Survey Methodology and Missing Data: Tools and Techniques for Practitioners

ISBN-13: 9783319790107 / Angielski / Twarda / 2018 / 224 str.

Seppo Laaksonen

Opis

This book focuses on quantitative survey methodology, data collection and cleaning methods. Providing starting tools for using and analyzing a file once a survey has been conducted, it addresses fields as diverse as advanced weighting, editing, and imputation, which are not well-covered in corresponding survey books.

Spis treści

1. Introduction

2. Concept of survey and key survey terms

2.1 What is survey?

2.2 Five populations in surveys

2.3 Purpose of populations

2.4 Cross-sectional survey micro data

2.5 X variables, auxiliary variables in more details

2.6 Summary of the terms and the symbols of Chapter 2

2.7 Transformations

3. Designing a questionnaire and survey modes

3.1 What is questionnaire design?

3.2 One or more modes in one survey?

3.3 Questionnaire and questioning

3.4 Designing questions for the questionnaire

3.5 Developing questions for the survey

3.6 Satisficing

3.7 Straightlining

3.8 Examples on questions and scales

4. Sampling principles and missingness mechanisms

4.1 Basic concepts, both for probability and non-probability sampling

4.2 Missingness mechanisms

4.3 Non-probability sampling cases

4.4 Probability sampling framework

4.5 Sampling and inclusion probabilities

4.6 Illustration of the stratified three-stage sampling

4.7 Basic weights of stratified three-stage sampling

4.8 Two types of sampling weights

5. Design effects at sampling phase

5.1 DEFF due to clustering = DEFFc

5.2 DEFF due to varying inclusion probabilities = DEFFp

5.3 The entire design effect – DEFF, and the gross sample size

5.4 How to decide the sample size and allocate the gross sample into strata?

 6. Sampling design data file

 6.1 Principles of the sampling design data file

 6.2 Test data used in several examples in the book 

7. Missingness, its reasons and treatment

7.1 Reasons for unit nonresponse

7.2 Coding of item nonresponse

7.3 Missingness indicator and missingness rate

7.4 Response propensity models

8. Weighting adjustments due to unit missingness

8.1 Actions of weighting and reweighting

8.2 Introduction to re-weighting methods

8.3 Post-stratification

8.4 Response propensity weighting

8.5. Comparisons of weights in other surveys

8.6 Linear calibration

8.7 Non-linear calibration

8.8 Summary of all the weights

9. Special cases in weighting

9.1 Sampling of individuals, estimates for clusters such as households

9.2 If analysis weights only are available but the proper weights are required

9.3 Sampling and weights for households, estimates for individuals or other lower level

9.4 Panel of two years

10. Statistical editing

10.1 Edit Rules and ordinary checks

10.2 Other edit checks

10.3 Satisficing in editing

10.4 Selective editing

10.5 Graphical editing

10.6 Tabular editing

10.7 Handling screening data in editing

10.8 Editing not always completely done for public use data

11. Introduction to statistical imputation

11.1 Imputation and its purpose

11.2 Targets for imputation should be specified clearly

11.3 What can be imputed due to missingness?

11.4 ‘Aggregate imputation’

11.5 Most common tools for missing item handling without proper imputation

11.6 Several imputations for the same micro data

12. Imputation methods for single variables

12.1 Imputation process

12.2. Imputation model

12.3. Imputation task

12.4. Nearness metrics of real-donor methods

12.5. Post-Editing after the model-donor method possibly

12.6 Single and multiple imputation

12.7 Examples of Deterministic imputation methods for a continuous variable

12.8 Example of deterministic imputation methods for a binary variable

12.9 Example of the continuous variable when the imputation model is poor

12.10 Interval estimates

13. Summary and key tasks of survey data cleaning

14. Basic survey data analysis

14.1 ‘Survey instruments’ in the analysis

14.2 Simple and demanding examples

14.2.1 The sampling weights vary much

14.2.2 Feeling about household's income nowadays with two types of weights

14.2.3 Examples based on the test data (Chapter 6)

14.2.4 Using sampling weights for cross-country survey data without country results

14.2.5 The PISA literacy scores

14.2.6 Multivariate linear regression with survey instruments

14.2.7 The binary regression model with logit link

14.3 Concluding remarks about the results based on simple and complex methodology

O autorze

Seppo Laaksonen is a professor of statistics at the University of Helsinki, Finland, and has worked at various survey institutes including Statistics Finland, Eurostat and The Finnish Center for Social and Health Research. The former scientific secretary (2001-2003) and vice president of the International Association of Survey Statisticians (2007-2009), he has been a member of the sampling expert team of the European Social Survey since 2001. He has also been involved in a number of European research projects and is a consultant for surveys in Moldova, Ethiopia, Slovenia, the United Kingdom and Hungary.

Blurb

This book focuses on quantitative survey methodology, data collection and cleaning methods. Providing starting tools for using and analyzing a file once a survey has been conducted, it addresses fields as diverse as advanced weighting, editing, and imputation, which are not well-covered in corresponding survey books. Moreover, it presents numerous empirical examples from the author's extensive research experience, particularly real data sets from multinational surveys.