The use of an appropriate model is not only important in all forms of statistical analysis but also it is central to any scientific enquiry. In fact, taking a broad view of the word ‘data’ as ‘whatever are given’ by way of figures, maps, signatures, images, and documents and the like to study a phenomenon and of the word ‘model’ to imply a construct that can be used to experiment on the phenomenon and to generate pertinent data to be used as a part of the premises in any exercise for inferencing about the phenomenon, scientists have often raised doubts about which to come first: the model or the data? The question is so basic that it has led to two distinct schools of thought, one arguing that a model be first considered and then used to generate data that can be analysed to throw light on the phenomenon, while in the other, data are first collected and subsequently analysed through an appropriately chosen model to make data-based conclusions valid beyond the observed data. In either case, use of a model remains important, though the acceptance of a model as ‘appropriate’ requires a different outlook in each of the two thoughts.