National Statistics Online - Articles - Publishing sampling errors for business surveys
Articles Database
Publishing sampling errors for business surveys
The extent to which estimates of sampling error are published; and the plans made to extend their publication for key statistics.
Most business statistics produced by the Office for National Statistics are based on stratified random samples. In stratified sampling the population is divided into non-overlapping subpopulations called strata. A random sample is selected in each stratum. The statistics are therefore subject to sampling error.
This article sets out the extent to which estimates of sampling error are published and what plans have been made to extend their publication for key statistics. It highlights some of the challenges that still exist in making such information available. Firstly, however, the article draws attention to the wider concepts of the quality of statistics and in particular their accuracy.
Download
