The Application of Deep Learning Using the Think-Pair-Share Cooperative Learning Model in Grade X

Cut Naila Zhaafira (Universitas Syiah Kuala)
Suhartati Suhartati (Syiah Kuala University)
Dwi Fadhiliani (Syiah Kuala University)

Mosharafa: Jurnal Pendidikan Matematika
Vol. 15 No. 2 (2026): April

Abstrak Siswa sering kali mengalami kesulitan dalam menyelesaikan soal kontekstual pada materi sistem pertidaksamaan linear dua variabel akibat kurangnya pemahaman konsep yang mendalam, sehingga berdampak pada rendahnya hasil belajar matematika. Penelitian ini bertujuan untuk mengetahui hasil belajar siswa melalui penerapan deep learning dengan model kooperatif tipe Think Pair Share (TPS) dan fase TPS yang paling diminati oleh siswa. Penelitian mixed method ini menggunakan desain explanatory sequential design. Populasi penelitian adalah siswa kelas X SMAN 8 Banda Aceh dengan sampel 28 siswa kelas X IPAS-1 yang dipilih melalui purposive sampling. Data dikumpulkan melalui tes esai dan kuesioner, serta dianalisis dengan teknik statistik inferensial dan deskriptif. Hasil penelitian menunjukkan bahwa penerapan deep learning dengan model TPS dapat mencapai Kriteria Ketercapaian Tujuan Pembelajaran (KKTP) dengan rata-rata nilai siswa sebesar 74,64. Integrasi prinsip deep learning yaitu mindful, meaningful, dan joyful ke dalam tahapan TPS dapat membangun konsep secara mendalam dengan metode diskusi yang terstruktur.   Abstract Students often face difficulties solving contextual problems involving systems of linear inequalities in two variables due to a lack of in-depth conceptual understanding, resulting in low mathematics learning outcomes. This study aims to determine students’ learning outcomes through the application of deep learning using the Think-Pair-Share (TPS) cooperative learning model and identify the most preferred TPS phase. This mixed-methods study employed an explanatory sequential design. The population consisted of grade X students at SMAN 8 Banda Aceh, with a sample of 28 students from class X IPAS-1 selected through purposive sampling. Data collected via essay tests and questionnaires were analyzed using inferential and descriptive statistics. The results show that implementing deep learning using the TPS model achieved the learning mastery criteria (KKTP), with an average score of 74.64. Integrating the principles of deep learning (mindful, meaningful, and joyful) into the TPS phase can foster deep conceptual understanding through structured discussion.