The general application of significance editing

Abstract

The significance editing technique reduces and prioritises the input editing workload in a sample survey. The method is based on estimating the impact of resolving individual edit queries on the survey estimates. Using these impacts it is possible to ignore queries of trivial value while making sure that the most serious queries are given the most attention. The result is efficient allocation of resources to editing without compromising data quality. Lawrence and McDavitt .1994) described an implementation of significance editing in the Australian Survey of Average Weekly Earnings. This article extends the previous work by describing the general principles underlying the significance editing technique, and offering suggestions for applying the technique to other surveys.