This practical guide on conducting power analyses using IBM SPSS was written for students and researchers with limited quantitative backgrounds. Readers will appreciate the coverage of topics that are not well described in competing books such as estimating effect sizes, power analyses for complex designs, detailed coverage of popular multiple regression and multi-factor ANOVA approaches, and power for multiple comparisons and simple effects. Practical issues such as how to increase power without increasing sample size, how to report findings, how to derive effect size expectations, and how to support null hypotheses, are also addressed. Unlike other texts, this book focuses on the statistical and methodological aspects of the analyses.
Performing analyses using software applications rather than via complex hand calculations is demonstrated throughout. Ready-to-use IBM SPSS syntax for conducting analyses are included to perform calculations and power analyses at http://www.psypress.com/applied-power-analysis . Detailed annotations for each syntax protocol review the minor modifications necessary for researchers to adapt the syntax to their own analyses. As such, the text reviews both power analysis techniques and provides tools for conducting analyses. Numerous examples enhance accessibility by demonstrating specific issues that must be addressed at all stages of the power analysis and providing detailed interpretations of IBM SPSS output. Several examples address techniques for estimation of power and hand calculations as well. Chapter summaries and key statistics sections also aid in understanding the material.
Chapter 1 reviews significance testing and introduces power. Chapters 2 through 9 cover power analysis strategies for a variety of common designs. Precision analysis for confidence intervals around mean difference, correlations, and effect sizes is the focus of chapter 10. The book concludes with a review of how to report power analyses, a review of freeware and commercial software for power analyses, and how to increase power without increasing sample size. Chapters focusing on simpler analyses such as t-tests present detailed formulae and calculation examples. Chapters focusing on more complex topics such as mixed model ANOVA/MANOVA present primarily computer-based analyses.
Intended as a supplementary text for graduate-level research methods, experimental design, quasi-experimental methods, psychometrics, statistics, and/or advanced/multivariate statistics taught in the behavioral, social, biological, and medical sciences, researchers in these fields also appreciate this book’s practical emphasis. A prerequisite of introductory statistics is recommended.
"This book presents concepts in a more accessible manner than the other books out there. … The step-by-step explanations should make it accessible to a wide range of readers, even advanced undergraduates…the inclusion of SPSS syntax…makes the material such that more advanced readers are still interested and engaged…I would consider using this book for a third course in statistics…I would also consider purchasing a copy for my own use." – Allen I. Huffcutt, Bradley University, USA
" The book provides users with the means to compute power accurately for many situations where no other methods are readily available… The SPSS syntax provides a framework that allows the user to see a range of possible outcomes—information that can help the user gain a better feel for the costs and benefits of various sample sizes…[it] provides methods for dealing with complex data with greater accuracy… appropriate for a short course called statistical power…[or] as a supplement to any multivariate course." – Dale Berger, Claremont Graduate University, USA
"An important addition to every applied worker’s tool chest…This book … allows the author to make a very important contribution to the science of research methodology…this would be a nice complement to our ANOVA/ANOCOVA course, MANOVA/MANCOVA course." - Shlomo Sawilowsky, Wayne State University, USA
"This book will meet a significant need in the market. I would consider adopting it and recommending it to colleagues. In fact, I would expect that it would become a required reading in our statistics sequence here at Claremont. I would definitely purchase a copy…" - Stewart Donaldson, Claremont Graduate School, USA
"The book will be highly relevant to upper-division undergraduate courses and (particularly) graduate-level courses in psychology and related fields. I would likely include the book as a recommended source for my graduate-level course in statistics. In addition to my teaching, the book would also be of use to me professionally." - P. Wesley Schultz, California State University - San Marcos, USA
1. What is Power? Why is Power Important? 2. Chi-Square and Tests for Proportions. 3. Independent Samples and Paired t-tests. 4. Correlations and Differences between Correlations. 5. Between Subjects ANOVA (One Factor, Two or more Factors). 6. Within Subjects Designs. 7. Mixed Model ANOVA and Multivariate ANOVA. 8. Multiple Regression. 9. Covariate Analyses and Regression Interactions. 10. Precision Analysis for Confidence Intervals. 11. Additional Issues and Resources.