Making Friends Makes the Biggest Difference In Student Retention, According to New Rhodes Study
Publication Date: 12/13/2011
Numerous studies have assessed reasons why students drop out of college, with most measuring socio-economic status, academic ability, ethnicity, college costs, financial aid and some even examining the number of activities students participate in.
But a new Rhodes College study has found that one’s social network in college is the most powerful factor in student attrition and retention. The study—conducted by Dr. Jay Eckles ’00, Rhodes director of information services, and Eric Stradley ’11—recently was published in Social Psychology of Education and is mentioned today in Inside Higher Education.
Eckles and Stradley studied the Rhodes Class of 2012 to determine why students sometime leave before graduation by investigating the social and academic connections among the students. Analyzing course registration records, team and club rosters, residence hall records, and more, the pair constructed a model of the social network of an entire first-year college class. The study was the largest of its kind, using more comprehensive data and a larger group of students than any previous social network analysis with a focus on college student retention.
The analysis uncovered the fact that social connections influence the likelihood of a student staying at Rhodes, even after controlling for more traditional explanations like grades, ethnicity, religion, ability to pay, and participation in athletics or Greek life. The study concluded that students on the outside of the social network are more at risk, and students who have friends that leave are themselves more likely to leave. Alternatively, having friends who stay make a student more likely to stay as well.
“The results add knowledge to the field of higher education and give administrators at Rhodes and elsewhere new tools to help identify and assist students who are at risk of not succeeding at college,” says Eckles. He says this study examines individual’s friends’ values, intentions, and choices, using social network analysis (SNA) to understand this complex web of social interaction and how it may contribute to an individual student’s choice to stay at or leave a college.
The study’s authors point out that every friend who leaves makes a student five times more likely to leave, but every friend who stays makes a student 2.25 times more likely to stay.
“What would be very interesting to see is whether the odds ratios of attrition and retention scores maintain similar relative proportions in a network where the overall retention rate is not as high,” the authors write in the article. “In this network, attrition of friends is a very powerful force, but it is fortunately a relatively rare event that is more than outweighed by the more common retention of friends.”
In previous studies, factors such as athletic participation, membership in a fraternity or sorority, religion, and ethnicity were significant contributing factors to retention. But the new study shows that such factors are not individually significant, but rather represent strong social connections among like students or proxies for social networks. For example, it is not being a fraternity or sorority member, per se, that has made students more likely to stay; rather, membership in such a society puts students in a dense social network that exposes them to more students choosing to stay.
The authors point out that dense social networks can work against a college. If for some reason a fraternity or sorority member decides to leave the college, there are a lot more students who are impacted by that decision. If one of them is sufficiently influenced to also leave, soon a vicious cycle may emerge and attrition can begin spreading through that dense network like a disease.