Multilevel Modeling

Douglas A. Luke is Professor and Director of the Center for Public Health Systems Science at the Brown School at Washington University in St. Louis. Dr. Luke is a leading researcher in the areas of public health policy, imple- mentation science, and systems science. Most of the work that Dr. Luke di- rects at the Center focuses on the evaluation, dissemination, and implemen- tation of evidence-based public health policies. During the past decade, Dr. Luke has worked on applying systems science methods to important public health problems, especially social network analysis. He has published two systems science review papers in the Annual Review of Public Health, and the first study to employ new statistical network modeling techniques on public health data was published in the American Journal of Public Health in 2010. He was also a member of a National Academy of Sciences panel that produced a recent report, Assessing the use of agent-based models for tobacco regulation, which provided the FDA and other public health scien- tists with guidance on how best to use computational models to inform to- bacco control regulation and policy. Dr. Luke directs the doctoral progam in Public Health Sciences at the Brown School, where he also teaches doctoral courses in multilevel and longitudinal modeling, social network analysis, and philosophy of social science. Dr. Luke received his Ph.D. in clinical and community psychology in 1990 from the University of Illinois at Urbana- Champaign

• Series Editor′s IntroductionAbout the AuthorPreface1. The Need for Multilevel ModelingBackground and RationaleTheoretical Reasons for Multilevel ModelsStatistical Reasons for Multilevel ModelsScope of BookOnline Book Resources2. Planning a Multilevel ModelThe Basic Two-Level Multilevel ModelThe Importance of Random EffectsClassifying Multilevel Models3. Building a Multilevel ModelIntroduction to Tobacco Voting Data SetAssessing the Need for a Multilevel ModelModel-building StrategiesEstimationLevel-2 Predictors and Cross-Level InteractionsHypothesis Testing4. Assessing a Multilevel ModelAssessing Model Fit and PerformanceEstimating Posterior MeansCenteringPower Analysis5. Extending the Basic ModelThe Flexibility of the Mixed-Effects ModelGeneralized ModelsThree-level ModelsCross-classified Models6. Longitudinal ModelsLongitudinal Data as Hierarchical: Time Nested Within PersonIntra-individual ChangeInter-individual ChangeAlternative Covariance Structures7.GuidanceRecommendations for Presenting ResultsUseful ResourcesReferences

With growing statistical software package costs, more researchers are using R than ever before. This book allows researchers to do more when using R.