This essay is excerpted from a new Chronicle special report, “The Future of Advising,” available in the Chronicle Store.
The goal of academic advising, whether it’s accomplished by faculty or professional staff advisers, is the same: to guide students through the curriculum in a manner that helps them realize their academic, career, and personal goals. At least, that is the way it is supposed to work.
All too often, we reach only the students whose outcomes will not be substantially changed by academic advising. Undergraduates who come in for advising tend to be either (a) good students wanting reaffirmation of pre-endorsed plans or (b) struggling students who have been compelled to meet with an adviser because their grade-point average has fallen below the requisite threshold. The cold, hard truth is that an institution’s completion rates are not raised by either of those two populations. Good students are very likely going to graduate with or without intervention, and the circumstances of failing students can be hard to turn around, even with assistance.
But what about the vast majority of students who don’t fall into either camp? What about those — including many low-income, first-generation, and minority students — who aren’t on track but aren’t failing either? Developing strategies to reach them is essential for any institution committed to increasing degree completion and equity.
At Georgia State University, we found a solution in data and technology. By providing academic advisers with the real-time information they need to tailor messages and advising support to a wide array of students, we have closed gaps. Technology can be impersonal, but, in this instance, data has helped us to create a personalized advising experience for every student. No more impersonal “Dear Student” communications. No more academic risks calculated solely on the blunt instrument of demographics. No more advisement just for the people who seek it out. Supporting all students works for all students.
The results have been significant. Data-driven, personal advising has been a major reason why Georgia State has increased its graduation rates and eliminated achievement gaps. Over the past 15 years, our graduation rate has gone up by more than 35 percent. And for the seventh consecutive year, we have seen no differences in graduation rates based on race or ethnicity — a rare accomplishment for a large public university.
In 2012, in partnership with the Education Advisory Board, now known as EAB, Georgia State undertook a major predictive-analytics project. Utilizing 10 years of data — nearly 144,000 student records and 2.5 million grades — the project identified 800 analytics-based academic alerts that indicate a student may be going off track for timely degree completion. The goal was to identify academic data that our advisers could use to help students progress. All of our undergraduates are monitored daily for these alerts. Last year, prompted by these notifications, our advisers had more than 106,000 in-person meetings with students.
Not only have our completion rates increased, but students are also taking less time to graduate than before. Median credit hours at graduation have dropped by nearly a semester (11 credit hours). This benefits students in terms of both time and money. We estimate that this reduction in credit hours saved $21 million for the Class of 2022, compared with the Class of 2012.
The results also confounded expectations. One of the frequent criticisms levied at an advising model based on predictive analytics is that it encourages students to move to easier majors. That has not been the case at Georgia State. On the contrary, giving students the information they need to make timely decisions has helped undergraduates from all backgrounds to succeed — especially in rigorous disciplines. Since 2011, we have increased the number of STEM degrees conferred to Black students by 158 percent, to Black male students by 216 percent, and to Hispanic students by 406 percent, far outpacing the enrollment changes for these groups.
Personal stories. Behind those numbers are plenty of individual stories. They show how data-based, predictive advising can help a student who exhibits early signs of academic trouble. For example, I met a student, the daughter of immigrants, who was working in her parents’ nail salon. At 13, she was functionally managing the salon, greeting the customers, translating English and Vietnamese, handling the financial transactions, and making sure her little brothers behaved. After her first semester at Georgia State, however, she was struggling academically and working every weekend.
One day at the nail shop, she confided to me that college was proving difficult. An academic adviser had reached out to meet, she said, and I encouraged her to do so. Her parents wanted her to be a doctor. She was a pre-med major but had earned a D in biology — a data point that would prompt a communication. Her adviser told her about academic support for STEM courses and resources for first-generation college students, and connected her to workshops on time-management and other such skills. She and her adviser also discussed how counselors might be able to help her communicate with her parents about what she needed from them to be successful in school.
In other words: When data identified that she was at academic risk, the university marshaled its resources to support her. And it was a great joy to see her graduate on time.
Internal barriers. Sometimes better data has enabled us to identify, and fix, an institutional roadblock in advising. One of the best examples involves the university’s nursing school. It has a 90-percent graduation rate, one of the highest on the campus. However, due to limited clinical placements, only 150 students can be admitted to the nursing program each year. That left almost 1,000 pre-nursing majors whom the program was unable to accommodate. Data revealed that the prerequisites for admission to the nursing program were not functioning as intended and that the university was not giving pre-nursing majors enough viable degree alternatives to pursue. The stated GPA for admission to the nursing program was 2.8. However, a review of the data indicated that students with less than a 3.75 were not competitive for admission to the program. Once that mismatch was identified, the faculty raised the GPA for admission to nursing and set standards around academic performance in first-year courses so that students could better understand — as early as that first year — their likelihood of getting into nursing, and then change majors without adding to their time to earning a degree.
The university also responded at an institutional level by developing new majors in health-oriented fields such as public health, health management, and health informatics. Many undergraduates — especially first-generation and low-income students — are not aware of the full range of career options in health. They don’t know that those jobs are plentiful or that, in some cases, they have higher starting salaries than nursing. Better data led to new programs that are thriving at Georgia State.
More-equitable outcomes. Integrating data and technology into our academic advising allows these services to operate as intended. More specifically:
- Analytics-based advising is an essential part of leveling an uneven educational playing field. Georgia State started this period with low graduation rates and significant equity gaps. Advising offices sat empty for most of the semester. There was almost no outreach. Visits were tracked with hash marks on pieces of paper. I reflect on this with shame. More than 80 percent of Georgia State’s undergraduate population is first-generation, nonwhite, and/or Pell eligible. Many of these students cannot call on an experienced network of family and friends to help them negotiate a university’s big, complicated bureaucracy. When faced with academic, financial, or personal difficulties, they often ended up leaving college with debt but no degree. We provided educational access to these students but did not have systems of support in place to help them through.
- Using data and analytics to identify students at academic risk is a much more equitable way to support them than relying on the students to know that help is available. Monitoring all students and then using data to develop personalized campaigns to improve their academic progression — before things reach a crisis point — are the kinds of interventions that can improve success outcomes. For example, our graduation rate for students with a first-year GPA of between 2.0 and 2.99 has increased by nearly 70 percent since we changed our advising.
- Colleges and universities are huge repositories of data. However, only recently are they working to systematically use the data to improve academic advising and increase student success. Are gateway and prerequisite courses adequately preparing students for subsequent classes? Are degree requirements overly complex? Are students withdrawing at an unusually high rate from courses they need? Are there unnecessary holds blocking registration? Do course-availability issues frustrate student progress? Historic failures to fix barriers to completion structure the very inequities that lead to achievement gaps.
Since deploying this strategy, Georgia State has increased the number of degrees it awards annually by 84 percent, and the number of degrees awarded to low-income and minority students has more than doubled. Good advising is really about giving students the information that they need — when they need it. Data and technology help ensure that we do a better job of accomplishing that goal.
