Tougher GCSEs widen gap between poorer and better-off pupils

From the Centre of Excellence


The Guardian has published an article that claims the introduction of new, tougher GCSE exams in England has led to a widening gap in the results achieved by disadvantaged pupils compared with their better-off peers, according to official figures.

The Department for Education’s (DfE) analysis of last summer’s GCSE exams reveals that the gap between disadvantaged pupils and others at secondary school grew by 0.6 percentage points, after two years in which the gap narrowed.

While disadvantaged pupils showed an improvement in the proportion gaining a grade 5 – previously a C – in the compulsory subjects of English and maths, others made more rapid progress, causing the gap to widen. Just over half of all other pupils gained a 5 or above in English and maths but fewer than one in four disadvantaged pupils did the same.

300 schools picked out in GCSE 'off-rolling' investigation
The DfE was quick to argue the data showed that disadvantaged pupils studying in multi-academy trusts were doing better than the national average, vindicating the government’s flagship policy.

“It’s been clear for some time that standards are rising in our schools and today’s data underlines the role academies and free schools are playing in that improvement, with progress above the national average and impressive outcomes for disadvantaged pupils,” said Nick Gibb, the minister for school standards in England.

However, the overall results for multi-academy trusts showed below-average progress compared with all state-funded mainstream schools. Some 40% of trusts – which operate more than one academy – had progress measures that were below or well bellow the national average, while 29% were similar to the average.

Based on progress, the two best state schools in England were Tauheedul Islam Girls High School in Blackburn and Wembley High Technology College in Brent, while the highest-achieving school for attainment was Henrietta Barnett, a selective girls’ school in north London.

The DfE’s analysis showed 346 schools were below the government’s floor standards in terms of results, compared with 309 last year.

The new figures come as school leaders advise parents not to give too much attention to the government’s preferred measure, used in its performance tables, which attempts to gauge the progress made by pupils from the end of primary school to key stage four, when they sit GCSEs.

“Performance tables can never tell the full story of a school and we urge parents and governors not to place too much weight on them,” said Geoff Barton, the general secretary of the Association of School and College Leaders.

Barton said the DfE’s performance tables were “long past their sell-by date,” and called for reforms to the way school statistics are presented.

“The secondary school performance tables are inherently flawed in that the headline measure of Progress 8, which is used to judge the performance of schools, effectively penalises schools which have a high proportion of disadvantaged children,” Barton said.

“The effect of this is to stigmatise these schools, making it more difficult to recruit headteachers and teachers and demoralising pupils, parents and communities.”

'Food-deprived, pulling all-nighters': A-levels during Ramadan
Data analysed by the FFT Education Datalab research unit, and a working paper using previous GCSE exam results published by the University of Bristol, suggest that schools with diverse intakes, especially those with a high proportion of pupils from disadvantaged backgrounds, are unfairly penalised by the government’s measure.

The Bristol paper, which looked at the progress scores of 3,000 secondary schools, concluded: “It seems clear from our results that the higher the proportion of disadvantaged pupils in a school, the more it will effectively be punished for the national underperformance of these pupil groups.”

The researchers have called for the government to also publish a revised progress measure that adjusts for pupil background.