Image
Loading...
Above: Courtnea Rainey, Ph.D., Assistant Dean for Assessment and Evaluation and Leo Biggs, Ph.D., Senior Program Coordinator, begin data analysis for the Spring 2025 spotlight survey.
Though it wasn’t always the case, graduate education now evolves at a rapid pace. Students today have vastly different needs than those of the past; federal legislation continues to shape and alter the future of higher education; academic research has embraced an interdisciplinary lens in responding to the world’s increasingly complex problems.
On a regular basis, Duke graduate programs conduct self-studies to evaluate and reorient based on the current best practices in graduate education. Decisions made are data-informed, but it can be challenging for individual programs to expand their evaluative scope and consider the larger trends in graduate education across Duke. In 2025, Duke houses 82 graduate programs, and 3,500 graduate students across schools that sometimes have little to no interaction.
To answer this need and provide data to drive insights at a larger scale, the central administrative teams in The Graduate School are organizing a new data collection tool, which when paired with an automated distribution process they hope will provide greater flexibility in collecting meaningful data to improve graduate education at Duke.
A Collaborative Data Trio
In late 2024, three of The Graduate School’s staff—often called the TGS “data gurus”—joined forces to discuss how expanding existing processes, and implementing new ones, might assist the Duke graduate community with new data insights.
The existing working group includes Courtnea Rainey, Ph.D., Assistant Dean for Assessment and Evaluation, Leo Biggs, Ph.D., Senior Program Coordinator, and Joel Salgado, Senior Business Systems Analyst. Interestingly, the collaborative trio represents different units across TGS: Academic Affairs, Graduate Student Affairs, and Finance.
In their initial discussions, the group realized the need for an established workflow that would allow for data collection through surveys on a wider variety of topics in order to keep up with the rapid pace of change in higher education.
The nimbleness of this instrument, after all the work that’s been put into it, is what’s really useful now and going forward. Leadership in The Graduate School can collect targeted, useful data about a specific topic, and pretty quickly.
Leo Biggs, Ph.D., Senior Program Coordinator
Take for example, the topic of this year’s pilot survey on generative AI use. A year or two ago, generative AI use was only just becoming commonplace, and now it has infiltrated many facets of our educational—and personal—lives.
Supporting Graduate Education Behind the Scenes
Rainey, who serves as the project lead, says the goal of creating this initial survey was not only to collect meaningful data about generative AI use, but to establish an infrastructure that will allow for an instrument to collect data on any topic, and make that data available quickly—potentially with no human interaction—to stakeholders who would benefit from its use.
“How can we use this tool effectively to help make graduate education better? That’s the part in the background that nobody sees, and the part that can take the most time,” Rainey says. “When a student fills out the exit survey, the next day it’s updated in a live dashboard. That data, and the data from alumni follow-up surveys, are used in our program reviews. We realized we don’t have a tool to provide a more holistic perspective of data for students who are here now, not just those that are on their way out the door.”
The existing tools Rainey mentions, the student exit survey and alumni career outcomes surveys, have automated workflows and live dashboards that present data to programs almost in real time. The working group hopes to model this format, but with data on topics that have become significant in the broader cultural and academic conversation at the time.
A Versatile Instrument
The versatility of this project, with myriad survey topics using an established, automated workflow, is what the team says is the most valuable and useful.
“The nimbleness of this instrument, after all the work that’s been put into it, is what’s really useful now and going forward,” says Biggs. “Leadership in The Graduate School can collect targeted, useful data about a specific topic, and pretty quickly.”
Image
This year’s spotlight survey, distributed in the spring of 2025, aimed to gather basic data about generative AI use in the Duke community and in what context, such as for teaching, completing coursework, conducting research, for social activities, or for professional development. Frequency of using generative AI for specific tasks, such as for scheduling time, problem solving, or writing, was also collected.
For this year’s spotlight survey, students, faculty, and staff were included in the sample.
“We were trying to get a snapshot of how our community is using these tools and the frequency of use, and what concerns they have about generative AI,” says Rainey. “We talked to a lot of faculty and staff at Duke who have generative AI experience, and to people doing research related to generative AI just trying to understand some of the different ways they were using it so we could come up with the survey questions.”
Using their expertise in survey implementation and data collection, the working group designed the survey to be as user-friendly as possible.
“We went through a lengthy process of culling the questions we wanted to ask based on our conversations with staff and faculty,” says Biggs. “We made the survey as concise as possible to lessen survey burden.”
Salgado, who manages the majority of data collection and reporting for student records in The Graduate School, says that the true value of the process has been ironing out the process and using this survey to create the workflow that will present the data via a live dashboard.
Image
So, What Did We Find Out? And What Comes Next?
Although the data team is still reviewing the results from this year’s survey, some initial insights are providing context to AI use in the graduate school community.
Out of the survey respondents, graduate students seem to be adopting generative AI at a faster pace than faculty and staff, with 75% of graduate students reporting that they used generative AI for Duke-related work in the spring 2025 semester. Comparatively, 53% of faculty and staff reported use of generative AI.
Frequency of use also showed some differing trends: 36% of graduate students shared that they used generative AI tools daily or almost daily, compared to 22% of faculty and staff.
As the team analyzes the data from the survey, they are more attentive to how and where it will be presented. Creating an interactive, auto updating Tableau dashboard has worked in the past and will provide access to the data for members of the Duke community, including students, who may have an interest in accessing the results.