The process of developing and testing a teaching model in teaching statistical concepts at the introductory level is discussed. The model comprised three components, provision of supplementary notes, working in pairs and teaching concepts in worked examples, which were tested using a 2 3 randomised factorial block design with confounding. The development of the teaching model is supported by an interview study of teachers and lecturers in Scotland. The results suggested that the concepts are

more »

... t presented within a concrete example but that the provision of notes is only useful when students work in pairs. The whole study provided information on the experimental paradigms and showed that it is possible to carry out a complex experiment within a teaching environment. The key teaching combinations of the model require further testing in a wider environment.