Selective Editing: A Quest for Efficiency and Data Quality

National statistical institutes are responsible for publishing high quality statistical information on many different aspects of society. This task is complicated considerably by the fact that data collected by statistical offices often contain errors. The process of correcting errors is referred to as statistical data editing. For many years this has been a purely manual process, with people checking the collected data record by record and correcting them if necessary. For this reason the data editing process has been both expensive and time-consuming. This article sketches some of the important methodological developments aiming to improve the efficiency of the data editing process that have occurred during the past few decades. The article focuses on selective editing, which is based on an idea rather shocking for people working in the production of high-quality data: that it is not necessary to find and correct all errors. Instead of trying to correct all errors, it generally suffices to correct only those errors where data editing has substantial influence on publication figures. This overview article sketches the background of selective editing, describes the most usual form of selective editing up to now, and discusses the contributions to this special issue of the Journal of Official Statistics on selective editing. The article concludes with describing some possible directions for future research on selective editing and statistical data editing in general.