<![CDATA[ABSTRACT Students’ learning outcomes in the Road and Bridge Construction component at vocational high schools (SMK) have not yet reached an optimal level. This condition is presumed to be related to the limited variation in instructional models used as well as differences in students’ learning interests. This study was conducted to examine the effects of implementing the SAVI learning model and Direct Instruction, students’ levels of learning interest, and the interaction between these two variables on the learning outcomes of Grade XI DPIB students at SMK Negeri 2 Medan. The study employed a quasi-experimental approach using a 2×2 factorial design and involved a comparison group that was administered both a pretest and a posttest. The research subjects were Grade XI DPIB students in the 2025/2026 academic year, selected through purposive sampling. Learning outcome data were collected through tests, while learning interest was measured using a questionnaire. Data analysis began with prerequisite testing, including tests of normality and homogeneity, followed by a two-way analysis of variance (ANOVA) at a significance level of 0.05. The results indicated that the data were normally distributed and homogeneous. Hypothesis testing showed that the calculated F value exceeded the critical F value, indicating significant differences in learning outcomes between students taught using the SAVI model and those taught using Direct Instruction. Significant differences in learning outcomes were also found when viewed in terms of students’ levels of learning interest, and an interaction was identified between the instructional model and students’ learning interest. The findings reveal that the implementation of the SAVI learning model yields more optimal results for students with high learning interest, whereas the Direct Instruction model demonstrates greater effectiveness for students with lower levels of learning interest. Therefore, alignment between the instructional model and students’ learning interest characteristics plays an important role in improving learning achievement. ABSTRAK Hasil belajar siswa pada elemen Konstruksi Jalan dan Jembatan di SMK belum optimal, hal ini diduga berkaitan dengan pemilihan model pembelajaran yang kurang variatif serta perbedaan minat belajar siswa. Penelitian ini dilakukan untuk mengkaji pengaruh penerapan model pembelajaran SAVI dan Direct Instruction, tingkat minat belajar siswa, serta interaksi antara kedua variabel tersebut terhadap hasil belajar siswa kelas XI DPIB di SMK Negeri 2 Medan. Penelitian ini dilaksanakan dengan pendekatan kuasi eksperimental, menerapkan rancangan faktorial 2×2, serta melibatkan kelompok pembanding yang diberikan tes awal dan tes akhir. Subjek penelitian adalah siswa kelas XI DPIB tahun ajaran 2025/2026 yang dipilih melalui teknik purposive sampling. Data hasil belajar diperoleh melalui tes, sedangkan minat belajar diukur menggunakan angket. Tahap analisis data dimulai dengan pengujian prasyarat berupa uji normalitas dan uji homogenitas, selanjutnya dilakukan analisis varians dua arah (ANAVA) dengan tingkat signifikansi sebesar 0,05. Hasil penelitian menunjukkan bahwa data berdistribusi normal dan homogen. Hasil pengujian hipotesis memperlihatkan bahwa nilai F hitung melampaui F tabel, sehingga dapat disimpulkan terdapat perbedaan yang bermakna pada hasil belajar siswa yang mengikuti pembelajaran dengan model SAVI dan Direct Instruction, terdapat perbedaan capaian belajar ditinjau dari tingkat minat belajar, serta ditemukan interaksi antara model pembelajaran dan minat belajar siswa. Temuan penelitian mengungkapkan bahwa penerapan model pembelajaran SAVI memberikan hasil yang lebih optimal pada siswa yang memiliki minat belajar tinggi, sementara model Direct Instruction menunjukkan efektivitas yang lebih besar pada siswa dengan tingkat minat belajar rendah. Oleh karena itu, kesesuaian antara model pembelajaran dan karakteristik minat belajar siswa berperan penting dalam upaya meningkatkan capaian hasil belajar.]]>
