The Early Learning Positive Deviance Initiative: summary of quantitative findings

Image source: DataDrive2030

DataDrive2030, Percept, and others | September 2023

The use of large datasets in an Early Childhood Development (ECD) setting has, to date, been limited. Using DataDrive2030’s first combined dataset, a Positive Deviance (PD) approach has been adopted to identify factors associated with young children in lower socio-economic circumstances who perform on or above par compared to their higher socio-economic peers. The purpose of this report is to document insights emerging from the quantitative analysis that will be explored in the next phase of the project – a qualitative and ethnographic study of positive deviant facilities.

A myriad of ways was used to define PD, and multiple teams were contracted to investigate the data. In parallel, conversations were engaged with practitioners working at Early Learning Programmes (ELPs) and an open data competition ran to attract fresh, new perspectives. A total of 12,719 children across 1,975 ELPs were included in the initial analysis (10,936 after filtering out children in higher socio-economic bands).

Phase one – the quantitative analyses – identified a range of potential drivers of PD at various levels (child, home, and facility). These include children’s socio-emotional functioning, ELP process quality (such as practitioner-child engagement, class grouping during activities and free play), cognition-related and accessible materials and variables associated with good ELP management and governance. In phase two of this project, we validate/invalidate emerging insights and further explore the mechanisms behind these drivers of deviance, through in-depth qualitative research. The qualitative phase is reported separately.