Perceived competence, self-efficacy and academic engagement in active learning methodologies, an evolutionary model with panel data

The lack of empirical evidence limits the rigorous evaluation of active methodologies’ effects on students’ competencies and skills. Without a solid research base, educators struggle to design effective strategies. Although theoretical frameworks support these methods, the gap between theory and practice hinders their real-world implementation. Therefore, the aim is to study the perceived competence to carry out a task from an active learning methodology, such as the development of a project, and the effects on academic self-efficacy and academic engagement in two moments in time. The participants are 140 students attending university master’s degree in occupational risk prevention. A longitudinal model of evolution is proposed with panel data using the PLS-SEM methodology. A variance-based structural equation model is carried out. The results obtained show both direct and indirect effects of perceived competence and self-efficacy on the academic engagement of the participants. A causal mediation effect of self-efficacy is observed (β = .05, p < .001), increasing engagement at the second measurement moment (ΔR2 = .30; p < .001). From the results, it can be deduced that designing significant activities for students, linked to the professional future and addressing them from active methodologies, improves perceived competence, levels of self-efficacy and academic engagement. Theoretical and practical implications are addressed in this study.