ON METHODS AND PURPOSES IN ECONOMETRIC MODEL BUILDING.
Abstract
The paper is concerned with the empirical foundations of the natural and social sciences, namely, the formulation and testing of theories based on logically objective analyses of empirical observations. In many instances, both in the experimental and non-experimental sciences, this involves the use of logical empiricism in a stochastic context. The manner in which a scientist begins to construct a model is described, and some relevant definitions are put forward. Illustrations are drawn from several fields, and in particular economics, in which the empirical observations are analyzed as a sample realization of a stochastic process. In this fashion the concepts of stochastic economics are introduced, together with an integrated system of definitions to cover the basic assumptions, the economic theory relevant under these assumptions, and the model into which both these are incorportated. The logical, empirical and operational characteristics of economic theories are next considered, and certain basic requirements are put forward for simultaneous advancement of the generality and validity of these theories, in order to increase their explanatory power. The paper concludes with a short analysis of the practical utility of models for purposes of explanation, prediction, and decision making.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1967
- Accession Number
- AD0652572
Entities
People
- Camilo Dagum
Organizations
- Princeton University