A university’s data-driven response to COVID
A Q&A with the CIO explores how a university is using data and analytics to navigate the challenges presented by COVID
Like with any significant moment in time, “Where were you in 2020?” will likely become a topic of conversation for higher ed leaders for years to come, quickly followed by, “What did you do?”
Beyond the comradery that comes of shared experiences, there is significant value in this reflection, not just in the future, but today as colleges and universities steel themselves for an uncertain spring and chart their path forward.
Given the complexities of COVID, developing a plan is not easy work. I’ve heard it described as we’re all weathering the same storm, but not everyone is in the same boat. While this crisis has highlighted how an institution’s level of preparedness, infrastructure, and culture impacts its ability to respond to the challenges presented by the pandemic, there are strategies producing results that are universally relevant. Hearing how others have responded can spark ideas and elevate opportunities to try similar tactics and evolve efforts.
I’m a believer in sharing information early and often to help others gain insights. After seeing how Clark University has been leveraging data to inform its response to COVID, I asked Joseph Kalinowski, the university’s CIO, to share more about their approach.
Darren Catalano: The pandemic has galvanized leaders from every corner of campus to work together to address the challenges created by COVID. The CIO’s office plays a critical role in providing key information and insights to inform planning and preparation. What has this looked like at Clark?
Kalinowski: In looking for a silver lining in all of this, our spirit of collaboration absolutely is one. As the CIO, I have been afforded the opportunity to partner with a wide range of people across the campus, usually working on problems that need a quick solution and are along a critical path. This was most true with our efforts to safely reopen our campus this fall. The quick pivot to remote learning in March, as well as efforts to minimize personal contact in the fall, led to us evaluate numerous business processes and solve a lot of challenges with technology.
We were also heavily involved in developing data management around our COVID testing operation. We developed processes to gather information from students throughout the summer about their intent to return to campus, then assigned them testing cadences based on that (our students were tested every three days). Once a student arrived and started testing, we needed to track when they were last tested, whether they received a valid result from that test, and what that result meant for their next test. Then we responded accordingly. For instance, we sent out automated notifications to students who fell out of cadence, reminding them of the urgency to resume the testing protocol. There were many variables in play to determine if a student should get a reminder. IT was employed to manage that data and communication, which required partnerships with the Dean of Students, our testing center and Marketing & Communications.
Catalano: Certainly, one of the most pressing aspects of any institution’s response has been navigating the unprecedented health and safety concerns. In what ways have you been using data analytics to support that effort?
Kalinowski: Even before we welcomed students back to campus, we were looking at publicly available data to help define our testing regimen as well as help understand and manage our risk around where students were arriving from before they aggregated in Worcester. It was important for us to consider the impact on our community of bringing thousands of students from across the world back to a central location. As I mentioned above, we use data extensively to manage our testing environment.
When it comes to contact tracing, performing it well requires a conversation. We use technology as a way to help people remember details of where they might have been or what they might have done the few days before a positive test. For example, when we begin the triage of a positive case, we inform our contact tracers about the student’s living situation (location and roommates), their course schedule, their campus work details, athletic affiliations and emergency contact information. By building a report to aggregate all that data, we’re able to provide a set of information immediately to a team of people. This process helps define the scope of the spread we may be dealing with, and allows us to get in touch with the right people as quickly as possible to minimize future spread.
Catalano: In addition to the health and safety implications, across higher ed we all have been doing our best to determine the impact of the pandemic on enrollment and retention. How has Clark approached this question? What measures have you been monitoring?
Kalinowski: We actually relied most heavily on frequent communications and intermittent surveys to gauge student intentions and plans regarding enrollment. That information activated and informed our plans and interventions accordingly. This data was then aggregated into our admissions tool to manage our incoming class.
Catalano: Is that information you already could easily access? If not, how did you gather that data?
Kalinowski: Using Qualtrics, we were able to quickly get mobile-friendly surveys into the hands of our students and import those results into other tools to manage them. In the spring, we’ll use some custom forms to bring this survey data right into Oracle for us, simplifying the merge of the data with all other information we have about the students and getting it into our enterprise analytics platform. In addition, we conducted multiple short surveys over the course of the semester to gauge reaction to our COVID response and compliance and get a sense of student, staff, and faculty’s feelings of connection to Clark. Tableau was used to help us analyze the results of these surveys.
Catalano: What did you learn? Were there any surprising insights? How has the data informed strategic decision making and actions the university has taken to advance student success and the admissions process?
Kalinowski: One piece of data that surprised us was the limited instance of on-campus transmission. We were expecting more spread of the virus on campus, specifically between roommates, and we didn’t see that. We did see an overwhelming desire for students to return to campus — as we started the fall semester, 85 percent of our undergraduates were back at Clark. This confirmed for us that while remote learning has its benefits, there is still a hunger for the traditional in-person experience, and we have to be ready to provide options in both modalities.
Catalano: To what extent did Clark already have the analytics infrastructure needed to navigate the challenges presented by the pandemic? What new capabilities did you have to cultivate quickly?
Kalinowski: Overall, we were lucky when it came to our data infrastructure. We were a couple years into our implementation of our enterprise analytics platform as well as a new operational reporting tool. Having deep experience with both of the systems allowed us to use them quickly and efficiently in response to the pandemic. What we did spend a lot of time on was developing a data model to help us administer our testing environment. We focused on advice from medical professionals on how to best use surveillance testing to minimize the risk to our community. This resulted in testing protocols that no single application could help us manage. We built data models to aggregate data from multiple systems around testing, and then tied that data to our other systems. This allowed us to set testing cadences and enforce compliance to different groups at a granular level and in a flexible way to adapt to changes over the course of the semester.