“Choose your friends wisely” may not only be good parental advice, but also a way to do better in college, a research study finds.

A trio of researchers put that advice to the test at Berea College, a small liberal arts school in Kentucky, by looking at how much friends actually influence study habits and grades. They found that students who befriended studious peers spent more hours studying themselves and posted higher grades during their freshman year.

“It’s no fun to study by yourself,” said Nirav Mehta, an economist at the University of Western Ontario and one of the study’s authors, explaining the intuition behind the study. “If you want to goof off, and your friends are at the library, then you’re going to go to the library, too. And while you’re there — hopefully you’re not drinking — you’re probably going to get some studying done too.”

Of course, it’s possible that studious people gravitate toward other studious people. They might have hit the books and racked up as many A’s no matter who their friends were. So the researchers checked to see if randomly assigned roommates also have a positive influence on study habits and grades. They found almost identical results: students who were assigned a studious roommate freshman year also studied more each day and had higher grade-point averages.

Unfortunately, the opposite is also true, the researchers found. If you have friends and roommates who don’t study a lot, you’re likely to get dragged down by their poor habits, studying less, and earning lower grades. It’s important to clarify that having smart friends isn’t as important as having studious friends in this study. The researchers didn’t find that friends’ grades mattered. What influenced a student’s college grades was his or her friends’ high school study habits. To be sure, students with higher grades tend to have better study habits, so studious friends are also likely to be smart ones.

The study, “Time-use and Academic Peer Effects in College,” is a working paper, meaning it hasn’t been published in a peer-reviewed journal, but it was circulated by the National Bureau of Economic Research in October 2018. Mehta worked with a prolific father-son duo, Ralph and Todd Stinebrickner, a mathematician and an economist, who have conducted many studies mining the data from Berea College, which gives free tuition to all of its students, many of whom are low-income, and collects a lot of information on its students.

Analyzing friends and study habits is usually difficult for researchers. But students at Berea College were asked to list their four best friends at the end of each semester, and they kept careful daily logs of their time, including time spent studying. At the beginning of freshman year, the students were surveyed on their high school study habits. The researchers also had access to roommate assignments, high school grades, and college grades.

From this information, the economists calculated the average amount of time each student’s college friends had reported studying in high school. They found that for every additional 10 hours a week that a student’s friends had spent studying in high school, on average, the student’s own study time in college would likely increase by almost 25 minutes a day, and the student’s own GPA would likely rise by almost a tenth of a point during freshman year. The researchers controlled for gender, race, and prior academic achievement and found that the positive influence of studious peers was similar for both high achieving and low achieving students.

They ran the same calculation on roughly 180 freshman roommates who were randomly assigned by the college. They found that for every additional 10 hours a week a student’s roommate studied in high school, the student’s own study time in college would increase by more than 13 minutes a day, and the student’s GPA would increase by a little more than a tenth of a point. The goal of the roommate analysis wasn’t to measure which is more important — friends or roommates — but to confirm that peers are actually influencing study habits and echoing the friendship analysis.

The difference between a B and B+ average is 0.3 points, so studious friends and roommates are only nudging grades up a bit. Peers are just one of many factors influencing how well students do in college. Family background, the rigor of a student’s high school classes, and the teaching abilities of college professors all play a big role. The relative importance of peer effects is unclear from this study.

So should we encourage parents to meddle and pick their kids’ friends in college? “This is one outcome — GPA. There are other things in life,” said Mehta, who nostalgically admits to selecting studious friends when he was in college. “We were a crew of dorks,” he said. “It was great.”