<![CDATA[This research aims to 1) Know the application of the use of the Creative Problem Solving model to improve students' creative thinking skills and to 2) Know the effect of using the Creative Problem Solving model to improve students' creative thinking skills at SMA Pertiwi 1 Padang. The research method used in this research is the Quasi Experiment Method with the research design of Nonequivalent Control Group Design. Sampling was done by Random Sampling technique. The samples in this study were students of class X.1 as an experimental class using the Creative Problem Solving learning model and class X.3 as a control class using Conventional learning media. Data collection uses pretest and posttest instruments in the form of multiple choices that have been tested for validity, reliability. The results of the study explained that 1) The application of the Creative Problem Solving model to improve students' creative thinking skills at Pertiwi 1 Padang High School in class X.1 as an experimental class is declared quite effective to be applied, this is evidenced by documentary evidence and observations related to the activeness of students in applying the Creative Problem Solving model, which of the 4 indicators of creative thinking ability, there are 3 indicators that students understand best including Fluency (the ability to express many ideas fluently and confidently related to atmospheric problems), flexibility (diversity, being able to produce many diverse ideas regarding solutions to reduce the impact of air pollution in the surrounding environment), originality (creating an answer from the direction of one's own views related to atmospheric problems in the environment) The success of this application is supported by school facilities and infrastructure such as internet connections (Wi-Fi), and students are allowed to use laptops and mobile phones. 2) The effect of the Creative Problem Solving learning model on students' creative thinking skills in class X.1 SMA Pertiwi 1 Padang with the average posttest of the experimental class of 87.24 and the average posttest of class X.3 as the control class of 71.94. Data analysis using the Independent Sample T-Test test from the calculated data obtained a significance level (Sig.) of 0.003 because the significance is smaller than 0.05 (0.003 <0.05), then Ho is rejected and Ha is accepted.]]>
