Multilevel Modeling is a concise, practical guide to building models for multilevel and longitudinal data. Author Douglas A. Luke begins by providing a rationale for multilevel models; outlines the basic approach to estimating and evaluating a two-level model; discusses the major extensions to mixed-effects models; and provides advice for where to go for instruction in more advanced techniques. Rich with examples, the Second Edition expands coverage of longitudinal methods, diagnostic procedures, models of counts (Poisson), power analysis, cross-classified models, and adds a new section added on presenting modeling results. A website for the book includes the data and the statistical code (both R and Stata) used for all of the presented analyses.

Assessing a Multilevel Model

There is no need to ask the question “Is the model true?” If “truth” is to be the “whole truth” the answer must be “No.” The only question of interest is “Is the model illuminating and useful?”

—George Box, 1978

In the previous chapter, we saw how to build and estimate a mixed-effects multilevel model. Once you have a fitted model, it is important to look at the model closely to determine if it is working as intended and assess how well the model is explaining the data. It is also usually important to move beyond the simple parameter estimates of the model, and use the full model to examine the important patterns implied by that model. For example, you might want ...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles