Editing Procedure for Anthropometric Survey Data
Abstract
Described in this report are two computer programs designed to edit large masses of anthropometric survey data. The XVAL (extreme Value) program scans for gross errors by sorting out the ten highest and ten lowest values in a given set of data for visual inspection. The EDIT (Editing) program, a technique for sifting data more finely, tests each data point for each subject by comparing the measured value with a predicted value obtained from regression equations and flagging those which deviate from set limits. Detailed descriptions of input routines and computer output are liberally illustrated at every step. Instructions for the use of these programs are given to enable programmers to apply them to their own data. Computer printouts of the entire XVAL and EDIT programs, applied to actual data, are included as appendices. The authors emphasize the limitations of these programs by pointing out that however sophisticated the editing routines, they can only serve to discover, identify and flag possible errors. It is left to the experienced professional to confirm the deviant value as an error, assess its nature and decide whether to eliminate, correct or replace it.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1978
- Accession Number
- ADA060393
Entities
People
- Paul Kikta
- Thomas Churchill
Organizations
- University of Dayton