FAU Study on Adolescent Friendships Confirms ‘Birds of a Feather Flock Together – Stay Together’

No one likes to lose a friend, especially adolescents. But why do friendships end? Researchers in the Charles E. Schmidt College of Science at Florida Atlantic University sought to answer this question with a study examining whether adolescent friendships end because of undesirable characteristics of friends, because of differences between friends, or both. Findings from this long-term longitudinal study of 410 adolescents involved in 573 friendships are published in the current issue of Psychological Science.

Adolescent friendships are fleeting. The majority dissolve after a year or two. Previous studies of friendship dissolution have focused on either levels of individual characteristics or differences between friends, but not both, so it was impossible to know which was responsible for a split.  Previous studies also were short-lived, focusing on stability over the course of a single year. What’s novel about this new study is that it tracked friendships over the course of six years, measuring the effect of both dissimilarities and undesirable individual attributes in predicting when an adolescent friendship would end.

“Findings from our study suggest that compatibility is a function of similarity between friends rather than the presence or absence of a particular trait,” said Brett Laursen, Ph.D., professor and graduate studies coordinator in FAU’s Department of Psychology. “Adolescents are most likely to enjoy successful, long-term friendships with those who share similar traits.” 

Participants of the study attended two middle schools in lower-middle and middle-class neighborhoods in a small city in the northeastern United States. All friendships originated in the seventh grade, and researchers followed the participants from grade seven through grade 12. They collected data annually in the spring semester during a required English class.

Fewer than 1 in 4 friendships that started in the seventh grade were maintained across the next school year, and fewer than 1 in 10 friendships that started in the seventh grade survived the transition from middle school to high school. Only one percent of friendships that began in the seventh grade continued to the 12^th grade.

The strongest predictors of friendship dissolution were differences in sex, differences in the degree to which children were liked by other children, differences in physical aggression, and differences in school competence. By far the strongest predictor was differences in sex; other-sex friendships were almost four times more likely to dissolve than same-sex friendships. The next strongest predictor was differences in physical aggression, followed by differences in school competence, and differences in being liked by other children.  Rates of dissolution increased by 25 percent to 43 percent for each unit of difference on these variables.

Of particular note, when individual characteristics were considered together with differences between friends on the same characteristics, the former did not predict the occurrence or timing of the end of the friendship. Thus, undesirable attributes are not necessarily bad for friendship stability, so long as the friends display similar levels of the same undesirable behavior. 

“We knew from previous studies that children prefer similar others as friends,” said Laursen. “Now we know why. Dissimilarity is bad for friendships. It causes conflict, it interferes with cooperative activities and shared pleasures, and it creates circumstances where one friend bears more costs, such as the friend who is less aggressive; or gets more benefits, such as the friend who has lower social status than the other. Dissimilarity disrupts relationship bonds.”

Amy Hartl, first author of the publication and a Ph.D. student working at FAU with Laursen notes,

“Our study will help our understanding of the complex social development of adolescents. The loss of friendships can be painful, even debilitating for middle school adolescents, because they are going through cognitive and emotional changes at the same time that they are establishing independence from their parents.”

Laursen and Hartl collaborated on this study with Antonius H.N. Cillessen, Ph.D., who was responsible for the data collection.  Cillessen is professor and chair of developmental psychology at Radboud University (Nijmegen, The Netherlands).