Written in a friendly, conversational style, this book offers a hands-on approach to statistical mediation and moderation for both beginning researchers and those familiar with modeling. Starting with a gentle review of regression-based analysis, Paul Jose covers basic mediation and moderation techniques before moving on to advanced topics in multilevel modeling, structural equation modeling, and hybrid combinations, such as moderated mediation. User-friendly features include numerous graphs and carefully worked-through examples; "Helpful Suggestions" about procedures and pitfalls; "Knowledge Boxes" delving into special topics, such as dummy coding; and end-of-chapter exercises and problems with answers . The companion website provides downloadable sample data sets that are used in the book to demonstrate particular analytic strategies, and explains how researchers and students can execute analyses using Jose's online programs, MedGraph and ModGraph. Appendices present SPSS, AMOS, and Mplus syntax for conducting the key types of analyses. "The examples are great and are used effectively. In the basic model chapters, readers can learn from these examples about how to actually fit these models, how to read and interpret the output, and, most important, the right language to report the findings. The book's emphasis on using graphs to visualize model results will make it easier for novices to grasp subtle concepts and avoid pitfalls. The author's Medgraph and Modgraph are great tools. I would consider using this book as a supplement in an advanced statistics class. Its conversational style will complement the more formal presentation of most textbooks."--Maria M. Wong, PhD, Department of Psychology, Idaho State University "The first volume to focus on mediation and moderation, this book explains their differences from historical, substantive, and modeling perspectives. It is abundant with 'how-to' examples and provides readers with modeling tools that can be used directly in research. This is a suitable supplemental text for intermediate or advanced courses dealing with regression, hierarchical linear modeling, or structural equation modeling. The inclusion of cutting-edge topics such as mediated moderation and moderated mediation will attract more advanced quantitative researchers as well."--Feifei Ye, PhD, Department of Psychology in Education, University of Pittsburgh |