Unpacking Students’ Modeling Practices During a Modeling-Based STEM Curriculum on Highway Route Selection: Comparing Between High- and Low-Spatial Ability Students

The intersection of modeling and STEM practices offers a promising avenue for creating an integrated STEM curriculum. However, research on designing modeling-based STEM (m-STEM) curricula is limited, particularly concerning how the curriculum affects the learning of students with different spatial abilities. This study developed a four-round modeling cycle using highway route selection as the topic to support the development of modeling practices for learners with diverse spatial abilities. Using a mixed-method research approach, this study collected and analyzed the modeling practices of 24 Taiwanese upper elementary school students with different spatial abilities by modeling practice worksheets. Further analysis of the modeling practices was conducted on students in the top third and bottom third of spatial abilities. Qualitative data, including interviews, classroom observations, and teacher reflections, were also analyzed to identify the curriculum factors influencing learning in students with different spatial abilities. Results revealed that all students increased their modeling practices from level 1 (single factor) to level 3 (relation) as the model complexity increased, indicating the effectiveness of the m-STEM curriculum. Additionally, the curriculum improved the equity of spatial ability in modeling practices. Low-spatial ability students benefited from hands-on practices and digital tools during the modeling selection phase. In contrast, high-spatial ability students benefited from analogies and experimental thinking during the model construction phase. This study highlights the potential for m-STEM curricula to promote learning equity and provides insights into effective and inclusive design practices for STEM educators.