An application of forecasting techniques for a small geographical area

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Larry Samuel Spainhour (Creator)
The University of North Carolina at Greensboro (UNCG )
Web Site:
Donald Jud

Abstract: It was the purpose of this thesis to formulate and test short-run economic forecasting methodologies that are useful for small geographical areas. The major concern was to derive an accurate monthly revenue forecast for the Chesapeake and Potomac Telephone Company of Washington, D.C. A survey was made of the literature dealing with techniques used to forecast telephone demand. This review suggested methodologies that were appropriate given the special problems of the Washington area. A narrative analysis of current economic trends affecting telephone demand in Washington further refined the development of a proper forecasting methodology. Next, an analysis of the available time series data was presented. This analysis provided an understanding of the underlying characteristics of the data and led to the formulation of specific forecasting models for testing. Five separate empirical models were developed for forecasting telephone demand in Washington, D.C. These models were analyzed for their statistical significance and tested for their ability to produce accurate forecasts. Monthly forecasts were generated with each model, and a final forecast was selected.

Additional Information

Language: English
Date: 1977
Economic forecasting
Telephone companies $x Econometric models

Email this document to