Iterative method for reducing the impact of outlying data points: Ensuring data quality

摘要

Data editing is essential to check the survey data for possible data problems. Outlying data values are frequently encountered in sample surveys. Consequently, in working with data, the correctness of the reported values must be verified, and if a reported value constitutes an outlier, its appropriate treatment needs to be considered. In this paper, the Iterative method for the reducing the impact of outlying data points is proposed. The novelty of the Iterative method for the reducing the impact of outliers is the following: an iterative approach for determining the outlying data points is proposed; outliers are determined considering the impact of conjoined factors; estimation of weight coefficients of the outliers and estimation of the total measurement error of the non-linear regression model is carried out.