Digital Literacy dan Deep Learning Approach sebagai Determinan Utama Higher Order Thinking Skills Siswa SMK: Bukti Empiris dari Analisis SEM-PLS

Abstract

Higher Order Thinking Skills (HOTS) have become a fundamental competency in 21st-century vocational education to prepare graduates capable of critical, analytical, and creative problem-solving. However, students' HOTS remain relatively low, while previous studies have predominantly emphasized the effectiveness of instructional models, particularly Project-Based Learning (PjBL), with limited attention to learner capability as the primary determinant of HOTS. Consequently, it remains unclear which factors contribute most substantially to HOTS development. Furthermore, empirical studies integrating Digital Literacy, Deep Learning Approach, and PjBL within a single structural model are still scarce. This study employed a quantitative explanatory design involving 240 vocational high school students. Data were analyzed using Structural Equation Modeling–Partial Least Squares (SEM-PLS), while the moderating role of Deep Learning Approach was examined through the Two-Stage Moderation approach. The findings reveal that Digital Literacy and Deep Learning Approach exert significant positive effects on HOTS, whereas Project-Based Learning does not significantly influence HOTS. In addition, Deep Learning Approach fails to moderate the relationships between Project-Based Learning and HOTS as well as between Digital Literacy and HOTS. These findings indicate that learner capability, represented by Digital Literacy and Deep Learning, plays a more decisive role in developing HOTS than instructional models alone. Therefore, strengthening students' internal learning capabilities should become a strategic priority for improving HOTS in vocational education.