The need for Pragmatic Experimentation in Educational Research

Abstract

The randomised controlled trial (RCT) is the best method for evaluating changes in educational practice. Nevertheless, it has been criticised because its results often have poor ecological validity. This valid criticism can be addressed by designing the RCT so that it mimics normal teaching practice as closely as possible, while retaining the benefits of random allocation. In this article, we discuss the issues of designing a pragmatic RCT. We also present an example of such a trial in computer-assisted learning in literacy.