Forecasting via the Box-Jenkins method
Journal of the Academy of Marketing Science · 1978 · Volume 6 · Issue 3
Abstract
In numerous marketing studies, there is a need for adequate and appropriate forecasting methods. The Box-Jenkins model-building technique for discrete, correlated data points is described for both univariate and multivariate input-output models, yielding minimum mean square error forecasts. The procedure is applicable where at least fifty observations occurring at equally spaced time intervals are available.
Keywords
Autocorrelation, Autocorrelation Function, Impulse Response Function, Advertising Expenditure, Managerial Forecast
Enable JavaScript for citation tools and additional paper interactions.