USING DATA BASE MANAGEMENT SYSTEMS IN STATISTICAL DATA PROCESSING

Abstract

National and international statistical bureaus produce ca. 25,000 tables for publication each year, based on hundreds of inter-related object-types with thousands of attributes. It would appear that this environment should be well suited to the application of data base management techniques for the administration of the data. This paper presents a data oriented model of the statistical production process which is used as a basis for a review of the state of experience within statistical offices with commercially available data base management systems. We conclude with a presentation of some important data management facilities which must be enhanced or developed in order to support statistical production processing.