| 摘要: | 學術英語(Academic English)對於研究者很重要，因為它是學術界的主流語言，與一般英語不同的是，學術英語被定義為一正式的語言。然而，學術英語涉及很多面向，其中，最重要的是學術寫作，因為學生需要將研究成果以論文的方式呈現，但是有些學生需要花費較多的時間和精力來撰寫，因為他們沒有足夠的課程來引導他們撰寫論文，所以近幾年提高學術寫作的課程越來越受到重視。
在學習學術寫作中，英文文法與英文邏輯是為基礎的能力，因為前者學生需要了解語言，如句子類型和結構，而後者則是學生必須具備語言組織的能力，傳達有意義的訊息，因此有必要提高學生的英語語法和英語邏輯的寫作能力。如何增強這些能力則是一重要的議題，為了解決此問題，本研究應用數位學習工具來幫助學生學習，其中一種數位工具為電子評量系統(E-assessment)，因為此數位工具可以提供交互的學習環境，並且讓學生了解如何改進。但是，電子評量系統也存在幾個缺點，例如，缺乏立即的引導，這可能會增加學生的認知負荷，使學生降低學習表現。因此，鷹架輔助將用於電子評量系統中，讓學生獲得立即的反饋。此外，這些鷹架輔助不僅可以提高學生的學習表現，還可以讓學生解決超出其能力的問題。因此，鷹架輔助可能是有效的工具，幫助學生深入學習的方式。
另一方面，儘管有鷹架輔助的電子評量系統可以為學生提供豐富的學習環境，但仍需要考慮個體差異，因為每個學生都有不同的特質，這可能會影響他們的學習。其中，認知風格對學習有重要的影響，這是由於認知風格被定義為學生如何處理信息或如何使用信息的方式，因此認知風格是影響學生學習的人因之一。其中一種認知風格為整體型(Holism)與序列型(Serialism)，兩者在資訊處理上有所不同。因此在本研究進行了實證研究，受試者皆為北部某大專院校之學生，並探討學生使用數位工具學習英語文法和英語邏輯時，不同認知風格如何影響學生的學習策略，並分析其學習成效、學習觀感與學習行為之差異。
實驗一為探討學生使用數位工具學習英語文法，其結果顯示，整體型與序列型之學習者有不同的學習行為，如整體型學習者偏好蒐集不同性質的資訊來了解任務的概念，且也會使用不同類型的提示來幫助自己完成任務，此顯示整體型學習者傾向蒐集大量資訊，並且會同時處理這些資訊。而序列型學習者則是採取線性的學習模式，他們偏好按照步驟完成任務，因此在獲得新信息或得到錯誤答案時，會偏好從頭開始，這也顯示序列型學習者傾向專注在一特定事物。此外在學習成效上發現，兩者間沒有顯著差異，其可能為不同認知風格之學習者會選擇自己適合的問題解決方式，因此在學習成效上沒有差異。另外，關於學習觀感，兩者皆認為位置提示是最可以幫助他們解決問題，其原因為可以提供最直接的訊息，使學生快速知道錯誤的訊息。
實驗二為探討學生使用數位工具學習英語邏輯，其實驗結果顯示，整體型學習者擅長自己的學習方式，因此當他們解決問題時，他們有不同的問題解決模式，所以他們傾向於使用不同的提示來蒐集各種資訊，這些資訊可以幫助他們建構學習的概觀圖，這也可以說明整體型學習者會表現非線性的學習行為。而序列型學習者則是偏好逐步的解決問題，因此，他們在進行學習任務時無法處理過多的資訊。此外，關於學習成效上兩者間有顯著差異，整體型學習者的後測成績高於序列型學習者，其可能原因為後測沒有的鷹架輔助，因此可以說明序列型學習者在學習英與邏輯上需要更多額外的輔助。然而，在學習觀感中，多數整體型學習者認為答案提示可以幫助他們完成任務，而序列型學習者則認為位置提示可幫助他們解決問題。
綜上所述，兩個的實驗結果顯示不同認知風格在不同學習項目中，有相似的問題解決方式。整體型學習者在學習英文文法與英語邏輯，皆會表現非線性的學習方式，並傾向從蒐集大量且不同性質的資訊來建立任務的概觀圖，且他們也擅長調整自己解決問題的方式，因此可以說明整體型學習者會展現跳躍性的學習模式。然而，序列型的學習者則是偏好按部就班的完成任務，且同時只能專注於一件事情，因此在獲得新資訊時，會偏好從頭開始，而非立即的調整解決方式。因此，根據本研究之結果，未來可針對不同認知風格之偏好，開發個人化的數位學習系統。
;Academic English is important for researchers because it is a main language in academic area. Unlike general English, academic English is defined as a formal language. However, academic English involves many aspects. Among them, academic writing is the most important because students need to present their empirical results as academic papers. However, several students need to spend more time writing papers because there are no enough courses to guide them on the academic writing. Thus, the courses for academic writing have been paid more attention for recent years.
However, they would meet some problems in academic writing. One is the English grammar while the other is English logic. It might be because the former is related to the syntax of the language while the latter is related to information organization. Such abilities could make readers easily understand the content of the paper easily. Thus, how to improve students’ English grammar and English logic is important. In order to address this issue, digital learning tools are applied to facilitate students to learn English grammar and English logic. One of the digital tools is E-assessment because such digital tool can provide interactive learning environments for students so that can control their learning paths by themselves. However, there are still several disadvantages in the E-assessment. For instance, insufficient guidance would make students feel frustrated and increase their cognitive loading. Due to such disadvantages, scaffoldings are required in the E-assessment.
On the other hand, e-assessment systems with scaffoldings provide students with a rich learning environment but individual differences need to be considered. This is due to the fact that each student has different characteristics that may affect their learning. Among them, cognitive style is one of key human factors which had great effects on students’ learning. This is because cognitive style affects students how to handle information or how to use information. One of these cognitive styles is Holism and Serialism, which demonstrate different information processing patterns. Due to such differences, this research attempt to measure how students learn English grammar and English logic from a cognitive style perspective. To this end, two experiments were conducted
The participants of these two experiments are the graduated institute students who come from the north university in Taiwan. Study One is to explore how students with different cognitive styles use digital tools to learn English grammar. The results indicated that several differences exist between Holists and Serialists. Holists prefer to collect different types of information to build the global picture of the tasks. Moreover, they would use various types of scaffoldings to support themselves to complete the tasks. In other word, they could process the large information at a time. On the contrary, Serialists showed the linear learning pattern when they solved the problems. In other word, they prefer to complete the tasks step by step. In addition, there is no significant differences between these two cognitive style groups on the learning performance. It may be because learners with different cognitive styles will select their own solutions to the problems. Therefore, they got the similar scores on the learning tasks. Regarding the learning perception, both of Holists and Serialists consider the location hint is the best way to help them solve problems. The reason is that it can provide the most direct information, from which students could know the mistakes quickly.
Study Two is to investigate how students with different cognitive styles use digital tools to learn English logic. The results showed that Holists were good at adjusting their solutions when they solved the problems. In other word, they tended to use different approaches. That is the reason why the non-linear learning pattern was presented by Holists. On the contrary, Serialists would follow a sequential learning pattern when undertaking the learning tasks. Moreover, they cannot handle much information when they are completing the learning tasks. In other word, Serialists could only focus on one thing at a time. On the other hand, significant differences exist between these two cognitive style groups on the learning performance. Holists obtain higher scores than Serialists on the posttest. It might be because there is no additional support on the posttest so they could not obtain the information that they required. Regarding the learning perception, they showed the differences. Most of Holists considered that the answer hint could help them solve the problem while Serialists thought that the location hint could provide useful information for them. Although the differences existed between Holists and Serialists on the learning perception, such findings showed that both of them relied on scaffoldings.
In summary, the two empirical experiments showed that students with different cognitive styles had similar problem solving patterns in different learning contexts, including English grammar and English. Holists showed an iterative learning approach on both learning contexts. Moreover, they preferred to collect large information in order to build the global picture of subject contents and they are also good at using such information to adjust the solutions by themselves. Conversely, Serialists showed the linear learning pattern in both English contexts. In other word, they would solve the problem step by step. In other word, they could only focus on one thing at a time. More specifically, when acquiring new information, they would prefer to start from the beginning of the tasks, instead of adjusting the solution immediately. The results from the two experiments can be applied to develop a personalized digital learning system that can accommodate the preferences of different cognitive style groups in the future. |
