Effects of students’ perceived linguistic competence on 7 th -grade science learning outcomes: Engagement profiles as mediators

Students' self-perceived competence in linguistic features has been linked to increased engagement in science learning, yet the underlying pathways remain complex and insufficiently explored. This study investigated both the direct and indirect effects of perceived linguistic competence on science learning outcomes via engagement profiles among seventh-grade students. Structural equation modelling (SEM) was employed with a sample of 500 students. The SEM results revealed that perceived lexical linguistic competence (PLLC) exerted significant direct effects on cognitive (β = 0.26, p < 0.01), emotional (β = 0.25, p < 0.01), and behavioural (β = 0.27, p < 0.01) engagement, as well as on science learning outcomes (β = 0.19, p < 0.05). Perceived semantic linguistic competence (PSLC) also showed a significant direct effect on cognitive (β = 0.47, p < 0.001), emotional (β = 0.44, p < 0.001), and behavioural ­(β = 0.35, p < 0.001) engagement, while influencing science outcomes indirectly through engagement dimensions. Perceived phonological linguistic competence (PPLC) showed a moderately direct effect on behavioural engagement (β = 0.13, p < 0.05), but not on cognitive (β = 0.09, p > 0.05) or emotional (β = 0.06, p > 0.05) engagement. All engagement dimensions positively and significantly influenced science learning outcomes, with cognitive engagement showing a strong direct effect (β = 0.27, p < 0.01). The model explained 69% of the variance in science learning outcomes, signifying the mediating role of engagement profiles in the link between linguistic competence and academic achievement. These findings underline the importance of fostering students' perceived linguistic competence to strengthen engagement in science education and provide evidence-based insights for educators to design targeted instructional strategies.