Takeaways from the IMS 2019 Analytics Summit

Published on: December 6, 2019
Linda Feng, Software Architect

The IMS Analytics Summit was held last month, hosted by Microsoft in Redmond, Washington.  The event itself was attended by close to 150 people. My former colleague and friend, Michael Feldstein, gave the opening keynote to kick off the day. In it, he encouraged us to expand our thinking about data interoperability to include intent, that is, both pedagogical and student intent.

Every year, IMS organizes a summit about analytics and this year was no different.  In many ways, this year was a lot like the previous few years. However, it was different in some key areas. In contrast to previous years, the number of panels was far fewer and what panels there were had a deliberate focus. One notable panel covered the breadth of K-20 in the State of Georgia, with a gentleman from the Georgia Department of Education sitting on stage with representatives from key universities in Georgia, including the University System of Georgia.

This year it also seemed there was more interactivity during the sessions. In particular, I was pleased to see good discussions during the Caliper Product Steering Group listening session as well as the App Vetting Privacy panel, both of which I helped to moderate. In the afternoon, Norman Biers of Carnegie Mellon conducted an interactive learning engineering workshop, working through a mock exercise in evaluating data to inform learning design.

Previous years also had more talking and less showing. This year, I felt like we got to see some very real solutions implemented at places like UC Berkeley, Katy ISD, and Georgia State U., with tangible impacts on student success.

Dr. Jamey Hynds from Katy ISD showed several examples of how they iterated from a more complex set of dashboard visualizations to something more simplified by learning what works.

Katy ISD also used data in their district to impact social aspects of learning. They designed consistent processes for bullying investigations which allowed them to spot trends and target assistance.

Other Key Takeaways:

Intent Matters

As Michael Feldstein pointed out, intent is a critical part of capturing data for analytics. Data and communication are not the same. Pedagogical and student intent need to be better understood for data interoperability to matter.

What Questions Are We Asking?

Largely as an industry, we haven’t clearly defined what questions are we asking. Certainly there are great examples of schools that are using data to drive impactful decisions on campus, however, most of today’s data collection within the education sector results in reports related to usage and adoption. This very topic is something that the IMS Learning Data and Analytics HED Innovation Leadership Network (ILN) has begun to focus on. Rob Abel also reflects on this question in his blog post here: https://www.imsglobal.org/blog/?p=258

Georgia State is a Good Example

Tim Renick closed out the summit by describing how Georgia State University has actively used data to improve student outcomes. The progress made over the last several years there can be a great example for others. For more information on their work, please see: https://agb.org/trusteeship-article/the-transformation/


New Tools and Techniques to Improve Learning Outcomes

Norman Bier and Erin Czerwinski, of Carnegie Mellon’s Simon Initiative, emphasized the benefit of ascribing learning to observable practice. They also announced support for Caliper data so that H5P OpenEdX and SAIL Platform data can be used by the Simon Initiative LearnSphere environment. (More background and a demo of LearnSphere can be found here.)

Privacy and Security Now More Important Than Ever


Now is a great time to foster an open discussion within our community regarding ways to build in secure and ethical practices related to data handling for student data. Members of the IMS community are working on an App Vetting Rubric, which can help create a common vocabulary for institutions to communicate with their supplier partners.

All in all, the day long event was engaging and a great way for the education industry as a whole to check in on our progress along our collective analytics journey. We are further along than when we started, and collaboration through standards are critical as we iterate through patterns as an industry. I hope to share more updates with you here as they emerge. The next IMS quarterly meeting will be held next February 2020, in Atlanta Georgia, with a summit focusing on Digital Credentials. Who is planning to attend?

Linda Feng photo

Linda Feng

Software Architect

Linda Feng is a software architect at Unicon, Inc., a leading provider of education technology consulting and digital services. Linda has deep experience in student information systems (SIS) integration, open standards, and big data/ learning analytics, most recently as Senior Product Manager for Canvas SIS Integrations and Canvas Data at Instructure. Prior to Instructure, Linda held the position of software architect for Oracle's Student Products Division. In the last several years, she served as co-chair of the IMS Global Learning Initiative Learning Information Services & Privacy Working Groups, helping to bring a new Enterprise interoperability standard to market.