Title: Quantifying the relationship between student enrollment patterns and student performance
Authors: Shahab Boumi, Adan Vela, Jacquelyn Chini
First author’s institution: University of Central Florida
Journal: arXiv pre-print 2003.10874 [open-access]
College is typically seen as a linear 4 year process, where students enter college after high school, study for four years, and then graduate. However, this is hardly the case. Many students take longer than 4 years to graduate and only 30% of students are enrolled full time throughout college. In fact, most students cannot be considered full time or part time students since their enrollment statuses change throughout their time at college.
Now, you might be thinking, what’s the big deal. After all, full time vs part time is based off an arbitrarily agreed upon number of credit hours. Yet, studies have found benefits of students enrolling full time. First year full time students are less likely to leave the university after their first year and tend to have higher GPAs. Further, enrollment status has been found to be more important than the student’s race, age, or financial aid in determining whether they stay enrolled in the university.
However, these studies have tended to ignore students with mixed enrollment strategies, in part because there isn’t a clear definition of what full time and part time students are. Should a student who enrolled in all but one of their semesters be considered full time or not? Would the found benefits of full time enrollment still extend to students like these? The authors of today’s paper wanted to find this out.
To answer their question, the authors first needed to come up with a way to determine the enrollment strategy of each student. To do so, the authors created a Hidden Markov Model, which uses only the information from the most recent time only, ignoring past events, to make a prediction. For their model, the researchers used the enrollment statuses of over 170,000 students from the University of Central Florida for each of their semesters. Based on all of the student’s enrollment patterns, the model determined whether a full time, part time, or mixed enrollment strategy could best explain how the student enrolled while in college.
As the authors note, the strategy refers to the overall pattern and hence, may not match what the student actually did for the semester. For example, a student that enrolled part time, part time, full time, part time during their first four semesters would have a part time enrollment strategy.
When the authors looked at the results of their model, they found that 75% of students maintain a constant enrollment strategy through college, with full time being the most common strategy. In addition, changing from a full time enrollment strategy to a part time enrollment strategy was the most common change, with 75% of the students who change strategies making this specific strategy change. The authors believed that this was likely due to scheduling issues where a student only needed to take an off semester course to graduate or entering the workplace through a co-op before graduating.
Next, the authors were interested in which students tend to use each enrollment strategy. They found that women are more likely to use a full time strategy than men are and lower income students are more likely to use a part time enrollment strategy than higher income students are. For other types of enrollment strategies, the authors found no differences between men and women or lower and higher income students. The enrollment strategies did not differ for students from different racial or ethnic groups. Regardless of whether the student identified as white, Black, or Hispanic slightly more than 50% used a full time enrollment strategy.
When it came to differences in outcomes based on enrollment strategies, the authors found what you might expect. Students using full time enrollment strategies had the highest GPAs, students with part time enrollment strategies had the lowest GPAs, and students using mixed enrollment strategies fell in the middle. The same was true for DFW rate, the fraction of students who earn a D or F or drop the course, except that students with part time enrollment strategies had the highest DFW rates.
Perhaps the most surprising finding was that changing from the full time enrollment strategy to the part time enrollment strategy was often a cause for concern. Students who switch from the full time enrollment strategy to the part time enrollment strategy were more likely to drop out than students who stayed full time.
Further, students who switch from a full time enrollment status to a part time enrollment status often face a grade penalty. The authors compared students in similar programs and GPA at the end of their third semester who changed from a full time enrollment strategy to a part time enrollment strategy to student who continued to use a full time enrollment strategy. They found that in the following semester, first time students who changed saw their GPA drop by -0.64 points while students who stayed saw their GPA increase by 0.44.
What’s the overall message of this work then? First, student enrollment is more complicated than full time and part time. Many students are more accurately described as mixed enrollment and change between full time and part time frequently. These students also have educational outcomes such as GPA and DFW rates that are different than students described as full time or part time.
Second, university administrators and advisers need to ensure that part time students or those who switch from full time to part time have the right resources to succeed. These students are especially likely to suffer negative educational outcomes such a low GPA or increased drop out rate. By paying attention to how students enroll in courses, we can make sure we are directing the necessary resources to the students who need them most to ensure they reach their educational goals.
I am a postdoc in education data science at the University of Michigan and the founder of PERbites. I’m interested in applying data science techniques to analyze educational datasets and improve higher education for all students
Switching from full time to part time seems likely to be a result of financial issues, especially given that lower income students were more likely to do so. Even if the university didn’t decrease aid, need might easily have increased. Did the authors mention investigating this at all?
It does not appear that the authors did investigate this. They do comment on the students who switch from full time to part time though. Most of these switches occurred during the final two semesters which the authors took to mean students were taking a few classes they missed to finish their degrees or taking classes while doing a co-op. However, you are absolutely correct that these could be financial issues as well.