Hartl, Amy C.

Friendships are important for adolescent adjustment and development; however, adolescent
friendships are fleeting. Friend dissimilarity and undesirable individual attributes have been
hypothesized to predict friendship dissolution. The present study tests each as predictors of adolescent
friendship dissolution. A sample of 410 U.S. adolescents participated in a total of 573 reciprocated
friendships originating in the 7th grade. These friendships were followed annually from 8th-12th grade
to determine when each friendship dissolved. In the 7th grade, participants completed a peernomination
inventory, and teachers completed a survey of each participant’s school competence.
Discrete-time survival analyses used 7th grade friend dissimilarity and individual characteristics of sex,
age, ethnicity, number of friends, peer acceptance, peer rejection, leadership, physical aggression,
relational aggression, peer victimization, and school competence as predictors of the occurrence and
timing of friendship dissolution. Friendships originating in the 7th grade were at greatest risk for
dissolution during the first year. Only 1 percent of friendships that started in the 7th grade lasted 5
years. Friend dissimilarity on sex, peer acceptance, physical aggression, and school competence
predicted friendship dissolution. At each grade, the odds of friendship dissolution were higher for
friends dissimilar on these characteristics. Individual characteristics failed to predict friendship
dissolution. The findings suggest compatibility is a function of similarity between friends rather
than the presence or absence of a specific individual trait. Adolescents seeking friendships with
individuals dissimilar from them on school-related characteristics risk suffering the downside of
dissimilarity, namely rapid friendship dissolution.

As children enter adolescence, social status within the peer hierarchy gains
importance. Variable-oriented research has linked adolescent popularity with both
positive and negative adjustment outcomes. Popularity may be better understood with
reference to types or subgroups of similar individuals, identified through person-oriented
approaches. Resource Control Theory (RCT: Hawley, 1999) posits three distinct types of
popular adolescents: coercive, prosocial, and bistrategic. The existence and adjustment
correlates of the prosocial and coercive groups have been well-established, but little
evidence supports the existence of a bistrategic popular group of adolescents, and even
less is known about their adjustment correlates. The present study aims to confirm the
existence of the popularity groups hypothesized by RCT and to identify group differences
in social adjustment and problem behaviors.
A sample of 568 adolescents (n = 288 girls, 280 boys; M age = 12.50) completed
peer nomination procedures and self-report questionnaires in the Fall and Spring of the
7th and 8th grades. Longitudinal latent profile analyses classified adolescents into profile groups on the basis of initial physical aggression, relational aggression, and prosocial
behavior, and four time points of popularity spanning the 7th and 8th grades. Repeated
measures ANOVAs examined profile group differences in social adjustment (peer
acceptance, peer rejection, physical victimization, relational victimization, and preference
for solitude) and problem behaviors (disruptiveness and delinquency) across the 7th and
8th grades.
Results indicate that adolescents fall into one of four distinct groups: aggressive
popular, prosocial popular, bistrategic popular, and average. Bistrategic popular
adolescents evinced positive social adjustment, exhibiting the highest levels of popularity
and peer acceptance and the lowest levels of peer rejection, victimization, and preference
for solitude. Despite their social skill advantages, bistrategic popular adolescents were
also at risk for problem behaviors. Bistrategic popular adolescents scored above average
on problem behaviors, including physical and relational aggression, disruptiveness, and
delinquency. Bistrategic popular adolescents successfully navigate the social world in a
manner that both offers hope for positive long-term adjustment and concern for the same.

Adolescent friendships are critical for adjustment but are extremely unstable.
Dyadic characteristics may put friendships at risk for dissolution, whereas individual
characteristics may put individuals at risk for participating in unstable friendships. The
present study examines whether dyadic or individual school-related characteristics
predict rates of adolescent friendship dissolution. A sample of 410 adolescents (n=201 males, 209 females; M age=13.20 years) participated in 573 reciprocated friendships originating in the 7th grade which were followed from 8th-12th grade. Discrete-time survival analyses evaluated grade 7 dyadic and individual characteristics (sex, age, ethnicity, number of friends, peer acceptance, peer rejection, leadership, and school competence) as predictors of the occurrence and timing of friendship dissolution.
Dissimilarity in sex, peer acceptance, and school competence and similarity in
leadership predicted higher rates of friendship dissolution; individual characteristics were not significant predictors. Adolescents seeking friendships with more skilled individuals
risk suffering the downside of dissimilarity, namely dissolution.