Description

Providing a clear explanation of the fundamental theory of time series analysis and forecasting, this book couples theory with applications of two popular statistical packages--SAS and SPSS. The text examines moving average, exponential smoothing, Census X-11 deseasonalization, ARIMA, intervention, transfer function, and autoregressive error models and has brief discussions of ARCH and GARCH models. The book features treatments of forecast improvement with regression and autoregression combination models and model and forecast evaluation, along with a sample size analysis for common time series models to attain adequate statistical power. To enhance the book's value as a teaching tool, the data sets and programs used in the book are made available on the Academic Press Web site. The careful linkage of the theoretical constructs with the practical considerations involved in utilizing the statistical packages makes it easy for the user to properly apply these techniques.

Key Features

@introbul:Key Features @bul:* Describes principal approaches to time series analysis and forecasting * Presents examples from public opinion research, policy analysis, political science, economics, and sociology * Free Web site contains the data used in most chapters, facilitating learning * Math level pitched to general social science usage * Glossary makes the material accessible for readers at all levels

Readership

Upper level undergraduate and graduate students, professors, and researchers studying: time series analysis and forecasting; longitudinal quantitative analysis; and quantitative policy analysis. Students, professors and researchers in the social sciences, business, management, operations research, engineering, and applied mathematics.

Table of Contents

Preface. Introduction and Overview: Purpose. Time Series. Missing Data. Sample Size. Representativeness. Scope of Application. Stochastic and Deterministic Processes. Stationarity. Methodological Approaches. Importance. Notation. Extrapolative and Decomposition Models: Introduction. Goodness-of-Fit Indicators. Average Techniques. Exponential Smoothing. Decomposition Methods. New Features of Census X-12. Introduction of Box-Jenkins Time Series Analysis: Introduction. The importance of Time Series Analysis Modeling. Limitations. Assumptions. Time Series. Tests for Nonstationarity. Stabilizing the Variance. Structural or Regime Stability. Strict Stationarity. Implications of Stationarity. The Basic ARIMA Model: Introduction to ARIMA. Graphical Analysis of Time Series Data. Basic Formulation of the Autoregressive Integrated Moving Average Model. The Sample Autocorrelation Function. The Standard Error of the ACF. The Bounds of Stationarity and Invertibility. The Sample Partial Autocorrelation Function. Bounds of Stationarity and Invertibility Reviewed. Other Sample Autocorrelation Funcations. Tentative Identification of Characteristic Patterns of Integrated, Autoregressive, Moving Average, and ARMA Processes. Seasonal ARIMA Models: Cyclicity. Seasonal Nonstationarity. Seasonal Differencing. Multiplica

Details

No. of pages:
528
Language:
English
Copyright:
© 2000
Published:
Imprint:
Academic Press
eBook ISBN:
9780080478708
Print ISBN:
9780127678702
Print ISBN:
9781493302185

About the authors

Robert Yaffee

Robert A. Yaffee, Ph.D., is a Senior Research Consultant/Statistican in the Statistics and Social Science Group of New York University's Academic Computing Facility as well as a Research Scientist/Statistician at the State University of New York Health Science Center in Brooklyn's Division of Geriatric Psychiatry. He received his Ph.D. in political science from Graduate Faculty of Political and Social Research of The New School for Social Research. He serves as a member of the editorial board of the Journal of Gambling Behavior and was on the Research Faculty of Columbia University's School of Public Health before coming to NYU. He also taught in the Statistical packages in the Computer Science Department and the Empirical Research and Advanced Statistics in the Sociology Department of Hunter College. He has published in the fields of statistics, medical research, and psychology.

Affiliations and Expertise

New York University, New York, U.S.A.

Monnie McGee

Monnie McGee, Ph.D. is an Assistant Professor of Mathematics and Statistics at Hunter College. She received her Ph.D. from Rice University and has worked as a bio-statistical consultant for The Rockefeller University and as a computational statistician for Electricité de France.

Affiliations and Expertise

Hunter College, City University of New York

Reviews

@qu:"Robert Yaffee has performed an invaluable service to students of time series analysis by preparing an introduction to methods for analyzing time series data that includes examples drawn from the social sciences, and demonstrates how to program the procedures in SPSS and SAS. Introduction to Time Series Analysis and Forecasting will be a standard reference for years to come." @source:--DAVID F. GREENBERG, New York University, New York