OUTLIERS AND SOME NON-TRADITIONAL MEASURES OF LOCATION IN ANALYSIS OF WAGES

Abstract

The paper deals with an analysis of how to use certain measures of location in analysis of wages. One of traditional measures of location – the mean should to offer typical value of variable, representing all its values by the best way. Sometimes the mean is located in the tail of the distribution and gives very biased idea about the location of the distribution. The removing of outliers, if any, or using of different measures of location could be useful in these cases. Outliers are characterized and some robust methods of their detecting are described in the paper. Then the trimmed mean and M-estimators are characterized. Computing of one-step M-estimator and modified one-step M-estimator of location is described. The possibilities of using these tools are illustrated on the analysis of the gross yearly wages of employers of one Slovak firm in the year 2013.