Lehtonen / Pahkinen Practical Methods for Design and Analysis of Complex Surveys

2. Auflage 2004
ISBN: 978-0-470-09163-0
Verlag: John Wiley & Sons
Format: PDF
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 360 Seiten, E-Book

Reihe: Statistics in Practice

Large surveys are becoming increasingly available for public use,and researchers are often faced with the need to analyse complexsurvey data to address key scientific issues. For proper analysisit is also important to be aware of the different aspects of thedesign of complex surveys. Practical Methods for Design andAnalysis of Complex Surveys features intermediate and advancedstatistical techniques for use in designing and analysing complexsurveys. This extensively updated edition features much newmaterial, and detailed practical exercises with links to a Website, helping instructors and enabling use for distancelearning.
* Provides a comprehensive introduction to sampling and estimationin descriptive surveys, including design effect statistic and useof auxiliary data.
* Includes detailed coverage of complex survey analysis, includingdesign-based ANOVA and logistic regression with GEEestimation.
* Contains much new material, including handling of non-samplingerrors, and model-assisted estimation for domains.
* Features detailed real-li fe case studies, such as multilevelmodeling in a multinational educational survey.
* Supported by a Web site containing software codes, real datasets, computerized exercises with solutions, and online trainingmaterials.
Practical Methods for Design and Analysis of Complex Surveysprovides a useful practical resource for researchers andpractitioners working in the planning, implementation or analysisof complex surveys and opinion polls, including business,educational, health, social, and socio-economic surveys andofficial statistics. In addition, the book is well suited for useon intermediate and advanced courses in survey sampling.

Autoren/Hrsg.

Lehtonen, Risto

Pahkinen, Erkki

Inhaltsverzeichnis

Preface.
1. Introduction.
2. Basic Sampling Techniques.
2.1 Basic definitions.
2.2 The Province'91 population.
2.3 Simple random sampling and design effect.
2.4 Systematic sampling and intra-class correlation.
2.5 Selection with probability proportional to size.
3. Further Use of Auxiliary Information.
3.1 Stratified sampling.
3.2 Cluster sampling.
3.3 Model-assisted estimation.
3.4 Efficiency comparison using design effects.
4. Handling Nonsampling Errors.
4.1 Reweighting.
4.2 Imputation.
4.3 Chapter summary and further reading.
5. Linearization and Sample Reuse in VarianceEstimation.
5.1 The Mini-Finland Health Survey.
5.2 Ratio estimators.
5.3 Linearization method.
5.4 Sample reuse methods.
5.5 Comparison of variance estimators.
5.6 The Occupational Health C are Survey.
5.7 Linearization method for covariance-matrix estimation.
5.8 Chapter summary and further reading.
6. Model-assisted Estimation for Domains.
6.1 Framework for domain estimation.
6.2 Estimator type and model choice.
6.3 Construction of estimators and model specification.
6.4 Further comparison of estimators.
6.5 Chapter summary and further reading.
7. Analysis of One-way and Two-way Tables.
7.1 Introductory example.
7.2 Simple goodness-of-fit test.
7.3 Preliminaries for tests for two-way tables.
7.4 Test of homogeneity.
7.5 Test of independence.
7.6 Chapter summary and further reading.
8. Multivariate Survey Analysis.
8.1 Range of methods.
8.2 Types of models and options for analysis.
8.3 Analysis of categorical data.
8.4 Logistic and linear regression.
8.5 Chapter summary and further reading.
9. More Detailed Case Studies.
9.1 Monitoring quality in a long-term transport survey.
9.2 Estimation of mean salary in a business survey.
9.3 Model selection in a socioeconomic survey.
9.4 Multi-level modelling in an educational survey.
References.
Author Index.
Subject Index.
Web Extension.
In addition to the printed book, electronic materials supportingthe use of the book can be found in the web extension.

Über Autor(innen)

Brand and marketing managers, advertisers, and general business readers.