The review also examines the impact of parenting programmes, practitioner development, and programme resources in promoting early learning. Adapted from Early Learning Programme Outcome Study Baseline Report by Andrew Dawes, Linda Biersteker, Elizabeth Girdwood, Matthew Snelling, and Jessica Horler (2018).

What becomes possible when nationally representative early learning data is released for public use?

 

On 25 March, DataDrive2030 released the Thrive by Five Index 2024 dataset, hosted via DataFirst at the University of Cape Town, alongside a six-paper working paper series and webinar:

Working with the Thrive by Five Index 2024: Explorations of Early Learning Systems in South Africa

 

Built on nationally representative data on 5,001 four-year-olds across 1,388 early learning programmes, the dataset provides a detailed view of how children’s development is shaped by household conditions, programme characteristics, climate exposure, and structural context.

Ahead of public release, DataDrive2030 invited a small group of researchers to work with the data in depth. Their sustained engagement forms the foundation of the working paper series – offering concrete examples of how the dataset can be interrogated rigorously.

 

Why This Matters Now

These papers provide timely evidence to inform national debates on climate vulnerability, programme quality, enrolment gaps, structural inequality, resilience under adversity, and subsidy design in early childhood development.

 

HOW TO READ THESE FINDINGS

As this working paper series shows, analysing nationally representative early learning data can sometimes produce findings that appear surprising – or even contradictory – at first glance.
 
This is partly because children’s development is shaped by multiple, overlapping influences, including household conditions, caregiver characteristics, program quality, and broader structural context. When these factors are analysed together using multivariate models, some variables may no longer appear as independently significant. This is not because they do not matter, but because their effects are closely connected to other factors in the system.
 
It is also important to note that the Thrive by Five Index 2024 dataset is cross-sectional, capturing a snapshot in time. This allows researchers to identify patterns and associations, but not to determine cause and effect.
 
Different analytical approaches may highlight different aspects of the data. Taken together, these perspectives strengthen the understanding of how early learning systems function, rather than pointing to a single defining factor.
 
The key takeaway: Children’s outcomes are shaped by multiple parts of their environment working together. These papers should be read as contributions to a broader evidence base – helping to identify patterns, refine questions, and inform how programs and policies can be strengthened in context.
 

The Working Papers

1. Early Learning Under Climate Stress: Extreme Rainfall and Child Development in South Africa

This paper examines whether exposure to extreme rainfall over the preceding 18 months is associated with differences in child development outcomes.

Findings indicate significant negative associations between rainfall exposure and emergent literacy as well as cognition/executive function, with no consistent effects observed for numeracy or social-emotional functioning. Effects are domain-specific rather than uniform across all measures.

 

2. Determinants of School Readiness in South African Preschoolers: A Multifaceted Analysis of Numeracy and Literacy Skills

This study compares the relative contribution of child skills, caregiver characteristics, and preschool programme measures in predicting early numeracy and literacy.

Results indicate that child-level skills remain the strongest predictors in multivariate models, with household conditions also consistently associated. Preschool quality indicators show more limited and domain-specific associations.

 

3. Cognitive Gains from Early Learning Programme Enrolment in Three South African Provinces

This paper examines associations between enrolment in early learning programmes and developmental outcomes.

Non-enrolled children score, on average, approximately 5.8 ELOM points lower than comparable enrolled peers – equivalent to roughly five to six months of developmental difference. Associations are strongest in cognition/executive function and emergent literacy domains.

However, enrolment effects are conditional on instructional quality. Children only benefit from early learning programmes when instructional quality meets at least a basic standard. When instructional practices are weak, enrolled children perform no better than comparable children who are not enrolled.

Gains are amplified in programmes demonstrating at least basic instructional quality and in cases of sustained exposure of two or more years. The findings are associational rather than causal.

 

4. The Relationship Between Risks to Caregiver Mental Health and Child Learning Outcomes

This study assesses whether proxy caregiver mental health risk indicators remain predictive once household socioeconomic context is included.

Household asset levels show strong associations with child outcomes. Proxy caregiver risk indicators do not retain independent significance once socioeconomic conditions are accounted for.

 

5. On-Track in Early Childhood: Machine Learning Prediction of Early Learning Status in the Context of Socioeconomic and Developmental Risk

This paper applies machine learning approaches to examine whether patterns of school readiness can be predicted among children facing socioeconomic adversity.

Using individual, household, and early learning programme variables, the analysis explores which combinations of factors are associated with developmental resilience. The study demonstrates how advanced modelling techniques can be applied to nationally representative early childhood data.

 

6. The Role of Government Subsidisation in the Early Learning Sector in South Africa: An Analysis Using Thrive by Five 2024

This paper links Thrive by Five Index 2024 data with administrative records to examine whether differential receipt of the ECD subsidy is associated with variation in developmental outcomes.

The analysis contributes to ongoing debates about financing, targeting, and equity within South Africa’s early learning system.

 

 

Access the Dataset & Launch Webinar

 

⌦ Access the dataset via DataFirst

⌦ Watch the launch webinar