BU Learning Analytics

Stepping Out From Emergency Remote Teaching: What Can Learning Analytics Tell Us?

Learning analytics is a young field rooted in multiple fields including business analytics, educational measurement, and educational data mining. Facilitated by the wide adoption of information technology infrastructure and rapid development of data science, it can provide actionable information for decision making of educational practitioners. It has been applied around various topics, including but not limited to learning analytics and instructional design, competence-based learning analytics, self-regulated learning analytics, interoperability and standardization, etc., most of which have a small-scale and interpretive nature (Learning Analytics: The End of the Beginning). 

COVID-19 and the Learning Analytics Landscape

While educational practitioners started to be cautiously optimistic about the potential of learning analytics, challenges from large-scale online learning striked the entire world. Since the World Health Organization declared COVID-19 a pandemic, government-mandated school closures have affected most student populations from K-12 to post-secondary levels across the world. In the United States, all 50 states closed K-12 schools and childcare centers in March 2020, followed by nearly all colleges and universities (UNESCO).

Most educational institutions adopt online learning as its primary, if not only, teaching and learning approach. Unlike online learning that is planned and designed to implement remotely from the very beginning, remote learning during the pandemic is referred to as emergency remote teaching (ERT), a temporary remote teaching solution under crisis circumstances that would otherwise be delivered face-to-face or as blended courses. Unlike other online learning programs, ERT will return to its original format immediately once the crisis or emergence has abated (Educause). Learning analytics topics are also changed by the ERT context in 2020: more learning analytics research and applications focused on student cheating, student stress and Learning management system (LMS) usage compared to the past years (Figure 1).

BU Learning Analytics
Figure 1. New research topics and applications emerged during the pandemic.

The BU Connection

Digital Learning and Innovation also developed a learning analytics platform for BU Online MBA program to support their data-informed decision making during the pandemic. We first designed a standardized learning analytics data model with data from multiple third-party vendors to maintain a stable data flow. Learning data from the learning management system, online learning discussion forum, instant messenger and Zoom are integrated in our data warehouse and then into the learning analytics dashboard. 

BU Learning Analytics
Figure 2. Online MBA Learning Analytics Data Flow

We also adopted a participatory design approach by inviting the users in the design process. After discussing with the learning facilitators, we developed and pruned analytics modules according to their needs in their daily work flow.

BU Learning Analytics
Figure 3. Trend of Usage from LMS, Discussion Forum and Zoom.

Our efforts to visualize the LMS usage are consistent with the trending topics in Figure 1, and we took one step further to build an integrated learning analytics dashboard for all primary online learning tools usage, including the trend of visits, content analysis of discussion, social network of the learning community and assessment analysis for each subgroup (Figure 3).

These information would otherwise be scattered around different platforms and hard to pull, customize and compare. Meanwhile, a personalized profile for each student is provided for the learning facilitators to view before their office hours (Figure 4). 

BU Learning Analytics
Figure 4. Personal Profile for Each Student.

From Emergency Use to Practical Use

We’ll eventually step out of the emergency remote teaching modality with an unprecedentedly large number of students and educational practitioners used to check all kinds of numbers and information with their online learning tools. Moving to a long-term paradigm for learning analytics strategy post-pandemic, we can take our experience with data architecture and participatory learning design to work with a more analytics ready educational population.


  1. Learning Analytics: The End of the Beginning
  2. UNESCO: COVID-19 impact on education
  3. The difference between emergency remote teaching and online learning.
  4. Learning Analytics: A Time to Shine.

Qiyuan LiAbout the Author: Qiyuan Li is the data modeler and developer at Educational Technology for Digital Learning & Innovation. She integrates and manages data from varied sources, and produces learning analytics dashboards to support the data-informed decision making. Qiyuan received her Ph.D. degree in Learning, Design and Technology with a minor in Statistics from the Pennsylvania State University.