NEW VERSION--Learning Analytics: Research on Student Engagement

How it begins: A Collaboration between Mailman School of Public Health and SOLER

[Video]

Student engagement is key to learning, which comes in many forms such as raising a hand and asking a question in class, skimming through an assigned reading, or reviewing old lecture slides (Johar et al., 2023). By actively interacting with instructors or instructor-provided materials, students improve their intellectual skills, practical competence, moral development, self-image, and grades (Quaye et al. 2019).

As a mediator between course content and students' learning, student engagement motivates instructors to be intentional about how they design their course content, course learning management system (CourseWorks powered by Canvas, Blackboard, Google Classroom, etc.), and learning activities to monitor and maximize engagement (Reeve, 2012) (Johar et al., 2023).

But where do instructors start in increasing student engagement in their classes?

How can instructors know what is working and what is not?

How can instructors find out which students need more of a push to engage?

If you are wondering about these questions as how to help student learning, you are not alone. Dr. Samantha Garbers, aka Sam, who is an associate professor at Columbia University, Mailman School of Public Health, had been thinking about exactly the same questions over and over again. 

The course Sam teaches is a qualitative methods course for Public Health, and part of the core curriculum. It is a hard course on statistics. Even though abundant supporting materials such as readings and videos had been provided to students, students didn’t much take advantage of them. Sam and colleagues had observed that there's a significant relationship between engaging with the course materials, and the grades on specific assessments that directly mapped onto these course materials. As an avid researcher and proponent of the scholarship of teaching and learning, Sam wanted to understand why students’ engagement with course material was low, and how to improve that. Specifically, being aware of reports on nudges, which seek to help people make beneficial changes in behavior without significantly changing incentives (Szaszi et al., 2018), she asked: Can incorporating nudges in courseworks site help my students to engage more?

With that question in mind, Sam reached out to the Science of Learning Research Initiative (SOLER), who is dedicated to support teaching faculty to attach evidence of efficacy and improvement to their work, demonstrating effective and skillful teaching. SOLER introduced Sam to learning analytics -- the collection, analysis, interpretation and communication of data about learners and their learning that provides theoretically relevant and actionable insights to enhance learning and teaching (Learning Analytics). In fact, one of SOLER’s objectives is to utilize learning analytics to identify strategies to increase student engagement. For example, by processing learning analytics data from Canvas access dataPanopto view records, and even Vocareum activity, we observe patterns to test hypotheses with a controlled experiment. This was exactly what Sam wanted.

With a research grant from SOLER, Sam together with colleague Roxanne, who is assistant dean of digital learning at Mailman School of Public Health, officially embarked on the learning analytics project with continuous support from SOLER along the way. Since its inception in 2021, this project has walked through three phases of investigations. As is the case with any scientific discoveries, the road to insights is never a straight line. Despite careful analysis of learning analytics data, phase I findings suggested that canvas nudges failed to boost student engagement (Garbers et al.,2023). The results were calling to take an unexpected turn, as Sam did. Taking a step back in phase II, Sam and colleagues interviewed students on their learning motivations and realized that student engagement is affected by learning, efficiency, academic expectations, and external rewards, based on which they classified students into learners, doers, and performers (Russell et al.,2025). Thus, in phase III, tailored nudges were designed to tap into students’ learning motivations. Learning analytics revealed significantly higher engagements among those who received them (Garbers et al.,2025). 

timeline

We also introduce follow-along tutorials to view Course Analytics for your own class(es), as well as an interactive demonstration interface to see which Canvas pages your students frequently visit and interact with. All instructions are provided below; these demonstrations are meant to be accessible and do not require prior coding knowledge.

Read the paper abstracts and access the papers below.

August 2025: Student engagement predicts favorable educational outcomes. Nudges are increasingly applied to higher education contexts to promote engagement. A randomized study in a large quantitative analysis public health course deployed nudges catered to students’ intrinsic motivations. Participants (n = 127) self-selected into motivation groups (Learners [n = 46], Doers [n = 48], or Performers [n = 33]); half were randomized to receive four nudges tailored to their motivation group. Read the full paper here.

January 2025: Learning analytics are often used as proxies for student engagement. More qualitative data on how post-secondary students engage with course elements are needed to guide the design, development and deployment of learning analytics information, particularly in the use of nudge techniques. In the context of a graduate-level quantitative course within a public health core curriculum, the following research questions were explored: What do students cite as their motivations when making decisions about whether, when or how to engage with course content and learning supports? and What are student reactions to visual prompts designed to activate these motivations? Read the full paper here.

March 2023: Digital advances in the learning space have changed the contours of student engagement as well as how it is measured. Learning management systems and other learning technologies now provide information about student behaviors with course materials in the form of learning analytics. In the context of a large, integrated and interdisciplinary Core curriculum course in a graduate school of public health, this study undertook a pilot randomized controlled trial testing the effect of providing a “behavioral nudge” in the form of digital images containing specific information derived from learning analytics about past student behaviors and performance. Read the full paper here.

How You Can Do It: Interactive Demonstrations of Course Analytics

Did you know that your students’ interactions with Canvas materials are right at your fingertips?

 Follow the GIF tutorial below to learn more about this Canvas feature. You will find out which Canvas materials your students engage with and and receive a report on students’ activities, grades, etc. We invite you to use SOLER’s interactive demonstration interface to get a more comprehensive report.

All instructions are provided and no prior coding experience is required.

Instructions for Viewing Canvas Learning Analytics

The GIFs above are organized by figures (1-7). Hover your cursor over each GIF and read the corresponding instructions below.
Figure 1: On your Canvas home page, you can access the learning analytics reports for your course by clicking on “Course Analytics” on the left-hand menu.
Figure 2: The default tab “Course Grade” in “Course Analytics” shows the average grade overall in your course, as well as on specific assignments when you hover your cursor over them.
Figure 3: “Weekly Online Activity" shows the number of times students have viewed or interacted with Canvas material each week. For more resolution, hover your cursor over the weekly data points to see how many interactions occurred in a given week.
Figure 4: Scroll down on the same page to see which Canvas pages or resources were interacted with the most or least cumulatively throughout the course.
Figure 5: In the “Students” tab, you can view the overall activity of each individual student on your Canvas, including the last view/participation, and total cumulative views and participation.
Figure 6: At SOLER, we want to do more advanced learning analytics, so we work with the raw Canvas activity data. You can get this data by going to the “Reports” tab, and getting the report from “Course Activity” by clicking “Run Report”.
Figure 7: Download the csv file, which contains students' Canvas activity from the past thirty days, and view it in applications such as Numbers or Microsoft Excel.
Finally: If you follow all of these steps, you may upload this file to our provided Python notebook (located further down the page) and analyze it yourself!

Interactive Demonstration

Try out this interactive demonstration to identify what Canvas pages your students are looking at the most frequently. 

Visit this interactive demonstration interface, then upload your course activity file (see "Instructions for Viewing Canvas Learning Analytics" Figures 6-7) and run the code. A GIF visualization of these steps is provided below.

If you would like to try out the demonstration without uploading your own course activity file, download this sample file and then upload it when prompted by the interface.