Several international large-scale assessments such as TIMSS (Trends in International Mathematics and Science Study) were set up in the past few decades to compare students' educational achievement across countries in one or more subjects. Research found that countries are varying in their level of achievement over time, sometimes leading to alarming signals about the cognitive performances of students in reading literacy, mathematics or science. However, differences between schools regarding their students' level of achievement within a country are manifest. This raises the question as to how these differences can be explained and how schools can anticipate declining cognitive performances of their students. While several studies have indicated that class and teacher level variables such as instructional quality have an impact on student achievement, little to no attention has been paid to the role of processes at a school level. Schools are indeed granted a high extent of autonomy, and based on seminal school effectiveness literature it is suggested that a school's capacity to make own school policies has an impact on the achievement levels of their students. Up until now, TIMSS has never included information about school level processes such as a school's policy making capacity in the data collection. This project aims to go beyond the traditional reporting in TIMSS, and make an extension. The central objective is to study the relation between the policy making capacity of schools and the cognitive performance of students by running explanatory analysis through multilevel structural equation modelling. By doing so, this project goes beyond the classic reporting about students' performance in TIMSS and expand our understanding of differences between schools in their students' achievement level.