Abstract
Decision making is an increasingly valued skill in both schools and workplace. Lectures and case analysis remain limited in fostering decision-making skills due to a lack of contextual simulation, insufficient interactivity, and low learning motivation. Utilizing game-based learning to develop decision-making skills may help overcome these limitations. Therefore, in this study, we designed a decision-making game with a story plot, generative AI (GAI) simulation character interaction, and a clue inference mechanism to solve the limitations. This study designed a Non-Player Character (NPC) in a simulated dialog style through GAI. Learners played the role of a police detective in the game and they could talk to the victim which played by a GAI NPC, which was scripted by our research team. They have to explore, collect, and analyze clues, and make decisions about the location of the robbers' hideout within a limited period of time. A total of 15 participants engaged in the empirical evaluation of this study. It was found that the learners had a high level of flow, and moderate anxiety, and a high level of acceptance of the game, and that they believed that the game could help them to develop their ability. In addition, 80% of these learners felt that the game experience was more anthropomorphic than common GPTs, and the dialogues were more like real-life interactions. Learners also mentioned that the clues provided by the NPC were helpful, including location characteristics, and experiences of the incident, which could help them reason during decision-making.
Author Information
Hung-Yu Chan, National Taiwan University of Science and Technology, Taiwan
Yu-Chi Chen, National Taiwan University of Science and Technology, Taiwan
Chih-Chung Chien, National Taiwan University of Science and Technology, Taiwan
Huei-Tse Hou, National Taiwan University of Science and Technology, Taiwan
Comments
Powered by WP LinkPress