If Josh's friends all jumped off a bridge, would he jump, too? If all Josh's friends are pulling Cs, will he let his 3.7 GPA slide?
A new study published this afternoon in the online journal PLOS-1—and conducted by a team of high school students partnering with researchers via a National Science Foundation outreach program—suggests grades of friends can rise or fall together over time.
Social contagion is something of a cognitive germ theory; it posits that concepts and behaviors can spread through a social network like a cold spreads through students in the same school. Prior research already suggests children and adults become more likely to "catch" behaviors like smoking or exercising the more people in their social circle are already doing it. The concept is even more appealing for those who study children and teenagers, who have been shown to be heavily influenced by peers.
The study, "Spread of Academic Success in a High School Social Network," is the first to attempt to use this social network approach to analyze changes in students' grades. The authors asked 160 juniors at Maine-Endwell High School in Endwell, N.Y. to identify, on a list of the rest of the class, the other students considered close friends, friends, acquaintances, relatives, or unknown to the student. The authors then linked the social circles with administrative data, including GPA (as translated into class ranking), attendance, and disciplinary actions, for the 2010 and 2011 school years.
They found that students whose friends' average GPA was higher than their own at the start of the study were more likely to improve their GPA, and students with a higher GPA than their friends were more likely to drop in grades.
Co-author Hiroki Sayama, the director of the collective dynamics of complex systems research group at Binghamton University in New York, cautioned that the findings don't prove that having friends whose grades are higher causes a student to improve his own grades; students could, for example, be seeking out other academically inclined students. "Psychologically, it makes sense, because if you are working to improve your grades, subconsciously you might see smarter students as your friends," he said.
There was some evidence that this isn't just a matter of birds of a feather flocking together, though. Close friends—the ones most likely to be chosen based on personality similarities—were less strongly related to changes in a student's GPA than were students considered friends, but not as close.
Network researcher Mark Pachucki, a senior scientist with the Mongan Institute for Health Policy in Boston, who was not involved in the study, said he considers the study a "good first step" in exploring social networks in education, though he looks forward to seeing more advanced statistical modeling of students' social networks over time. "As we know, students form and drop friendships very quickly over the school year," Mr. Pachucki said. "While it's only a pre-/post- analysis, they [the study authors] are right to point out that you need to look at finer details of those relationships."
Most of the researchers themselves had a firsthand look at those high school social dynamics: Authors Deanna Blansky, Christina Kavanaugh, Cara Boothroyd, and Brianna Benson were students at Maine-Endwell High when the study was conducted, while co-authors Julie Gallagher is the school's principal and John Endress is on the information technology staff. The study was part of NetSci High, a pilot outreach project by the National Science Foundation intended to teach students and educators about emerging science methods.
"It was fun to see the students grow through the process, even something as simple as realizing, 'I'm going to be a colleague with my professor,'" Ms. Gallagher said.
All four of the student authors have since graduated high school and entered science fields in college, according to Ms. Gallagher.