Enhancing students’ understanding of multivariate relationships through simulation-based learning: A quasi-experimental study

Abstract

Interpreting how multiple variables interact is often challenging for students, especially when statistical learning goes beyond calculation to explanation. In many cases, students can apply procedures correctly but still struggle to make sense of relationships among variables. Although simulation tools have been introduced to support learning, less is known about how they contribute to students’ understanding of multivariate relationships when combined with clear instructional design.

This study explored the use of a simulation based learning approach with first year pharmacy students. A quasi experimental pre test–post test design was conducted with 85 students, together with an attitudinal survey involving 229 participants. Students in the experimental group worked with a simulation environment that allowed them to explore how variables interact under different conditions, while the control group followed conventional instruction. Students who engaged with the simulation achieved higher post test scores than those receiving conventional instruction (t(83) = 2.096, p = 0.039, d = 0.46). When pre test differences were taken into account, the instructional effect remained significant (F(1,82) = 22.55, p < 0.001, η² = 0.216). Students in the experimental group also reported greater engagement and confidence in interpreting statistical relationships. Students who worked with the simulation were better able to interpret how multiple variables interacted than those in the control group. From a practical standpoint, the study suggests that using simulation based environments in statistics instruction can support learning activities that foreground interpretation and conditional reasoning rather than routine procedural work, particularly in applied contexts such as pharmacy education.