Meta-Analysis: A Structural Equation Modeling Approach

Preface

“If all you have is a hammer, everything looks like a nail.”

—Maslow's hammer

Purpose of This Book

There were two purposes of writing this book. One was personal and the other was more “formal.” I will give the personal one first. The primary motivation for writing this book was to document my own journey in learning structural equation modeling (SEM) and meta-analysis. The journey began when I was a undergraduate student. I first learned SEM from Wai Chan, my former supervisor. After learning a bit from the giants in SEM, such as Karl Jöreskog, Peter Bentler, Bengt Muthén, Kenneth Bollen, Michael Browne, Michael Neale, and Roderick McDonald, among others, I found SEM fascinating. It seems that SEM is the statistical framework for all data analysis. Nearly all statistical techniques I learned can be formulated as structural equation models.

In my graduate study, I came across a different technique—meta-analysis. I learned meta-analysis by reading the classic book by Larry Hedges and Ingram Olkin. I was impressed that a simple yet elegant statistical model could be used to synthesize findings across studies. It seems that meta-analysis is the key to advance knowledge by combining results from different studies. As I was trained with the SEM background, everything looks like a structural equation model to me. I asked the question, “could a meta-analysis be a structural equation model?” This book summarized my journey to answer this question in the past one and a half ...