Statistical Methods for Handling Incomplete Data

Cover
Half Title
Title Page
Copyright Page
Dedication
Contents
List of Figures
List of Tables
Preface
1. Introduction
	1.1. Introduction
	1.2. Outline
	1.3. How to Use This Book
2. Likelihood-Based Approach
	2.1. Introduction
	2.2. Observed Likelihood
	2.3. Mean Score Function
	2.4. Observed Information
3. Computation
	3.1. Introduction
	3.2. Factoring Likelihood Approach
	3.3. EM Algorithm
	3.4. Monte Carlo Computation
	3.5. Monte Carlo EM
	3.6. Data Augmentation
4. Imputation
	4.1. Introduction
	4.2. Basic Theory
	4.3. Variance Estimation after Imputation
	4.4. Replication Variance Estimation
5. Multiple Imputation
	5.1. Review of Bayesian Inference
	5.2. MI: Bayesian Justification
	5.3. MI: Frequentist Justification
	5.4. MI Using Mixture Models
	5.5. MI for General Purpose Estimation
6. Fractional Imputation
	6.1. Parametric Fractional Imputation
	6.2. Nonparametric Approach
	6.3. Semiparametric Fractional Imputation
	6.4. FI Using Mixture Models
	6.5. FI for Multivariate Categorical Data
	6.6. Model Selection
7. Propensity Scoring Approach
	7.1. Introduction
	7.2. Regression Weighting Method
	7.3. Propensity Score Method
	7.4. Optimal PSEstimation
	7.5. Maximum Entropy Method
	7.6. Doubly Robust Estimation
	7.7. Empirical Likelihood Method
	7.8. Nonparametric Method
8. Nonignorable Missing Data
	8.1. Model Identification
	8.2. Conditional Likelihood Approach
	8.3. Pseudo Likelihood Approach
	8.4. GMM Approach
	8.5. Exponential Tilting Model
	8.6. Latent Variable Approach
	8.7. Callbacks
	8.8. Capture-Recapture Experiment
9. Longitudinal and Clustered Data
	9.1. Ignorable Missing Data
	9.2. Nonignorable Monotone Missing Data
		9.2.1. Parametric Models
		9.2.2. Nonparametric p(y | x)
		9.2.3. Nonparametric Propensity
	9.3. Past-Value-Dependent Missing Data
		9.3.1. Three Different Approaches
		9.3.2. Imputation Models under Past-Value-Dependent Nonmonotone Missing
		9.3.3. Nonparametric Regression Imputation
		9.3.4. Dimension Reduction
		9.3.5. Simulation Study
		9.3.6. Wisconsin Diabetes Registry Study
	9.4. Random-Effect-Dependent Missing Data
		9.4.1. Three Existing Approaches
		9.4.2. Summary Statistics
		9.4.3. Simulation Study
		9.4.4. Modification of Diet in Renal Disease
10. Application to Survey Sampling
	10.1. Introduction
	10.2. Calibration Estimation
	10.3. Propensity Score Weighting Method
	10.4. Multiple Imputation
	10.5. Fractional Imputation
	10.6. Fractional Hot Deck Imputation
	10.7. Imputation for Two-Phase Sampling
	10.8. Synthetic Data Imputation
11. Data Integration
	11.1. Mass Imputation
	11.2. Propensity Score Method
	11.3. Nonparametric Propensity Score Approach
	11.4. Doubly Robust Method
	11.5. Statistical Matching
	11.6. Mass Imputation Using Multilevel Models
	11.7. Data Integration for Regression Analysis
	11.8. Record Linkage
12. Advanced Topics
	12.1. Smoothing Spline Imputation
	12.2. Kernel Ridge Regression Imputation
	12.3. KRR-Based Propensity Score Estimation
	12.4. Soft Calibration
	12.5. Penalized Regression Imputation
	12.6. Sufficient Dimension Reduction
	12.7. Neural Network Model
Bibliography
Index