Use of a Score Function to Prioritize and Limit Recontacts in Editing Business Surveys

Michel Latouche and Jean-Marie Berthelot

Abstract:
In order to reduce survey costs, large statistical agencies intend to concentrate respondent follow-ups to suspicious units that may have an important effect on the estimates. This paper first presents a summary of the overall strategy for recontact and follow-up in economic surveys. Then, it discusses and compares three score functions used to classify suspicious units according to their potential effect on the estimates. Finally, a trial application of a score function to the Canadian Annual Retail Trade Survey is presented.

Keywords:
Data collection and capture; data quality; follow-up; recontact; respondent burden; survey costs.