ABSTRACT
This chapter considers the role of randomised trials in education as a necessary for drawing conclusions about effective educational practice. Using a toolbox metaphor, it identifies what kinds of questions are appropriate for different Randomised controlled trial designs in terms of supporting causal inference as part of a wider set of tools for educational inquiry. The chapter argues that access to and engagement with 'scientific' knowledge is an essential condition for the democratic participation of teachers in making judgements about educational practice. It argues that the role of evidence is more crucial for practice here than for policy. Impact evaluation in education usually means assessing the effects of policies and initiatives or other approaches to bring about intentional change in terms of valued educational outcomes for learners. One of the challenges in identifying and understanding learning classrooms is the complexity involved in the interactions between learners and the teacher in relation to intended educational aims and outcomes.
