Can A/B Testing at Scale Accelerate Learning Outcomes in Low- and Middle-Income Environments?

On current trends the world will fail to reach the objectives set in the UN’s Sustainable Development Goals for Education by 2030 or even within the 21st century. Changing this trend will require a significant acceleration in learning outcomes. Digital personalised learning (DPL) tools are a potentially cost-effective intervention that can contribute to this acceleration. In particular, the continuous experimentation afforded by these tools through software A/B testing, has considerable potential to create compounding improvements in learning outcomes. This paper provides an overview of EIDU, an educational platform combining student focused DPL content with digital structured pedagogy programmes in public pre-primary schools in Kenya. Collection of student’s longitudinal unsupervised assessment data at scale creates the possibility of learning outcome focused A/B testing. This is a novel contribution to the development and research field as up until now this type of capability has largely been confined to students in high-income environments.