Online Program

WITHDRAWN: An Assessment of Medical Expenditure Panel Survey Sampling and Estimation Procedures through Benchmarking with the National Health Interview Survey

*Sadeq R Chowdhury, Agency for Healthcare Research and Quality 

Keywords: Evaluation, MEPS, NHIS, Sampling, Weighting

A new sample for the Medical Expenditure Panel Survey (MEPS) is selected each year from the responding households of the previous year’s National Health Interview Survey (NHIS). The sample is selected by first creating a frame of households using approximately three-eighths of the NHIS sample. Since the size of the NHIS sample is about five to six times of the responding sample size in a MEPS panel, it gives an opportunity to benchmark the estimates from the MEPS with that of the NHIS. Moreover, since the data collected in the NHIS are also available for the MEPS frame and the sample, the estimates based on the same variables but from different subsets allow a controlled comparison among estimates that avoids any distortion due to differences in data collection and processing between the surveys. This paper evaluates the effectiveness of the MEPS frame creation, sampling and weighting procedures by comparing estimates produced from the MEPS frame and the sample with the NHIS estimates.