Sustainability literacy has become a critical component in advancing Education for Sustainable Development (ESD). Integrating ESD into the design of global engineering curricula involves embedding sustainability literacy as a core element. As a result, teaching and assessing sustainability knowledge in higher education institutions is vital to raising sustainability awareness and promoting ESD.
This study aims to identify the most influential predictors of sustainability literacy using a Random Forest regression model. The sample consisted of 105 undergraduate engineering students from a university in Bogotá, Colombia. Sustainability literacy was measured using the Sustainability Literacy Test (Sulitest), a validated international assessment instrument. The Random Forest Analysis included nine predictors mixing between four numerical variables such as Sustainable Humanity, Transition Towards Sustainability, Global–Local Human Interactions, and Role to Play and five categorical variables like academic program, age, years of working experience, semester, gender.
The model achieved strong predictive performance (Test R² = 0.878; RMSE = 7.94; MAE = 5.95), explaining 86.61% of variance in literacy scores. Variable importance metrics (%IncMSE and IncNodePurity) revealed that Sustainable Humanity (31.26%) and Transition Towards Sustainability (26.63%) were the most influential predictors, while demographic variables showed negligible impact. Consequently, this research demonstrated Sustainable Humanity was the strongest predictor of sustainability literacy, followed by Transition Towards and Global–Local Human, underscoring the importance of human awareness, transitional dynamics, and global-local perspectives in shaping sustainability knowledge.
Are you a researcher? Would you like to cite this paper? Visit the ASEE document repository at peer.asee.org for more tools and easy citations.