Direction des Statistiques

STES Timeliness Framework: Selective (or Significance) Editing


Back to STES Timeliness Framework

This page contains links to detailed documentation on improving timeliness or reducing costs in the production of short-term economic statistics through selective (or significance) editing.   The papers below provide detailed information on methods relevant to this topic which have been implemented within statistical organisations.  Issues covered include:

•   Methodological frameworks and the application of selective editing, which ensures that potentially erroneous survey data is only queried if the suspected error would have a significant impact on estimates
•   Efficient approaches to editing survey data at both the data receipt and estimation stages

Primary papers

The papers below focus primarily on the issue of selective (or significance) editing.  They may also contain information on other statistical processes defined in the STES Timeliness Framework.

General Application of Significance Editing in Australia (2005)

Score Functions to Reduce Business Survey Editing at the U.K Office for National Statistics (2003)

Retaining Quality Output Whilst Reducing Validation Costs at the ONS (2006)

Significance Editing in the Australian Survey of Average Weekly Earnings (1994)

Significance Editing in the Australian Survey of Employment and Earnings (1999)

The General Application of Significance Editing (2000)

Developments in Data Editing, Imputation, Respondent Follow Up and using VAT Data at Statistics Netherlands (2003)

Spanish Methods to Improve Timeliness in the Industrial Production Indices (2002)

Developing Selective Editing Methodology for Surveys with Varying Characteristics at U.K. ONS (2003)

Secondary papers

The papers below refer to the issue of selective (or significance) editing to some extent.  They also provide more detail on other statistical processes defined in the STES Timeliness Framework.

An Integrated System for Data Input Processing at the U.S. Bureau of Census (2000)

Queries and the submission of new papers

Questions on the content of or how to use this framework should be sent to





Also Available