Data Editing for Complex Surveys in Presence Of Administrative Data: An Application to Fss 2013 Livestock Survey Data Based on The Joint Sequential Use Of Different R Packages

Abstract

Data editing and imputation (E&I) in complex sample business surveys is a task which is usually split into two steps to gain efficiency in terms of time and human resources: first selective editing techniques are applied to the primary target estimates variables in order to identify a potential set of influential errors that require usually manual editing and a second part of automatic identification and imputation of inconsistencies and missing values. Within this framework, the present paper reviews the Italian top-down data editing strategy adopted and automated imputation showing the experience applied to 2013 Farm Structure Survey livestock data. In this edition this process has been entirely carried out in the R environment by means of different R packages.

Keywords

• Automated Edit Rules
• Imputation of Missing Values
• Compositional Data
• Random vs. Systematic Errors
• Influential non Influential Errors
• Statistical software R
• Selective editing
• data editing
• Business Surveys