A functional evaluation of edit and imputation tools

Article type: Research Article

Authors: Poirier, Claude^{; 1}

Affiliations: Statistics Canada, 120 Parkdale Ave., Ottawa, Canada K1A 0T6. E-mail: [email protected]

Note: [1] Paper presented to the Work Session on Statistical Data Editing in Rome, Italy, 2–4 June, 1999.

Abstract: In the development of new surveys or the redesign of existing surveys, edit and imputation applications are usually prepared. Many statistical agencies develop and maintain generalized systems to offer managers basic tools for each survey step, including the edit and imputation process, but the choice of a system may sometimes be difficult. Internal systems may not offer the required functionality, which make softwares that are developed outside their own agencies potential candidates for their applications. The money and time anybody would invest to acquire, understand and adapt such systems must be considered in the choice. The paper presents a functional evaluation of four edit and imputation systems. They are the Generalized Edit and Imputation System from Statistics Canada, the New Imputation Methodology also from Statistics Canada, the Standard Economic Processing System from the U.S. Bureau of the Census, and Solas from Statistical Solutions – Ireland. The evaluation tries to identify the best contexts in which each system can serve survey statisticians. Strengths and weaknesses are detailed to help future developers in the implementation of any new system.

DOI: 10.3233/SJU-2000-17101

Journal: Statistical Journal of the United Nations Economic Commission for Europe, vol. 17, no. 1, pp. 1-11, 2000