<![CDATA[Science learning in the modern era demands innovative teaching strategies that focus not only on conceptual mastery but also on developing students' science process skills and scientific attitudes. This study analyzes the effectiveness of a deep approach to learning—characterized by higher-order thinking and constructivist strategies—on these competencies in physics education. A quasi-experimental Nonequivalent Pretest-Posttest Control Group Design was employed, involving two groups of 200 students each: the experimental group received the deep learning intervention, while the control group used conventional methods. Data from written tests, observations, and attitude questionnaires were analyzed using ANCOVA to control for pretest differences between non-equivalent groups. Results showed significant improvements in the experimental group. Science process skills advanced across all indicators (observation: mean 78.12; classification: 77.45; prediction: 75.90; measurement: 76.30; concluding: 77.80; communicating: 78.50), with overall posttest mean of 76.97 (SD=7.35; t=2.95; p=0.006 <0.05) versus control's 69.37 (SD=8.12). Scientific attitudes also improved (curiosity: 77.20; openness to evidence: 76.80; objectivity: 75.90; critical thinking: 76.50; collaboration: 77.00), yielding mean 76.09 (SD=6.84; t=2.04; p=0.047 <0.05) against control's 75.00 (SD=7.20). The constructivist-based deep approach to learning effectively fosters essential scientific competencies, providing practical recommendations for meaningful physics instruction in Indonesia]]>
