3 On Forecasting Techniques In Socio-Economic Environment

ABSTRACT. Several forecasting methods for a target (or response) variable Y is considered. Of these, even simple explanatory-variable-based forecast, let alone multiple explanatory-variable-based causal-chain forecasting models, should provide superior, to most “expert ” judgment and various extrapolation-based, time series analyzing methods of Box-Jenkins type and other “black box ” related forecasts. Steps in the development of the relevant theoretical apparatus for judging the quality of data base which, in turn, affects the qualities of individual forecasting methods are described in the text to prove the point. All the investigated forecasting methodologies have been tested for accuracy and compared with actual data for the forecast period. The data series chosen for the experiment is the annual series “Total International Passengers ” at the Jeddah King Abdulaziz International Airport (KAIA) for the years 1975-1987 of which 1975-1981 is considered “historical ” data and, based on this interval, forecasts were made for the year