Missing data and small-area estimation : modern analytical equipment for the survey statistician / Nicholas T. Longford.

Includes bibliographical references (pages [337]-352) and index.

Part I Missing data -- 1 Prologue -- 1.1 Terminology. Some basics. -- 1.2 Populations and variables -- 1.3 Missing data -- 1.4 Suggested reading -- 1.5 Exercises -- 2 Describing incompleteness -- 2.1 The problem of incompleteness -- 2.2 The extent of missing data and the response pattern -- 2.3 Sampling and nonresponse processes -- 2.4 Exercises -- 3 Single imputation and related methods -- 3.1 Data reduction -- 3.2 Data completion -- 3.3 Models for imputation -- 3.4 EM algorithm. -- 3.5 Suggested reading -- 3.6 Exercises -- 4 Multiple imputation -- 4.1 The consequences of imperfect imputation -- 4.2 The method -- 4.3 Conditional distributions -- 4.4 From theory to practice -- 4.5 NMAR and sensitivity analysis -- 4.6 Other applications of MI -- 4.7 Suggested reading -- 4.8 Exercises -- 5 Case studies -- 5.1 The UK Labour Force Survey -- 5.2 The National Survey of Health and Development -- 5.3 The International Social Survey Programme -- 5.4 The Scottish House Condition Survey -- 5.5 Suggested reading. Software -- Part II Small-area estimation -- 6 Introduction -- 6.1 Preliminaries. -- 6.2 Choosing the estimator -- 6.3 Composition -- 6.4 Estimating the district-level variance -- 6.5 Spatial similarity -- 6.6 Suggested reading -- 6.7 Exercises -- 7 Models for small areas -- 7.1 Analysis of variance. -- 7.2 Auxiliary information -- 7.3 Computational procedures -- 7.4 Model selection issues -- 7.5 District-level models -- 7.6 Generalised linear models. -- 7.7 Suggested reading -- 7.8 Exercises -- 8 Using auxiliary information -- 8.1 From models to small-area estimates -- 8.2 Composite estimation -- 8.3 Multivariate composition -- 8.4 Applications -- 8.5 Planning and design for small-area estimation -- 8.6 Suggested reading -- 8.7 Exercises -- 9 Using small-area estimators -- 9.1 Non-linear transformations of the estimates -- 9.2 Ranking and ordering -- 9.3 Estimating many variances and precisions -- 9.4 Suggested reading -- 9.5 Exercises -- 10 Case studies -- 10.1 The UK Labour Force Survey -- 10.2 Samples of Anonymised Records -- 10.3 Norwegian municipalities -- 10.4 The Scottish House Condition Survey -- 10.5 Suggested reading -- Part III Combining estimators -- 11 Model selection -- 11.1 The problem. -- 11.2 Why model selection fails -- 11.3 Synthetic estimation -- 11.4 Analysis of variance. -- 11.5 Ordinary regression -- 11.6 Discussion -- 11.7 Other applications of synthesis -- 11.8 Suggested reading -- 11.9 Exercises -- References