Instructors: Why should you use this book?
How Multilevel Model Foundations differs from other introductory books on multilevel models
Multilevel Model Foundations is different from these other introductory texts in many ways. For more details, see pages 1 - 2 and 7 - 8 in the text. Most importantly:
​
(1) Multilevel Model Foundations proceeds at a slow pace. It takes a lot of time, and pictures, to ensure you get the fundamental ideas behind this approach. It transitions slowly from monolevel to multilevel models.
​
(2) Multilevel Model Foundations relies exclusively on a very small dataset for all the examples. This allows you, the reader, to get close to the data and the model results, and see in granular detail what is happening. With things like empirical Bayes adjustments of neighborhood means, you can literally see what is happening on each of the four sides of the playing board.
​
(3) The volume illustrates how multilevel models can apply to longitudinal data. Not all volumes cover this topic.
(4) It carefully explains some pesky points of confusion not treated in other introductory texts, for example, a) how is it possible to get a negative level-2 R squared, b) what is actually controlled for, and not controlled for, in random intercepts models, and c) how to systematically troubleshoot hidden problems in mixed models.
Meeting students' needs
Starting around 2000, I began teaching graduate students how to multilevel model. All of the students in this course had previously taken a basic course in statistics that included multiple regression. Despite that exposure, some students felt challenged and confused by many multilevel modeling ideas, operations, and results. Needless to say, the instructor of course bore some responsibility for the situation.
​
Over two decades I tried out different well known textbooks that might assist students. I was looking for a textbook that had three things. First, was it clear and correct? Second, was it accessible and user-friendly, or would mildly numbers-phobic students break out in hives upon opening the book? Third, did it show how multilevel analyses could provide clear answers to specific theory, policy, or evaluation questions?
​
I never found a text that met these requirements.
In the last couple of iterations of this second level graduate statistics course, in 2018 and 2020, I started building examples using the Monopoly dataset. I provided students with weekly workbook chapters and programs that relied on the Monopoly dataset. The tiny dataset, and students familiarity with the properties, the game board, and the game, facilitated their readily grasping some key multilevel modeling ideas. Introducing the random coefficients regression model by asking if they thought rent impacts of a hotel varied by neighborhood is a key case in point.
​​
In short, this book should provide your students a user-friendly, easily understandable introduction to fundamental ideas in multilevel modeling, with examples from the Monopoly dataset that make sense to the students right from the get-"GO!"