The aim of this paper is to model evidence-informed design based on a selective critical analysis of research articles. The authors draw upon findings from an investigation into practitioners' use of educational technologies to synthesise and model what informs their designs. They found that practitioners' designs were often driven by implicit assumptions about learning. These shaped both the design of interventions and the methods sought to derive evaluations and interpret the findings. The authors argue that interventions need to be grounded in better and explicit conceptualisations of what constitutes learning in order to have well-informed designs that focus on improving the quality of student learning.