Research strongly suggests that low socioeconomic position [SEP] in childhood is related to poorer mental health and behavioural outcomes, particularly in childhood and adolescence. Evidence suggests that low SEP is related to internalising and externalising symptoms, antisocial behaviour, and substance use behaviours. Interestingly, recent studies suggest that the relationship between childhood SEP and these emotional and behavioural outcomes are changing across generations.

Yet, these relationships vary considerably when different SEP indicators are examined. Traditional SEP indicators include income, education, occupational social class, and/or housing tenure; however, SEP is a multi-dimensional construct that is related to an individual's access to material, educational, and social resources. SEP indicators are limited in their ability to capture the multi-dimensionality of SEP when used in isolation, but using multiple SEP indicators can cause issues for statistical analysis if used improperly. Researchers have demonstrated that regression-based models may be inappropriate for investigating the effect of SEP on an outcome. Therefore, alternative approaches are needed and person-centred approaches, such as latent class analysis, are favoured in the literature.

Additionally, cross-cohort comparisons of emotional and behavioural outcomes necessitates retrospective harmonisation for the scales used. Retrospective harmonisation allows researchers to compare different scale measures that measure the same underlying construct in different cohorts, for example. Without conducting retrospective harmonisation, researchers cannot be assured that the findings observed not due to systematic differences between cohorts. This project will utilise both statistical processes to analyse differences in emotional and behavioural outcomes across three British birth cohorts.