Triggered by the mediocre results of German students in the TIMS-Study, a new type of comprehensive model experiments was promoted in Germany that should lead to lasting changes in the educational system. Therefore the models were to be scientifically founded and monitored. Additionally, they were to be implemented in a substantial number of schools from a minimum of ten of the German federal states to gain results that could be generalised for Germany. The first programme later known as SINUS was established in 1998 running up to 2007. It aimed to increase the efficiency of mathematics and science instruction and to foster mathematics and science literacy in a broad sense. The SINUS programme draws on research from various sources (Prenzel, 2000; Prenzel & Ostermeier, 2006). It aims to encourage collaboration between teachers within and between the participating schools (secondary level) and processes of quality assurance and quality development. A total of 180 schools organised into 30 networks of six schools each participated in the first five year phase of the programme. This phase was followed by two transfer phases where the programme was disseminated to numerous other schools (first transfer phase 800, second transfer phase 1800 schools). At the core of the programme are eleven “modules” addressing key problem areas of science and mathematics instruction in Germany (Baumert et al., 1997). They concern certain conditions and methods of teaching and learning which can be successfully changed and indicate possible directions for quality development. The modules can be combined or restricted to certain aspects; in any case, they have to be adapted to the particular subject matters and problems at hand. The teachers collaborate jointly on the development of new ideas and materials as well as the evaluation and documentation of results. Collaboration is offered both within schools and within small networks of schools in which every school is embedded. The teacher groups and school networks are supported by coordinators, materials explaining the modules, consultations with researchers from education and science education and in-service training.

The SINUS programme was evaluated in a systematic way. Different approaches were applied, serving different functions (i.e. formative and summative) at different levels (teachers, students, schools). One approach focussed on the coordination of the programme and the provision of support. Surveys were conducted to find out to what extent the teachers identify themselves with the aims of the programme and accept the provided support. A summative evaluation was conducted to reveal the success and effectiveness of the programme. This evaluation used the PISA assessments in the years 2000 and 2003 and additional instruments developed by the German national PISA consortium. The first data collection served as “baseline”, investigating the schools “starting conditions” compared to a representative national reference group. The data collection in 2003 served to estimate developments in student performance and interest, teacher professionalism, and processes of quality development.

All applied approaches to evaluate the SINUS programme deliver evidence for the expected impact on teachers, students and schools. The teacher surveys for instance show that coordination of the school networks plays a crucial role in the enrolment of teacher groups in the programme and the experienced development (Ostermeier, 2004). The summative evaluation drawing on PISA instruments shows for instance positive effects of SINUS on students’ attitudes towards science and mathematics, and on students’ competencies compared to the representative sample (Ostermeier, Carstensen, Prenzel, & Geiser, 2004; Prenzel, Carstensen, Senkbeil, Ostermeier, & Seidel, 2005). Until now the model of SINUS has been transferred to numerous schools in Germany. The need for educational innovation and change in Europe (Rocard et al., 2007) made it move across boundaries: In the EU funded project “Mind the Gap” the question is examined of how this model can be exported to other countries.