This study examined the effect of game-based design strategies and auxiliary mechanisms on learners with varied cognitive styles (field dependence-independence, FD-I). It carried out an empirical study on the sixth graders of an elementary school in southern Taiwan. The following research questions were investigated through analyzing gaming behavioral data: (1) With respect to the improvement of reasoning ability, has the reasoning ability of the experimental group significantly improved compared to that of the control group? (2) Based on the game design mechanisms used in this study, does the design mechanisms of adaptive reasoning games be able to correspond with the gaming behaviors of FD-I learners? The results showed that in the reasoning ability pre-tests and post-tests, there was significant improvement in the reasoning ability of the FD (N=30, p=.017<.05) and FI (N=29, p=.014<.05) groups that had participated in the reasoning activities, while that of those who did not take part in the activities showed no significant improvement. These results confirm that the main principle, FD-I-oriented design strategy, formulated in planning and developing the reasoning game of “Dio's Wonderland Adventure” have substantial contribution in improving students' reasoning ability. With respect to the design strategies, they were mainly based on the behavioral preferences of FD-I students in information acceptance, information experience, and organizing information; and the seven design strategies were employed to develop a game model suitable for FI and FD.