Introduction Scottish Teachers for a New Era (STNE) is a collaborative, six-year pilot project focusing on the research and development of a primary teacher education programme at Aberdeen University. In line with the global interest on evidence-based decision, STNE research is expected to generate data to inform developments. This linear model however has proved limited – both in theory and in practice – to grapple with the complexities of teacher education and the multiple interacting components and relationships amongst the stakeholders that combine to create the learning experience for student, beginning and experienced teachers. Recognition of such complexities eschews the idea of certainty and predictability, but focuses rather on understanding and responsiveness. An important feature of such a system is that direct experimentation is not reversible: there is no way to run a second test. Rather, such systems carry history, which influences their evolutionary dynamics, and they are capable of learning (Dupuy, 2004). In this view, a model based on causal relationships and prediction is challenged by a process of search for meaning in a cultural context, whereby culture is understood to be shared, interconnected, patterned and rooted in practices (Das, 2006). STNE research embarked in the piloting of a model of research that would move away from more conventional linear models of evidence-driven decisions, to support internal reflection and evaluation. In this process some methodological challenges were encountered: what shape would the research design, methods and relationships amongst the stakeholders need to take to support and be aligned to the dynamic processes at play?

Methodology Acknowledging that complex systems are multi-layered and multi-scaled and that there are non-linear feedback loops and multiple relationships amongst parts of a system requires a mixed method approach and the development of an integrated methodology, with a shift away from the idea of prediction and control to that of participation and learning (Solè and Goodwin, 2000). This equates to an approach that can assemble meaning from the many perspectives, facets, and relationships, a form of bricolage (Kincheloe & Berry, 2004). Data from a wide range of sources – students’ characteristics on entry and exit, structure and delivery of courses, school experiences, beginning teachers’ experiences, teacher, tutor and student perspectives, pupil data, teacher action research experiences and so on are collected for the purpose of discussion and search for a shared understanding amongst all concerned.

Results and implications Examples will be given of the exercise of dealing with spurious data in triangulation processes for the purpose of reflecting on the challenges of keeping an open-ended and flexible approach to the results which are obtained. A critical reflection is due on the nature of democratic relationships surrounding the project and how far these can be realigned with notions of power equality. In fact, these are the premises to a decision-making process which, by means of actively acknowledging uncertainty, seeks to multiply possibilities and actively searches for the conditions for such diversity to operate