ABSTRACT
For any research design, but perhaps especially for a trial, it is important to collect more than the presumed outcome data. In-depth and contextual data can help explain why an intervention does or does not work, how to improve it, or which sub-groups of learners it is most appropriate for. This chapter presents the findings of a large number of attempted robust evaluations in education. Each one involved a process evaluation, and in each study these were very similar. Process evaluations were generally light-touch and conducted in co-operation with the developers of each intervention. Schools agreed to be part of the evaluation when agreeing to be part of the intervention. A substantial part of the evaluation fieldwork was to assess how closely schools adhered to the intended intervention, and what the short term or intermediate impacts were. For each trial, one should look at prior evidence relevant to the intervention, and at how the topic had previously been investigated.
