The analysis of the GCSE scores of more than 700,000 girls taught in the state sector concludes that those at girls' schools consistently made more progress than those in co-ed secondaries.
The fact that pupils with the lowest test scores when they started secondary made the biggest leap in girls' school will reopen the debate about whether more children should have access to a single-sex education in order to drive up results.
The number of girls schools has dwindled in the state sector since the 1970s and has been dropping more recently among private schools, as more and more parents demand co-ed schools. Only 221,000 girls and 160,000 boys are now taught in state single sex secondaries out of a total school population of more than 3.5 million.
The research, conducted on behalf of the Good Schools Guide, looked at the "contextually value added" scores for every girl who took GCSEs in the state sector between 2005 and 2007. Grammar schools were excluded. The government introduced value added scores to rate the progress pupils make between the end of primary school and GCSEs taking into account their socio-economic backgrounds.
Of the 71,286 girls who sat GCSEs in single-sex schools over the three-year period, on average all did better than predicted on the basis of their end of primary Sats results. By comparison, of the 647,942 who took exams in mixed-sex schools, 20% did worse than expected.
In the value added score, in which a zero score indicates a child achieving the GCSE results expected on the basis of their Sats results at 11, the lowest 10% achievers who went to girls' schools scored on average 17. Among the lowest 10% achievers in mixed schools that score was -10. It means girls who