本研究想要透過問題解決模式探討學生在學習上的行為模式，會經由討論板來收集學生討論的資料，記錄學生的討論過程。由於edx的討論板無法擷取討論數據，且有各國不同的學生線上討論，故實驗對象無法明確定義，因此需要有個平台來讓特定的實驗對象(學生)做線上討論，並且記錄學生的討論內容，於是本實驗就開發一個討論板及聊天室並擷取edx的課程到系統中，方便學生點選課程相關的討論板進行討論。 本實驗建立一個學習活動，參與者為資工系大學三年級四十四位學生，實驗時間為一百零三年三月二十四至五月二十一日，共六周的實驗周數，實驗期間，學生可以在Moocs上課並在本實驗所開發的討論板進行討論，最後收集到的文章主題篇數共128篇，而回應文章數共有384篇，其中有9篇是同時有二個編碼，接著將收集到的文章，透過問題解決模式的內容分析及後滯序列分析及分析工具，產生各個行為模式之間的頻率數據，並推導出z-score的值，最後推導成事件轉換圖。 結果會發現P1(提出問題)→P2(提供解決方法)是最有顯著性。也就是幾乎所有「提出問題」的行為，比較容易進入到「提供解決方法、提供意見」的行為；而P2(提供解決方法)→P3(討論其它人的意見)，則可以看出，回應者「提供解決方法」，而下一個回應，很容易進入到「討論回應者所給予的意見及看法」；P2(提供解決方法)→P4(產生結論)也列為顯著性之一，學生會將大家所提供的方法及回應，選擇比較好的回應並產生結論；P3(討論其它人的意見)→P3，當回應者針對其它回應者回答進行回應，下一個回應者又會針前前一個回應者回答內容產生疑惑進而提出自己的想法，代表大家都很熱烈討論。但也發現網路資源眾多學生可以快速找到答案的現象，及學生沒辦法在討論中選擇出比較好的討論結果，造成持續不斷的討論。 透過實驗結果可以推導出本實驗的行為模式是跟舊有模式相符的，以及從中發現老師可以在討論過程中介入討論，引導學生進入更深層的討論，並幫助學生從討論中產生結論，除了鼓勵學生經常使用討論板之外，也要鼓勵學生產生結論。 ;This research discusses students’ behavior model through problem-solving model to study students’ on-line learning behavior model. I collected students’ discussing data from forum and record discussing process. Edx’forum cannot gather discussing data and there are students from different countries discuss on-line. Therefore experiment targets cannot clearly be defined. There should be a platform to let experiment targets(students) discuss on-line and record students’ discussing content. Therefore this research developed a forum and chat room then gathered edx’s lessens into system. Students choice related forum to discuss. This research starts up a learning activity. Participants are 44 computer science department’ third grade students. Experiment period during 2014/03/24-5/21, total six weeks. During experimental period, students can study at Moocs and discuss at forums that developed by this research. At last, total articles collected are 128.Responses are 384, 9 are two-code simultaneously. After articles collected, get z-score value than produce frequency data between behavior models through problem-solving model’s content analysis and lag-sequential analysis. At last infer it to event transfer chart. Results reveals that P1(bring up question) →P2(provide solution) is most significant. Almost all of “bring up question” behaviors are easier to become “provide solutions, provide opinion” behaviors. P2(provide solutions) →P3(discuss others’ opinions) can discovered that respondent provides “solutions” and next response is very easy to become “discuss opinions that respondents give”. P2(provide solutions) →P4(conclusions) is also significant. Students will choose better responses and get conclusions from all of methods and responses. P3(discuss others’ opinions) →P3, while respondents answer others’ responses and next respondent also questions and provides opinion means discussions is very hot. But also find out that resources at internet are abundant and students can find out solutions quickly. And students cannot select better results in discussion, though causes constantly discussion process. Experimental results show that behaviors models are correspondents to old ones. Also discovered that teacher can discuss during discussion process and guide students to deeper discussions. Then help students conclude from discussions. Encourage students use forum often and conclude.
