Interaction Effects in Linear and Generalized Linear Models

Robert Kaufman (PhD University of Wisconsin, 1981) is professor of sociology and the Chair of the Department of Sociology at Temple University. His substantive research focuses on economic structure and labor market inequality, especially with respect to race, ethnicity, and gender. He has also explored other realms of race-ethnic inequality, including research on wealth, home equity, residential segregation, traffic stops and treatment by police, and media portrayals of crime. More abstract statistical issues motivate some of his current work on evaluating different methods for correcting for heteroskedasticity using Monte Carlo simulations. Dr. Kaufman has published papers on quantitative methods in American Sociological Review, American Journal of Sociology, Sociological Methodology, Sociological Methods and Research, and Social Science Quarterly. He served on the editorial board of Sociological Methods and Research for 15 years and has taught graduate-level statistics courses nearly every year for the past 30 years.

"This book is remarkable in its accessible treatment of interaction effects. Although this concept can be challenging for students (even those with some background in statistics), this book presents the material in a very accessible manner, with plenty of examples to help the reader understand how to interpret their results."

• Series Editor’s IntroductionPrefaceAcknowledgmentsAbout the Author1. Introduction and BackgroundOverview: Why Should You Read This Book?The Logic of Interaction Effects in Linear Regression ModelsThe Logic of Interaction Effects in GLMsDiagnostic Testing and Consequences of Model MisspecificationRoadmap for the Rest of the BookChapter 1 NotesPART I. PRINCIPLES2. Basics of Interpreting the Focal Variable’s Effect in the Modeling ComponentMathematical (Geometric) Foundation for GFIGFI Basics: Algebraic Regrouping, Point Estimates, and Sign ChangesPlotting EffectsSummarySpecial TopicsChapter 2 Notes3. The Varying Significance of the Focal Variable’s EffectTest Statistics and Significance LevelsJN Mathematically Derived Significance RegionEmpirically Defined Significance RegionConfidence Bounds and Error Bar PlotsSummary and RecommendationsChapter 3 Notes4. Linear (Identity Link) Models: Using the Predicted Outcome for InterpretationOptions for Display and Reference ValuesReference Values for the Other Predictors (Z)Constructing Tables of Predicted Outcome ValuesCharts and Plots of the Expected Value of the OutcomeConclusionSpecial TopicsChapter 4 Notes5. Nonidentity Link Functions: Challenges of Interpreting Interactions in Nonlinear ModelsIdentifying the IssuesMathematically Defining the Confounded Sources of NonlinearityRevisiting Options for Display and Reference ValuesSolutionsSummary and RecommendationsDerivations and CalculationsChapter 5 NotesPART II. APPLICATIONS6. ICALC Toolkit: Syntax, Options, and ExamplesOverviewINTSPEC: Syntax and OptionsGFI Tool: Syntax and OptionsSIGREG Tool: Syntax and OptionsEFFDISP Tool: Syntax and OptionsOUTDISP Tool: Syntax and OptionsNext StepsChapter 6 Notes7. Linear Regression Model ApplicationsOverviewSingle-Moderator ExampleTwo-Moderator ExampleSpecial TopicsChapter 7 Notes8. Logistic Regression and Probit ApplicationsOverviewOne-Moderator Example (Nominal by Nominal)Three-Way Interaction Example (Interval by Interval by Nominal)Special TopicsChapter 8 Notes9. Multinomial Logistic Regression ApplicationsOverviewOne-Moderator Example (Interval by Interval)Two-Moderator Example (Interval by Two Nominal)Special TopicsChapter 9 Notes10. Ordinal Regression ModelsOverviewOne-Moderator Example (Interval by Nominal)Two-Moderator Interaction Example (Nominal by Two Interval)Special TopicsChapter 10 Notes11. Count ModelsOverviewOne-Moderator Example (Interval by Nominal)Three-Way Interaction Example (Interval by Interval by Nominal)Special TopicsChapter 11 Notes12. Extensions and Final ThoughtsExtensionsFinal Thoughts: Dos, Don’ts, and CautionsChapter 12 NotesAppendix: Data for ExamplesChapter 2: One-Moderator ExampleChapter 2: Two-Moderator Mixed ExampleChapter 2: Two-Moderator Interval ExampleChapter 2: Three-Way Interaction ExampleChapter 3: One-Moderator ExampleChapter 3: Two-Moderator ExampleChapter 3: Three-Way Interaction ExampleChapter 4: Tables One-Moderator Example and Figures Example 3Chapter 4: Tables Two-Moderator ExampleChapter 4: Figures Examples 1 and 2Chapter 4: Figures Example 4Chapter 4: Tables Three-Way Interaction Example and Figures Example 5Chapter 5: Examples 1 and 2Chapter 5: Example 3Chapter 5: Example 4Chapter 6: One-Moderator ExampleChapter 6: Two-Moderator ExampleChapter 6: Three-Way Interaction ExampleChapter 7: One-Moderator ExampleChapter 7: Two-Moderator ExampleChapter 8: One-Moderator ExampleChapter 8: Three-Way Interaction ExampleChapter 9: One-Moderator ExampleChapter 9: Two-Moderator ExampleChapter 10: One-Moderator ExampleChapter 10: Two-Moderator ExampleChapter 11: One-Moderator ExampleChapter 11: Three-Way Interaction ExampleChapter 12: Polynomial ExampleChapter 12: Heckman ExampleChapter 12: Survival Analysis ExampleReferencesData SourcesIndex