ABSTRACT
Early developments in econometrics were mainly concerned with the problems of estimation and inference from a sample of economic data. The main objective of such estimation and inference was to obtain consistent and efficient estimates of the parameters and to construct better confidence intervals and hypothesis testing procedures. In an early stage of econometric development, models were typically formulated considering highly parsimonious relationships (according to the present-day economic theories), then estimated and tested statistically for model adequacy. Additional terms in the models were added to those models found inadequate and the process kept repeating. This procedure is often known as being specific to general methodology. Later, in contrast to this idea, the use of Hendry methodology (called general to specific methodology) became popular. It involves the formulation of a general model and then sequentially testing the model for various parameter restrictions until a desirable model is chosen. Recently, the attention of the subject has been diverted to the problems of diagnostic checking, specification testing, semiparametric estimation and model selection.
