Asaph Young Chun, PhD, is Director-General, Statistics Research Institute, Statistics Korea, Republic of Korea.Michael D. Larsen, PhD, is Professor and Chair, Department of Mathematics and Statistics, Saint Michael’s College, Vermont, USA.Gabriele Durrant, PhD, is Professor, Department of Social Statistics and Demography, University of Southampton, UK.Jerome P. Reiter, PhD, is Professor and Chair, Department of Statistical Science, Duke University, North Carolina, USA.

Select Guide Rating

• Preface xvAcknowledgments xxiList of Contributors xxiiiPart I Fundamentals of Administrative Records Research and Applications 11 On the Use of Proxy Variables in Combining Register and Survey Data 3Li-Chun Zhang1.1 Introduction 31.1.1 A Multisource Data Perspective 31.1.2 Concept of Proxy Variable 51.2 Instances of Proxy Variable 71.2.1 Representation 71.2.2 Measurement 101.3 Estimation Using Multiple Proxy Variables 121.3.1 Asymmetric Setting 131.3.2 Uncertainty Evaluation: A Case of Two-Way Data 151.3.3 Symmetric Setting 171.4 Summary 20References 202 Disclosure Limitation and Confidentiality Protection in Linked Data 25John M. Abowd, Ian M. Schmutte, and Lars Vilhuber2.1 Introduction 252.2 Paradigms of Protection 272.2.1 Input Noise Infusion 292.2.2 Formal Privacy Models 302.3 Confidentiality Protection in Linked Data: Examples 322.3.1 HRS–SSA 322.3.1.1 Data Description 322.3.1.2 Linkages to Other Data 322.3.1.3 Disclosure Avoidance Methods 332.3.2 SIPP–SSA–IRS (SSB) 342.3.2.1 Data Description 342.3.2.2 Disclosure Avoidance Methods 352.3.2.3 Disclosure Avoidance Assessment 352.3.2.4 Analytical Validity Assessment 372.3.3 LEHD: Linked Establishment and Employee Records 382.3.3.1 Data Description 382.3.3.2 Disclosure Avoidance Methods 392.3.3.3 Disclosure Avoidance Assessment for QWI 412.3.3.4 Analytical Validity Assessment for QWI 422.4 Physical and Legal Protections 432.4.1 Statistical Data Enclaves 442.4.2 Remote Processing 462.4.3 Licensing 462.4.4 Disclosure Avoidance Methods 472.4.5 Data Silos 482.5 Conclusions 492.A.1 Other Abbreviations 512.A.2 Concepts 52Acknowledgments 54References 54Part II Data Quality of Administrative Records and Linking Methodology 613 Evaluation of the Quality of Administrative Data Used in the Dutch Virtual Census 63Piet Daas, Eric S. Nordholt, Martijn Tennekes, and Saskia Ossen3.1 Introduction 633.2 Data Sources and Variables 643.3 Quality Framework 663.3.1 Source and Metadata Hyper Dimensions 663.3.2 Data Hyper Dimension 683.4 Quality Evaluation Results for the Dutch 2011 Census 693.4.1 Source and Metadata: Application of Checklist 693.4.2 Data Hyper Dimension: Completeness and Accuracy Results 723.4.2.1 Completeness Dimension 733.4.2.2 Accuracy Dimension 753.4.2.3 Visualizing with a Tableplot 783.4.3 Discussion of the Quality Findings 803.5 Summary 813.6 Practical Implications for Implementation with Surveys and Censuses 813.7 Exercises 82References 824 Improving Input Data Quality in Register-Based Statistics: The Norwegian Experience 85Coen Hendriks4.1 Introduction 854.2 The Use of Administrative Sources in Statistics Norway 864.3 Managing Statistical Populations 894.4 Experiences from the First Norwegian Purely Register-Based Population and Housing Census of 2011 914.5 The Contact with the Owners of Administrative Registers Was Put into System 934.5.1 Agreements on Data Processing 934.5.2 Agreements of Cooperation on Data Quality in Administrative Data Systems 954.5.3 The Forums for Cooperation 964.6 Measuring and Documenting Input Data Quality 964.6.1 Quality Indicators 964.6.2 Operationalizing the Quality Checks 974.6.3 Quality Reports 994.6.4 The Approach is Being Adopted by the Owners of Administrative Data 994.7 Summary 1004.8 Exercises 101References 1045 Cleaning and Using Administrative Lists: Enhanced Practices and Computational Algorithms for Record Linkage and Modeling/Editing/Imputation 105William E. Winkler5.1 Introductory Comments 1055.1.1 Example 1 1055.1.2 Example 2 1065.1.3 Example 3 1075.2 Edit/Imputation 1085.2.1 Background 1085.2.2 Fellegi–Holt Model 1105.2.3 Imputation Generalizing Little–Rubin 1105.2.4 Connecting Edit with Imputation 1115.2.5 Achieving Extreme Computational Speed 1125.3 Record Linkage 1135.3.1 Fellegi–Sunter Model 1135.3.2 Estimating Parameters 1165.3.3 Estimating False Match Rates 1185.3.3.1 The Data Files 1185.3.4 Achieving Extreme Computational Speed 1235.4 Models for Adjusting Statistical Analyses for Linkage Error 1245.4.1 Scheuren–Winkler 1245.4.2 Lahiri–Larsen 1255.4.3 Chambers and Kim 1275.4.4 Chipperfield, Bishop, and Campbell 1285.4.4.1 Empirical Data 1305.4.5 Goldstein, Harron, and Wade 1325.4.6 Hof and Zwinderman 1335.4.7 Tancredi and Liseo 1335.5 Concluding Remarks 1335.6 Issues and Some Related Questions 134References 1346 Assessing Uncertainty When Using Linked Administrative Records 139Jerome P. Reiter6.1 Introduction 1396.2 General Sources of Uncertainty 1406.2.1 Imperfect Matching 1406.2.2 Incomplete Matching 1416.3 Approaches to Accounting for Uncertainty 1426.3.1 Modeling Matching Matrix as Parameter 1436.3.2 Direct Modeling 1466.3.3 Imputation of Entire Concatenated File 1486.4 Concluding Remarks 1496.4.1 Problems to Be Solved 1496.4.2 Practical Implications 1506.5 Exercises 150Acknowledgment 151References 1517 Measuring and Controlling for Non-Consent Bias in Linked Survey and Administrative Data 155Joseph W. Sakshaug7.1 Introduction 1557.1.1 What is Linkage Consent? Why is Linkage Consent Needed? 1557.1.2 Linkage Consent Rates in Large-Scale Surveys 1567.1.3 The Impact of Linkage Non-Consent Bias on Survey Inference 1587.1.4 The Challenge of Measuring and Controlling for Linkage Non-Consent Bias 1587.2 Strategies for Measuring Linkage Non-Consent Bias 1597.2.1 Formulation of Linkage Non-Consent Bias 1597.2.2 Modeling Non-Consent Using Survey Information 1607.2.3 Analyzing Non-Consent Bias for Administrative Variables 1627.3 Methods for Minimizing Non-Consent Bias at the Survey Design Stage 1637.3.1 Optimizing Linkage Consent Rates 1637.3.2 Placement of the Consent Request 1637.3.3 Wording of the Consent Request 1657.3.4 Active and Passive Consent Procedures 1667.3.5 Linkage Consent in Panel Studies 1677.4 Methods for Minimizing Non-Consent Bias at the Survey Analysis Stage 1687.4.1 Controlling for Linkage Non-Consent Bias via Statistical Adjustment 1697.4.2 Weighting Adjustments 1697.4.3 Imputation 1707.5 Summary 1727.5.1 Key Points for Measuring Linkage Non-Consent Bias 1727.5.2 Key Points for Controlling for Linkage Non-Consent Bias 1727.6 Practical Implications for Implementation with Surveys and Censuses 1737.7 Exercises 174References 174Part III Use of Administrative Records in Surveys 1798 A Register-Based Census: The Swedish Experience 181Martin Axelson, Anders Holmberg, Ingegerd Jansson, and Sara Westling8.1 Introduction 1818.2 Background 1828.3 Census 2011 1838.4 A Register-Based Census 1858.4.1 Registers at Statistics Sweden 1858.4.2 Facilitating a System of Registers 1868.4.3 Introducing a Dwelling Identification Key 1878.4.4 The Census Household and Dwelling Populations 1888.5 Evaluation of the Census 1908.5.1 Introduction 1908.5.2 Evaluating Household Size and Type 1928.5.2.1 Sampling Design 1928.5.2.2 Data Collection 1938.5.2.3 Reconciliation 1948.5.2.4 Results 1948.5.3 Evaluating Ownership 1958.5.4 Lessons Learned 1988.6 Impact on Population and Housing Statistics 1998.7 Summary and Final Remarks 201References 2039 Administrative Records Applications for the 2020 Census 205Vincent T. Mule Jr, and Andrew Keller9.1 Introduction 2059.2 Administrative Record Usage in the U.S. Census 2069.3 Administrative Record Integration in 2020 Census Research 2079.3.1 Administrative Record Usage Determinations 2079.3.2 NRFU Design Incorporating Administrative Records 2089.3.3 Administrative Records Sources and Data Preparation 2109.3.4 Approach to Determine Administrative Record Vacant Addresses 2129.3.5 Extension of Vacant Methodology to Nonexistent Cases 2149.3.6 Approach to Determine Occupied Addresses 2159.3.7 Other Aspects and Alternatives of Administrative Record Enumeration 2179.4 Quality Assessment 2199.4.1 Microlevel Evaluations of Quality 2199.4.2 Macrolevel Evaluations of Quality 2219.5 Other Applications of Administrative Record Usage 2249.5.1 Register-Based Census 2249.5.2 Supplement Traditional Enumeration with Adjustments for Estimated Error for Official Census Counts 2249.5.3 Coverage Evaluation 2259.6 Summary 2269.7 Exercises 227References 22810 Use of Administrative Records in Small Area Estimation 231Andreea L. Erciulescu, Carolina Franco, and Partha Lahiri10.1 Introduction 23110.2 Data Preparation 23310.3 Small Area Estimation Models for Combining Information 23810.3.1 Area-level Models 23810.3.2 Unit-level Models 24710.4 An Application 25210.5 Concluding Remarks 25910.6 Exercises 259Acknowledgments 261References 261Part IV Use of Administrative Data in Evidence-Based Policymaking 26911 Enhancement of Health Surveys with Data Linkage 271Cordell Golden and Lisa B. Mirel11.1 Introduction 27111.1.1 The National Center for Health Statistics (NCHS) 27111.1.2 The NCHS Data Linkage Program 27211.1.3 Initial Linkages with NCHS Surveys 27211.2 Examples of NCHS Health Surveys that Were Enhanced Through Linkage 27311.2.1 National Health Interview Survey (NHIS) 27311.2.2 National Health and Nutrition Examination Survey (NHANES) 27411.2.3 National Health Care Surveys 27411.3 NCHS Health Surveys Linked with Vital Records and Administrative Data 27511.3.1 National Death Index (NDI) 27611.3.2 Centers for Medicare and Medicaid Services (CMS) 27611.3.3 Social Security Administration (SSA) 27711.3.4 Department of Housing and Urban Development (HUD) 27711.3.5 United States Renal Data System and the Florida Cancer Data System 27811.4 NCHS Data Linkage Program: Linkage Methodology and Processing Issues 27811.4.1 Informed Consent in Health Surveys 27811.4.2 Informed Consent for Child Survey Participants 27911.4.3 Adaptive Approaches to Linking Health Surveys with Administrative Data 28011.4.4 Use of Alternate Records 28111.4.5 Protecting the Privacy of Health Survey Participants and Maintaining Data Confidentiality 28211.4.6 Updates Over Time 28311.5 Enhancements to Health Survey Data Through Linkage 28411.6 Analytic Considerations and Limitations of Administrative Data 28611.6.1 Adjusting Sample Weights for Linkage-Eligibility 28711.6.2 Residential Mobility and Linkages to State Programs and Registries 28811.7 Future of the NCHS Data Linkage Program 28911.8 Exercises 291Acknowledgments 292Disclaimer 292References 29212 Combining Administrative and Survey Data to Improve Income Measurement 297Bruce D. Meyer and Nikolas Mittag12.1 Introduction 29712.2 Measuring and Decomposing Total Survey Error 29912.3 Generalized Coverage Error 30212.4 Item Nonresponse and Imputation Error 30512.5 Measurement Error 30712.6 Illustration: Using Data Linkage to Better Measure Income and Poverty 31112.7 Accuracy of Links and the Administrative Data 31212.8 Conclusions 31512.9 Exercises 316Acknowledgments 317References 31713 Combining Data from Multiple Sources to Define a Respondent: The Case of Education Data 323Peter Siegel, Darryl Creel, and James Chromy13.1 Introduction 32313.1.1 Options for Defining a Unit Respondent When Data Exist from Sources Instead of or in Addition to an Interview 32413.1.2 Concerns with Defining a Unit Respondent Without Having an Interview 32513.2 Literature Review 32613.3 Methodology 32713.3.1 Computing Weights for Interview Respondents and for Unit Respondents Who May Not Have Interview Data (Usable Case Respondents) 32713.3.1.1 How Many Weights Are Necessary? 32813.3.2 Imputing Data When All or Some Interview Data Are Missing 32813.3.3 Conducting Nonresponse Bias Analyses to Appropriately Consider Interview and Study Nonresponse 32913.4 Example of Defining a Unit Respondent for the National Postsecondary Student Aid Study (NPSAS) 33013.4.1 Overview of NPSAS 33013.4.2 Usable Case Respondent Approach 33313.4.2.1 Results 33313.4.3 Interview Respondent Approach 33513.4.3.1 Results 33613.4.4 Comparison of Estimates, Variances, and Nonresponse Bias Using Two Approaches to Define a Unit Respondent 33813.5 Discussion: Advantages and Disadvantages of Two Approaches to Defining a Unit Respondent 34013.5.1 Interview Respondents 34013.5.2 Usable Case Respondents 34113.6 Practical Implications for Implementation with Surveys and Censuses 34213.A Appendix 34313.A.1 NPSAS:08 Study Respondent Definition 34313.B Appendix 343References 348Index 349

Asaph Young Chun, PhD, is Director-General, Statistics Research Institute, Statistics Korea, Republic of Korea.Michael D. Larsen, PhD, is Professor and Chair, Department of Mathematics and Statistics, Saint Michael’s College, Vermont, USA.Gabriele Durrant, PhD, is Professor, Department of Social Statistics and Demography, University of Southampton, UK.Jerome P. Reiter, PhD, is Professor and Chair, Department of Statistical Science, Duke University, North Carolina, USA.

Select Guide Rating

• Preface xvAcknowledgments xxiList of Contributors xxiiiPart I Fundamentals of Administrative Records Research and Applications 11 On the Use of Proxy Variables in Combining Register and Survey Data 3Li-Chun Zhang1.1 Introduction 31.1.1 A Multisource Data Perspective 31.1.2 Concept of Proxy Variable 51.2 Instances of Proxy Variable 71.2.1 Representation 71.2.2 Measurement 101.3 Estimation Using Multiple Proxy Variables 121.3.1 Asymmetric Setting 131.3.2 Uncertainty Evaluation: A Case of Two-Way Data 151.3.3 Symmetric Setting 171.4 Summary 20References 202 Disclosure Limitation and Confidentiality Protection in Linked Data 25John M. Abowd, Ian M. Schmutte, and Lars Vilhuber2.1 Introduction 252.2 Paradigms of Protection 272.2.1 Input Noise Infusion 292.2.2 Formal Privacy Models 302.3 Confidentiality Protection in Linked Data: Examples 322.3.1 HRS–SSA 322.3.1.1 Data Description 322.3.1.2 Linkages to Other Data 322.3.1.3 Disclosure Avoidance Methods 332.3.2 SIPP–SSA–IRS (SSB) 342.3.2.1 Data Description 342.3.2.2 Disclosure Avoidance Methods 352.3.2.3 Disclosure Avoidance Assessment 352.3.2.4 Analytical Validity Assessment 372.3.3 LEHD: Linked Establishment and Employee Records 382.3.3.1 Data Description 382.3.3.2 Disclosure Avoidance Methods 392.3.3.3 Disclosure Avoidance Assessment for QWI 412.3.3.4 Analytical Validity Assessment for QWI 422.4 Physical and Legal Protections 432.4.1 Statistical Data Enclaves 442.4.2 Remote Processing 462.4.3 Licensing 462.4.4 Disclosure Avoidance Methods 472.4.5 Data Silos 482.5 Conclusions 492.A.1 Other Abbreviations 512.A.2 Concepts 52Acknowledgments 54References 54Part II Data Quality of Administrative Records and Linking Methodology 613 Evaluation of the Quality of Administrative Data Used in the Dutch Virtual Census 63Piet Daas, Eric S. Nordholt, Martijn Tennekes, and Saskia Ossen3.1 Introduction 633.2 Data Sources and Variables 643.3 Quality Framework 663.3.1 Source and Metadata Hyper Dimensions 663.3.2 Data Hyper Dimension 683.4 Quality Evaluation Results for the Dutch 2011 Census 693.4.1 Source and Metadata: Application of Checklist 693.4.2 Data Hyper Dimension: Completeness and Accuracy Results 723.4.2.1 Completeness Dimension 733.4.2.2 Accuracy Dimension 753.4.2.3 Visualizing with a Tableplot 783.4.3 Discussion of the Quality Findings 803.5 Summary 813.6 Practical Implications for Implementation with Surveys and Censuses 813.7 Exercises 82References 824 Improving Input Data Quality in Register-Based Statistics: The Norwegian Experience 85Coen Hendriks4.1 Introduction 854.2 The Use of Administrative Sources in Statistics Norway 864.3 Managing Statistical Populations 894.4 Experiences from the First Norwegian Purely Register-Based Population and Housing Census of 2011 914.5 The Contact with the Owners of Administrative Registers Was Put into System 934.5.1 Agreements on Data Processing 934.5.2 Agreements of Cooperation on Data Quality in Administrative Data Systems 954.5.3 The Forums for Cooperation 964.6 Measuring and Documenting Input Data Quality 964.6.1 Quality Indicators 964.6.2 Operationalizing the Quality Checks 974.6.3 Quality Reports 994.6.4 The Approach is Being Adopted by the Owners of Administrative Data 994.7 Summary 1004.8 Exercises 101References 1045 Cleaning and Using Administrative Lists: Enhanced Practices and Computational Algorithms for Record Linkage and Modeling/Editing/Imputation 105William E. Winkler5.1 Introductory Comments 1055.1.1 Example 1 1055.1.2 Example 2 1065.1.3 Example 3 1075.2 Edit/Imputation 1085.2.1 Background 1085.2.2 Fellegi–Holt Model 1105.2.3 Imputation Generalizing Little–Rubin 1105.2.4 Connecting Edit with Imputation 1115.2.5 Achieving Extreme Computational Speed 1125.3 Record Linkage 1135.3.1 Fellegi–Sunter Model 1135.3.2 Estimating Parameters 1165.3.3 Estimating False Match Rates 1185.3.3.1 The Data Files 1185.3.4 Achieving Extreme Computational Speed 1235.4 Models for Adjusting Statistical Analyses for Linkage Error 1245.4.1 Scheuren–Winkler 1245.4.2 Lahiri–Larsen 1255.4.3 Chambers and Kim 1275.4.4 Chipperfield, Bishop, and Campbell 1285.4.4.1 Empirical Data 1305.4.5 Goldstein, Harron, and Wade 1325.4.6 Hof and Zwinderman 1335.4.7 Tancredi and Liseo 1335.5 Concluding Remarks 1335.6 Issues and Some Related Questions 134References 1346 Assessing Uncertainty When Using Linked Administrative Records 139Jerome P. Reiter6.1 Introduction 1396.2 General Sources of Uncertainty 1406.2.1 Imperfect Matching 1406.2.2 Incomplete Matching 1416.3 Approaches to Accounting for Uncertainty 1426.3.1 Modeling Matching Matrix as Parameter 1436.3.2 Direct Modeling 1466.3.3 Imputation of Entire Concatenated File 1486.4 Concluding Remarks 1496.4.1 Problems to Be Solved 1496.4.2 Practical Implications 1506.5 Exercises 150Acknowledgment 151References 1517 Measuring and Controlling for Non-Consent Bias in Linked Survey and Administrative Data 155Joseph W. Sakshaug7.1 Introduction 1557.1.1 What is Linkage Consent? Why is Linkage Consent Needed? 1557.1.2 Linkage Consent Rates in Large-Scale Surveys 1567.1.3 The Impact of Linkage Non-Consent Bias on Survey Inference 1587.1.4 The Challenge of Measuring and Controlling for Linkage Non-Consent Bias 1587.2 Strategies for Measuring Linkage Non-Consent Bias 1597.2.1 Formulation of Linkage Non-Consent Bias 1597.2.2 Modeling Non-Consent Using Survey Information 1607.2.3 Analyzing Non-Consent Bias for Administrative Variables 1627.3 Methods for Minimizing Non-Consent Bias at the Survey Design Stage 1637.3.1 Optimizing Linkage Consent Rates 1637.3.2 Placement of the Consent Request 1637.3.3 Wording of the Consent Request 1657.3.4 Active and Passive Consent Procedures 1667.3.5 Linkage Consent in Panel Studies 1677.4 Methods for Minimizing Non-Consent Bias at the Survey Analysis Stage 1687.4.1 Controlling for Linkage Non-Consent Bias via Statistical Adjustment 1697.4.2 Weighting Adjustments 1697.4.3 Imputation 1707.5 Summary 1727.5.1 Key Points for Measuring Linkage Non-Consent Bias 1727.5.2 Key Points for Controlling for Linkage Non-Consent Bias 1727.6 Practical Implications for Implementation with Surveys and Censuses 1737.7 Exercises 174References 174Part III Use of Administrative Records in Surveys 1798 A Register-Based Census: The Swedish Experience 181Martin Axelson, Anders Holmberg, Ingegerd Jansson, and Sara Westling8.1 Introduction 1818.2 Background 1828.3 Census 2011 1838.4 A Register-Based Census 1858.4.1 Registers at Statistics Sweden 1858.4.2 Facilitating a System of Registers 1868.4.3 Introducing a Dwelling Identification Key 1878.4.4 The Census Household and Dwelling Populations 1888.5 Evaluation of the Census 1908.5.1 Introduction 1908.5.2 Evaluating Household Size and Type 1928.5.2.1 Sampling Design 1928.5.2.2 Data Collection 1938.5.2.3 Reconciliation 1948.5.2.4 Results 1948.5.3 Evaluating Ownership 1958.5.4 Lessons Learned 1988.6 Impact on Population and Housing Statistics 1998.7 Summary and Final Remarks 201References 2039 Administrative Records Applications for the 2020 Census 205Vincent T. Mule Jr, and Andrew Keller9.1 Introduction 2059.2 Administrative Record Usage in the U.S. Census 2069.3 Administrative Record Integration in 2020 Census Research 2079.3.1 Administrative Record Usage Determinations 2079.3.2 NRFU Design Incorporating Administrative Records 2089.3.3 Administrative Records Sources and Data Preparation 2109.3.4 Approach to Determine Administrative Record Vacant Addresses 2129.3.5 Extension of Vacant Methodology to Nonexistent Cases 2149.3.6 Approach to Determine Occupied Addresses 2159.3.7 Other Aspects and Alternatives of Administrative Record Enumeration 2179.4 Quality Assessment 2199.4.1 Microlevel Evaluations of Quality 2199.4.2 Macrolevel Evaluations of Quality 2219.5 Other Applications of Administrative Record Usage 2249.5.1 Register-Based Census 2249.5.2 Supplement Traditional Enumeration with Adjustments for Estimated Error for Official Census Counts 2249.5.3 Coverage Evaluation 2259.6 Summary 2269.7 Exercises 227References 22810 Use of Administrative Records in Small Area Estimation 231Andreea L. Erciulescu, Carolina Franco, and Partha Lahiri10.1 Introduction 23110.2 Data Preparation 23310.3 Small Area Estimation Models for Combining Information 23810.3.1 Area-level Models 23810.3.2 Unit-level Models 24710.4 An Application 25210.5 Concluding Remarks 25910.6 Exercises 259Acknowledgments 261References 261Part IV Use of Administrative Data in Evidence-Based Policymaking 26911 Enhancement of Health Surveys with Data Linkage 271Cordell Golden and Lisa B. Mirel11.1 Introduction 27111.1.1 The National Center for Health Statistics (NCHS) 27111.1.2 The NCHS Data Linkage Program 27211.1.3 Initial Linkages with NCHS Surveys 27211.2 Examples of NCHS Health Surveys that Were Enhanced Through Linkage 27311.2.1 National Health Interview Survey (NHIS) 27311.2.2 National Health and Nutrition Examination Survey (NHANES) 27411.2.3 National Health Care Surveys 27411.3 NCHS Health Surveys Linked with Vital Records and Administrative Data 27511.3.1 National Death Index (NDI) 27611.3.2 Centers for Medicare and Medicaid Services (CMS) 27611.3.3 Social Security Administration (SSA) 27711.3.4 Department of Housing and Urban Development (HUD) 27711.3.5 United States Renal Data System and the Florida Cancer Data System 27811.4 NCHS Data Linkage Program: Linkage Methodology and Processing Issues 27811.4.1 Informed Consent in Health Surveys 27811.4.2 Informed Consent for Child Survey Participants 27911.4.3 Adaptive Approaches to Linking Health Surveys with Administrative Data 28011.4.4 Use of Alternate Records 28111.4.5 Protecting the Privacy of Health Survey Participants and Maintaining Data Confidentiality 28211.4.6 Updates Over Time 28311.5 Enhancements to Health Survey Data Through Linkage 28411.6 Analytic Considerations and Limitations of Administrative Data 28611.6.1 Adjusting Sample Weights for Linkage-Eligibility 28711.6.2 Residential Mobility and Linkages to State Programs and Registries 28811.7 Future of the NCHS Data Linkage Program 28911.8 Exercises 291Acknowledgments 292Disclaimer 292References 29212 Combining Administrative and Survey Data to Improve Income Measurement 297Bruce D. Meyer and Nikolas Mittag12.1 Introduction 29712.2 Measuring and Decomposing Total Survey Error 29912.3 Generalized Coverage Error 30212.4 Item Nonresponse and Imputation Error 30512.5 Measurement Error 30712.6 Illustration: Using Data Linkage to Better Measure Income and Poverty 31112.7 Accuracy of Links and the Administrative Data 31212.8 Conclusions 31512.9 Exercises 316Acknowledgments 317References 31713 Combining Data from Multiple Sources to Define a Respondent: The Case of Education Data 323Peter Siegel, Darryl Creel, and James Chromy13.1 Introduction 32313.1.1 Options for Defining a Unit Respondent When Data Exist from Sources Instead of or in Addition to an Interview 32413.1.2 Concerns with Defining a Unit Respondent Without Having an Interview 32513.2 Literature Review 32613.3 Methodology 32713.3.1 Computing Weights for Interview Respondents and for Unit Respondents Who May Not Have Interview Data (Usable Case Respondents) 32713.3.1.1 How Many Weights Are Necessary? 32813.3.2 Imputing Data When All or Some Interview Data Are Missing 32813.3.3 Conducting Nonresponse Bias Analyses to Appropriately Consider Interview and Study Nonresponse 32913.4 Example of Defining a Unit Respondent for the National Postsecondary Student Aid Study (NPSAS) 33013.4.1 Overview of NPSAS 33013.4.2 Usable Case Respondent Approach 33313.4.2.1 Results 33313.4.3 Interview Respondent Approach 33513.4.3.1 Results 33613.4.4 Comparison of Estimates, Variances, and Nonresponse Bias Using Two Approaches to Define a Unit Respondent 33813.5 Discussion: Advantages and Disadvantages of Two Approaches to Defining a Unit Respondent 34013.5.1 Interview Respondents 34013.5.2 Usable Case Respondents 34113.6 Practical Implications for Implementation with Surveys and Censuses 34213.A Appendix 34313.A.1 NPSAS:08 Study Respondent Definition 34313.B Appendix 343References 348Index 349

Asaph Young Chun, PhD, is Director-General, Statistics Research Institute, Statistics Korea, Republic of Korea.Michael D. Larsen, PhD, is Professor and Chair, Department of Mathematics and Statistics, Saint Michael’s College, Vermont, USA.Gabriele Durrant, PhD, is Professor, Department of Social Statistics and Demography, University of Southampton, UK.Jerome P. Reiter, PhD, is Professor and Chair, Department of Statistical Science, Duke University, North Carolina, USA.

Select Guide Rating

• Preface xvAcknowledgments xxiList of Contributors xxiiiPart I Fundamentals of Administrative Records Research and Applications 11 On the Use of Proxy Variables in Combining Register and Survey Data 3Li-Chun Zhang1.1 Introduction 31.1.1 A Multisource Data Perspective 31.1.2 Concept of Proxy Variable 51.2 Instances of Proxy Variable 71.2.1 Representation 71.2.2 Measurement 101.3 Estimation Using Multiple Proxy Variables 121.3.1 Asymmetric Setting 131.3.2 Uncertainty Evaluation: A Case of Two-Way Data 151.3.3 Symmetric Setting 171.4 Summary 20References 202 Disclosure Limitation and Confidentiality Protection in Linked Data 25John M. Abowd, Ian M. Schmutte, and Lars Vilhuber2.1 Introduction 252.2 Paradigms of Protection 272.2.1 Input Noise Infusion 292.2.2 Formal Privacy Models 302.3 Confidentiality Protection in Linked Data: Examples 322.3.1 HRS–SSA 322.3.1.1 Data Description 322.3.1.2 Linkages to Other Data 322.3.1.3 Disclosure Avoidance Methods 332.3.2 SIPP–SSA–IRS (SSB) 342.3.2.1 Data Description 342.3.2.2 Disclosure Avoidance Methods 352.3.2.3 Disclosure Avoidance Assessment 352.3.2.4 Analytical Validity Assessment 372.3.3 LEHD: Linked Establishment and Employee Records 382.3.3.1 Data Description 382.3.3.2 Disclosure Avoidance Methods 392.3.3.3 Disclosure Avoidance Assessment for QWI 412.3.3.4 Analytical Validity Assessment for QWI 422.4 Physical and Legal Protections 432.4.1 Statistical Data Enclaves 442.4.2 Remote Processing 462.4.3 Licensing 462.4.4 Disclosure Avoidance Methods 472.4.5 Data Silos 482.5 Conclusions 492.A.1 Other Abbreviations 512.A.2 Concepts 52Acknowledgments 54References 54Part II Data Quality of Administrative Records and Linking Methodology 613 Evaluation of the Quality of Administrative Data Used in the Dutch Virtual Census 63Piet Daas, Eric S. Nordholt, Martijn Tennekes, and Saskia Ossen3.1 Introduction 633.2 Data Sources and Variables 643.3 Quality Framework 663.3.1 Source and Metadata Hyper Dimensions 663.3.2 Data Hyper Dimension 683.4 Quality Evaluation Results for the Dutch 2011 Census 693.4.1 Source and Metadata: Application of Checklist 693.4.2 Data Hyper Dimension: Completeness and Accuracy Results 723.4.2.1 Completeness Dimension 733.4.2.2 Accuracy Dimension 753.4.2.3 Visualizing with a Tableplot 783.4.3 Discussion of the Quality Findings 803.5 Summary 813.6 Practical Implications for Implementation with Surveys and Censuses 813.7 Exercises 82References 824 Improving Input Data Quality in Register-Based Statistics: The Norwegian Experience 85Coen Hendriks4.1 Introduction 854.2 The Use of Administrative Sources in Statistics Norway 864.3 Managing Statistical Populations 894.4 Experiences from the First Norwegian Purely Register-Based Population and Housing Census of 2011 914.5 The Contact with the Owners of Administrative Registers Was Put into System 934.5.1 Agreements on Data Processing 934.5.2 Agreements of Cooperation on Data Quality in Administrative Data Systems 954.5.3 The Forums for Cooperation 964.6 Measuring and Documenting Input Data Quality 964.6.1 Quality Indicators 964.6.2 Operationalizing the Quality Checks 974.6.3 Quality Reports 994.6.4 The Approach is Being Adopted by the Owners of Administrative Data 994.7 Summary 1004.8 Exercises 101References 1045 Cleaning and Using Administrative Lists: Enhanced Practices and Computational Algorithms for Record Linkage and Modeling/Editing/Imputation 105William E. Winkler5.1 Introductory Comments 1055.1.1 Example 1 1055.1.2 Example 2 1065.1.3 Example 3 1075.2 Edit/Imputation 1085.2.1 Background 1085.2.2 Fellegi–Holt Model 1105.2.3 Imputation Generalizing Little–Rubin 1105.2.4 Connecting Edit with Imputation 1115.2.5 Achieving Extreme Computational Speed 1125.3 Record Linkage 1135.3.1 Fellegi–Sunter Model 1135.3.2 Estimating Parameters 1165.3.3 Estimating False Match Rates 1185.3.3.1 The Data Files 1185.3.4 Achieving Extreme Computational Speed 1235.4 Models for Adjusting Statistical Analyses for Linkage Error 1245.4.1 Scheuren–Winkler 1245.4.2 Lahiri–Larsen 1255.4.3 Chambers and Kim 1275.4.4 Chipperfield, Bishop, and Campbell 1285.4.4.1 Empirical Data 1305.4.5 Goldstein, Harron, and Wade 1325.4.6 Hof and Zwinderman 1335.4.7 Tancredi and Liseo 1335.5 Concluding Remarks 1335.6 Issues and Some Related Questions 134References 1346 Assessing Uncertainty When Using Linked Administrative Records 139Jerome P. Reiter6.1 Introduction 1396.2 General Sources of Uncertainty 1406.2.1 Imperfect Matching 1406.2.2 Incomplete Matching 1416.3 Approaches to Accounting for Uncertainty 1426.3.1 Modeling Matching Matrix as Parameter 1436.3.2 Direct Modeling 1466.3.3 Imputation of Entire Concatenated File 1486.4 Concluding Remarks 1496.4.1 Problems to Be Solved 1496.4.2 Practical Implications 1506.5 Exercises 150Acknowledgment 151References 1517 Measuring and Controlling for Non-Consent Bias in Linked Survey and Administrative Data 155Joseph W. Sakshaug7.1 Introduction 1557.1.1 What is Linkage Consent? Why is Linkage Consent Needed? 1557.1.2 Linkage Consent Rates in Large-Scale Surveys 1567.1.3 The Impact of Linkage Non-Consent Bias on Survey Inference 1587.1.4 The Challenge of Measuring and Controlling for Linkage Non-Consent Bias 1587.2 Strategies for Measuring Linkage Non-Consent Bias 1597.2.1 Formulation of Linkage Non-Consent Bias 1597.2.2 Modeling Non-Consent Using Survey Information 1607.2.3 Analyzing Non-Consent Bias for Administrative Variables 1627.3 Methods for Minimizing Non-Consent Bias at the Survey Design Stage 1637.3.1 Optimizing Linkage Consent Rates 1637.3.2 Placement of the Consent Request 1637.3.3 Wording of the Consent Request 1657.3.4 Active and Passive Consent Procedures 1667.3.5 Linkage Consent in Panel Studies 1677.4 Methods for Minimizing Non-Consent Bias at the Survey Analysis Stage 1687.4.1 Controlling for Linkage Non-Consent Bias via Statistical Adjustment 1697.4.2 Weighting Adjustments 1697.4.3 Imputation 1707.5 Summary 1727.5.1 Key Points for Measuring Linkage Non-Consent Bias 1727.5.2 Key Points for Controlling for Linkage Non-Consent Bias 1727.6 Practical Implications for Implementation with Surveys and Censuses 1737.7 Exercises 174References 174Part III Use of Administrative Records in Surveys 1798 A Register-Based Census: The Swedish Experience 181Martin Axelson, Anders Holmberg, Ingegerd Jansson, and Sara Westling8.1 Introduction 1818.2 Background 1828.3 Census 2011 1838.4 A Register-Based Census 1858.4.1 Registers at Statistics Sweden 1858.4.2 Facilitating a System of Registers 1868.4.3 Introducing a Dwelling Identification Key 1878.4.4 The Census Household and Dwelling Populations 1888.5 Evaluation of the Census 1908.5.1 Introduction 1908.5.2 Evaluating Household Size and Type 1928.5.2.1 Sampling Design 1928.5.2.2 Data Collection 1938.5.2.3 Reconciliation 1948.5.2.4 Results 1948.5.3 Evaluating Ownership 1958.5.4 Lessons Learned 1988.6 Impact on Population and Housing Statistics 1998.7 Summary and Final Remarks 201References 2039 Administrative Records Applications for the 2020 Census 205Vincent T. Mule Jr, and Andrew Keller9.1 Introduction 2059.2 Administrative Record Usage in the U.S. Census 2069.3 Administrative Record Integration in 2020 Census Research 2079.3.1 Administrative Record Usage Determinations 2079.3.2 NRFU Design Incorporating Administrative Records 2089.3.3 Administrative Records Sources and Data Preparation 2109.3.4 Approach to Determine Administrative Record Vacant Addresses 2129.3.5 Extension of Vacant Methodology to Nonexistent Cases 2149.3.6 Approach to Determine Occupied Addresses 2159.3.7 Other Aspects and Alternatives of Administrative Record Enumeration 2179.4 Quality Assessment 2199.4.1 Microlevel Evaluations of Quality 2199.4.2 Macrolevel Evaluations of Quality 2219.5 Other Applications of Administrative Record Usage 2249.5.1 Register-Based Census 2249.5.2 Supplement Traditional Enumeration with Adjustments for Estimated Error for Official Census Counts 2249.5.3 Coverage Evaluation 2259.6 Summary 2269.7 Exercises 227References 22810 Use of Administrative Records in Small Area Estimation 231Andreea L. Erciulescu, Carolina Franco, and Partha Lahiri10.1 Introduction 23110.2 Data Preparation 23310.3 Small Area Estimation Models for Combining Information 23810.3.1 Area-level Models 23810.3.2 Unit-level Models 24710.4 An Application 25210.5 Concluding Remarks 25910.6 Exercises 259Acknowledgments 261References 261Part IV Use of Administrative Data in Evidence-Based Policymaking 26911 Enhancement of Health Surveys with Data Linkage 271Cordell Golden and Lisa B. Mirel11.1 Introduction 27111.1.1 The National Center for Health Statistics (NCHS) 27111.1.2 The NCHS Data Linkage Program 27211.1.3 Initial Linkages with NCHS Surveys 27211.2 Examples of NCHS Health Surveys that Were Enhanced Through Linkage 27311.2.1 National Health Interview Survey (NHIS) 27311.2.2 National Health and Nutrition Examination Survey (NHANES) 27411.2.3 National Health Care Surveys 27411.3 NCHS Health Surveys Linked with Vital Records and Administrative Data 27511.3.1 National Death Index (NDI) 27611.3.2 Centers for Medicare and Medicaid Services (CMS) 27611.3.3 Social Security Administration (SSA) 27711.3.4 Department of Housing and Urban Development (HUD) 27711.3.5 United States Renal Data System and the Florida Cancer Data System 27811.4 NCHS Data Linkage Program: Linkage Methodology and Processing Issues 27811.4.1 Informed Consent in Health Surveys 27811.4.2 Informed Consent for Child Survey Participants 27911.4.3 Adaptive Approaches to Linking Health Surveys with Administrative Data 28011.4.4 Use of Alternate Records 28111.4.5 Protecting the Privacy of Health Survey Participants and Maintaining Data Confidentiality 28211.4.6 Updates Over Time 28311.5 Enhancements to Health Survey Data Through Linkage 28411.6 Analytic Considerations and Limitations of Administrative Data 28611.6.1 Adjusting Sample Weights for Linkage-Eligibility 28711.6.2 Residential Mobility and Linkages to State Programs and Registries 28811.7 Future of the NCHS Data Linkage Program 28911.8 Exercises 291Acknowledgments 292Disclaimer 292References 29212 Combining Administrative and Survey Data to Improve Income Measurement 297Bruce D. Meyer and Nikolas Mittag12.1 Introduction 29712.2 Measuring and Decomposing Total Survey Error 29912.3 Generalized Coverage Error 30212.4 Item Nonresponse and Imputation Error 30512.5 Measurement Error 30712.6 Illustration: Using Data Linkage to Better Measure Income and Poverty 31112.7 Accuracy of Links and the Administrative Data 31212.8 Conclusions 31512.9 Exercises 316Acknowledgments 317References 31713 Combining Data from Multiple Sources to Define a Respondent: The Case of Education Data 323Peter Siegel, Darryl Creel, and James Chromy13.1 Introduction 32313.1.1 Options for Defining a Unit Respondent When Data Exist from Sources Instead of or in Addition to an Interview 32413.1.2 Concerns with Defining a Unit Respondent Without Having an Interview 32513.2 Literature Review 32613.3 Methodology 32713.3.1 Computing Weights for Interview Respondents and for Unit Respondents Who May Not Have Interview Data (Usable Case Respondents) 32713.3.1.1 How Many Weights Are Necessary? 32813.3.2 Imputing Data When All or Some Interview Data Are Missing 32813.3.3 Conducting Nonresponse Bias Analyses to Appropriately Consider Interview and Study Nonresponse 32913.4 Example of Defining a Unit Respondent for the National Postsecondary Student Aid Study (NPSAS) 33013.4.1 Overview of NPSAS 33013.4.2 Usable Case Respondent Approach 33313.4.2.1 Results 33313.4.3 Interview Respondent Approach 33513.4.3.1 Results 33613.4.4 Comparison of Estimates, Variances, and Nonresponse Bias Using Two Approaches to Define a Unit Respondent 33813.5 Discussion: Advantages and Disadvantages of Two Approaches to Defining a Unit Respondent 34013.5.1 Interview Respondents 34013.5.2 Usable Case Respondents 34113.6 Practical Implications for Implementation with Surveys and Censuses 34213.A Appendix 34313.A.1 NPSAS:08 Study Respondent Definition 34313.B Appendix 343References 348Index 349