Case Study: A Longitudinal Data Study on the Ages and Stages Questionnaire 3
HDRC and BIHR research highlights inequalities in early development and school readiness: Boys and children living in more deprived areas are less likely to achieve GLD; notable differences by ethnicity.
The longitudinal study used routine data to analyse the association between the Ages and Stages Questionnaire 3 assessment at age 2 and the Early Years Foundation Stage at age 5.
A HDRC and BIHR collaboration
This project aimed to use linked routine data from the Connected Bradford database to better understand early child development and school readiness. Specifically, the study examined whether developmental outcomes measured at age 2 using the Ages and Stages Questionnaire-3 (ASQ-3) were associated with school readiness at age 4 - 5, measured using the Early Years Foundation Stage Profile (EYFSP). It also explored which sociodemographic factors, including sex, ethnicity and deprivation, were associated with poorer outcomes.
ASQ-3 is a developmental assessment used in the Healthy Child Programme in England at around 24 to 30 months. It assesses five domains: communication, gross motor, fine motor, personal-social and problem solving. In this study, ASQ-3 was used to derive a Good Level of Development (GLD) measure based on whether children were below the cut-off in any domain.
EYFSP is a statutory assessment completed at the end of the Reception year, when children are aged 4–5. It provides a measure of school readiness, including whether a child has achieved a Good Level of Development across key early learning areas such as communication and language, physical development, personal, social and emotional development, literacy and mathematics.
The study was carried out collaboratively by Professor Josie Dickerson (BIHR), Dr Kate Mooney (University of York), Dr Yanhua Xu (HDRC), Rob Shore (HDRC) and Hollie Henderson (HDRC, Bradford University).
The power of routine data linkage
We used Connected Bradford’s linked, pseudonymised routine data to bring together information from health and education sources and create a longitudinal view of children’s early development and later school readiness. The work included both data preparation and statistical analysis. Data preparation involved cleaning and quality checking records, standardising variables across sources, and linking area-level deprivation information by mapping children’s Lower Layer Super Output Area (LSOA) to the Index of Multiple Deprivation (IMD).
As part of this process, we created analysis-ready datasets for three groups: children with an ASQ-3assessment, children with an EYFSP assessment, and the subgroup with both measures linked across time. The analysis focused on children with an ASQ-3 completed between 24 and 30 months and, where available, an EYFSP record at the end of Reception.
We created the ASQ-3 Good Level of Development (GLD) measure using the Public Health England (PHE) convention. Children were classified as at risk if they were recorded as “Below Cut Off” in any of the five ASQ-3 domains. Children were classified as having achieved GLD if all five domains were recorded as either “No Risk” or “Monitor.” We also calculated a simple continuous ASQ-3 composite score by summing binary scores across the five domains, and used the total EYFSP score where available for secondary analyses.
Using the cleaned and linked dataset, we carried out multivariable logistic regression analyses to examine the relationship between ASQ-3 at age 2 and EYFSP at age 4–5, while adjusting for sex, ethnicity and deprivation. We also conducted sensitivity analyses using continuous scores and tested whether the strength of the association varied across different sociodemographic groups through interaction models.
Key Findings and Impact
The project produced local evidence on the relationship between developmental outcomes at age 2 and school readiness at age 5.
Among 47,046 children with an ASQ-3 record, 83.3% achieved a Good Level of Development.
Among 9,469 children with an EYFSP record, 67.7% achieved GLD. In the linked sample, children who achieved GLD on ASQ-3 had more than three times the odds of achieving GLD on EYFSP. The study also highlighted inequalities in early development and school readiness. Boys and children living in more deprived areas were less likely to achieve GLD, and there were important differences by ethnicity.
These findings provide useful evidence for early years services, public health and local partners in thinking about where earlier and more targeted support may be needed. More broadly, the work demonstrated the value of linked routine data for answering practical policy questions across service boundaries. It showed how combining health and education data can support local decision-making, improve understanding of inequalities, and strengthen the evidence base for early intervention and school readiness planning.
What's next?
The next step is to use these findings to inform future discussions on how ASQ-3 information collected at age 2 could be used more proactively to identify children who may benefit from earlier support before school entry. This includes considering how linked data can help services better understand which children may be at greater risk of poorer school readiness outcomes.
There is also scope to build on this work by improving completeness of linked records, refining data flows, and extending the analysis to more recent cohorts as additional data becomes available. This would strengthen the local evidence base and support ongoing work on reducing inequalities in child development and school readiness.
The study has been submitted to BMJ's Archives of Disease in Childhood. You can read the preprint.
“Working with the Bradford HDRC and receiving support from their analyst has enabled this vital research to be undertaken. Note these findings will have direct implications for how local authorities across England design their Best Start in Life plans, which are being developed now to meet the government's 2028 school readiness target.
Rather than waiting until children start school to identify those who need additional support, these findings suggest that improving how we deliver, and act upon, ASQ-3 reviews could help thousands of children to succeed in the future.”
Professor Josie Dickerson, BIHR