Differences in exam performance between school types mirrors genotype differences between them

On average, independent and grammar schools out-perform their state-funded, non-selective counterparts in national exams, including the General Certificate of Secondary Education (GCSEs), taken in the UK at age 16. These... [ view full abstract ]

On average, independent and grammar schools out-perform their state-funded, non-selective counterparts in national exams, including the General Certificate of Secondary Education (GCSEs), taken in the UK at age 16. These differences are often attributed to selection factors such as student ability or socio-economic status, however the possible role of DNA differences has not yet been considered. Here, for the first time, we use a UK-representative sample of 5,825 unrelated, genotyped individuals to investigate GCSE and genotype differences between three types of schools in the UK: state-funded schools that are non-selective, state schools that are selective (‘grammar’), and private schools that are selective (‘independent’). We created a polygenic score derived from a 2016 genome-wide association study of years of education and found that GCSE differences between schools were mirrored in the polygenic score differences between schools, with non-selective state schools students showing significantly lower polygenic scores compared with those in grammar schools (d = 0.37) and independent schools (d = 0.35). Furthermore, we find that after controlling for student selection factors (general intelligence, prior ability in English and Maths, and SES), school type explains less than 1% of variance in GCSE scores. These results suggest that GCSE differences between school types primarily reflect genetically driven selection factors between schools, rather than educational value added by the school types.