Multilevel analysis : an introduction to basic and advanced multilevel modeling

摘要: Preface second edition to first Introduction Multilevel analysis Probability models This book Prerequisites Notation Theories, Multi-Stage Sampling and Models Dependence as a nuisance an interesting phenomenon Macro-level, micro-level, cross-level relations Glommary Statistical Treatment of Clustered Data Aggregation Disaggregation The intraclass correlation Within-group between group variance Testing for differences Design effects in two-stage samples Reliability aggregated variables Within-and Regressions Correlations Estimation within-and between-group correlations Combination within-group evidence Random Intercept Model Terminology notation A regression model: fixed only Variable intercepts: or random parameters? When use coefficient Definition the intercept model More explanatory regressions Parameter estimation 'Estimating' effects: posterior means Posterior confidence intervals Three-level Hierarchical Linear slopes Heteroscedasticity Do not force ?01 be 0! Interpretation slope variances Explanation intercepts Cross-level interaction general formulation parts Specification Centering with slopes? Three more levels Tests parameters Multiparameter tests Deviance powerful Other part Confidence specification Working upward from level one Joint consideration level-one level-two Concluding remarks on How Much Does Explain? Explained Negative values R2? proportion explained two-level three-level Components at functions Quadratic two Missing General issues missing data Implications design dependent variable Full maximum likelihood Imputation imputation method Putting together multiple results Multiple imputations by chained equations Choice Assumptions hierarchical linear Following logic Include contextual Check whether have heteroscedasticity What do case Inspection residuals Residuals Influence units distributional assumptions Designing Studies Some introductory notes power Estimating population mean Measurement subjects association Allocating treatment groups individuals Exploring structure Variance Methods Bayesian inference Sandwich estimators standard errors Latent class Imperfect Hierarchies crossed factor Crossed membership classification Survey Weights Model-based design-based Descriptive analytic surveys Two kinds weights Choosing model-based Inclusion probabilities informativeness sampling Example: Metacognitive strategies measured PISA study divided into assign multilevel Appendix. Matrix expressions single-level Longitudinal Fixed occasions compound symmetry fully multivariate Multivariate occasion designs Populations curves Explaining 27415.2.4 Changing covariates Autocorrelated Why analyze simultaneously? Discrete Dependent Variables generalized logistic Heterogeneous proportions logit function: Log-odds empty Further topics Representation threshold Residual Consequences adding Ordered categorical event history Poisson Software Special software modeling HLM MLwiN MIXOR suite SuperMix Modules general-purpose packages SAS procedures VARCOMP, MIXED, GLIMMIX, NLMIXED R Stata SPSS, commands VARCOMP MIXED PinT Optimal MLPowSim Mplus Gold REALCOM WinBUGS References Index