Skip to main content
University of Michigan Press
Fulcrum logo

You can access this title through a library that has purchased it. More information about purchasing is available at our website.

Share the story of what Open Access means to you

a graphic of a lock that is open, the universal logo for open access

University of Michigan needs your feedback to better understand how readers are using openly available ebooks. You can help by taking a short, privacy-friendly survey.

  1. Home
  2. Modeling and Interpreting Interactive Hypotheses in Regression Analysis

Modeling and Interpreting Interactive Hypotheses in Regression Analysis

Cindy D. Kam and Robert J. Franzese, Jr. 2007
Restricted You do not have access to this book. How to get access.
Social scientists study complex phenomena about which they often propose intricate hypotheses tested with linear-interactive or multiplicative terms. While interaction terms are hardly new to social science research, researchers have yet to develop a common methodology for using and interpreting them. Modeling and Interpreting Interactive Hypotheses in Regression Analysis provides step-by-step guidance on how to connect substantive theories to statistical models and how to interpret and present the results.

"Kam and Franzese is a must-have for all empirical social scientists interested in teasing out the complexities of their data."

---Janet M. Box-Steffensmeier, Ohio State University

"Kam and Franzese have written what will become the definitive source on dealing with interaction terms and testing interactive hypotheses. It will serve as the standard reference for political scientists and will be one of those books that everyone will turn to when helping our students or doing our work. But more than that, this book is the best text I have seen for getting students to really think about the importance of careful specification and testing of their hypotheses."

---David A. M. Peterson, Texas A&M University

"Kam and Franzese have given scholars and teachers of regression models something they've needed for years: a clear, concise guide to understanding multiplicative interactions. Motivated by real substantive examples and packed with valuable examples and graphs, their book belongs on the shelf of every working social scientist."

---Christopher Zorn, University of South Carolina

"Kam and Franzese make it easy to model what good researchers have known for a long time: many important and interesting causal effects depend on the presence of other conditions. Their book shows how to explore interactive hypotheses in your own research and how to present your results. The book is straightforward yet technically sophisticated. There are no more excuses for misunderstanding, misrepresenting, or simply missing out on interaction effects!"

---Andrew Gould, University of Notre Dame

Cindy D. Kam is Assistant Professor, Department of Political Science, University of California, Davis.

Robert J. Franzese Jr. is Associate Professor, Department of Political Science, University of Michigan, and Research Associate Professor, Center for Political Studies, Institute for Social Research, University of Michigan.

For datasets, syntax, and worksheets to help readers work through the examples covered in the book, visit: www.press.umich.edu/KamFranzese/Interactions.html

Read Book Buy Book
ISBN(s)
  • 978-0-472-06969-9 (paper)
  • 978-0-472-02299-1 (ebook)
  • 978-0-472-09969-6 (hardcover)
Subject
  • Political Science:Political Methodology
  • Economics
  • Psychology
  • Sociology
Citable Link
  • Table of Contents

  • Stats

  • Cover
  • Title
  • Copyright
  • Dedication
  • Preface
  • Contents
  • 1. Introduction
  • 2. Interactions in Social Science
  • 3. Theory to Practice
    • Specifying Empirical Models to Reflect Interactive Hypotheses
    • Interpreting Coefficients from Interactive Models
    • Linking Statistical Tests with Interactive Hypotheses
    • Presentation of Interactive Effects
  • 4. The Meaning, Use, and Abuse of Some Common General-Practice Rules
    • Colinearity and Mean-Centering the Components of Interaction Terms
    • Including x and z when xz Appears
  • 5. Extensions
    • Separate-Sample versus Pooled-Sample Estimation of Interactive Effects
    • Nonlinear Models
    • Random-Effects Models and Hierarchical Models
  • 6. Summary
  • Appendix A. Differentiation Rules
  • Appendix B. Stata Syntax
    • Marginal Effects, Standard Errors, and Confidence Intervals
    • Predicted Values, Standard Errors, and Confidence Intervals
    • Marginal Effects, Using “lincom”
  • References
  • Index
72 views since November 02, 2018
University of Michigan Press logo

University of Michigan Press

Powered by Fulcrum logo

  • About
  • Blog
  • Feedback
  • Contact
  • Contribute
  • Accessibility
  • Preservation
  • Privacy
  • Terms of Service
  • Log In
© University of Michigan Press 2020
x This site requires cookies to function correctly.