150,000 Sats tests re-marked after fiasco

Thousands of schools requested that more than 150,000 tests papers be re-marked following last year's Sats fiasco, official data showed today.

Record numbers of schools sent papers back, with a high proportion requesting re-marks for English tests.

One teacher's leader said the figures were "appalling" and said they showed that schools had no confidence in the quality of marking last year.

The statistics, published by the Qualifications and Curriculum Authority, show that 4,628 primary schools asked for 25,142 English papers to be re-marked. This is almost six times as many schools as in 2007, when 784 schools requested 1,456 reviews.

And 261 primary schools requested a re-mark for a whole group of pupils.

At secondary school level, 1,001 schools returned English tests for 11,217 pupils, while 684 asked for a group review.

In 2007, just 412 schools requested 3,346 individual English papers be reviewed.

Mick Brookes, general secretary of the National Association of Head Teachers (NAHT), said: "I think the figures are appalling.

"For every paper that's sent back, that's an increased workload for people who don't believe in the system at all, but are determined that the children will have a fair assessment of their work at the end of each key stage."

A separate report published by the QCA today has concluded, based on a small sample of papers marked as part of a quality assurance process, that there is "no particular cause for concern" about the marking of last year's tests.

Last summer the Sats results of more than a million schoolchildren were delayed after a series of administrative failures by the QCA's contractor ETS Europe.

An independent inquiry into the fiasco, led by Lord Sutherland, heaped blame on the QCA and ETS.

The figures show that 1,234 primary schools sent back maths papers for 1,719 pupils. This is almost four times as many schools as in 2007, when 332 schools requested 373 reviews.

For primary school science, 1,643 schools sent 2,977 papers back - three times as many as in 2007, when 560 schools returned 777 tests.

Four schools requested group re-marks in maths and four asked for group reviews in science.

In secondary schools, where Sats tests were taken by 14-year-olds, 2,221 maths papers were returned by 764 schools. In 2007, 248 schools requested 384 reviews.

And in science, 1,255 schools asked for 7,704 papers to be reviewed, compared with 590 schools returning 1,605 scripts in 2007.

In total, 150,980 papers were returned by schools last year.

More English papers were returned than in the other two subjects because they are more difficult, and take longer to mark.

Last summer there were longer delays in returning English papers to schools than in maths and science.

In total, individual reviews were requested for 4.4 per cent of primary school English papers and 2 per cent of secondary school English papers.

Mr Brookes said: "Teachers don't believe the marking is good.

"This is only the tip of the iceberg. Teachers don't tend to send back papers if they don't appear to cross a threshold boundary into another level."

He added that, because of the "high-stakes nature" of the testing, schools were unlikely to send back papers they believed had over-estimated a child's ability.

Mr Brookes again called on ministers to scrap Sats tests for 11-year-olds, following Schools Secretary Ed Balls's decision to abolish them for 14-year-olds last October.

The QCA said that part of the reason the number of reviews would be higher this year is because the practice of borderlining has been abandoned.

Under this practice, any paper which fell two or three marks below a grade boundary would have automatically been re-marked.

QCA acting chief executive Andrew Hall said: "QCA is taking a lead in investigating and reporting on issues around the quality of marking and is committed to making continuous improvements.

"Further improvements can be made to marking quality through the introduction of on-screen marking, which allows for more frequent and robust assurance checks to be carried out during the marking process and for some scripts to be double-marked.

"The move towards on-screen marking was one of the recommendations of Lord Sutherland's report into national curriculum tests and QCA fully supports this view."

Christine Blower, acting general secretary of the National Union of Teachers, said: "It is obvious from the QCA reports on marking quality that the QCA has had to deal with an impossible set of circumstances. We can see from the research that there is an issue over the marking of the English tests at Key Stage 3 and nothing to suggest that there are not similar concerns over Key Stages 1 and 2.

"The fact is that last summer's crisis was not just about maladministration but about the nature of marking quality itself.

"The results of this research, combined with the fact that the QCA now says that it cannot guarantee the delivery of the test results for 2009 on time, simply underlines the utter unreasonableness of Ed Balls' determination to continue with an unreliable, invalid and damaging testing system for primary schools."

A spokeswoman for the Department for Children, Schools and Families admitted that it was likely that the increase in requests for reviews was due, "in large part", to last year's failures.

She said: "Given delays over the summer, it is no surprise that some schools had concerns about their results. The contractor who was delivering the test has since had its contract terminated.

"Whilst it is an important part of the quality assurance process that schools are able to request reviews if they have concerns about marking, Ofqual advise that the quality of marking in 2008 was at least as good as in previous years.

"The overall proportion of review applications is very small when compared with the number of pupils taking the test."

Around 1.24 million children in approximately 20,600 schools took Sats tests in English, maths and science last year.

Kathleen Tattersall, chair of exams watchdog Ofqual, said: "Ofqual has now come to a final judgment about the quality of marking of national curriculum tests in 2008.

"Based on our extensive monitoring, all the evidence presented - including the reports published by QCA today - and the views expressed to us, we have concluded that the quality of marking in 2008 was at least as good as in previous years.

"A number of measures were introduced last year to try to improve marking quality. The impact of these measures is unclear, but it is important that marking quality initiatives continue to be used.

"While there was a rise in the number of review requests in 2008, Ofqual's own research shows that test review information alone is an unreliable barometer of test marking quality."

Liberal Democrat schools spokesman David Laws said the "shocking figures" confirmed concerns about the quality of marking.

He said: "Although ministers have said that they believe marking quality has not suffered, the statistics don't seem to bear out this complacency.

"Not only have the number of appeals quadrupled, but the number of successful appeals has also significantly increased.

"It is unacceptable to have national testing which has such high stakes if it also has such low quality marking and if schools cannot be confident about the results."

Chris Keates, general secretary of the NASUWT teaching union, said: "Given the staggering level of incompetence displayed by the QCA and the ETS contractors in administering the Sats, it is no surprise that schools' faith in the marking of the tests was shaken.

"However, the teachers and markers did a thoroughly professional job and it is a shame that they were let down by the mismanagement of the responsible exam bodies."

The QCA's research report into the quality of marking reveals that in 2006 and 2007 almost half of English writing exams were wrongly marked.

The study looked again at the results of a small sample of tests taken by 14-year-olds in those years to see if the levels awarded were accurate.

It shows that around 44 per cent of levels awarded in English writing were wrong in 2006 and in 2007.

More than a third of levels awarded in English reading were wrong.

Maths marks were found to be broadly accurate, and in science around one in six (13 per cent) was wrongly marked.

Tests for 11-year-olds were not included.