School leaders are central to this development research. The central research problem is the need to strengthen the quality of school leadership as a means of improving education quality in challenging school contexts. This exploratory multi-country research adopted a realistic evaluation framework to explore the use of a specific continuous improvement approach—improvement science (IS)—with groups of school leaders in Chile, Kenya and the Philippines. The multidisciplinary research team, comprising international researchers from education, leadership, and international development, collaborated with in-country partners to undertake the work in the highly fluid and dynamic environments of the pandemic.

In each country, the in-country partner cooperated with local authorities to select a small group of school leaders to invite to participate in the project. These school leaders and other local education actors formed an improvement community under the guidance of the in-country partner. The improvement communities identified a local issue—a problem of practice, which they would like to improve. These problems were initially related to a project focus on improving the pedagogic use of ICTs during the prolonged school closures. Members of the community analyzed the system (school and context) to understand how local conditions contributed to the problem, and developed and tested their hypothesis through Plan, Do, Study, Act (PDSA) cycles. The ambition was for the school leaders to develop a problem aim statement and measurements for success, thereby creating local ownership of the solutions to identified issues and collective responsibility for testing them and spreading successes through the improvement community.

The project offers a contribution to knowledge on how an amended IS methodology might be successfully utilized by school leaders working in under-resourced contexts to support changes in institutional practices, attitudes, and relationships. It offers a comparative analysis across different geographies and systems of the conditions that support or hinder the use of the IS methodology, draws attention to the importance of partnerships (through improvement communities) and documents promising emerging changes when conditions are favorable. Many of these findings echo those of high-income contexts but are complemented by previously unreported insights on the importance of contextual factors, in spite of the relatively small scale of the research.

From a development perspective, the use of IS methodology in education appears to offer the potential to shift towards more equitable dialogue between education partners, moving away from the idea of problem-solving through implementation of external “what works” solutions in schools, and towards a legitimizing of the use of small changes or local adaptations that respond to highly specific local conditions and capabilities. It moves discussion from “this isn’t being implemented properly” to “what works for my institution.” The sharing of these local improvements with peers and other actors starts to point towards ways in which school leader collaboration could be harnessed for improvements in the quality of local education provision, in particular pedagogic change.