This study explores the effectiveness of a negotiation simulator bot in teaching and assessing integrative, distributive, and compatible strategies for multi-issue scorable negotiations. Students, acting as company representatives, negotiated with a bot to purchase a car by balancing priorities across eight issues of varying importance. Using the simulator, they generated three Multiple, Equivalent, and Simultaneous Offers (MESOs), while the bot provided real-time counter-MESOs, enabling iterative learning and strategy refinement. Four strategy clusters were identified using k-modes clustering, based on 733 negotiation attempts by 55 students. Cluster 2 strategies yielded the highest median joint points (Md = 24.0) and average joint points (M = 24.2), approaching the joint optimum with a focus on mutual gains. Cluster 4 adopted a competitive but less effective approach, prioritizing buyers’ outcomes (Md = 19.8, M = 18.9) at the expense of sellers, resulting in lower joint outcomes (Md = 21.6, M = 22.3). Cluster 2 achieved the highest seller points (Md = 8.95), while Cluster 3, a self-focused strategy, resulted in the lowest seller points (Md = 6.20). This study highlights the potential of computational negotiation pedagogy to enhance adaptability and scalability. Future research will assess skill retention and explore AI for dynamic learning experiences.
