RUME 28
2026
Alexandria, Virginia
Measuring Undergraduate STEM Reasoning: Graphical, Covariational and Proportional Skills as Predictors of Course Performance
Page: 591
This study examined the role of students’ graphical, covariational, and proportional (GCP) reasoning skills in predicting performance in introductory STEM courses. Specifically, it investigates whether GCP proficiency predicts overall course performance. Data from 382 students revealed significant course-level differences, F(2, 379) = 10.27, p < .001. Students in Introductory Chemistry for Engineers scored highest (EMM = 59.0), significantly outperforming peers in Biology (EMM = 49.5) and Chemistry (EMM = 43.9). GCP scores were moderately correlated with course performance (r = .56). Mixed-effects modeling confirmed that GCP scores were a strong predictor of final performance (β = 0.025, p < .001), even when accounting for variation across majors. This study has implications for STEM instructors. Assessments should be designed to evaluate students' application of knowledge, problem-solving skills, and ability to formulate coherent arguments. This would give a more comprehensive picture of a student's academic potential that goes beyond memory recall.