Imagine you are in front of a four piece jigsaw puzzle that has to fit perfectly (sounds easy): 1) sampling, 2) instruments, 3) data management and 4) Field work.
This is how, in the very first session on surveys and sampling of the Taller Internacional de Encuestas y Evaluación de Políticas Públicas, Juan Munoz, a founder of Sistemas Integrales, presented very intuitively the total quality concept as it applies to surveys, integrating these four pieces in order to have a reliable and decent data base.
The keys to total quality:
Apart from the logistical details involved in the preparation of a survey, Juan shared a couple of suggestions:
- Choose a firm, with clear terms of reference
- Acknowledge that sampling errors are as serious as non sampling errors
- Make sure field personnel are self sufficient
- Collect data, be realistic
- Assure permanent data quality by integrating quality control into data collection
Not integrating means:
- The long and winding road to “data cleaning” is unavoidable and data might no longer be relevant for decision making
- Data quality is not assured. “Cleaning” might lead to, in the best case scenario, to a data base that is internally consistent, but does not reflect reality
- The “cleaning” implies a miriad of decisions, which are generally not documented and users might not trust the data. A problem.
And finally, he explained the benefits of integration
- Reliable and timely data bases
- Constant field monitoring
- All field personnel applies consistent criteria in data collection
- Inconsistencies are resolved at the field, and not on assumptions
- Total quality control
In the second week, participants will field surveys, and see if Juan’s ideas work.