Estimates published by NCHS must meet reliability criteria based on the relative standard error (RSE or coefficient of variation) of the estimate and on the number of sampled records on which the estimate is based. RSE is a measure of variability and is calculated by dividing the standard error of an estimate by the estimate itself. The result is then converted to a percentage by multiplying by 100. Guidelines used by NCHS authors to determine whether estimates should be presented in tables of NCHS published data reports include:
If the estimate is based on fewer than 30 sampled cases, the value of the estimate is not reported. This is usually indicated with an asterisk (*).
If the estimate is based on 60 or more sampled cases and the RSE is less than 30%, the estimate is reported and is considered reliable.
All other reported estimates should not be assumed to be reliable. These include estimates with an RSE of 30% or more and estimates based on 30-59 cases, regardless of RSE.