近年來，遊戲式學習已經被廣泛地應用在教育領域上，並且也被視為一個有效的教學方法，因為其擁有了許多傳統教學所沒有的優點。然而，先前的研究也指出雖然遊戲式學習擁有了許多優勢，但是同樣也存在一些問題，例如遊戲式學習可能會增加某些學生在學習時的認知負荷，所以並非所有學生都可以克服這些問題，因此，遊戲式學習是否適合所有的學習者還是一個尚待研究的議題。有鑑於此，本研究之目的乃係探討人因(Human factors)如何影響學生對於具有不同設計的數位遊戲式學習環境所產生的反應，為了達到此目的，本研究進行了兩個子研究。在研究一中，探討了人因(性別差異、先備知識及認知風格)如何影響學生一般的遊戲偏好，而研究二則是著重在人因如何影響學生對於兩款具有不同設計特點的教育遊戲的反應，兩款遊戲分別為Machinarium (機械迷城) 和CSI: WebAdventures。本研究使用資料探勘技術來進行資料分析，以發現人因和學生對於遊戲式學習的反應之間的關係。研究結果指出，人因確實在遊戲式學習中扮演了影響學生反應的角色，例如，具有較高的遊戲經驗和遊戲頻率的學生會忽略遊戲中所提供的文字提示，而具有較低的遊戲經驗和遊戲頻率的學生則會需要更多的提示來幫助他們進行遊戲。於本篇論文的最後，也建立了一個圖形化的模型來總結本研究的發現，此圖形化模型可以幫助遊戲設計者和教師根據每一個學生的特質選擇適合的教育遊戲，以達到適性化學習的目標。In the past decades, game-based learning has been widely applied by educational institutes as an effective learning tool because it can provide many benefits to students’learning. However, previous researches also indicated that game-based learning possesses some problems, such as the increase of learners’ cognitive overload. However, not all of learners can overcome these problems. Thus, whether game-based learning approach is suitable for all learners is inconclusive. In this vein, the research presented in this dissertation aims to examine how human factors affect learners’ responses to design features provided by game-based learning by conducting two empirical studies. Study One examines the influences of human factors (i.e., genderdifferences, prior knowledge and cognitive styles) on learners’ preferences for general game design features while Study Two investigates how human factors influencelearners’ reactions to design features provided by two particular games, i.e. Machinarium and CSI: Web Adventures. A data mining approach was applied to explore the relationship between human-factors and learners’ reactions in this research. The results showed that human factors play an important role in the learners’ reactions to game-based leaning. In brief, males tended to play a game with a highlevel of difficulties and ignored additional support provided by games while females showed an opposite preference. Moreover, learners with high levels of FPG and EPG attempted to ignore the text-based hints in games while learners with low levels of FPG and EPG need more hints to help them play the game. Finally, a framework was developed to summarize the findings of these two studies. The ramework can be applied to support designers and instructors to choose suitable design features for each learner based on his/her characteristics.
