<![CDATA[Undifferentiated learning causes a delay in students' learning process skills so that students tend to behave negatively such as being lazy to learn, less motivated and less focused in learning. This study aims to: (1) test the effectiveness of the Predict–Observe–Explain (POE) Differentiation method on the science learning outcomes of elementary school students, (2) analyze the impact of the companion method on students' process skills and scientific attitudes, and (3) prove the significance of the influence of the Differentiation POE method compared to conventional learning methods. The research method used was explanatory quantitative with multiple linear regression design and MANOVA, involving 108 students from six elementary schools. Data were collected through process skills tests, scientific attitude questionnaires, and empirically validated POE Differentiation implementation instruments. Data analysis in this study used multiple linear regression and MANOVA. The results of the analysis showed that the Differentiated POE model had a significant effect on process skills (R² = 0.837) and scientific attitudes (R² = 0.783), as well as improving overall science learning outcomes. The implications of this study confirm that the application of differentiated POE is effective in building students' conceptual understanding, scientific ability, and character, so that it is suitable for adoption as an inquiry-based and differentiated learning strategy at the basic education level.]]>
