Title: |
|
An Examination of State Estimates Using Multiple Years of Data from the Medical Expenditure Panel Survey, Household Component |
Description: |
|
This paper first examines the decrease in errors by using two or three consecutive years of data from the MEPS-HC rather than a single year. Due to sample overlap between years in the MEPS-HC, data from year to year are correlated and additional sample has less effect on total error than it would if the years had independent samples. The loss of sample efficiency depends on the correlations of the variables estimated. |
Agency: |
|
Rockville (MD): Agency for Healthcare Research and Quality |
Back
Email Updates
