Non-response and measurement errors are among the most important non-sampling errors instatistical surveys. Calculation of these errors and providing procedures for reduction of them leads to the improvement of survey results and more efficient estimates. Since researches and survey methodologists are trying to find more efficient methods to assess data quality and results of the surveys, identification of the non-response and measurement errors sources and provision of procedures for measuring and reducing these errors are very important in the surveys.
For this purpose, considering its mission on conducting researches aimed at increasing statistics quality, Statistical Research and Training Centre has put forward the implementation of a research as “Identifying and Controlling Non-Response and Measurement Errors of Statistical Surveys”. This research has been planned with three following cases:
- The impact of non-response errors and the results of testing some methods to compensation of non-response
- Verbatim questioning and using “cognitive” methods in designing survey questionnaire as an effective method for uniformity of measurement errors and test results of this method
- Using "cognitive" methods in designing a questionnaire as an effective method for uniformity of measurement errors and test results of this method
- Using Hidden Markov Models to assess measurement errors in Panel or Semi-Panel surveys and their performance results.