comparison between of estimators Robust and Classical in repeated measurement experiments analysis

Abstract

The main objective of the present study is the use of known statistical methods to overcome the problems in the data and the comparison between them and also to discover the best method and apply it on factual data. This represents the influence of time on children weights, to do that the study considered taking four measurements of weights; they are (the weight in the first month, the weight in the third month, the weight in sixth month, and the weight in the ninth month) from which is one of the governmental health centers in Najaf governorate. The data have been analyzed with different methods including the (Classical) method which represents (Linear Regression Model) and the (Robust Method). Outliers were added to the factual data then analyzed again with the same abovementioned methods. To find out which of the two methods is the best, a comparison has been conducted in which different measurements are used, and they are: Mean Square Error (MSE), Coefficient of Determination (R2), and (P-Value). The comparison showed that Linear Regression Model is better that Robust Model in case of factual data free of outliers, whereas the Robust Model is found to be better than Linear Regression Model in case of factual data with outliers