A Sats test that left children in tears was “unduly hard” for some pupils to understand, according to an official report by the exams watchdog.
Around a quarter of pupils did not finish the reading test, taken by more than 550,000 10- and 11-year-olds in May 2016, Ofqual found.
And its report into the exam concluded that: "On the balance of evidence presented, it seems plausible that the combined impact from multiple ostensibly negligible challenges – stemming from both question and text factors – may have rendered the 2016 reading test unduly hard to access for at least some pupils."
The Sats had been made tougher in 2016 to reflect the new curriculum. The tests included questions designed to “stretch” the highest achieving pupils, who in previous years could have taken a separate test.
The reading test consisted of three passages of text, including a story about two children sneaking away from a party to row across a lake in the grounds. Children had one hour to answer 33 questions.
The report found that standards in the 2016 test were set appropriately – but the test did seem to be more challenging than the sample materials provided.
“As soon as the reading test had been sat, teachers began to express concerns over its accessibility,” the review said.
When the results were published, just two-thirds of pupils had reached the expected standard in reading. Although the new tests are not directly comparable to the previous Sats, the government had said that the new expected standard would be similar to a level 4b – a level that 80 per cent of children had reached in reading in 2015.
The Ofqual review used test data from the Standards and Testing Agency (STA), the government agency that is responsible for the tests. It also looked at independent research, its own investigation into the tests and even comments published on social media.
It found that teachers’ concerns that the hardest questions on both the first and second texts occurred mid-way through the set of questions seemed to be supported by STA test data.
And it heard evidence that the tests may have proved particularly hard for pupils with special educational needs and disabilities (SEND) because all of the texts were continuous passages and challenging vocabulary was used.
Ofqual concluded that there were a number of “important questions” for the STA including whether pupils were given enough time to complete the reading test, if concerns about this could have been flagged up earlier during the development process, and how potential biases against certain groups of pupils – such as those with SEND – could be better identified.
The report on the key stage 2 reading test was published alongside an evaluation of the STA's approach to developing the reading and maths tests.
This report found that the agency's approach was robust and compared favourably to similar tests in other countries. But it did warn that some of the items in the national curriculum, such as the need to “discuss and evaluate how authors use language”, were not amenable to testing – and there were concerns that these requirements might not be taught.
“We are reassured that the STA’s approach to sampling from the national curriculum is robust. We have identified specific questions that we will continue to discuss with the STA to help them enhance the validity of the reading and maths tests, over time, for all pupils,” Michelle Meadows, deputy chief regulator of Ofqual, said.
A Department for Education spokesperson said: “A good primary education lays the foundations for success at secondary school and beyond. This year’s key stage 2 results show that our curriculum reforms are starting to raise standards. It is important to have an assessment system that helps to drive up academic standards.
“That’s why we have reformed the primary assessment system in England to free up teachers to educate and inspire young children, while holding schools to account in a proportionate and effective way.”