<![CDATA[Abstract: This research was aimed at searching and finding: 1) the most effective mathematics learning model among TPS learning model with PMR, TPS learning model, and direct learning model, 2) the level of student’s spatial intelligence having the highest achievement among students with high, average, and low spatial intelligence, 3) the most effective learning model among TPS learning model with PMR, TPS learning model, and direct learning model towards student’s achievement on each level of spatial intelligence, and 4) the level of students having the highest achievement in every learning model. This was a quasy-experimental research with a 3x3 factorial design. The population was students of grade X of state senior high school in Surakarta in 2013/2014. Stratified random sampling and cluster random sampling techniques were applied. The samples in this research were: 1) experiment group 1, consisting of 91 students; 2) experiment group 2, consisting of 90 students; 3) control group, consisting of 99. The data collecting instruments were student’s spatial intelligence test and achievement test in the form of multiple choices. Balance test with unbalanced one-way anova test, analysis prerequisite tests (normality test with Liliefors test and homogenity test with Bartlett test) and hipothesis test (unbalanced two-way anova test) were conducted. It can be concluded that: 1) TPS model with PMR is more effective towards student’s achievement than TPS model and direct model, and TPS model is as effective as direct model towards student’s achievement; 2) students with high spatial intelligence gain higher achievement than those with average and low spatial intelligence, and students with average spatial intelligence gain higher achievement than those with low spatial intelligence; 3) to students with high and average spatial intelligence, TPS model with PMR, TPS model and direct model give the same achievement. For students with low spatial intelligence, TPS model with PMR and TPS model give the same achievement, but TPS model with PMR gain higher achievement than those with direct model, and TPS model and direct model give the same achievement; and 4) dealing with TPS model with PMR, students with high and average spatial intelligence gain the same of achievement, but students with high spatial intelligence gain higher achievement than those with low spatial intelligence, and students with average and low spatial intelligence gain the same of achievement, while dealing with with TPS model, students with high and average spatial intelligence gain the same of achievement but students with high spatial intelligence gain higher than those with low spatial intelligence, while students with average spatial intelligence gain higher achievement than those with low spatial intelligence, and dealing with direct model, students with high spatial intelligence gain higher achievement than those with average and low spatial intelligence, but students with average and low spatial intelligence gain the same of achievement.Key Words: learning model, TPS, PMR, spatial intelligence ]]>
