Penerapan Model Pembelajaran Berdasarkan Masalah (Pbm) untuk Meningkatkan Hasil Belajar IPS Siswa Sekolah Dasar (Studi Eksperimen Kelas IV Sdn 005 Banjar Guntung Kabupaten Kuantan Singingi)

The study was motivated by the low learning outcomes of students in learning IPS. The main objective of this study was to determine differences in learning outcomes learning outcomes between classes receive instructional learning model is based on a problem with the class who received conventional learning. The benefits for the students in this study is the application of learning models based issues can help improve student learning outcomes, and can increase the understanding of the material IPS. Experimental research methods were used in this study was a pretest-posttest control Rondomized Group. This research was conducted in the second semester of the school year 2014/2015 elementary school students 005 Guntung Banjar Regency Kuantan Singingi. Subjects of research are two classes of class IVa the number of students 25 students and class IVb with the number of students 25 students. The data collected in this study is the data of student learning outcomes in social studies by way of social studies test results are used in the form of an objective matter totaling 30 items. Based on the analysis of data, obtained by the difference in improving student learning outcomes among students memoeroleh learning based on problems with students who received conventional learning. Can be seen from the increase in the average initial test results of the experimental class 44.44 into 71.48 on the final test with the average index reached 0.42 with a gain medium category, while the average results of the initial test control class 42.84 risen to 65, 08 on the final test with the index reached an average gain of 0.31 including medium category. It can be concluded that an increase in student learning outcomes in social studies teaching fourth grade students of SD Negeri 005 Banjar Guntung. IPS learning outcomes of students who use the learning model based on the problem of having a significant difference compared with students who received conventional learning. This difference detected by the test of the average difference between the index gain experimental class and control class by using the t test and obtained t_hitung 2.2 while t_tabel 2.01