The report about an empirical study in self-managing schools focuses mainly on the relationship between team work and teacher collaboration on one side and perception of teacher efficacy and teaching patterns on the other side. The study is embedded in a large accompanying research project, which evaluates the program for widening the autonomy of primary and secondary schools in the German federal state of North Rhine-Westphalia. The theoretical framework regards to the approach of learning organizations. The first assumption of our analyses is that organizational learning in schools is primarily based on the intensity of teacher collaboration in teams as professional learning communities (PLC) (c.f. Rosenholtz, 1991; Seashore Louis & Kruse, 1995; Leithwood, 2000; Stoll & Seashore Louise, 2007) . The second assumption is that teacher collaboration in PLC-teams influences teacher efficacy as an aspect of teacher effectiveness and also shows impact on teaching development as well as on quality of teaching in classrooms and classroom management of teachers (Bonsen & Rolff, 2005; Holtappels, 2007). We distinguish between two forms of teacher collaboration teams: a) classroom teams and year (grade) teams with focus on pedagogical work in the classroom and cooperation to create convenient learning climate as well as differential teaching patterns and learning support for students, b) subject teams with focus on curriculum, subject oriented didactic and diagnosis for learning process. The examinations in this presentation will indicate forms and intensity of teacher cooperation and also identify and compare different kinds and levels of team work as characteristics of PLC. Further analysis draw focus on relationships between levels of professional learning communities (PLC) and intensity of teacher collaboration and the perception of teacher efficacy and teaching patterns.

The analyses are based on a longitudinal data set over three time points (2003, 2005, and 2007) which is taken from the research evaluation of the development of self-managing schools. The data sets consists information of headmasters (=246), steering committees (n=1146), teachers (n=1824) and students (n=9407 in n = 80 classrooms). We create a merged longitudinal data file from several files containing the student level data, the teacher data and the school data. The data collection includes reliable and valid scales to measure teacher efficacy (Schmitz, 1999), the quality of instruction and classroom management (Gruehn, 2000) as well as the teacher cooperation (Holtappels, 1997). In addition to descriptive findings of the data, our fore cited assumption of the impact of teachers’ collaboration will be modelled by means of confirmatory factor analysis as well as path analysis regarding the hierarchal and longitudinal structure of the data. For the distinction and a typology of the different kinds and levels of team work and teacher collaboration we will use latent class analysis.

Following the theoretical approaches of Bandura (1977, 1998, 1998a) we expect an influence of the teacher collaboration on teacher efficacy as well on teaching patterns and quality of teaching and classroom management. In particularly we assume that institutionalized forms of the teacher collaboration (PLC) show higher impact on the observed variables teacher efficacy and teaching quality and teaching settings and classroom management. What kind of the two PLC teams is the most influencing form of collaboration needs to be ascertained.