Data Quality Assessment System
Author: Arturo Y. Pacificador, Jr.
Philippine government agencies generate statistical data for a variety of reasons. One of which is to assess the effectiveness of projects and programs being implemented. For data to be useful for such purposes, it must satisfy standards of quality. Hence, the quality of data is directly associated with fitness of use. It is emphasized, however, that there is no one acceptable measure of data quality since it is considered multi-dimensional. This paper presents a framework for data quality assessment of the Bureau of Labor and Employment Statistics (BLES) surveys and labor statistics generated from administrative reports. Guidelines were formulated to ensure data quality for BLES surveys, as well as enable other agencies under the Department of Labor and Employment (DOLE) to develop parameters of data quality for administrative-based statistics. The quality measures and indicators used for this study were guided by Statistics Canada (StatCan) to ensure their efficiency. StatCan presented six dimensions of quality namely: Relevance, Accuracy, Timeliness, Accessibility, Interpretability, and Coherence. Slight modifications were introduced as each of the data system within DOLE sees fit. A questionnaire was created to gain insights from interviews as basis for the modifications in the StatCan’s framework to suit the agency’s conditions. Moreover, each dimension was made up of elements from which quality measures and indicators were defined. The framework proposed, however, is not final. It will evolve over time with the addition of new dimensions and modification of elements, measures and indicators within.