Socioeconomic inequalities in vocabulary and their implications for educational attainment and mental health in adolescence

This thesis examines the extent to which childhood vocabulary can be used to predict adolescent mental health and educational outcomes, and to which differences in early vocabulary are associated with socioeconomic circumstances (SEC). These questions arise out of an increased interest in the relation between SEC and vocabulary, and efforts to develop interventions to promote vocabulary prior to school entry, which have so far been met with mixed success. Since vocabulary is thought to predict later mental health and education outcomes, and there are social inequalities in language, education and mental health, this thesis sought to provide insight into whether early vocabulary interventions are likely to benefit children in the longer term, through secondary data analysis of two large, nationally representative UK datasets: 1) The 1970 British Cohort Study (BCS1970), and 2) The Millennium Cohort Study (MCS2001). I first examined whether early child vocabulary is related to adolescent mental health, finding small effects, which importantly differed as a function of reporter: when adolescent self-report was considered, better childhood vocabulary skill predicted poorer adolescent mental health outcomes. I next looked more closely at the relation between SEC and vocabulary, by investigating socioeconomic inequalities in vocabulary throughout childhood and into adolescence. I found multiple SEC indicators (most notably parent education, income and occupation) uniquely predicted child and adolescent vocabulary. Inequalities persist from ages 3 to 14 years, with effects being most pronounced at the start and end of formal schooling. I finally investigated whether vocabulary at school entry predicted educational outcomes at the end of secondary school. Here, unlike for mental health, there was a clear relation: better childhood vocabulary predicted better educational outcomes on GCSE or equivalent examinations. This effect was substantially moderated by SEC. Thus, not all children benefit from strong early vocabulary skills in the same way. Overall, the findings of this thesis suggest that good vocabulary skill, as measured by standardised tests, is important for educational attainment but not internalising mental health. Effective interventions are potentially well placed to improve educational outcomes. However, moderation effects suggest such early interventions alone may not suffice, since the educational benefits of good early vocabulary do not appear stable over SEC strata. In sum, although early language interventions are well placed to improve educational outcomes, in order to improve wider functioning in adolescence, we also need to directly target internalising mental health and structural inequalities