xAPI for Actionable Learner Data Collection

Concurrent Session 5

Session Materials

Add to My Schedule

Brief Abstract

As more learning occurs beyond the traditional LMS digital spaces, stakeholders wonder, “How do I know what students are doing online?” With xAPI and some planning, it’s possible to track learner engagement with online content like articles, videos, quizzes, Storyline objects, etc., and generate actionable data for learners, faculty, and instructional designers.


Katrina is an Instructional Designer at Penn State University and recently began her journey to obtain her doctorate in learning, design and technology. She has instructional design experience in industry, K-12 and higher education. One of her core beliefs is that technology can transform education for both learners and instructors through the careful selection and application of educational technology solutions with respect to specific course outcomes and objectives.

Additional Authors

Bryan is a programmer at Penn State and the project lead for ELMS: Learning Network, an open-source Next Generation Digital Learning Environment (NGDLE) platform. He has been building educational technology solutions since 2007 and is a tireless advocate for the transformative nature of open source communities. Bryan is a leading voice in the Drupal in Education community and has helped Penn State adopt and grow a vibrant Drupal community.

Extended Abstract

This session will begin with a brief overview of what xAPI is and how it can be used to track learner interactions in digital spaces. We will explain the necessary components for getting started with xAPI and how these statements are sent to Learning Record Stores. We will show examples from past courses of learner-generated xAPI statements and how we used that data to answer questions about the course. We will use the system we currently implement for tracking to do a live demonstration with attendee generated statements during our presentation, delivered through the system we use for tracking. We will also illustrate the differences between real-time reporting during course deployment, and why it is useful, versus overall “big picture” data after course conclusion.

After providing this foundation knowledge for attendees, our session will shift to exploring ways to generate actionable data to make meaning for different stakeholder groups from the same dataset. For example, time on task can be shown to students based on timestamps they are accessing materials, videos they are watching, links they click, and pages they view. This can then be presented relative to the time we think it takes to engage with material to suggest a degree of attention they are paying to the online materials.

This same data can be constructed for faculty to say what % of the class is paying attention and visualized per lesson. For instructional designers, we could use data about a lack of attention paid to certain videos to suggest revisions to their format, or judge whether they are valuable at all. For program managers, reviewing engagement time across a portfolio could suggest courses that need revision.

We will wrap up the session with a brief discussion of the potential xAPI has for NGDLEs as they expand and toolsets grow. xAPI and distributed data analytics help ensure “future-proof” data collection going forward.