Linking local authority data to understand and respond to the school attendance challenge

ObjectivesPupil absence rates in Newcastle are among the highest in the country. Teachers, local authority attendance officers and community partners face significant challenges. While these professionals possess tacit knowledge about the reasons behind pupil absenteeism, supporting data is lacking. MethodsWe linked data from the school census, child in need census and attendance reports for pupils in years 1-11 in Newcastle from 2020/21 to 2022/23. Using multi-level negative binomial regression models, we examined how factors like Free School Meal eligibility, Special Educational Needs, Child in Need status, and socioeconomic conditions, influence likelihood of school absence. We estimated the variation in pupil absence across Newcastle attributable to school-level factors and identified outlier schools with the highest and lowest population-adjusted absence rates for subsequent qualitative research. This research was conducted by researchers within <Redacted>, facilitating effective collaboration with local stakeholders. ResultsResults will be available at the ADR UK Conference 2025. Administrative data will be descriptively analysed and graphically displayed to characterise the school attendance challenge in Newcastle and identify trends over recent years. We will present adjusted associations between pupil factors and absence, as well as estimates of the school environment’s influence on pupil absence. The limitations of our analysis will be discussed, along with how the research was enriched through focus groups, interviews and surveys with children, families and relevant professionals. ConclusionMissing school can adversely affect mental wellbeing and educational attainment, which can impact employment opportunities and health in the long term. By identifying those at higher risk, we aim to support local stakeholder discussions and inform the development of early interventions to improve outcomes for children and their families.