Precision Standards Guidelines for Reporting MEPSIC Descriptive Statistics
The following statistical standards are applied for tabular estimates that can be obtained from the MEPS website and are recommended
as general rules for researchers analyzing MEPSIC data. All tables and table cells on the MEPS website
(https://meps.ahrq.gov/mepsweb/data_stats/quick_tables.jsp)
must satisfy U.S. Census Bureau disclosure review rules. These guidelines are based upon table cell size standards as well as relative standard error (RSE) standards,
where the relative standard error of an estimate is defined as the standard error of the estimate divided by that estimate:
A. Private Sector tabulations
 n ≥ 20: Published estimates should be based on an unweighted sample of at least 20 observations for the subgroup of interest.
 RSE ≥ .30: Estimate can be reported but flagged with an * to indicate that its precision is not reliable.
B. Public Sector (State and Local Governments) tabulations
 n ≥ 20: Published estimates should generally be based on an unweighted sample of 20 or more observations for the subgroup of interest.
 n < 20 and RSE ≤ .10: When there are fewer than 20 observations for the subgroup of interest, estimates may be published if the relative standard error is 10 percent or less.
 RSE ≥ .30: Estimate can be reported but flagged with an * to indicate that its precision is not reliable.
Note: In the MEPSIC tables on the AHRQ website, some rows or columns of estimates have been suppressed if a large portion of the entire row or
column have unreliable estimates (RSE ≥ .30). However, for these rows and columns, the U.S. total estimate is usually provided.
