This 4 page paper considers the statement that the Box Jenkins methodology of forecasting is not an explanatory method but a simple statistical method. The bibliography cites 10 sources.
Name of Research Paper File: TS14_TEboxjen.rtf
Unformatted Sample Text from the Research Paper:
model, described by some as a statistician dream (Pankratz, 1993). However, it has been argued that the model may not undertake all we may wish and that Box Jenkins is
not an explanatory method but a simple statistical method. To consider if this statement is true, and this is only a statistical tool the method itself need to be examined.
In looking at this model it has a wide range of uses, from science to business. Logically there must be a value to this tool. This model is a
type of ARMA or ARIMA, meaning auto regressive integrated moving average (Brewley, 2003). This starts to indicate there is more than simply statistical analysis, as their is an element of
regressive analysis included in the model (Brewley, 2003). This model makes use of the wide range of ARMA models that are available,
and then, when the model is chosen subjects the model to a range of diagnostic tests (Vandaele, 1983, Hill, 1975). The tests will either be passed in which case
the chosen model was the appropriate model. Or where models do not pass the test a new model will be suggested and then the diagnostic tests will be carried out
again. This may occur repeatedly until a suitable model is found (Brewley, 2003, Hill, 1975). Therefore, there is a system of
identifying the process to be used, estimation and then the diagnostic testing of the model (Pankratz, 1983). Therefore, at first glance it appears that this model is based
on statistical data, as statistics are analysis, however, there is also the use of regression. However the regression is not conducted in