ABSTRACT This study was a quasi-experimental study employing the pretest-postest nonequivalent control group design. This study used two groups: experimental group and control group. The research population comprised all year VIII students of SMPN 1 Wates consisting of seven classes. The sample selected by purpose sampling as much two classes that are VIIID and VIIIG. Mathematics instructions in VIIIG used the TGT model and that in VIIID used the conventional model. The data collecting instruments were a questionnaire of students’ learning motivation and test of mathematics problem solving abilities. To know whether there was a different effect between the group taught using the TGT that taught using model and conventional model, the researcher applied the manova test and the dependent sample was tested in order to observe the variable that contributed to the differences. The research results show that (1) there are significant differences in the TGT model and conventional model in terms of learning motivation and mathematics problem solving abilities of the students of SMPN 1 Wates, (2) the TGT model has a positive effect in terms of learning motivation and mathematics problem solving abilities of the students of SMPN 1 Wates. KEY WORDS: Teams Games Tournament (TGT) model, motivation to learn, mathematics problem solving abilities
                        
                        
                        
                        
                            
                                Copyrights © 2016