Can randomised controlled trials revolutionise educational research?

Until recently, very few randomised controlled trails (RCTs) were conducted in educational contexts in the UK. However in the last three years things have changed.

In 2011, the Education Endowment Foundation (EEF) was created with a £125 million grant from the UK government to fund educational innovation to raise the attainment of disadvantaged children in England, and to evaluate these innovations using RCTs. We are one of 15 groups on the EEF evaluator's panel, and are currently evaluating a set of projects around oracy, reading for pleasure and mentoring. The use of RCT based evaluations has now spread beyond the EEF and we are involved in two projects for Department for Education.

RCTs work by randomly allocating children and young people into different groups, some of whom receive an 'intervention' and some of whom don't - then comparing the outcomes to see if the intervention has an impact. RCTs provide strong evidence about whether an intervention actually works to accomplish the aims that it sets out to achieve - something teachers are keen to know. However, RCTs on their own provide limited detail on why an intervention has a positive (or negative) impact, or whether specific aspects of a complex intervention are more (or less) effective than others. Because of this, our RCT evaluation designs incorporate a process evaluation that mixes qualitative and quantitative research approaches to explore issues such as

  • how consistently the intervention is implemented across different schools
  • how much of the intervention the participants actually receive ('dosage')
  • participant/teacher/school engagement with the intervention under evaluation.

Findings from this are then linked with those from the RCT.

RCTs and their accompanying process evaluations are an important part of the researchers' toolkit, and we are committed to their use in educational contexts. We welcome the additional rigour that an RCT methodology can bring to evaluation research. But RCTs are not a panacea. Other research methods and approaches are needed to help uncover how interventions work - or not - and can explore a wide range of complex outcomes. And in educational contexts, the importance of educational values, professional choices, and the complexity of the teaching and learning process cannot be ignored.