I School Team Aspires to Democratize Education Through MOOCs
Improved Personalization in MOOCs to Enhance Student Learning is the May 2018 Dr. James R. Chen Award-winning MIMS final project in the Data, Algorithms, and People track. The award went to team members Sana Iqbal, Suchismita Padhy (Suchi), Priyanka Patki, and Nianthrini Vivekanandan (Nia). The team was advised by Assistant Professor Zachary Pardos.
The research aims to understand the problems that MOOC (Massive Open Online Courses) users face, such as lack of pedagogical support, and aspires to create a solution to address those problems and to improve learning. Toward this goal, the I School team created a review page compatible with the edX platform. The page gives personalized feedback to students, guiding and motivating them like a teacher would towards the completion of their online courses.
What inspired your project?
Sana: We all wanted to work towards solving a real world problem in our final project. We believe in the strength of education as a facilitator of social development; this belief, along with our past experiences, inspired this project. Some of us have experience in teaching and understand the importance of catering to individual needs in a classroom. We had all participated in MOOCs and believe they have the (unmet) potential to provide important opportunities in education worldwide.
Nia: Each of us wanted to change or advance in our careers and started with MOOCs, but were discouraged by the lack of support they offered. We ended up dropping out of many MOOCs courses.
Sana: We aspired to work towards closing this gap in MOOCs, and ultimately towards democratizing education.
Priyanka: Poor user interface and an impersonal learning environment also impact engagement with MOOCs platforms. We approached Professor Zach Pardos, who was working with the edX platform on the application of Machine Learning in edX, to work with us on this project.
What was the timeline or process like from concept to final project?
Suchi: When we first started out, we weren’t clear on what particular problem we wanted to focus our attention on. We began by compiling the findings from past literature studies, and conducted multiples surveys and interviews to better understand user pain points. Before trying to develop a solution, we wanted to ensure that we fully understood the intricacies of this complex space, and the various actors and how they engage and interact with one another.
Priyanka: All throughout the execution of this project we have strived to include the learners (our users) at the center, as we wanted to create a thoughtful solution which would be effective for them. So everytime we came up with an idea, the next step would be to validate it from user research.
Nia: After we reached to a conclusion that a review page could be a potential solution to a learners problems, the next step was to design and develop this page. There were multiple decisions that needed to be taken, such as placement of this review page, frequency at which it would be displayed to user, and how much information to display so that it was not too overwhelming.
Sana: Our review page used machine learning models in the back-end which was developed in three different phases that built upon each other. We went through many brainstorming sessions, incremental-iterative rounds, back and forth with designs and models, multiple meetings among ourselves and with our advisor.
How did you work as a team?
Nia: Each of us had individual strengths and diverse fields and backgrounds ranging from Civil Engineering and Chemical Engineering to Electrical and Electronics Engineering, and hail from the four corners of India. The common thread was that we were passionate about doing something that matters.
Sana: We would brainstorm collectively while developing ideas and to make ideas work we assigned tasks to each team member.
Priyanka: We always provided feedback and suggestions for improvement on each others work, which resulted in multiple iterations. Questioning our assumptions was vital in gaining clarity, and bringing multiple perspectives to the table led to a more thought through outcome every time.
Suchi: Our project had many working components and that required us to prioritize the tasks, outlining to-do items for every week, and also at times we would maintain trackers of completed, in-progress and yet to start tasks. We often debated among ourselves while determining what features should make our “must-have” list (versus “nice-to haves”) and this exercise was helpful in planning and scoping what could be achieved given the time frame we had to work under.
How did your I School curriculum help prepare you for this project?
Nia: Various I School courses like Information Organization and Retrieval, User Interface Design and Development, Needs and Usability Assessment, Web Architecture, Data Mining and Analytics, Machine Learning in Education, Information Visualization and Presentation, Lean/Agile Product Management, Applied Behavioral Economics along with courses from other departments like Haas, IEOR, and EECS, were crucial in shaping our project as we directly and indirectly applied the knowledge gained to our research.
The I School curriculum prepared us to think about a problem holistically, and carefully consider the long term consequences a solution might have by examining it through different lenses, and assessing the potential social, legal, ethical, and privacy implications. – Suchi Padhy
Priyanka: During the past two years at the I School we have been encouraged to think creatively. Complementing a technical solution with user-focused designs and methodologies has the potential to create a huge social impact.
Do you have any future plans for the project?
Nia: We want to further extend our idea to incorporate learner motivations while enrolling in a course and provide topic/course suggestions based on that.
Sana: We believe motivation has a huge influence on the learning style adopted by different users and in the future when we take motivation into account, we can improve personalized recommendations by tailoring it to individual needs.
How could this project make an impact?
Nia: We think that this could potentially make the process of learning through MOOCs platform more valuable and efficient.
Priyanka: We believe that this can be done by giving the student personalized, direct feedback and suggestions to improve understanding of concepts or topics in the course.
Nia: It will also motivate a student to come back to complete the course rather than giving up.
Priyanka: In addition, this is the first time a new machine learning model was implemented in the online education space, opening new avenues in research for applying deep learning techniques in this realm. This research could also potentially prove beneficial in other situations where teachers are unable to help students individually, as we feel that personalization in MOOCs will help students to learn better by taking into consideration their educational backgrounds, strengths and weaknesses in a given topic.
Additional info to share?
Priyanka: There are many people whose input and cooperation made this project successful. First, our advisor, Professor Pardos, whose expertise and guidance was vital in shaping our project. He is going to continue working on this research study and we would encourage anyone who is interested in contributing to this project to contact him. We also want to extend our gratitude to the I School community for their support. Lastly, this project benefited greatly by the people who wished to participate in our user research activities, and we want to thank them for their time. Looking at the great work done by our peers through their final projects makes us feel proud to be part of this community.
