Peer Status in Early Adolescence: A Longitudinal Study of Relational Aggression, Physical Aggression

1
Peer Status in Early AdolescenceA Longitudinal
Study of Relational Aggression, Physical
Aggression, and Prosocial Behavior

• Melanie Zimmer-Gembeck
• Griffith University, Gold Coast
• Australia
• Tasha Geiger Nicki Crick
• University of Minnesota
• USA

2
This work was completed while the first author
was a NIMH postdoctoral fellow in the Life
Course Center, U of Minnesota and supported by
NIMH grant R29MH53524 awarded to Nicki
Crick.Thanks to all the members of theProject
Kids lab!
3
BackgroundPhysical Aggression and Peer Rejection

• Children who are physically aggressive are more
  likely to be rejected by their classmates.
• Boys are more likely nominated as physically
  aggressive than girls.
• Many studies focus on boys.

4
BackgroundRelational Aggression and Peer
Rejection

• Behaviors that can harm others through damage (or
  threat of damage) to relationships, feelings of
  acceptance, friendship, or group inclusion.
• Relational aggression significantly associated
  with rejection by peers.

5
BackgroundProsocial Behavior and Peer Acceptance

• Should also be considered.
• Associated with acceptance by classmates.

6
Study Purpose

• Investigate how childrens behaviors
• physical aggression
• relational aggression
• prosocial behavior
• impact their acceptance and rejection by peers
  over a short (1-year, grade 3 to 4) and a longer
  (3-year, grade 3 to 6) span of time.

7
Additional Study Aims

• Examine stability of relational aggression
• Investigate moderating role of gender
• Gender nonnormative hypothesis?
• Relational aggression has greatest impact on
  females peer relationships?

8
Participants

• Children/adolescents participated in assessments
  in classrooms up to 3 times over 4 years.
• Grade 3 (T1), N 2335, 95 classrooms (85 in
  Grade 3 - some mixed grade classes).
• Grade 4 (T2)
• Grade 6 (T3)
• Grade 3 to 4 SHORT, n 1119, 51 female
• Grade 3 to 6 LONG, n 464, 54 female, 94
  transitioned to middle school.

9
Subsamples Representative?

• Compared to children who only participated at
  Time 1, SHORT more accepted and more likely to be
  of white race/ethnicity.
• Compared to children who only participated at
  Time 1, LONG were more accepted, prosocial, and
  more likely to female and white.

10
Procedure

• Peer nominations in (northern hemisphere) Spring
  (February to May).
• 70 or more of children in each class
  participated.
• Two interviewers in each class. One read
  assessment, other helped students.
• Alphabetized rosters of classmates with id
  numbers.
• Nominate up to 3 classmates (male or female) for
  each item.

11
Measurement

• Peer Preference
• All students nominated 3 classmates liked most
  and 3 classmates liked least.
• Physical aggression, relational aggression,
  prosocial behavior
• Nominate 3 classmates
• Physical aggression 3 items, all overt physical
  behaviors (e.g., kids who hit other kids)
• Relational aggression 5 items (e.g., kids who
  try to make other kids not like someone by
  spreading rumors or talking behind the kids
  back)
• Prosocial behavior 3 items (e.g, people who
  say or do nice things for other classmates)

12
Results
13
Stability in Relational Aggression

• SHORT
• r .47 (males .50, females, .46)
• LONG
• r .44 (males .47, females .43)

14
Cross-Lag Models

• Path modeling (with SEM) used to examine
  cross-lag associations between childrens
  aggressive and prosocial behaviors in grade 3 and
  peer status in grade 4 (or grade 6) while also
  estimating
• Stabilities in childrens behaviors and peer
  preference across time.
• Concurrent associations between all constructs.

15
Data Analytic Procedure

• Four Major Steps
• 1. Estimated all longitudinal stabilities and
  all correlations between constructs within each
  wave of measurement.
• 2. Added all possible cross-lag paths for one
  construct (e.g., associations of relational
  aggression in grade 3 with peer acceptance, peer
  rejection, physical aggression, and prosocial
  behavior in grade 4)
• 3. Fixed nonsignificant associations to 0 for
  the focal construct in Step 2.
• 4. Repeated steps 2 and 3 for all five
  constructs.

16
Final Path Model Illustration
17
SHORT
LONG
18
SHORT
LONG
Grade 4
Grade 3
Grade 6
Grade 3
Peer Acceptance
Peer Acceptance
.16
.21
Peer Rejection
Peer Rejection
.26
.29
.19
.12
.11
Relational Aggression
Relational Aggression
.08
Physical Aggression
Physical Aggression
Prosocial Behavior
Prosocial Behavior
19
SHORT
LONG
Grade 3
Grade 4
Grade 3
Grade 6
Grade 3
Grade 4
Peer Acceptance
Peer Acceptance
.20
.21
Peer Rejection
Peer Rejection
.11
.09
-.14
.22
-.14
Relational Aggression
Relational Aggression
.23
.21
Physical Aggression
Physical Aggression
.32
.29
Prosocial Behavior
Prosocial Behavior
.28
.25
20
SHORT - Moderating Effects of Gender
Grade 4
Grade 3
Males, Females
.38,.37
Peer Acceptance
Peer Acceptance
.23,.19
-.16,-.15
.25,.24
-.19,-.15
.43,.30
Peer Rejection
Peer Rejection
.33,.19
.19,.12
.26,.19
-.14,-.15
.10ns,.13
Relational Aggression
Relational Aggression
.36,.38
.24,.19
Model fit Statistics c2(22) 47.0, p lt.001 GFI
1.00 NFI .99 RMSEA .03 c2/df 2.1
.02ns,.02ns
Physical Aggression
Physical Aggression
.36,.30
.28,.16
-.08,.04ns
Prosocial Behavior
Prosocial Behavior
.16,.26
.21,.27
21
ConclusionsGender and Relational Aggression

• Gender differences in relational aggression
  depend upon grade level (all available study
  participants)
• No gender differences in the stability of
  relational aggression (longitudinal subsamples).

22
ConclusionsPath Models Prediction Over Time

• Relational and physical aggression had
  independent positive effects on peer rejection.
• Prosocial behavior was not associated with later
  peer status.
• In LONG, relational aggression also positively
  associated with acceptance.

23
ConclusionsGender

• Few gender differences
• Little support for either the Gender Nonnormative
  Hypothesis or the hypothesis that relational
  aggression is most associated with peer
  preference among girls.