<![CDATA[ABSTRACT Mathematical representation skills play a crucial role in enabling students to comprehend, express, and communicate mathematical ideas effectively. However, students’ representation abilities in the topic of Systems of Linear Equations in Two Variables (SLETV) remain insufficient, indicating the need for more engaging and innovative instructional approaches. This study aimed to examine the effect of a Deep Learning approach supported by Desmos on students’ mathematical representation skills. The novelty of this research lies in integrating Deep Learning principles with the Desmos application in SLETV instruction at the junior high school level. A quasi-experimental method employing a pretest–posttest control group design was implemented. The participants were 30 eighth-grade students from SMP IT Permata Hati Tebing Tinggi, divided into an experimental group and a control group. Data were collected through mathematical representation tests administered before and after the intervention and analyzed using normality, homogeneity, and Independent Samples t-tests. The findings revealed that the experimental group achieved a significantly higher posttest mean score than the control group, with t(28) = 16.8 and p < .001. These results indicate that Deep Learning instruction supported by Desmos has a positive and significant impact on enhancing students’ mathematical representation skills in learning SLETV. ABSTRAK Kemampuan representasi matematis merupakan keterampilan penting yang membantu siswa memahami dan mengomunikasikan konsep matematika. Namun, kemampuan tersebut pada materi Sistem Persamaan Linear Dua Variabel (SPLDV) masih relatif rendah sehingga diperlukan pembelajaran yang lebih inovatif. Penelitian ini bertujuan menganalisis pengaruh pembelajaran Deep Learning berbantuan Desmos terhadap kemampuan representasi matematis siswa. Kebaruan penelitian terletak pada integrasi pendekatan Deep Learning dan aplikasi Desmos pada pembelajaran SPLDV di tingkat SMP. Penelitian menggunakan metode eksperimen semu dengan desain pretest-posttest control group. Subjek penelitian terdiri atas 30 siswa kelas VIII SMP IT Permata Hati Tebing Tinggi yang dibagi ke dalam kelas eksperimen dan kelas kontrol. Data dikumpulkan melalui tes kemampuan representasi matematis sebelum dan sesudah pembelajaran, kemudian dianalisis menggunakan uji normalitas, homogenitas, dan Independent Samples T-Test. Hasil penelitian menunjukkan bahwa rata-rata posttest kelas eksperimen lebih tinggi dibandingkan kelas kontrol dengan nilai t(28) = 16,8 dan p < 0,001. Temuan ini menunjukkan bahwa pembelajaran Deep Learning berbantuan Desmos berpengaruh positif dan lebih efektif dalam meningkatkan kemampuan representasi matematis siswa pada materi SPLDV.]]>
