Abstract:

Developing students’ ability to reason about multivariable causality (MVC)—understanding how multiple interacting factors jointly produce outcomes—remains a persistent challenge in secondary science education. While interactive simulations offer powerful opportunities for inquiry-based learning, less is known about how different instructional sequences within simulation-based environments shape the development and transfer of such complex reasoning, particularly in underserved educational contexts.

This study examines how two learning designs—exploration-first (students engage with simulations prior to instruction) and instruction-first (students receive instruction before engaging with simulations)—influence students’ MVC reasoning and its transfer across contexts. Drawing on perspectives from inquiry-based learning and the Composite Instructional Design framework, the study also investigates the role of intermediate knowledge as a mechanism linking learning design and transfer outcomes.

A total of 197 secondary students from two underserved school pathways—a mainstream underserved school and a youth advancement alternative school—participated in a controlled intervention. Students were randomly assigned to one of the two learning sequences while engaging with interactive simulations on predator–prey systems and virus transmission. MVC reasoning was assessed through open-ended tasks administered at pretest, intermediate, posttest, and transfer phases. Quantitative analyses, including ANCOVA and mediation modeling, were used to examine learning gains, transfer performance, and underlying mechanisms.

Findings indicate that both learning sequences supported significant development of MVC reasoning, with no overall differences in posttest performance between approaches. Students from the alternative school pathway demonstrated greater learning gains despite lower initial performance, highlighting the potential of simulation-based environments to support underserved learners. In the transfer phase, no significant differences were found between learning sequences and schools; however, important differences emerged at the process level. Specifically, intermediate knowledge played a stronger direct and indirect role in supporting transfer in the exploration-first condition, suggesting that early exploratory engagement may shape how knowledge is structured and later applied.

These findings contribute to ongoing debates about the sequencing of exploration and instruction by showing that differences may not always appear in immediate learning outcomes, but rather in how knowledge supports transfer. For educational practice, the study highlights the importance of designing simulation-based learning environments that strategically integrate exploration and instruction to foster deeper, transferable scientific reasoning. For underserved contexts, the results suggest that well-designed technology-enhanced learning can support equitable development of complex reasoning skills across diverse educational pathways.

Keywords:

Interactive simulations, Learning Approaches, Multivariable causality reasoning, Learning transfer, Underserved students.