Record check, record linkage, nonsampling error, response bias
The SIPP Record Check Study uses administrative record data to describe and model errors in SIPP responses for a variety of income sources and transfer programs. The project uses record linkage techniques to identify SIPP sample persons in four states who are on record as having received payments from any of nine state or Federal programs, and then compares survey-reported dates and amounts of payments with official record values. The paper describes basic considerations in designing the project and presents some early findings describing errors in reported program participation and amounts of payments received. Results of the study are also used to shed light on the “seam bias” problem in SIPP.
Moore, Jeffrey C. and Kent H. Marquis. (2010). Using Administrative Record Data to Describe SIPP Response Errors. Statistical Research Division Research Report Series (Survey Methodology #2010-11). U.S. Census Bureau. Available online at <http://www.census.gov/srd/papers/pdf/rsm2010-11.pdf>.