誰會為線上開放式課程公司付費? 基於理論的預測與純數據分析預測的比較

Who is willing to pay for MOOC? Theory-based prediction V.S Pure-data prediction

何文暘(Wen-Yang He)

指導教授 ： 許裴舫

國立清華大學/科技管理學院/服務科學研究所/碩士(2020年)

摘要

大規模開放在線課程(MOOC)近年來已逐漸成熟，人們開始關注這種新型態的學習方法是否是成功的。現有文獻使用了數據分析方法來定義MOOC的成功，例如:預測學生何時輟學，預測學生學習成績和預測用戶滿意程度。但是，對於預測誰將為MOOC付費的研究很少。我們的研究建立了預測模型的研究程序，以識別願意為不同的MOOC付費方式（即一次性付費，每月付費，完成證明付費，證書證明付費）付費的客戶。
此外，一般的純數據預測分析模型要求公司收集盡可能多的數據，以建立功能強大的預測模型，這非常耗時，而且缺乏理論依據。在沒有理論依據的情形下，這種預測模型有時很難清楚地證明結果的合理性。因此，我們提出了一種基於理論的預測方法，該方法基於理論來預測使用者是否會付費。在此篇研究中，我們使用TAM(技術接受模型)建立基於理論的預測模型。結果顯示在MOOC中，基於理論的模型可以在兩個付費方式(即每月付費和證書證明付費)中獲得與純數據預測分析相同的預測能力。我們的研究還討論並了解如何基於預測結果來擴展TAM理論，以及向MOOC的公司提出意見，即不同的MOOC平台應採用不同的支付策略以最大化收益。

關鍵字

大規模開放線上課堂 ； 技術接受模型 ； 支付意願 ； 純數據 ； 基於理論

並列摘要

MOOC, massive open online course, has gradually matured in recent years. Extant literature has spent efforts on predicting the success of MOOCs, such as dropout rate, academic performance, and satisfaction, using data mining methods. However, the attention on predicting who will pay for MOOCs is scarce. Our study demonstrates a research routine to build predictive models in order to identify customers who are willing to pay for different MOOCs package (i.e. adoption, monthly, achievement, certificate). In addition, pure data mining modelling requires company to collect as much data as possible in order to build a powerful prediction model, which is time-consuming. Without a theoretical guidance, this predictive modelling sometimes is very difficult to justify the results clearly. Therefore, we propose a theory-based prediction method that builds on theory to predict future observations. Our findings show that, in MOOC, theory-based models can gain the same predictive power as pure-data models can in two payments Monthly and Achievement. Our study also discusses how we can extend the TAM theory based on predictive results as well as signals to MOOCs practitioners that different MOOCs platforms should pursue different payment strategies to maximize the revenue.

並列關鍵字

MOOC ； TAM ； Willingness to Pay ； Pure-data ； Theory-based

參考文獻

1.Astin, A. W. (1993). What matters in college? Four critical years revisited. San Fran.

2.Bayer, J., Bydzovská, H., Géryk, J., Obsivac, T., & Popelinsky, L. (2012). Predicting Drop-Out from Social Behaviour of Students. International Educational Data Mining Society.

3.Berinsky, A. J., et al. (2011). "Using Mechanical Turk as a subject recruitment tool for

experimental research."

4.Broadbent, J. (2017). Comparing online and blended learner’s self-regulated learning