Helping School Leaders See Variation and Build the Capability to Improve Complex Systems

Mary-Louise Leger ^* Louis M Gomez

• University of California, Los Angeles, Los Angeles, United States

This study investigates how educational leaders within a research-practice partnership used improvement science thinking and tools to address declining attendance rates following the COVID-19 pandemic. Utilizing time-series data and run charts, central office leaders and school administrators visualized system performance, allowing them to see variations and patterns over time. The partnership's focus on variation helped leaders challenge their initial biases, leading to a more accurate understanding of the factors driving attendance improvement. Cypress Elementary, identified as a positive outlier, demonstrated that small-scale, context-specific interventions, including personalized outreach and incentives, played a crucial role in enhancing attendance. The study also highlights the importance of building leadership capability to interpret data effectively and implement sustainable, locally relevant solutions. Through this process, leaders developed greater proficiency in diagnosing system-level issues and enacting systemic improvements, contributing to the broader literature on data-driven decision-making, sensemaking, and educational equity. This research demonstrates how improvement science can support leaders in understanding data more clearly and building their capacity to lead system-wide improvements.