National Statistics Online - Articles - Implementing selective editing in a monthly business survey
Articles Database
Implementing selective editing in a monthly business survey
How selective (or significant) editing may be used to reduce the amount of editing required, without adversely affecting the quality of the outputs.
National Statistical Institutes (NSIs) have traditionally believed that focusing a large proportion of resources on data editing produces high quality survey data. However, over the last decade, this notion has been questioned by some NSIs and research has shown that over-editing induces high editing costs, high respondent burden, and possibly delivers data of a lower quality than anticipated.
A new philosophy in data editing is therefore emerging which supports reducing the levels of data editing whilst maintaining data quality. By implementing systems which support the new philosophy, resources could be freed up and redirected towards other areas of the survey process that may require development or more input.
The paper discusses how the Office for National Statistics has recently embraced this new approach to editing, and describes the research undertaken into a new selective editing system for one monthly business survey. The results are discussed from a trial run on the Monthly Inquiry into the Distribution and Services Sector, and these show that selective editing reduces the amount of editing required, without adversely affecting the quality of the outputs.
Download
