Comparative Analysis of the Feature Extraction Approaches for Predicting Learners Progress in Online Courses: MicroMasters Credential versus Traditional MOOCs

Although MicroMasters courses differ from traditional undergraduate level MOOCs in student demographics, course design, and outcomes, the various aspects of this type of program have not yet been sufficiently investigated. This study aims to pave the path towards enhancing the design of constituent courses of MicroMasters programs with the focus on the application of Machine Learning algorithms. Thereby, we use a large-scale clickstream edX database to explore the trends in the online engagement of learners in a MicroMasters program, detect clickstream events that are highly correlated with the students' progress, and investigate how the engagements differ from those in a classic individual MOOC. Contrary to the previous application of machine learning algorithms in learning analytics, we implement various well-known machine learning approaches such as stepwise regression and tree-based algorithms, evaluate their performance, and propose the best-performed approach. We elaborate on noticeable differences between the engagements of the considered two groups.

[1]  R. Tibshirani,et al.  Regression shrinkage and selection via the lasso: a retrospective , 2011 .

[2]  W. F. van Valkenburg,et al.  MicroMasters: the pursuit of the Holy Grail in online learning , 2017 .

[3]  William D. Berry,et al.  Multiple regression in practice , 1985 .

[4]  Andy Laws,et al.  Machine learning approaches to predict learning outcomes in Massive open online courses , 2017, 2017 International Joint Conference on Neural Networks (IJCNN).

[5]  Joseph Jay Williams,et al.  HarvardX and MITx: Two Years of Open Online Courses Fall 2012-Summer 2014 , 2015 .

[6]  Hansol Lee,et al.  Evaluation of Fairness Trade-offs in Predicting Student Success , 2020, ArXiv.

[7]  René F. Kizilcec,et al.  Scaling Effective Learning Strategies: Retrieval Practice and Long-Term Knowledge Retention in MOOCs , 2018, J. Learn. Anal..

[8]  Kerrie A. Douglas,et al.  Surveying Motivation and Learning Outcomes of Advanced Learners in Online Engineering Graduate MOOCs , 2020, 2020 IEEE Frontiers in Education Conference (FIE).

[9]  Meina Zhu,et al.  A systematic review of research methods and topics of the empirical MOOC literature (2014-2016) , 2018, Internet High. Educ..

[10]  Zenun Kastrati,et al.  MOOC dropout prediction using machine learning techniques: Review and research challenges , 2018, 2018 IEEE Global Engineering Education Conference (EDUCON).

[11]  Niels Pinkwart,et al.  Predicting MOOC Dropout over Weeks Using Machine Learning Methods , 2014, EMNLP 2014.

[12]  Bart Rienties,et al.  Investigating variation in learning processes in a FutureLearn MOOC , 2019, J. Comput. High. Educ..

[13]  Jeonghyun Lee,et al.  Predicting Applicant Admission Status for Georgia Tech's Online Master's in Analytics Program , 2020, L@S.

[14]  A. Atkinson.  Subset Selection in Regression , 1992 .

[15]  K. Ingolfsdottir.  Winds of Change in Higher Education. , 2016, Trends in pharmacological sciences.

[16]  Chao Li,et al.  Machine learning application in MOOCs: Dropout prediction , 2016, 2016 11th International Conference on Computer Science & Education (ICCSE).

[17]  Andy Liaw,et al.  Classification and Regression by randomForest , 2007 .

[18]  Reginald DesRoches,et al.  Identification of the Significant Uncertain Parameters in the Seismic Response of Irregular Bridges , 2017 .

[19]  Justin Reich,et al.  Evaluating access, quality, and equity in online learning: A case study of a MOOC-based blended professional degree program , 2020, Internet High. Educ..

[20]  Miguel Morales,et al.  The Experience of Designing and Developing an edX's MicroMasters Program to Develop or Reinforce the Digital Competence on Teachers , 2018, 2018 Learning With MOOCS (LWMOOCS).

[21]  Jingjing Zhang,et al.  Learner support in MOOCs: Identifying variables linked to completion , 2018, Comput. Educ..