<![CDATA[ABSTRACT Students’ mathematical Creative Problem Solving (CPS) ability at the junior high school level remains relatively low, as indicated by a preliminary study involving 32 seventh-grade students at SMP Negeri 3 Bae Kudus, who obtained a mean score of 13.34. This study aimed to analyze differences in CPS ability between students taught using the CPS model assisted by MathShape App and those taught using direct instruction, the proportion of students achieving learning mastery after implementation of the model, and the improvement in CPS ability before and after the intervention. This quantitative study employed a Quasi-Experimental Design with a Nonequivalent Pretest-Posttest Control Group Design. The sample consisted of 32 students from class VII D as the experimental group and 32 students from class VII E as the control group, selected through cluster random sampling. Data were collected using a CPS ability essay test and analyzed using normality and homogeneity tests, an independent samples t-test, a one-proportion z-test, and N-Gain analysis. The results showed that the CPS ability of the experimental group was significantly higher than that of the control group (p < 0.001). The proportion of students achieving classical mastery was 87.5%, while the mean CPS ability exceeded the mastery criterion of 75. The improvement in CPS ability was indicated by an N-Gain score of 0.52, categorized as moderate. These findings indicate that the CPS model assisted by MathShape App can serve as an alternative learning approach to support the development of junior high school students’ creative problem-solving abilities. ABSTRAK Kemampuan Creative Problem Solving (CPS) matematis siswa SMP masih tergolong rendah, sebagaimana ditunjukkan oleh studi pendahuluan terhadap 32 siswa kelas VII SMP Negeri 3 Bae Kudus dengan rata-rata skor 13,34. Penelitian ini bertujuan menganalisis perbedaan kemampuan CPS antara siswa yang memperoleh model CPS berbantuan MathShape App dan pembelajaran langsung, ketuntasan siswa setelah penerapan model, serta peningkatan kemampuan CPS sebelum dan sesudah pembelajaran. Penelitian menggunakan pendekatan kuantitatif dengan metode Quasi Experimental Design dan rancangan Nonequivalent Pretest-Posttest Control Group Design. Sampel terdiri atas 32 siswa kelas VII D sebagai kelompok eksperimen dan 32 siswa kelas VII E sebagai kelompok kontrol yang dipilih melalui cluster random sampling. Data dikumpulkan menggunakan tes uraian kemampuan CPS dan dianalisis dengan uji normalitas, homogenitas, independent samples t-test, uji z proporsi, dan N-Gain. Hasil penelitian menunjukkan bahwa kemampuan CPS kelompok eksperimen lebih tinggi daripada kelompok kontrol dengan nilai p < 0,001. Proporsi siswa yang mencapai ketuntasan klasikal sebesar 87,5%, sedangkan rata-rata kemampuan CPS telah melampaui kriteria ketuntasan sebesar 75. Peningkatan kemampuan CPS ditunjukkan oleh N-Gain sebesar 0,52 dalam kategori sedang. Temuan ini menunjukkan bahwa model CPS berbantuan MathShape App dapat menjadi alternatif pembelajaran untuk mendukung pengembangan kemampuan pemecahan masalah kreatif siswa SMP.]]>
