智慧型手機的普及，降低系統輔助教學硬體設備需求，本次實驗第一項活動設計使用MyET線上英語學習平臺練習基本會話跟讀作為第二項口說練習活動設計的模板，第二項活動參考Toastmasters的即席演講練習(Table topic)及口說講評機制設計開放式問答題並公告於Google Classroom，要求學生於課堂上完成答題並上傳錄音檔，講評者透過語音轉文字(Otter.ai)、音頻相似度比較(blue2digital.com)及文法分析系統(GEC)及人工講評等方式完成評分及線上講評給予學生反饋。 本次活動總共有30名學生參與研究，活動一，學生登入MyET實施15分鐘基本會話練習，控制組學生較實驗組學生以角色扮演的型式，多練習乙次基本會話加深印象(控制組學生共練習30分鐘)，系統評分機制讓學生即時修正口語表達上的缺點；活動二，實驗組學生收到6組與MyET練習主題相關的問答題，答題時間計40分鐘，控制組收到3組，答題時間計20分鐘，學生以錄音完成作答並上傳至Google Classroom，實驗組將於課程結束後一週內收到5分鐘左右的個別講評(Individual Evaluation)錄音檔，兩週後針對個別講評的缺失重新完成同一組題的作答；控制組則每週收到3組問答並於Google Classroom上收聽針對組內學生答題共同缺失所錄製10分鐘左右的綜合講評(General Evaluation)，兩組學生每堂課皆為60分鐘，為期六週，作答題數皆為12題，前、後測均以自我介紹方式評估學生口語表達能力，線上平臺的應用讓學生能依照個人學習進度實施答題並收聽講評，縮減傳統課堂上學生需要等待老師個別指導的時間。 研究結果顯示，控制組學生雖然在第一項活動較實驗組多複習乙次會話，但字詞句的重複練習並無法提升口語表達的流暢度及發音準確度；傳統Toastmasters於即席演講活動所使用的綜合講評方式雖然能讓參與者知道演講者的共同缺失，個別講評的設計更能提升學習成效；另外，學習過程中有部分學生因為程度較差無法獨立完成作答，授課教官安排較早完成答題的學生透過同儕協助學習策略(Peer-Assisted Learning Strategies)輔助同學完成第二階段答題，因此，此項活動設計應用於課堂學習較使用於課後練習更能提升學習效率。 ;Ever since smartphones were universally used, the usage of academic props for in-class teaching has been replaced by smartphone applications. Activity One, an online English learning platform (MyET) was implemented for follow-up dialogue practice as the template for the second activity. Activity Two, students were asked to answer open-ended questions on Google Classroom and uploaded their recorded answers. This theme-based activity design is originated from impromptu speech session of Toastmasters international (Table Topics). Evaluators utilized voice-to-text system (Otter.ai), voice similarity comparing system (blue2digital.com), and Grammar Checker to help them score and send back recorded evaluations. There were thirty students participated in this study. Activity One, students logged into MyET real-time scoring system and practiced follow-up dialogues for fifteen minutes. Students from CG would practice those dialogues twice with role-playing function to deepen their impression. Real-time scoring mechanism gave the students’ feedbacks of their oral speaking skills in terms of intonation and fluency. As for activity two, students in the experimental group were given six topics while students from control group received three topics. The formers were given forty minutes while the laters got twenty minutes to complete the questions. Students recorded their answers then uploaded on Google Classroom. The difference was that experimental group received individual evaluations for five minutes within one week after the course completed. On week three and four, they were asked to amend the same questions with those individual evaluations. Students from CG received three questions per week and listens to general evaluation for fifteen minutes on Google Classroom. The time spent of each group were both sixty minutes for six weeks. Both the numbers of open-ended questions were twelves in total. On pre-test and post-tests ,students were asked to do self-introduction as the benchmark for evaluating students′ oral proficiency. Google Classroom allowed students to answer questions and listen to the comments in accordance with their personal learning portfolio. It reduced the time the students spent awaiting for individual guidance from the teacher inside traditional classrooms. The results indicated that repeated dialogues practice with machines cannot promote students learning achievements in speaking fluency and intonation. Instead, learning behavior in either group on post-test were all better than that of pretest while students in EG who received individual evaluations (IE) performed better than those who received general evaluations (GE). We concluded that students who were given individual evaluations would show better learning achievement. In addition, the field instructor applied peer-assisted learning strategies by arranging those who completed the training faster to help their classmates. As a result, these activity designs would have better proficiency on in-class study instead of after-class study.
