Repository logo

Testing the specificity question - do environmental factors have broad or specific associations with key developmental outcomes?



Change log


Bignardi, Giacomo 


This PhD thesis investigates how adversity impacts a broad range of child development outcomes – ranging from depression, drug use, and cognitive development. The first study is a validation study for a series of novel cognitive assessments created for the Cambridge-based RED longitudinal study, which are utilized in later chapters (adapted from Bignardi et al. 2020). The second study investigates whether ratings of anxiety and depression symptoms have changed relative to before the COVID-19 lockdown, in a sample of 9-11 year-olds (adapted from Bignardi et al. 2020). The final two studies investigate how adversity impacts child development through the lens of specificity theories. These theories propose that specific aspects of the environment (e.g. exposure to a richer linguistic environment) have specific consequences for children’s development (e.g. better language skills). The third study investigates whether socioeconomic status, a widely used marker of potential adversity, is associated with specific cognitive skills (e.g. language skills, executive functions). SEM models are used to test if SES predicts specific skills after controlling for general cognitive ability (GCA). This study utilizes data from the three large-scale cohort studies, including the RED study first introduced in the thesis. The final study takes a data-driven approach, to test whether different risk factors measured in infancy are associated with specific developmental outcomes in adolescence, using longitudinal data from the UK Millennium Cohort Study. Across both studies, we found evidence of strong, broad effects of SES and other risk factors across most outcomes in addition to more minor, specific associations with outcomes. Whilst SES is broadly associated with cognition and behaviour, we found in chapter 3 that vocabulary skills are particularly susceptible, even when controlling for GCA. In the final chapter, we found that SES and other risk factors have broad associations with outcomes, a different set of risk factors better predicts adolescent drug use and self-rated wellbeing.





Astle, Duncan
Dalmaijer, Edwin


development, psychology


Doctor of Philosophy (PhD)

Awarding Institution

University of Cambridge