1352.0.55.104 - Research Paper: The Use of Scores to Detect and Prioritise Anomalous Estimates (Methodology Advisory Committee), Nov 2009  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 23/11/2010  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All
  • About this Release

This paper provides an overview of scores used in macro-editing and presents some new scores based on significance criteria. Problems with macro-editing scoring methodologies due to the effect of swamping and masking are discussed. A review of the well-known Hidiroglou–Berthelot edit is provided within a significance editing context. After a brief summary of work done by the U.S. Census Bureau on several score-based methods, the paper introduces the concept of significance for macro-editing and outlines a macro significance editing framework based on an extension of the existing micro significance editing framework used within the Australian Bureau of Statistics (ABS). Some results from empirical comparisons between a proposed macro significance editing application called hierarchical macro-editing and several variants of the Hidiroglou–Berthelot macro-edit are discussed. The paper finishes with a summary of findings and recommendations for developing score-based macro-editing for business surveys conducted by the ABS.