| Abstract: | 過去有許多研究將數位遊戲式學習(Digital Game-Based Learning, DGBL)與英語口說結合，創造一個數位遊戲式英語口說學習環境，過去的研究表明DGBL的環境有利於學習，且為了在DGBL中也能學習英語口說，將語音辨識(Automatic Speech Recognition, ASR)功能融入系統中的做法也越來越常見，透過語音辨識的即時回饋能夠幫助學習者學習時了解自身的口說準確度。而DGBL中的遊戲特徵能夠激勵學習者學習，提高學習者的專注力，常見的遊戲特徵例如：獎勵機制，遊戲中給予學習者積分、獎章、排行榜等獎勵，能夠激勵學習者積極參與更多的學習內容。然而獎勵機制不一定總是給予學習者正面的影響，當學習者了解得到獎勵的規則時，學習者可能會為了得到獎勵而學習，不是為了提升自我，過去的研究也指出學習者在遊戲中可能得到不確定的獎勵時，會更吸引學習者學習並提升學習的持久性與參與度。
雖然DGBL對學習者有許多的正面影響，但並非適合所有的學習者，因為各個學習者之間存在的個體差異會影響學習者的學習表現及遊戲表現，而不同先備能力是影響學習的一個重要因素，過去有許多研究將英語口說能力作為劃分不同先備能力的依據，但很少將不同英語口說能力者與確定獎勵及不確定獎勵一起探討，因此將確定獎勵與不確定獎勵同時設計在DGBL中，希望在這兩種獎勵方式下，探討不同英語口說能力學習者英語口說學習成效、遊戲成效、學習感知及行為模式的影響與相關性。
本研究參考彭琮樺(2021)數位遊戲式英語口說學習系統-「傳奇島冒險」，並將確定獎勵機制與不確定獎勵機制結合至系統中，學習者在學習過程中透過角色扮演的方式與遊戲中的各個角色進行互動，藉由進行任務的過程中學習英語口說，完成挑戰任務後可以自行決定獲得確定獎勵或不確定獎勵。本研究主要目的為探討不同英語口說能力學習者在具不確定性獎勵機制下其學習成效、遊戲成效、學習感知及行為模式的影響，進一步探討不同英語口說能力學習者在具不確定性獎勵機制下其英文口說學習成效與學習感知之間的相關性。
本研究共有55名實驗者，依照前測成績分為兩組，高英語口說能力組別有26人，低英語口說能力有29人，研究工具包含英語口說學習成效測驗卷、學習感知量表及系統紀錄檔。本實驗總計花費約150分鐘，在測驗開始前，受測者須填寫英語口說能力測驗(前測)及英語程度測驗，填寫時間約30分鐘，接著進行「傳奇島冒險」學習系統操作說明約10分鐘，然後讓受測者玩遊戲90分鐘，最後進行20分鐘填寫後測， 包括英語口說能力測驗(後測)、學習感知問卷。
不同英語口說能力學習者在具不確定性獎勵機制的數位遊戲式學習環境下，本研究主要發現以下5點：(1)不同英語口說能力學習者其英語口說學習成效皆有顯著進步；在前測比較及後測比較中，高英語口說能力組高於低英語口說能力組；在進步分數的比較中，則是低英語口說能力組的進步幅度顯著大於高英語口說能力組。(2)在遊戲成效方面，高英語口說能力組在與口說相關的學習類遊戲成效(挑戰任務完成個數、選擇確定與不確定獎勵的次數、練習任務進行次數、準確率達到90%的次數等)、遊戲類遊戲成效(練習任務得到的金幣數量)及輔助類遊戲成效(查看任務列表)上顯著優於低英語口說能力組；而低英語口說能力組則在輔助類遊戲成效(查看個人檔案)顯著高於高英語口說能力組。(3)在學習感知方面，高英語口說能力組與低英語口說能力組皆無呈現顯著差異。(4)高英語口說能力組與低英語口說能力組的行為轉移大致上相同，比較不同的是高英語口說能力組較注重於與口說相關的學習上，較常出現確認口說的準確率以及查看各個任務完成的狀態的行為轉移；而低英語口說能力組較注重於在遊戲中獲得的遊戲回饋，因此較常出現查看個人檔案的行為轉移。(5)高英語口說能力組的學習成效與學習感知的有用性、易用性及整體性，皆無顯著相關性；低英語口說能力組的學習成效與學習感知的有用性、易用性及整體性呈顯著正相關。
本研究結果有助於提供未來數位遊戲式學習結合口說學習的研究者在設計遊戲中獎勵機制時的參考。此外，本研究所開發的英語口說學習系統，也能提供未來教學者應用不確定獎勵機制於不同的教學環境中。
;In the past, many studies have combined Digital Game-Based Learning (DGBL) with English speaking to create a digital game-based English speaking learning environment. Past research has shown that DGBL’s environment is conducive to learning. In order to learn spoken English in DGBL, it is becoming more and more common to integrate speech recognition (Automatic Speech Recognition, ASR) functions into the system. Real-time feedback through speech recognition can help learners understand their speaking accuracy while learning. The game features in DGBL can motivate learners to learn and improve their concentration. Common game features include: reward mechanism. The game gives learners points, medals, rankings and other rewards, which can encourage learners to actively participate in more learning content. However, reward mechanisms may not always have a positive impact on learners. When learners understand the rules for receiving rewards, learners may learn to receive rewards rather than to improve themselves. Past research has also pointed out that when learners may receive uncertain rewards in games, they will be more attracted to learning and improve learning persistence and participation.
Although DGBL has many positive impacts on learners, it is not suitable for all learners because individual differences among learners will affect learners′ learning performance and game performance. Different prior abilities are an important factor affecting learning. In the past, many studies have used English speaking ability as a basis for classifying different prerequisite abilities, but few have discussed people with different English speaking abilities together with certain rewards and uncertain rewards. Therefore, this study designs both certain rewards and uncertain rewards in DGBL. Under these two reward methods, the impact and correlation of English speaking learning effectiveness, game effectiveness, learning perception and behavior patterns of people with different English speaking abilities are explored.
This research refers to Peng Conghua′s (2021) digital game-based English speaking learning system-"Legendary Island Adventure", and combines the certain reward mechanism and the uncertain reward mechanism into the system. Learners interact with various characters in the game through role-playing during the learning process. By learning spoken English while completing the tasks, learners can decide whether to receive certain rewards or uncertain rewards after completing the challenge tasks. The main purpose of this study is to explore the impact of people with different English speaking abilities on their learning effectiveness, game effectiveness, learning perception and behavior patterns under an uncertain reward mechanism, and further explore the correlation between English speaking learning effectiveness and learning perception of people with different English speaking abilities under an uncertain reward mechanism.
There were 55 experimenters in this study. According to the pre-test scores, they were divided into two groups, with 26 people in the high English speaking ability group and 29 people in the low English speaking ability group. Research tools include English speaking learning effectiveness test papers, learning perception scales and system log files. This experiment took about 150 minutes in total. Before the test starts, the test taker must fill in the English speaking ability test (pre-test) and English proficiency test, which takes about 30 minutes. Then proceed to "Legendary Island Adventure" to learn system operation instructions for about 10 minutes. Then let the subjects play the game for 90 minutes, and finally take 20 minutes to fill in the post-test, including the English speaking ability test (post-test) and the learning perception questionnaire.
For learners with different English speaking abilities in a digital game-based learning environment with an uncertain reward mechanism, this study mainly found the following five points: (1) The English speaking learning effectiveness of learners with different English speaking abilities improved significantly. In the pre-test comparison and post-test comparison, the high English speaking ability group was higher than the low English speaking ability group. In the comparison of improvement scores, the improvement of the low English speaking ability group was greater than that of the high English speaking ability group. (2) In terms of game effectiveness, the high-English speaking ability group’s performance in oral-related learning games (the number of challenge tasks completed, the number of times to choose certain and uncertain rewards, the number of practice tasks performed, and the accuracy rate reached 90% times, etc.), game results (the number of gold coins obtained from practicing tasks) and auxiliary game results (viewing the task list) were significantly better than the low English speaking ability group. The low English speaking ability group′s performance in auxiliary games (view profile) was significantly higher than that of the high English speaking ability group. (3) In terms of learning perception, there was no significant difference between the high English speaking ability group and the low English speaking ability group. (4) The behavioral transfer between the high English speaking ability group and the low English speaking ability group was roughly the same. The different was that the high-English speaking ability group focuses more on oral-related learning. They often shifted their behavior to confirm the accuracy of spoken words and check the completion status of various tasks. The low English speaking ability group paid more attention to the game feedback obtained in the game, so the behavior shift of checking personal files was more common. (5) There was no significant correlation between the learning effectiveness of the high-English speaking ability group and the perceived usefulness, ease of use, and integrity of learning. There was a significant positive correlation between the learning effectiveness of the low English speaking ability group and the perceived usefulness, ease of use and integrity of learning.
The results of this study will help provide a reference for future researchers who combine digital game-based learning with oral learning when designing reward mechanisms in games. In addition, the English speaking learning system developed in this research can also provide future teachers with the ability to apply uncertain reward mechanisms in different teaching environments. |
