Tracking Student Progress
Doctoral students Corin M. Campbell and Jessica Mislevy at the University of Maryland at College Park recently presented a paper on student retention and attrition at the annual meeting of the Association for Institutional Research. Their study found that the best way to predict whether a student will stay at their college or university is to simply ask them. Or, more precisely, ask them how they feel about their school. Students who express strongly positive feelings toward their college or university early in their first term are most likely to stick around. Students who express strongly negative feelings have a tendency to transfer or to drop out.
Campbell and Mislevy based their research on the theory of 'student swirl,' which frames student persistence as a circular process. The theory posits four different enrollment outcomes:
- Continuous enrollment: Remaining continuously enrolled at one school from matriculation to graduation.
- Stop-out: Dropping temporarily out of school to return later to either the same or another university.
- Transfer-out: Transferring directly from one institution to another.
- Drop-out: Leaving higher education altogether.
In order to understand how educators might predict what leads students to these different outcomes, the study focused on eight key factors that have previously been linked to student persistence:
- Student involvement and engagement
- Academic ability
- Financial constraints
- Sense of belonging
- Educational and degree aspirations
- Race and ethnicity
- Residency, or local student status
The researchers drew from two existing sets of data that had been collected on students at the University of Maryland (UM). The first is the Beginning Student Survey (BSS), which is given to first-time UM freshmen eight weeks into their fall semester. Campbell and Mislevy used data gathered in the 2002 BSS, which asked students a broad range of questions about their expectations, attitudes and behaviors. They then compared this information to data from the National Student Clearinghouse (NSC), which collects continuing collegiate enrollment and degree information, in order to connect relevant issues covered on the BSS '02 with student outcomes.
Of the 2,084 usable surveys, 51% were male, 49% were female and the mean age was 18 years old. Sixty-four percent were white, 13% were Asian American, 12% were black / African American, 6% were Hispanic, under 1% were American Indian and 5% were of unknown race or ethnicity. Sixty-eight percent of respondents had entered UM as in-state residents and 32% matriculated as out-of-state residents.
Attitude and Academic Ability Play Key Roles
Overall, Campbell and Mislevy found that respondents' general attitude toward the university had the strongest relationship to their subsequent enrollment pattern. The researchers note that this finding suggests that students can detect early in their educational experience whether their school is a good fit. This replicates previous research that connected a sense of belonging and a positive first year experience with student persistence.
Breaking the data down between men and women, Campbell and Mislevy noticed that enrollment patterns seem to be more complex for women than for men, meaning that there are more factors that may help identify who is less likely to stay continuously enrolled. Female students who lack a strong sense of future direction are at a higher risking of 'stopping out,' which may indicate that they're choosing to take time off to get a better sense of their educational goals.
The researchers also found that the higher female students scored on the academics measure, the more likely they were to stop out and transfer. They wonder what it means for both student motivation and UM that women who perceive themselves as highly academic are more at risk to transfer or simply pause their education. Interestingly, female students who indicated that UM was not their first choice school were more likely to stop out and return, but not to transfer out. Are these students still failing to get into their preferred schools, or is there something that attracts them back to the campus? Finally, non-resident female students were also at a greater risk of transferring out. The researchers suggest that this could be due to higher tuition, distance from home or difficulty settling into the campus culture.
The data for male students was much less complex. Other than general attitude and race/ethnicity, Campbell and Mislevy found only one predictive factor: study skills. The lower men scored on the study skills measures, the more likely they were to drop out. This finding is consonant with previous research that shows that academic ability matters in student retention, although it doesn't identify whether the negative relationship is due to a lack of confidence or a true lack of ability.
The study concludes with a few brief suggestions on how administrators might use this information to predict - and improve - student persistence:
- Collect data early. Most of the key issues are identifiable a mere eight weeks into a student's first term. Even schools that don't have instruments like the BSS could encourage advisors or resident assistants to query students on their feelings about the school. Information about academic ability and study skills could be gathered by looking at institutional records and ongoing student progress.
- Be proactive. Administrators should not only ask students what their attitudes are toward their institutions, but also what might enhance their experiences. This information could potentially be used to better engage both current and future students.
- Utilize your resources. Actively guiding students toward school resources can help them feel more enfranchised, get guidance toward developing a future direction and get help with academic issues and study skills.
Campbell and Mislevy do caution educators that the associations they found are not necessarily causal, suggesting that future research should be pursued to see if their recommendations have an actual impact on enrollment behavior.