<![CDATA[This study aims to see the high level of learning achievement of second year students of SMPN 3 Singosari using the REACT learning model compared to the discovery learning model on material pressure and its application in everyday life. The problem experienced by students in understanding material pressure material that is abstract is one of the factors that influence the low learning achievement. Students also have difficulty connecting their concepts with real-world phenomena. Therefore, we need a meaningful learning reform to improve learning achievement. One of the reforms is to choose a contextual learning model REACT to overcome these problems. The quasi-experimental method was used for this study, with the Pretest-Posttest Control Group Design in the experimental class and the control class. The research population is second year students of SMPN 3 Singosari 2018/2019 academic year consisting of 9 classes. The technique for sampling in this study is purposive sampling. The sample chosen to be the subject of the study is SMPN 3 Singosari students in the 2018/2019 academic year class VIII I as an experimental class and class VIII H as a control class, with 31 students per class. Analysis of students' initial abilities through pretest results, while high or not student achievement is seen through posttest. The T test gives the results of > = 2.3057 > 1.6706, which indicates an inequality in learning achievement, between students learning through REACT and conventional learning models. After that, Tukey test was conducted to produce the value of > = 3.289 > 2.89, higher learning achievement using the REACT model compared to conventional models. The REACT model is effective as an effort to solve the problem of low learning achievement.]]>
