<![CDATA[Abstract: The objectives of the research were to find out: (1) which one providing better mathematics learning achievement the cooperative learning model of the jigsaw type, TGT or direct learning model, (2) in the student mathematical reasoning abilities, which one having better mathematics learning achievement, the students with the high, moderate or low mathematical reasoning abilities, (3) in each learning models (jigsaw, TGT or direct learning model) which one providing better mathematics learning achievement, the students with the high, moderate or low mathematical reasoning abilities, (4) in each student mathematical reasoning abilities (high, moderate, or low) which one providing better mathematics learning achievement, the cooperative learning model of the jigsaw type, TGT, and the direct learning model. This research used the quasi experimental research. Its population was all of the students in Grade VIII of State Junior Secondary Schools in Metro City. The samples of the research were taken by using the stratified cluster random sampling technique and consisted of 243 students. The instruments to collect the data were test of Mathematics learning achievement and test of achievement in mathematical reasoning abilities. Technique of analyzing data that used was unbalanced two way analysis of variance. The results of the research were as follows. 1) The cooperative learning model of the jigsaw type results better Mathematics learning achievement than the cooperative learning model of the TGT type and the direct learning model, the cooperative learning model of the TGT type results better learning achievement in Mathematics than the direct learning model. 2) The students with the high mathematical reasoning abilities have better learning achievement in Mathematics than those with the moderate mathematical reasoning abilities and those with the low mathematical reasoning abilities, the students with the moderate mathematical reasoning abilities have better learning achievement in Mathematics than those with the low mathematical reasoning abilities. 3) In the cooperative learning model of the jigsaw type and TGT, the students with the high mathematical reasoning abilities have same achievement in Mathematics as those with the moderate mathematical reasoning abilities and those with the low mathematical reasoning abilities. In the direct learning model, the students with the high mathematical reasoning abilities have better learning achievement in Mathematics than those with the low mathematical reasoning abilities. 4) In each of the mathematical reasoning abilities of students which are high and moderate, the cooperative learning model of the jigsaw type results same learning achievement in Mathematics as the cooperative learning model of the TGT type and the direct learning model. In addition, in the low mathematical reasoning abilities, the cooperative learning model of the jigsaw type results better learning achievement in Mathematics than the direct learning model.Keywords: Jigsaw, TGT, Direct Learning, and Mathematical Reasoning Abilities]]>
