There are several methods available to evaluate an edit system. One method is to compare the raw (unedited) data file with the edited file for each question. By ordering the changes by descending absolute magnitude, the cumulative impact of the editing changes to the total editing change or to the estimated item total can be displayed in graphs and tables. This technique can be used to identify questionnaire problems and respondent errors. When forms are reviewed before data capture or when general editing changes are noted on the questionnaires, evaluation studies can be carried out by selecting a sample of forms and analysing the effect of the editing procedures on individual data items. A widely used technique to evaluate new editing methods is to simulate the new process using a raw data file from the survey. By replacing values flagged according to the new method with the values from the tabulation file containing the data edited by the alternative editing process, it becomes possible to compare estimates from this newly edited file with estimates from the tabulation file.
UNECE Data Editing Group
06 Dec 2013