Record Linkage Techniques-1997 -- Copyright -- Preface -- CONTENTS -- COPY PREPARATION AND REVIEWS -- OBTAINING RECORD LINKAGE TECHNIQUES-1985 -- OTHER SOURCES OF RELATED INFORMATION -- SOURCES OF SUPPORT -- Acknowledgments -- Contents -- Chapter 1 Keynote Address: Record Linkage and Public Policy- A Dynamic Evolution -- Record Linkage and Public Policy-A Dynamic Evolution -- Abstract -- Introduction -- Four Critical Factors Shaping the Evolution of Record Linkage -- A Historical Digression -- The Statistical Domain-A Model? -- The Dog That Does Not Bark: Public Administration and Public Unconcern -- Out of Control-Linkage in the Private Sector -- Conclusion -- References -- Chapter 2 Invited Session on Record Linkage Applications for Epidemiological Research -- OX-LINK: THE OXFORD MEDICAL RECORD LINKAGE SYSTEM -- Abstract -- Medical Record Linkage -- Matching and Linking -- Unique Personal Identifiers -- Personal Identifying Variables -- File Ordering and Blocking -- File Blocking -- Phonemic Name Compression -- Soundex Code and the Oxford Name Compression Algorithm (ONCA) -- ORLS File Blocking Keys and Matching Variables -- Generating Extra Records Where a Number of Name Variants Are Present -- Matching Methods -- Match Weights -- Calculating the Weights for the Names Items -- Calculation of Weights for Surnames -- Forenames -- Calculating the Weights for the Non-Names Items -- Threshold Weighting -- Algebraic Summation of the Individual Match Weights -- Linking -- Results -- Acknowledgments -- References -- COMPLEX LINKAGES MADE EASY -- Abstract -- Introduction -- The Example Database -- Some Definitions -- Database Structure -- Types of Linkage (with Examples) -- Simple Linkage -- Auto-Linkage Within a File (Associations Within a File) -- Complex Linkages -- User-Friendly Linkage Software -- Finding Subsets -- The Output.

Discussion and Conclusions -- References -- TIPS AND TECHNIQUES FOR LINKING MULTIPLE DATA SYSTEMS: THE ILLINOIS DEPARTMENT OF HUMAN SERVICES CONSOLIDATION PROJECT -- Abstract -- Chapter 3 Contributed Session on Applications of Record Linkage -- LINKING ADMINISTRATIVE RECORDS OVER TIME: LESSONS FROM THE PANELS OF TAX RETURNS -- Abstract -- Introduction -- Overview of the Statistics of Income (SOI) Panel Studies -- Problems Created by Incorrect SSNs -- Identifying Incorrect SSNs -- The SOI Editing Strategy -- Limitations of the Editing Strategy -- Validating SSNs Against IRS/SSA Records -- Use of the Return Name Control -- Use of Full Name Lines -- Strategies When Name Lines Were Not Available -- Multiple Occurrences within Filing Year -- Findings -- Conclusions and Recommendations -- Acknowledgments -- RECORD LINKAGE OF CENSUS AND ROUTINELY COLLECTED VITAL EVENTS DATA IN THE ONS LONGITUDINAL STUDY -- Abstract -- What Is the ONS Longitudinal Study? -- Creation of the Sample and Methods of Linkage -- The Creation of the Original LS Sample -- The Linkage of Census Data to the LS -- How Good Was the Linkage Between Censuses? -- Vital Events Linkage -- How the Linkage Process Works -- How Good Is the Linkage of Events? -- Confidentiality Issues -- Conclusion -- USE OF PROBABILISTIC LINKAGE FOR AN ANALYSIS OF THE EFFECTIVENESS OF SAFETY BELTS AND HELMETS -- Abstract -- Introduction -- CODES -- Use of Linked Data to Standardized Non-Uniform Data for Analysis -- Outcome Analysis Using "As Reported" Data -- Definition of the Occurrence of an Injury and Injury Severity -- Use of Linked Data to Expand Existing Data -- Identifying Injuries Not Documented by the Police -- Identifying Financial Outcome -- Identifying the Type of Injury -- Barriers to Linkage of Crash and Injury Data -- Benefits of Data Linkage.

MULTIPLE CAUSES OF DEATH FOR THE NATIONAL HEALTH INTERVIEW SURVEY -- Abstract -- Introduction -- Methods -- Weights -- Calibration Samples -- Classification of Potential Matches -- Results -- Subgroup Biases in Classification -- Discussion -- References -- Chapter 4 Invited Session on Record Linkage Methodology -- A METHOD FOR CALIBRATING FALSE-MATCH RATES IN RECORD LINKAGE -- 1.AN OVERVIEW OF RECORD LINKAGE AND THE PROBLEM OF CALIBRATING FALSE-MATCH RATES -- 1.1 General Description of Record Linkage -- 1.2 The Need for Better Methods of Classifying Records as Matches or Nonmatches -- 1.3 A Proposed Solution to the Problem of Calibrating Error Rates -- 2. CALIBRATING FALSE-MATCH RATES IN RECORD LINKAGE USING TRANSFORMED-NORMAL MIXTURE MODELS -- 2.1 Strategy Based on Viewing Distribution of Weights as Mixture -- 2.2 Using a Training Sample to Estimate Global Parameters -- 2.4 False-Match Rate Estimates and Standard Errors with Fixed Global Parameters -- 2.5 Reflecting Uncertainty in Global Parameters -- 3. PERFORMANCE OF CALIBRATION PROCEDURE ON CENSUS COMPUTER-MATCHING DATA -- 3.1 Results From Test-Census Data -- 3.2 A Limitation in the Extreme Tails -- 4. SIMULATION EXPERIMENT TO INVESTIGATE PROPERTIES OF CALIBRATION PROCEDURE -- 4.1 Description of Simulation Study -- 4.2 Results -- 5.DISCUSSION -- APPENDIX: IMPLEMENTATION OF THE SEM ALGORITHM -- REFERENCES -- MODELING ISSUES AND THE USE OF EXPERIENCE IN RECORD LINKAGE -- Abstract -- Introduction -- Mixture Models -- Record Linkage Theory -- Procedure -- Application -- Conclusion -- Acknowledgments -- References -- REGRESSION ANALYSIS OF DATA FILES THAT ARE COMPUTER MATCHED-PART I -- ABSTRACT -- 1 INTRODUCTION -- 2 RECORD LINKAGE BACKGROUND -- 2.1 Fellegi-Sunter Record Linkage Model -- 2.2 Handling Potential Links -- 3. REGRESSION WITH LINKED DATA -- 3.1 Motivating Theory.

3.2 Simulated Application -- 4 SOME HIGHLIGHTS AND LIMITATIONS OF THE SIMULATION RESULTS -- 4.1 Illustrative Results for Mediocre Matching -- 4.2 Overall Results Summary -- 4.3 Some Simulation Limitations -- 5 CONCLUSIONS AND FUTURE WORK -- 5.1 Improvements in Linkage Adjustment -- 5.2 Handling Erroneous Nonlinks -- 5.3 Concluding Comments -- ACKNOWLEDGMENTS AND DISCLAIMERS -- APPENDIX -- A.1 Univariate Regression Model -- A.2 Multiple Regression with Independent Variables from One File and Dependent Variables from the Other File -- A.3 Multiple Regression with Independent Variables from Both Files -- A.4 Rubin-Belin Model -- REFERENCES -- REGRESSION ANALYSIS OF DATA FILES THAT ARE COMPUTER MATCHED-PART II -- Abstract -- Introduction -- Application Setting -- Relation To Earlier Work -- Basic Approach -- Structure of What Follows -- EI and RL Methods Reviewed -- Edit/Imputation -- Record Linkage -- Simulation Setting -- Matching Scenarios -- "First Poor" Scenario (Figure 1a) -- "Second Poor" Scenario (Figure 1b) -- Summary of Matching Scenarios -- Quantitative Scenarios -- Simulation Results -- Initial True Regression Relationship -- Regression after Initial RL` RA Step -- Regression After First Combined RL ` RA ` EI ` RA Step -- Second True Reference Regression -- Regression After Second RL ` RA Step -- Regression After Second Combined RL- RA- EI- RA Step -- Additional Iterations -- Further Results -- Comments and Future Study -- Overall Summary -- Planned Application -- Expected Extensions -- "Limited Accomplishment" -- Guiding Practice -- References -- Chapter 5 Contributed Session on Confidentiality and Strategies for Record Linkage -- USING RECORD LINKAGE TO THWART THE DATA TERRORIST -- Abstract -- Micro- AND Tau-ARGUS: SOFTWARE FOR STATISTICAL DISCLOSURE CONTROL -- Abstract -- Introduction -- SDC for Microdata at Statistics Netherlands.

Re-identification -- SDC Techniques -- The Development of µ-ARGUS -- Metadata -- Generation of Tables and Marginals and Identification of the MINUCs -- Global Recoding and Local Suppression -- Final Steps -- The Development of Tau-ARGUS -- References -- INVESTIGATING AUTO INJURY TREATMENT IN A NO-FAULT STATE: AN ANALYSIS OF LINKED CRASH AND AUTO INSURE DATA -- Abstract -- Background -- Data -- The Matching Procedure -- Match Results -- Findings -- Who Goes to the Chiropractor? -- What Is the Relationship Between Occupant, Vehicle, and Crash Characteristics and Choice of Care? -- Summary and Discussion -- References -- QUANTITATIVE EVALUATION OF THE LINKAGE OPERATIONS OF THE 1996 CENSUS REVERSE RECORD CHECK -- Abstract -- Introduction -- The Reverse Record Check -- The Role of Probabilistic Linkage -- Checking Procedure -- Definition of the Reliability of a Link -- Reliability Evaluation Procedure -- Results -- Linkage Results and Conclusion -- References -- Chapter 6 Invited Session on Business and Miscellaneous Record Linkage Applications -- LINKING FEDERAL ESTATE TAX RECORDS -- Abstract -- Introduction -- The Importance of Inherited Wealth -- The Original Estate Tax Data: Saved in the Nick of Time -- Linking the Data: Overlapping Estate Tax Returns -- Particular Features of Estate Tax Data -- What Lies Ahead -- Acknowledgments -- Footnotes -- References -- POST-MATCHING ADMINISTRATIVE RECORD LINKAGE BETWEEN SOLE PROPRIETORSHIP TAX RETURNS AND THE STANDARD STATISTICAL… -- Abstract -- Introduction -- Context and Motivation for the Matching -- Description of the Files for Matching -- Variables for Matching and their Comparability -- Software Used for the Matching -- Blocking Criteria -- Matching Variables and Weights -- How the Model for Relating the Payroll Variables was Determined -- Match Results -- The Parser -- Unduplication -- Match Error Rates.

Modeling the Error.