One of the most important topics that are discussed in survey methodology is the accuracy of statistics or survey errors that may occur in the parameters estimation process. In statistical literature, these errors are grouped into two main categories: sampling errors and non-sampling errors. Measurement error is one of the most important non-sampling errors. Since estimating of measurement error is more complex than other types of survey errors, much more research has been done on ways of preventing or dealing with this error. The main problem associated with measurement error is the difficulty to measure or estimate this error in surveys. Various methods can be used for estimating measurement error in surveys, but the most appropriate method in each survey should be adopted according to the method of calculating statistics and the survey conditions. This paper considering some practical experiences in calculating and evaluating surveys results, intends to help statisticians to adopt an appropriate method for estimating measurement error. So to achieve this aim, after reviewing the concept and sources of measurement error, some methods of estimating the error are revised in this paper. Also, the advantages and disadvantages of measurement error estimation methods are discussed and some examples of estimating methods using surveys real data are shown in this paper. It should be noted that if estimating the measurement error with an acceptable accuracy is impossible in practice, it should be ensured based on statistical methods that this error does not have a large value or any increasing trend over time.
Zahedian, Alireza and Saba, Roshanak A.
Measurement Error Estimation Methods in Survey Methodology,
Applications and Applied Mathematics: An International Journal (AAM), Vol. 11,
1, Article 5.
Available at: https://digitalcommons.pvamu.edu/aam/vol11/iss1/5