<![CDATA[Lack of student learning outcomes in science subjects especially physics due to the lack of involvement or the involvement of the student during the learning process. Therefore, research aimed at improving learning outcomes through learning generative models with the contextual approach. The specific objectives of the study to describe: (1) lesson plan implementation, (2) student learning outcomes, (3) student misconceptions, (4) students' response to learning generative models with the contextual approach. This research is a model TOD Hopkins consisting of three cycles. The results showed an increase (1) in the first cycle lesson plan implementation meeting 1sebesar 87% (good), the first cycle of meetings 2 by 81% (good), the second cycle is 95% (very good), the third cycle of 96.00% ( very good), (2) mastery learning outcomes of students in the classical cycle I encounter 1sebesar 54.05% (not finished), I cycle 2 meeting at 81.57% (complete), on the second cycle of 83.78% (complete) and the third cycle of 86.48% (complete), (3) misconceptions cycle I for one meeting at 27.02%, the first cycle for the meeting 2 at 52.26%, on the second cycle of 18.91%, and cycle III amounted to 13.51%, most students are difficult to analyze sums and examples, (4) students' response to generative learning contextual learning approach, where the average attention (attention) amounted to 79.31% of students with a category quite well, relavace (linkage ) amounted to 77.93% of students with good category, confidence (confidence) of 71.60% of students with both categories, and satisfaction (satisfaction) for 74.13% of students with either category. Retrieved conclusion that through learning generative models with the contextual approach to improving student learning outcomes of class VII H SMP Negeri 9 Banjarmasin on the subject of heat. ]]>
