Katrin Rääk ^* Eve Eisenschmidt

• Tallinn University, Tallinn, Estonia

This study examines how five Estonian schools perceive evidence-driven school improvement in a three-year school-university partnership program. In each school, the principal and teachers collaborated with an external mentor. Supported by university experts, the school improvement teams worked on projects aimed at enhancing student learning in their schools and fostering a collaborative, evidence-driven school culture. Data was collected through focus group interviews with the school teams and analyzed using thematic content analysis. The findings reveal that schools view data as connected to accountability and decision-making, with considerably less emphasis on instructional improvement. School organizational and teacher-related factors, together with data overload, hindered systematic data use. Notably, the school improvement program's effectiveness was most evident in the final year, with the sustainability of improvement largely dependent on collaborative routines.school contexts and external pressures to schools (Grützmacher et al., 2023;Kyriakides et al., 2024). Another issue is the short duration of improvement programs that might not be sufficient in reaching set goals or supporting new initiatives in becoming routine actions within the organization (Conley and Enomoto, 2009). To provide schools with enough time to initiate and sustain necessary activities to reach goals, schools in Estonia are required by national policy to adopt evidence-driven approaches in their development plans (Eisenschmidt et al., 2023). These plans, designed for a minimum of three years, must outline how data will be used to monitor school's progress and to make informed decisions on possible improvement practices (Estonian Parliament, 2010). In addition to school-collected data, other data sources are provided by the government, including results from exams or school satisfaction surveys. Data use helps schools reach key goals, such as to enhance instruction, create strong learning communities and to potentially improve student outcomes (Levin, 2010). However, previous research has found that teachers often lack the skills and knowledge about how to use data effectively for school improvement (