Data Input Error Detection and Correction Procedures

摘要

The study is an examination of the input data error problem in computerized information systems. The area of concern is the detection and correction of input data errors resulting from human recording during the initial collection of the data. This research attempts to remove some of the mystery surrounding the input error problem. A system for classifying errors by type is developed; attention is paid to the kinds of errors which can be made or introduced at various levels in the data generation-data processing chain. More important, these levels and their potential use to managers and researchers alike provide a conceptual framework in which intelligent discussions concerning the error process can be formulated. The concept of data worth alone provides a significant step forward in building an intelligent detection and correction process. The study develops a systematic procedure--a model--for evaluating the various detection and correction alternatives. The final evaluation of the detection and correction procedures to be used in the system is based on cost. This is not displacement cost, but cost associated with improved operations through more accurate information. The value of information is the worth of the data, and the worth of the data is the data accuracy problem. The major contributor to data accuracy is formal procedures for input error detection and correction. This study has developed these formal procedures.