How teens choose their friends

But a national study led by a Michigan State University scholar finds that the courses students take have powerful effects on the friendships they make. The study was funded by the National Science Foundation and the National Institutes of Health.

The findings, published in the American Journal of Sociology, indicate the pattern of course-taking is distinctive to each high school. In one school, for example, friendships may form among students taking woodshop, Spanish and European history, while in another it may be among students taking agricultural business management, advanced accounting and calculus.

"People generally want to think that kids are choosing their friends from the well-known categories like jocks and nerds -- that it's like "The Breakfast Club" and the same at every school," said Kenneth Frank, professor in MSU's College of Education.

"But our argument is that the opportunities an adolescent has to choose friends are guided by the courses the adolescent takes and the other students who take the courses with them. Moreover, the pattern of opportunities differs from school to school."

Frank and colleagues analyzed survey data and academic transcripts from some 3,000 students at 78 high schools across the United States. The researchers developed a new computer algorithm and software to identify the unique sets of students and courses from the transcripts in each school.

Students were more likely to make friends in small classes, often electives, which set them off from the general student population. Friendships were more likely to be created in Latin 4 and woodshop, for example, than in a large physical education class that is required of everyone in a particular grade.

Students who take the same set of courses tend to get to know each another very well and focus less on social status, such as how "cool" someone is. They're also less likely to judge classmates on visible characteristics like race and gender.

In addition, Frank said girls are more likely to take more demanding math classes if other girls in their shared sets of courses took advanced math. "In other words," he said, "the peer groups that formed around shared courses had implications for students' academic effort as well as their social world."

The findings have implications for school administrators as well. Schools that simply offer classes without thought to mixing up high- and low-achieving students run the risk of driving them apart socially and academically, Frank said.

To combat this, he said schools could better highlight the value of certain academic pursuits -- such as math -- and also group students together in ninth grade so the low-achievers have high-achievers in their classes potentially throughout high school.

"This would give the students in the lower group a 'beacon' of sorts -- or others who could be there as a marker to help them move along."