4. THE SIMULATION STUDY We make Mote Carlo simulation compare. Least Squares Estimators (OLS), M-Huber (MH), M-Bisquare, MM- Estimator robust regression, S-Estimator, and MM based initials of coefficient S-Estimator, (MM(S). We use R language to create our program to set up Monte Carlo simulation and this program is available if requested. 4.1 Design of the Simulation Monte Carlo experiments were carried out based on the following data-generating process: Obtain the error term (8) using normal distribution (n, 0, o). o is stander deviation of Normal distribution, o = 1,5. X is distributed Uniform distribution on interval (0,1), (1,3), (2,4), (3,6), and (0,6) where (k-1) is number of the Variables of X Selecting K = (3,6), samples of size n = 50, 100, and 150 and consider that these samples may contain outliers, To investigate the robustness of the methods against outliers, we randomly generate different percentages of outliers (P= 5%, 10%, 15%, 20%, 25% and 30%). Setting the coefficients B equal 1, all simulation results are based on 1500 replications. All computations are obtained based on the R language. The simulation methods are compared using the criteria of estimation method parameters, bias and mean square errors (MSE). When comparing to the MSE of the OLS for such robust methods. MSE = Mean(B - B)2 (12) where B is the estimated value of B