold may not solve. Comparisons between schools become very risky.”
The NAHT says it is happy for schools to share their results with parents and for the results to be available on the RAISEonline website, which is accessible to Ofsted. But it does not think they should be publically available.
The union also wants the government to allow teachers “sensible flexibility” when assessing children’s writing. At the moment, to be judged as “working at the expected standard”, children must meet all of the government’s 18 criteria, including spelling.
But there has been concern that this process will incorrectly label children as “working below the expected standard”, because they miss out on just one or two items. There is particular concern that no allowance has been made on this rule for dyslexic children.
The letter points out that the top grades at GCSE, A level and degree level are all attainable with a score below 100 per cent, and yet children at the end of Year 2 and Year 6 must get 100 per cent on their assessments to meet the expected standard.
As a solution, the assocation proposes moving from a “secure-fit” approach to judgements in the assessment of writing to a “best-fit”.
Review of assessment
The letter comes after the NAHT yesterday announced it was setting up an independent panel to review primary and key stage 3 assessment, which aims to publish interim recommendations by the end of this term.
The association has also repeated its call for the Department for Education to undertake a fundamental review of primary assessment.
A DfE spokesperson said: “We have reformed the primary curriculum to help ensure all children leave primary school having mastered the basics, and the support and hard work of teachers is key to making this happen.
“We are determined to get this right and remain committed to working with teachers and headteachers as we continue with our primary assessment reform. We will respond to this letter in due course.”
The open letter in full…
Following Sats week, we have collected the feedback of members and urge you to consider some changes to the current and future arrangements for assessment. The experience in a large majority of schools has not been a positive one.
Teachers and headteachers all agree that a thorough review of assessment is necessary. We hope that you will commit to a fundamental review of assessment to avoid further problems next year.
In the meantime, there are two pressing concerns and possible solutions:
- Hold off on publication of any 2016 test data;
- A commitment to changing from “secure-fit” to “best-fit” judgements in the assessment of writing.
Publication of data
Given concerns about both the design and administration of the new assessments, the lack of preparation for schools, the inadequate time to implement the new curriculum for the current cohort, and the variations in approaches between schools resulting from delayed and obscure guidance, it is hard to have confidence in the data produced by this round of assessments.
It is not just that the marks may be lower overall, which could be addressed, but that they will vary in unpredictable ways. We know of widely different approaches to writing assessment across the country, for example. And the content and sequencing of the reading test meant that lower attaining pupils had little opportunity to show their progress. This may result in a skewed distribution of marks that simply setting a lower threshold may not solve. Comparisons between schools become very risky.
School level data should not be externally published under these circumstances. Assessment data should still be available on RAISEonline, which summarises a school’s performance at the end of each Key Stage, and could be shared with parents, but the aggregated school-level scores should not be published externally. We understand that Ofqual is already mandated to conduct a review of this year’s data. In our view, a hold on external publication, until we can be sure what the data is telling us, would be a sensible step. In this interim year, we should be cautious about the data that’s been collected.
Problems have arisen with the new “secure-fit” model; teachers need some sensible flexibility when assessing children’s writing and would be happier with a “best-fit” model. This would give a more accurate reflection of whether or not a child has grasped the overall skills of writing.
Children who are clearly excellent writers will be incorrectly labelled as working below the expected standard this year simply because teachers are not permitted to use their own judgement about their balance of abilities. We are particularly concerned about the impact on the thousands of dyslexic children in school.
There are few other tests or examinations at any other stage of education where a student is judged by “secure fit”. The top grades at GCSE, A level and degree level are all attainable with a score below 100 per cent, and yet only 100 per cent will do if our six and 10-year-olds are to meet the required standard.
A move from “secure fit” to “best fit” would remove some problems. However, it is clear that the interim framework is not working and needs a sustainable long-term replacement.
Serious problems have emerged in the planning and implementation of tests this year, with a negative effect on schools. We believe that the suggestions we have outlined above would go some way towards settling growing disquiet about assessment and demonstrate a clear faith in the profession to deliver the government’s reforms.
Russell Hobby, general secretary
Kim Johnson, president
James Bowen, NAHT Edge director
Amanda Hulme, chair of NAHT’s assessment group