U.S. flag

An official website of the United States government, Department of Justice.

Applied Time Series Analysis for the Social Sciences

NCJ Number
R McCleary; R A Hay; E E Meidinger; D McDowall
Date Published
325 pages
This textbook, intended for use in a graduate seminar in time series analysis, explains the practical aspects of time series analysis, with emphasis on applications of interest to economists, political scientists, psychologists, and sociologists.
The book assumes no training in statistics beyond intermediate statistical methods. Following an overview of time series analysis, the basic concepts of univariate Box-Jenkins time series analysis are presented. The univariate ARIMA (Autoregressive Integrated Moving Average) model is used as the baseline building block for impact assessment, forecasting, and causal modeling. The third chapter presents a general impact assessment model for the analysis of an interrupted time series quasi-experiment. This model has been widely used to assess the effects on social systems of planned and unplanned interventions. In the fourth chapter, the use of ARIMA models for forecasting future values of a time series is explained. The fifth chapter extends the Box-Jenkins approach to multivariate time series analysis. A final chapter focuses on the use of computer software and numerical routines to conduct Box-Jenkins time series analysis. The likelihood function is derived, and the solution procedure is illustrated. Related topics that may affect the analysis are discussed. The book reviews several available software packages for the analysis of time series data and the use of interactive software. In each chapter, practical examples are used to illustrate the principles presented. Tables, figures, footnotes, an index, a bibliography listing 72 references, and appendixes listing symbols and conventions and providing data sets, are included. (Author abstract modified)