Learning Through Consequence: Entrepreneurial Learning in a Games-Based Simulation

Abstract

Games-based learning is increasingly embedded within entrepreneurship education; however, limited empirical attention has been given to how business simulations support entrepreneurial learning. This study examines how a business simulation operates as a learning environment in which students learn through the consequences of their decisions. A primarily quantitative, cross-sectional survey design was adopted, with a limited qualitative element provided by open-ended questions. Survey data were collected from 33 postgraduate students participating in an 18-credit business analytics simulation module using SimVenture Evolution. Most attitudinal and evaluative items used five-point response scales. Measures captured perceptions of realism, decision-making confidence, preparedness for entrepreneurial challenges, risk orientation, and mindset development, while six substantive open-ended responses provided contextual insight into decision-making and feedback. Findings indicate that students perceived the simulation as a realistic and immersive entrepreneurial environment supporting experimentation and iterative decision-making. Students reported increased confidence under uncertainty and greater awareness of risk, trade-offs, and strategic consequences. A central feature associated with perceived learning was the ability to observe and respond to decision outcomes over multiple simulated periods, enabling reflection and adaptation. One open-ended response illustrated how the absence of real-world risk could support experimentation. The findings suggest that entrepreneurial learning in business simulations may be supported by consequence-based feedback and iterative decision cycles. The study proposes the “learning through consequence loop” as an evidence-informed interpretation of how decision, consequence, feedback, reflection, and adaptation may interact in the development of decision-making capability. Given the post-only, self-reported design, the model is presented for further testing rather than as a demonstrated causal mechanism.