Generalizing the Regression Model
Techniques for Longitudinal and Contextual Analysis
- Blair Wheaton - University of Toronto, Canada
- Marisa Young - McMaster University, Canada
This comprehensive text introduces regression, the general linear model, structural equation modeling, the hierarchical linear model, growth curve models, panel data, and event history models, and includes discussion of published implementations of each technique showing how it was used to address substantive and interesting research questions. It takes a step-by-step approach in the presentation of each topic, using mathematical derivations where necessary, but primarily emphasizing how the methods involved can be implemented, are used in addressing representative substantive problems than span a number of disciplines, and can be interpreted in words. The book demonstrates the analyses in STATA and SAS. Generalizing the Regression Model provides students with a bridge from the classroom to actual research practice and application.
A website for the book at https://edge.sagepub.com/wheaton1e (coming soon!) includes resources for instructors.
Quantitative analyses are so often relegated to OLS techniques when they should not be. The authors more than adequately demonstrate the why, what, and how other procedures (GMM, SEM, panel regression, event history analysis to name a few) are far superior to the OLS approaches widely but inappropriately found in published research or used in practice. Kudos to them.
Generalizing the Regression Model is a highly accessible textbook that covers a remarkable array of complex material with ease. Its applications and examples make the material intuitive and interesting for students to learn.
This is an excellent textbook, but more appropriate for a different course design. We will consider it for FTEC 6319 in the future, but before then, we will see how "Data Analysis for Business, Economics, and Policy" works out.