Case Study: It Takes More Than Just Sunshine

One of Hardwick-Day’s California clients has a mission of serving first-generation students and students of color. Hardwick-Day analyzed hundreds of variables and combinations of variables to find out if they were meeting their students’ needs—as indicated through retention. Many of the variables were interdependent and/or variations on one another, such as the timing of applications, admissions, and deposits. This is where multivariate analysis comes in handy—to tease out the subtleties.

Overall, this college retained 80 percent of its freshman class. Since average retention for comparable institutions is around 73 percent, you could say they are on top of things, especially since race and first-generation status did not influence retention rates.

But how could they aim higher and be even more competitive? First-year GPA was the best predictor of retention. (Interestingly, since ACT conducts the annual National Collegiate Retention and Persistence to Degree Rates report and other retention research, so far Hardwick-Day has found little correlation in our analyses between ACT/SAT scores and retention, once high-school GPA is accounted for.)

We further analyzed incoming characteristics that predicted poor performance in college, and incoming characteristics that predicted retention among students with a GPA above 2.7 and for those earning a GPA below 2.7.

The following variables emerged as significant for all students:

• class load (the heavier the better);
• work-study (improved retention for students who performed well in high school while reducing it for poor performers);
• aid application status (those who indicated intent and applied retained at 82%, vs. 63% for those who did not apply. Figures for those who did not indicate intent were similarly low, whether they actually applied or not.)
• distance from campus (85% retention rate for those who live within 10 miles of campus vs. 72% for those living more than 40 miles away); and
• how many students were admitted from the same high school (88% retention for students with 13 or more fellow students from their high school, down to 74% if they are the only one).

For students earning college GPAs below 2.7, retention was highest among those from middle-income families and students who had low SAT scores. Retention was low among the poor-performing students who changed majors.

For students earning college GPAs above 2.7, retention decreased if they had high high-school GPAs and if they were gapped by $13,500 or more. Retention increased if these students made their deposit early or if they received Pell grants. Among athletes with good grades, women retained much better than their non-athlete counterparts while for men it was just the opposite.

These are just the top-line findings of an 86-page, heavily-annotated presentation which Hardwick-Day delivers on campus and which involves a give-and-take discussion among our principals and campus enrollment leaders. We hope it gives you an idea of how data analysis can help you sift through the hundreds of possible retention factors in play on your campus and zone in on particular populations with concrete interventions.

As for recommendations, our first one for this school was to reduce or limit gap in the admissions process. We also suggested they consider a policy of reducing gap for sophomores earning a certain minimum first-year GPA, both to encourage better performance and to improve retention among high-achieving students. Of course, we acknowledge the effects of this strategy on discount rate and net revenue targets; we discuss various trade offs and scenarios which challenge all of us to find “the sweet spots.”

There are also operational improvements this school can make to their early recruitment efforts in the fall to reduce their dependence on late applicants and deposits. They can intervene with non-aid applicants, and, if needed, help to walk them through the aid application process retroactively. And obviously, they can focus their interventions on students with the characteristics identified in this report.

We offer this case study as an example of how data and advice from “people who’ve been there” can help target retention interventions. There are spillover benefits for admissions as well. Hardwick-Day works with another college on the other side of town. Same city, same sunshine, but after analyzing 170 variables, first-year GPA was no where near as strong a predictor of retention for this school. Their story is their story, and it is about campus activities, high-school performance, sports, and geography. These insights gleaned from Hardwick-Day’s retention and other analyses can help you tell your recruitment and admissions story and dispel those high schoolers’ impressions that “all these colleges are alike.”