An efficient editing and imputation strategy within a corporate-wide data collection system at INE Spain: a pilot experience
David Salgado. Methodology Unit. National Statistic Office- Spain.
Data editing is a crucial step in the survey statistics production process. It affects several dimensions of survey quality such as accuracy, timeliness, response burden and cost effectiveness. On the other hand, a step-by-step transition from the stovepipe production model to an industrialized production process has been recently initiated at INE Spain with the design of a corporate-wide metadata-driven production process having the GSBPM as a general framework. In this context INE Spain has begun to revise the editing and imputation strategies of most business surveys in order to increase the efficiency of this production phase while, at the same time, implementing international standards in the design of these strategies. We show how we have undergone this updating process with the Spanish Industrial Turnover Index and New Orders Received Index survey. As a prominent result in this example, we have reduced the recontact rate from roughly 55% to about 15% while keeping the same precision in the indexes. This will serve as a pilot experience to generalize this process to other surveys.