Comparing CAPI Trace File Data and Quality Control Reinterview Data as Methods of Maintaining Data Quality

• Activity Number: 383
• Type: Invited
• Date/Time: Tuesday, August 2, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Matt Jans*+ and Robyn Sirkis and Christina Schultheis and Renee M. Gindi and Jim Dahlhamer
• Companies: U.S. Census Bureau and U.S. Census Bureau and U.S. Census Bureau and Centers for Disease Control and Prevention and Centers for Disease Control and Prevention
• Keywords: paradata ; CAPI ; data quality ; interviewers ; NHIS ; Census Bureau

Abstract

Quality control is paramount to survey operations. Two methods currently used at the U.S. Census Bureau to measure data quality are the Performance and Data Analysis (PANDA) system and Quality Control (QC) Reinterview. PANDA uses CAPI trace files, data files, and other case information as indicators of cases or interviewers that might be at risk for lowering overall data quality. The QC Reinterview is a verification interview with respondents that asks questions about the interview experience to detect falsification. This paper explores whether these systems capture the same cases and interviewers. Both systems are used on the National Health Interview Survey, a nationwide face-to-face CAPI survey sponsored by the National Center for Health Statistics. Using data quality results from 2009 and 2010, we analyze which sample units and interviewers are identified by PANDA, by QC Reinterview, or by both. We use a multinomial logistic regression predicting identification by PANDA, Reinterview, both, or neither to see if any sample or caseload factors predict identification. Our results suggest independence of the systems with some qualifications. We discuss avenues for further research.