<![CDATA[This study aims to determine the effectiveness of the example non example learning model with a realistic approach in improving students' problem solving abilities. The subjects of this study were fourth grade students at SDN Ngempon 2 and fourth grade students at Ngempon 1 SDN. The type of data in this research was quantitative with an experimental method in the form of cluster random sampling. Data collection techniques in this study were tests, observations and questionnaires. Data analysis techniques in this study were the normality test, homogeneity test and to test the hypothesis using the independent sample t-test, simple linear regression and paired sample t-test. The results showed that: (1) There were differences in the use of the Example Non Example learning model with a realistic approach with the average value of the experimental class being higher than that of the control class, namely 80.00 > 73.68. As well as a significant value in the experimental class 0.011 <0.05 and in the control class 0.008 <0.05. (2) There is an effect of using the Example Non Example learning model with a realistic approach to improve problem solving skills in flat shape material for fourth grade students of SD N Ngempon 2 with values a = 11.354 and b = 0.783, so the regression equation is: ẏ = 11.354 + 0.783 x . while the F value is 5.441 and sig = 0.031 and the R square value is 0.232 = 23.2%. (3) There is an increase in the problem solving abilities of fourth grade students at SD N Ngempon 2 by applying the influence of the Example Non Example learning model with an average pretest score of 35.50 increasing to 80.00. The correlation value is 0.342. Next, the results of the paired sample test show that the significance value is 0.001.]]>
