<![CDATA[The author carried out this research based on the results of the initial observations of the researchers found that the long jump learning outcomes of class X TKJ SMK Muhammadiyah 1 Sumedang had not shown good results. One effort that can be done is to use a direct instruction model to improve learning outcomes in the squat style long jump. In general, the formulation of the problem in this study is: can the use of direct instruction (direct instruction) model improve the learning outcomes of the squat style long jump in class X TKJ SMK Muhammadiyah 1 Sumedang?, and to what extent can the use of direct instruction (direct instruction) model increase learning outcomes of squat style long jump in class X TKJ SMK Muhammadiyah 1 Sumedang? In this study, the authors used an experimental method with 24 students in class X TKJ SMK Muhammadiyah 1 Sumedang as a sample from a population of seven classes. The data for this study were collected through observation sheets and squat style long jump tests. Data processing procedures concerning whether the hypothesis is accepted or not according to the significance proposed are: finding the average, finding the standard deviation, the liliefors normality test, and the increase test. Based on the results of the research and statistical calculations on the hypothesis testing of the use of direct instruction models in improving the learning outcomes of the squat style long jump (experimental research on class X students of SMK Muhammadiyah 1 Sumedang, the authors can conclude as follows: (1) from the average value -the average of the initial test and posttest is 17.92 <23.33.Thus the average value of the final test is greatly increased (significantly) (2) from the increase test the value of t count is 23.48 which is in the area of higher hypothesis rejection the value of the t table is 2,500.Thus, the increase in learning outcomes in the long jump squat style by using a direct instruction model for class X students at SMK Muhammadiyah 1 Sumedang is very significant (significant).]]>
