The focus of this symposium is to discuss challenges in linking process and product data in order to understand instructional practices in junior secondary classrooms. Based on the experiences from the Norwegian PISA+ we will discuss how analyses of process data can shed light on achievement patterns within the Norwegian PISA test. The PISA+ study was designed for pursue the problematic findings within the Norwegian PISA score. The achievement patterns for Norway for all three domains (reading, mathematics and science) are near the OECD average, and have declined in 2003 (Kjærnslie 2004). The PISA + was designed as a comprehensive video study with a three camera solution (teacher, class, pair of students) and with an ambition to capture offered and experienced learning activities in maths-, science and reading classrooms. By extensive documentation of classroom activities in the three PISA domains we wanted to shed light on how we can understand the general achievement patterns in Norwegian classrooms as well as the pedagogical processes that shape this reality. The main corpus of data is videodata supported with video- and audio based interviews. How can process data (such as videodata) shed light on achievement patterns? Alongside the ’rapid and powerful rise’ of major international quantitative studies (such as PISA and TIMSS) a steady growth of more qualitative comparative approaches has developed (Osborn 2004). We recognize today a new interest in examining common factors in teaching that are critical for success, and which draw on both process data and product data. How to combine these two sources of data in a comprehensive methodological design and framework is however still under dispute. The following four papers present how in depths studies of teaching and learning activities from respectively math, science and reading classrooms, and developed within a common methodological design can expand our understanding of relevant factors of relevance when examining and analyzing teaching and learning in classrooms. Two of the papers draw on analyses from reading classrooms. One paper demonstrates contradictions in instructional practices in science classrooms and the forth paper discuss how individualized teaching methods in math education reduce students' opportunities to learn math. Together the four papers show how studies of instructional practices could profit from appropriate blending of a rich variety of available methodologies, and which can range from complex statistical analyses based on quantitative data at the one side, through intensive interaction analyses based on videodocumentation on the other.

The PISA+ study was designed as a comprehensive video study capturing teaching and learning activities in all three PISA domains. We have video recordings, observations and interviews, and copies of student work from six science, math and reading classrooms. The video design focused on observing sequences of lessons rather than single lessons (and with minimum six lessons in sequence). The six schools/ classrooms were chosen on the basis of demographic and pedagogical criteria linked to variation as the overall sampling strategy. We spent three weeks in at each classroom. In sum we have videorecordings from 136 reading-, math, and science classrooms backed up with student- and teacher interviews. Videodata could be treated as raw data (Klette 2009) and provides for analyses from multiple perspectives and approaches. In our analyses we have combined systematic coding procedures with with more qualitative and interpretative approaches.

The video design gave possibilities for layers of interpretations and where we could combine systematic analyses of instructional activities with analyses of dialogues and interaction that accompanied the same activities. The performed analyses indicate huge variations in the observed teachers' professional competence and subject specific expertise. The use of instructional repertoires, accompanying relevant learning tools and discursive practices are often poorly integrated. The interaction patterns in the observed classrooms gave ample room for student initiatives, and our analyses suggest a rather active and autonomous student role. A close examination of content and subject matter at stake in student initiated utterances and questions suggest however that these initiatives mainly are concerned with practical matters and only to a limited degree are aligned with subject related topics. Individualized teaching methods (i.e. work plans) are frequently used in the observed classrooms. These reduce students' opportunity to learn from their peers and class mates.