Spectral Analysis of Time-Series Data

Rebecca M. Warner

Hardcover
Hardcover
May 22, 1998
ISBN 9781572303386
Price: $65.00
225 Pages
Size: 6" x 9"
order
bookProfessors: request an exam copy

This book provides a thorough introduction to methods for detecting and describing cyclic patterns in time-series data. It is written both for researchers and students new to the area and for those who have already collected time-series data but wish to learn new ways of understanding and presenting them. Facilitating the interpretation of observations of behavior, physiology, mood, perceptual threshold, social indicator variables, and other responses, the book focuses on practical applications and requires much less mathematical background than most comparable texts. Using real data sets and currently available software (SPSS for Windows), the author employs extensive examples to clarify key concepts. Topics covered include research design issues, preliminary data screening, identification and description of cycles, summary of results across time series, and assessment of relations between time series. Also considered are theoretical questions, problems of interpretation, and potential sources of artifact.

This title is part of the Methodology in the Social Sciences Series, edited by Todd D. Little, PhD.


“This is an excellent book for behavioral and social scientists seeking a quick but thorough introduction to spectral analysis. Rigorous in presenting basic equations, it also features practical examples that facilitate the learning process. The author's clear exposition and use of commonly accessible software to illustrate analyses will help readers make the leap from reading this book to actually analyzing their own time-series data.”

—Randy J. Larsen, PhD, Department of Psychology, University of Michigan


“A wonderful book, filled with clear language and interesting examples. Warner helps us understand how many of the enduring features of life are repetitive ones that cannot be described in terms of means and static relationships. A common-sense guide to cyclical patterns in time-series data, the volume is both practical and intellectually stimulating.”

—James M. Dabbs, PhD, Department of Psychology, Georgia State University

Table of Contents

Contents

1. Research Questions for Time-Series and Spectral Analysis Studies

2. Issues in Time-Series Research Design, Data Collection, and Data Entry: Getting Started

3. Preliminary Examination of Time-Series Data

4. Harmonic Analysis

5. Periodogram Analysis

6. Spectral Analysis

7. Summary of Issues for Univariate Time-Series Data

8. Assessing Relationships between Two Time Series

9. Cross-Spectral Analysis

10. Applications of Bivariate Time-Series and Cross-Spectral Analyses

11. Pitfalls for the Unwary: Examples of Common Sources of Artifact

12. Theoretical Issues

Appendix A. Raw Time-Series Data

Appendix B. Critical Values for the Fisher Test of Significance for Periodogram Analysis


About the Author

Rebecca M. Warner, PhD, is Professor of Psychology at the University of New Hampshire. Her research interests include communication style, cardiovascular reactivity and modulation of physiological rhythms in social interactions, and coordination of talk patterns in conversation.

Audience

Researchers and graduate students in a range of behavioral and social science disciplines, including developmental, clinical, physiological, and social psychology; sociology; and communication.

Serves as a text in advanced/special topics statistics courses and in seminars that deal with analysis of dyadic social interaction data or other research problems involving interrelated time series.

Course Use

Serves as a text in advanced/special topics statistics courses and in seminars that deal with analysis of dyadic social interaction data or other research problems involving interrelated time series.