As rich as the PISA database is, it does not allow by itself the apprehension of such complex phenomena as the ones this project has set out to explore. For this reason, this twofold research design aimed at shedding light on aspects that the statistical analysis was not able to account for. In order to further “unpack” the effects of the above listed governance factors, approximately 60 semi-structured interviews were completed in Chile and Uruguay (Table 5.3). Two types of interviews were carried out: one the one hand, we interviewed teachers and heads of selected schools, and on the other, we interviewed key informants with high levels of expertise in our subject matter. Data were collected using a theoretical purposeful non-random sample. The educational centers were selected following two main criteria: (a) type of center provision, allowing us to grasp different institutional formats and assess their impact on provision, and (b) educational outcomes (measures in achievement). In the case of Uruguay, where information is not readily and publicly available, we resorted to key informants.