Effective data management can be a challenge for any organization. In the health-care industry, the complexity of the information and critical need for privacy make it even harder. 

A group of Master of Science in Information Management students used the skills they’ve gained at the iSchool to help a company that works with health-care providers improve its data quality. Chance Landis, Riddhi Mistry, Kayvon Tari and Tanvi Save (pictured, left to right) did the work as their Capstone project, a chance to test their skills on a real-world problem. 

The team worked with Qualtrics, which helps organizations get feedback on the services they provide. This includes supporting health-care providers’ patient satisfaction surveys, which can help providers find out where they’re doing well and where they could improve. 

A challenge Qualtrics faces is that the data it gets from customers varies in content and in quality. The company wanted to be able to look at all of the data and see how a single client stacks up against the average. Having that understanding could help Qualtrics give its health-care clients suggestions to improve their data. 

Some of the students on the team have a background in health care, which made them particularly excited to work on this project. 

“This project has honestly been one of the most valuable and inspiring experiences I’ve had during my time at the iSchool,” Save said. “Diving into health-care data sparked a genuine interest in health care. I realized how much impact data can have in improving real-world outcomes, and it’s definitely a direction I’d love to explore further in my career.”

Before the team could create a benchmarking tool, however, they had to deal with the prickly problem of patient privacy. The students needed sample data, but Qualtrics couldn’t provide them with real data due to privacy concerns. Additionally, not every health-care provider asked the same set of questions — or sometimes they asked the questions in different ways or provided answers in different formats. 

“All the different clients have different standards in their project,” Tari said. “Some have robust data sets. Some have handwritten notes. There’s a huge shift from client to client how the data will look.” 

It took creative problem solving and collaborating with their sponsor at Qualtrics, but the team created a Python script that allowed them to determine key information about various columns, such as how many characters were in the answer. Once they had that profile, they were able to create sample data. 

While the team was initially experimenting with a system to create AI generated responses, they ultimately determined it was out of scope of the project. While the feature was not included in the final version of the tool, the team gained valuable insights into developing AI driven solutions through their efforts, and they appreciated that their iSchool education had given them a background on AI and the ethics of using it. 

Using their sample data, which turned out to be a critical part of their project, the team worked to build a tool to help Qualtrics identify problems with the information it’s getting from clients. 

“I enjoyed doing data quality work, which is what I truly enjoy,” Landis said. “Good data powers people to make smart decisions.”

The team appreciated putting what they’ve learned in their coursework to use in a real-world problem. 

“On a technical level, I gained hands-on experience with Python, especially in building scalable validation systems and learning how to approach messy, real-world data,” Save said. “But more importantly, I learned how to approach problem-solving when there isn’t a clear starting point — and how to work through ambiguity in a structured and strategic way.”

Ultimately, the team worked to create a scalable tool that will fit in with the systems Qualtrics is already using.

Zoë Diener, a data quality lead at Qualtrics and also a Mid-Career MSIM student at the iSchool, worked with the group throughout the project. She works in regulatory data and says that getting good quality information from the surveys is important to Qualtrics’ clients because it can help them provide better care. It’s also tied to reimbursements for Medicaid and Medicare, and having better data can help improve reimbursements. 

“The students were well-equipped to handle the problem,” Diener said. “The project was made more difficult because we could not give them real data. They were really flexible and really easy to work with.”