Impact of a bilingual STEM activity on students’ learning performance and motivation toward STEM and English

Purpose Taiwan’s ‘Bilingual Nation 2030’ policy encourages the integration of English and Mandarin instruction in primary and secondary education. This study combined the Predict-Observe-Explain-Do (POED) inquiry model and the Content and Language Integrated Learning (CLIL) approach to design a bilingual STEM activity on ‘Sensor Lights.’ The purpose was to examine the effects of this activity on fifth-grade students’ STEM and English learning performance and learning motivation.Design This study was conducted at an elementary school in Taiwan, and purposive sampling was used to select four intact classes, totaling 95 fifth-grade students. Using intact classes, a non-equivalent group quasi-experimental design compared the experimental group (EG, n = 47, POED + CLIL) and the control group (CG, n = 48, POED only). Data on students’ STEM and English learning performance and learning motivation were collected through pretests and posttests. In addition, teacher interviews were conducted to examine the factors affecting the implementation of the bilingual STEM activity.Findings Results indicated that the EG outperformed the CG in both STEM and English learning performance. However, no significant difference in motivation was observed between the two groups. Teacher interviews identified implementation tensions involving teacher preparation, instructional resources, and institutional support.Value This study provides a specific implementation model that coordinates collaborative planning, differentiated teacher roles, and language scaffolding within POED-based bilingual STEM learning. It offers practical considerations for adapting such arrangements in other bilingual and CLIL settings, while identifying conditions that may constrain their implementation and sustainability.