This study discusses marching skills for mildly retarded students. The aim is to prove that effective modeling techniques are used in improving the line-marching skills of mildly retarded students. This type of research is an experimental method in the form of pre-experimental with the type of one group pretest-posttest design. The population in this study were nine junior high school students at SLB N 2 Padang, all populations were sampled called saturated sampling. The technique of collecting data through an act test in the form of a pretest to determine the initial conditions of students in line marching movements, then given treatment using modeling techniques and continued by providing posttest processed and compared using the Wilcoxon Rank Tast test. From the results of the study the data obtained on average at the time of the pretest is 45.00 while for the posttest there is an increase that is to 81.67. Data is processed more scientifically using Wilcoxon test obtained value of tast rank between pretest and posttest 2.754 with Asymp. Sig (2-tailed) 0.006. Predetermined probability α = 0.05. The alternative hypothesis is accepted because the probability 
                        
                        
                        
                        
                            
                                Copyrights © 2019