Data Analytics Approaches to Understanding Student Enrollment and Learning Outcomes in an Online MicroMasters Program

Concurrent Session 2

Brief Abstract

In this presentation, we discuss examples of data analytics we used to explore trends in student enrollments, demographics, and learning outcomes in an online MicroMasters program. Based on our findings, we invite attendees to in-depth conversations about leveraging large-scale learning data to gain insights about online students’ needs and experiences.

Extended Abstract

Online micro-credential programs designed to support workforce development and lifelong learning have become increasingly popular for their flexibility and affordability. This presentation will discuss a case study of leveraging data analytics to gain insights into the enrollment trends and learning progress of students who participated in an online STEM MicroMasters program. Our case study sought to examine large data generated from this online professional program to profile students' demographics and actions from enrollment to course completion to provide insights into the patterns of students' persistence, progress, and success. Our presentation will provide valuable information to the OLC community by sharing illustrative examples of data-driven practices to understand the needs of mid-career professionals and lifelong learners.

As an overview of this STEM MicroMasters program, it is a one-year program in which learners can access graduate-level course content for free of charge, while those who pay 500 USD are eligible to earn a verified certificate for each course. Students are able to earn the final program certificate if they successfully completed and earned a verified certificate for all three of the MicroMasters program core courses. To pass the course, students’ grades must meet or exceed the passing grade, determined by an instructor, which is generally around 60 out of a total score of 100. In this program, there are two different types of students, audit and verified. Audit students refer to those who have not paid the fee for earning a certificate. Verified students are those who verified their identities and paid required fees to become eligible to earn the certificate.

Data source of our study came from an encrypted edX data package for an engineering-focused research university that offers the MicroMasters program. Collected data reflect information about students (both audit and verified students) who were enrolled in at least one of the three MicroMasters courses that opened (or will open) between 2019 and 2023. Data analysis focused on conducting descriptive statistics to explore and compare trends in student enrollment, learning progress (e.g., course completion and in-progress rates), and demographics over time. In terms of enrollment trends, we examined enrollments of audit, verified, and overall students at both the program and course level (across/within courses). For demographic information, students’ gender distribution, prior education level, and geographic distribution were specifically examined. As our main data processing and visualization tools, we used various Python packages including numpy, pandas, matplotlib, plotly, scikit-learn, and pytorch.

During the presentation, we will share some of our key findings from our analysis of the student data. With respect to enrollment trends, we observed that the number of enrollments peaked during the spring and summer of 2020, with high registrations observed particularly in one of the three core courses designed for instructing essential analytics models and methods and how to apply them appropriately. Next, even though less than 10% of the total verified students earned the program certificate (i.e., having successfully completed all three core courses), more than 60% of enrolled verified students successfully completed individual core courses. However, our findings showed that the number of students whose course progress is less than 20% is still found substantial, suggesting that providing extra support at the beginning of the course module would be beneficial. Regarding demographic trends, typical verified students came from the US and they were mostly male with an age between 30-35 and holding a bachelor’s degree, consistently across time. Also, students older than 40 were more likely to hold master's or higher degrees than those younger than 30. Overall, our data implied that successful Micromasters students are professionals who might intentionally enter the program to reskill or upskill to promote career development.

How do these findings inform future data analytics research and what are some important takeaways from this study? First, our learning analytics approach offers some possibilities to inform marketing efforts and student recruitment strategies for the online MicroMasters program. One example would be to use the data to outreach and promote the program enrollment toward a post-bachelor community including a substantial body of the alumni. Another example would be to link the data to student profiles to identify which student type will benefit most and how students will be successful in the program depending on their potential life circumstances or employment status. Additionally, our study findings inform areas of opportunities to support student engagement and success in an online micro-credential program environment. For instance, as indicated by a significant proportion of students having completed only less than 20% of the course module (i.e., students who either do it or don’t do it), certain students appear to need more assistance at earlier stages of learning. Application of learning analytics will help instructors or program staff to promptly identify at-risk students and intervene to help these students get back on track. Finally, our analysis sets the initial stage for automating ways to understand students’ learning experiences. Establishing an automated system to curate and present student trends will enable researchers to gain valuable data insights about potential needs and areas for support in an efficient manner.

Level of Participation:

In this discovery session, the presenter will be given five to seven minutes to share their ongoing work and illustrate an example of using data analytics to provide insights about student learning in the online STEM MicroMaster’s program. During this conversational presentation, the presenter will occasionally pause to invite attendees to exchange any comments or ideas with both the presenters and other members in the audience. After the presentation, the session attendees will be invited to in-depth conversations during which the presenters will guide the audience to collaboratively build ideas about the benefits and future opportunities of leveraging learning analytics toward supporting lifelong learning.

Session Goals:

Individuals attending this discovery session will be able to discuss a wide range of strategies to leverage big data generated from various online education programs that offer workforce development training or lifelong learning opportunities for effectively supporting learning, teaching, and advising. They will be able to describe currently used data analytics practices in micro-credential or other online credential programs and then pose any issues or challenges that need to be addressed to ensure that these practices will make positive and meaningful impacts.