Description
Series: Quantitative Applications in the Social Sciences
This monograph is designed to serve as an in-depth introduction to a variation of the basic regression model that utilizes data of a special type-time series. A collection of data X, (r= 1, 2,..., 7) with the interval between X, and Xat being fixed and constant is referred to as a time series. In short, the order of the observations is of extreme importance-we are interested not only in the particular values of the observations, but also in the order in which they appear. For example, series relating to presidential popularity, U.S. defense expenditures, amount of war in the international system, and unemployment meet the requirements of time series analysis.
Given data in a specified temporal ordering, it is possible to raise questions concerning how variables behaved in the past and how they are likely to behave in the future. The great advantage of time series regression analysis is the possibility for both explaining the past and predicting the future behavior of variables of interest. Thus, the history of a time series is called on to do double duty (Nelson, 1973:19): "first, it must inform us about the particular mechanism which describes the evolution through time and second, it allows us to put that mechanism to use in forecasting the future." As can be seen, both of these efforts are predicated upon being able to correctly postulate a model and estimate its parameters. For example, the decision by the U.S. regarding how much to spend for defense is of great concern to the President, Congress, and the public. As a result, understanding how such a decision is made and what the future ramifications of the decision mechanism are likely to be is imperative. In an effort to provide continuity to the technical discussion, we shall return to this example throughout this monograph.
ISBN:803931355863
