The relation between theory and application in statistics
General comments on the relation between theory and application in statistics are made and emphasis placed on issues and principles of model formulation. Three examples are described in outline. Criteria for the choice of models are discussed.