About 120 college campuses across the United States are testing a new artificial intelligence tool designed to make transfer course equivalencies clearer and more standardized.
Credit transfer is a sticking point for students who move between institutions and for administrators who manage it, according to Daniel Knox, director of the Center for Data & Analytics at the National Association of Higher Education Systems (NASH), a partner organization in the CourseWise pilot. Knox, who previously served as vice provost at the State University of New York (SUNY), said assessment practices vary from campus to campus and are not always consistently followed.
CourseWise, a new tool developed by the University of California, Berkeley lab that studies computational approaches to human learning (CAHL), uses AI to learn from institutions’ past credit transfer decisions and recommend which courses at one college are most equivalent to courses at another. Administrators can then approve or decline suggestions, minimizing time spent searching through course catalogs.
THE DATA LANDSCAPE
Heather Adams, a transfer consultant at higher education consulting firm Sova Solutions who has worked for years on transfer research and policy efforts, described the nation’s higher education data landscape as “a real mess.”
“Everyone organizes it differently, everyone collects it differently, everyone defines it differently and almost no one collects it during a transfer,” she said.
Angikaar Singh Chana, chief operating officer of CourseWise, pointed out that data can be housed in different formats: for example, a course catalog in a physical book, student transcripts in PDF format that are difficult for machines to read, and past course equivalency decisions in Excel files with different column headers. While staff were previously responsible for comparing information between these systems, he said, CourseWise aims to unify data on a single screen with simple suggestions.
The process of comparing two courses and determining whether they are equivalent, called course articulation, can be subjective.
For example, Texas A&M University requires students to take a course on Texas government, according to the university’s assistant vice chancellor in charge of advising Isaiah Vance. A faculty member’s opinion on what is most important for students to take away from the class—specific events and practices of Texas government, or a broader consideration of how state governments operate—may impact whether or not credit is accepted for a prior course.
“I know Georgia has a state government class, too,” Vance said. “Yes, the exact laws are a little different, but the way state governments work in general is pretty similar no matter what state you’re in.”
Articulating courses can be time-consuming for faculty and staff, and ultimately frustrating for students who struggle to navigate or end up repeating courses with similar content, Adams said. According to a report According to the U.S. Government Accountability Office, which analyzed Department of Education data on student transfers from 2004 to 2009, on average, students who transferred lost about 43 percent of the credits they had previously earned during the transfer process.
HOW COURSEWISE WORKS
CourseWise builds on more than a decade of research conducted in the CAHL Lab on how humans trust and interact with AI-generated course recommendations. The partners say its origin in research makes it unique among education technology tools, providing more transparency than a proprietary company and ensuring its features are based on documented trends and user feedback.
“Articles were published before there was a platform or anything that people could use,” Knox said.
Zachary Pardos, the UC Berkeley lead researcher who created the tool, initially built models trained on millions of existing equivalencies and enrollment records in the SUNY system. Early versions produced 10 to 12 suggested course equivalencies and sometimes showed bad matches, according to Vance.
“I remember there were some classes where physics included woodworking classes,” he said.
As it stands, the tool offers a main suggestion for the course taught at a host institution that best matches the course a student is taking elsewhere. Through validation testing on SUNY data, the model significantly improved its suggestions.
“This was what the algorithm needed to achieve accuracy comparable to human mappings,” Knox said.
To use CourseWise, institutions upload their own articulation histories and course information. Chana helps schools organize their information to ensure it is formatted appropriately, and CourseWise has published suggested standardization practices which make the articulation data work with the platform.
At Texas A&M, three of the system’s 12 universities are preparing to move beyond the typing stage and test the tool directly, Vance said. At the university system or state level, users hope that using the tool can reduce duplicate work.
“We have 37 public universities in Texas,” he said. “If each of them reviews lessons independently, that’s a lot of wasted resources. »
THE FUTURE OF COURSEWISE
CourseWise developers are looking to expand its features in the future, including expanding the types of data schools can enter to include PDF files and developing a tool for students to help them plan their degrees.
“It kind of helps resolve that difficult handoff that sometimes happens with the board,” Pardos said. “[A student could] continue to use the planner once they are at the receiving school because the planner knows the degree requirement rules of the receiving school.
Partners testing CourseWise also see broader implications.
CourseWise data can help institutions better understand students’ academic journeys and identify trends in the courses they take, Vance said. He also sees an opportunity to expand reciprocal transfers, thinking beyond moving from two-year to four-year institutions to include how college credits would transfer to a community college, for example.
Simplifying the student experience could also have ripple effects on concerns about enrolling in higher education.
“One of the things we’ve really discovered is that speed matters. Things like admissions decisions, credit evaluations, transfer between admissions and advising, those kinds of things, take a lot of time,” Knox said. “We’ve taken processes around this stuff, say, on a campus where it would take an average of 45 days, and we’ve reduced it to 48 hours or 72 hours, and we’re seeing enrollment increase. [up].”
