Description: Statistics often seems to be a daunting subject for students and researchers. Thus, the simplest analyses are often conducted and crucial analyses overlooked. For many reasons, researchers tend to focus on the individual level of analysis, but much of our rich data contains multiple levels. This book addresses these multilevel analyses and their usefulness in applied research.
Purpose: The main purpose is to provide a practical introduction to multilevel analyses for nonexperts in statistics with examples from standard statistical software packages such as SPSS.
Audience: The target audience is mainly students of social sciences, but the book also is appropriate for researchers who are delving into multilevel analyses for the first time. The author has many years of experience as a researcher, analyst, and professor.
Features: Although this book was intended for nonexperts in statistics, it is clear from the beginning that a graduate-level understanding of statistics will be necessary to fully reap the benefits. For those who have a graduate statistics background, the terminology is immediately familiar and should resurrect buried statistical foundations. The book begins with an introduction of what multilevel analysis is and the special circumstances (e.g., nested designs) in which it is used. This is followed by a chapter about nested designs and is furthered by a discussion of contextual variables. The book then progresses to more specific applications of OLS and multilevel analysis. Each chapter begins with an introduction and ends with a summary. In between, the chapters are filled with detailed examples, graphs, tables, and equations. There are also step-by-step instructions for running these analyses in SPSS. The book is arranged to progress through topics with increasing sophistication, so readers will need to devote some time to reading it; those who try to reference a topic of interest in the middle will probably end up more confused than when they started.
Assessment: This is a worthwhile endeavor to make multilevel analyses more accessible to nonexperts in statistics. It certainly is more comprehensible than some jargon-encumbered texts, but still requires a fair amount of statistical savvy and the techniques will not be learned in a few hours of leisurely reading. For readers willing to take the time to carefully read through this book, the rewards should be plentiful, especially as part of a statistics course.