Physical Aggression and Self-Injury in Juvenile Delinquent

1

Physical Aggression and Self-Injury in Juvenile
Delinquent Nikki J. Deaver University of
Nebraska-Lincoln

Results Univariate statistics for variables
collected in the survey are shown in Table 1.
Mixed groups ANOVAs with follow-up analyses using
the LSD procedure (p.05) were used to examine
the main effects and interactions of prior
criminal offenses (yes vs. no), removal from
treatment by judge (kept vs. removed), and
monthly scores (Month 1 vs. 3 vs. 6) as they
relate to physical aggression scores and as they
relate to self-injury scores. Physical
Aggression Figure 1 shows physical aggression
scores for each of the design conditions. There
was a significant 3-way interaction,
F(2,78)6.001,p.004, MSe2.565. Examination of
the cell means (using LSDmmd1.381) reveals that
the pattern of interaction was that those kept by
judge for continuing treatment and with no prior
criminal offenses and those with prior criminal
offenses showed no differences in scores across
months in treatment. The pattern of interaction
for those removed from treatment by judge and
with no prior criminal offenses showed increasing
physical aggression scores from months 1-3, but
decreasing scores from 3-6, and for those removed
from treatment by judge with prior criminal
offenses there was no physical aggression score
difference from months 1-3, but increasing scores
from 3-6. Those who were removed from treatment
with no prior criminal offenses showed higher
physical aggression for month 3 and month 6 than
those with prior criminal offenses. There
was a significant 2-way interaction of monthly
physical aggression scores and removal from
treatment, F(2,78)3.529, p.034 (r.208). The
pattern of interaction (using LSDmmd.976) was
that there was no score difference between those
kept in treatment and those removed in month 1,
however for months 3 and 6 those removed from
treatment had higher physical aggression scores
than those kept in treatment. For those who were
kept in treatment there were no monthly score
differences. For those who were removed from
treatment scores were highest for months 3 and 6.
This pattern of interaction is descriptive for
each successive month among those with no prior
criminal offenses who were kept in therapy, for
months 3 vs. 6 in those with no prior offenses
and removed from treatment, and finally for all
months when comparing those with no prior
criminal offenses who were kept in treatment vs.
those removed from treatment. This pattern is
also descriptive for those with prior criminal
offenses and kept in therapy for each month
comparison, but not for those removed from
therapy. Self-Injury Figure 2 shows
self-injury scores for each of the design
conditions. Analyses revealed there was not a
significant 3-way interaction, F(2,78)2,.874,
p.062, MSe.316 (r.189). There was no
significant 2-way interaction of monthly
self-injury scores and removal from treatment,
F(2,78)1.119, p.332 (r.119).

Introduction Physical aggression and
self-destruction are common behaviors found in
youth, especially those who are delinquent.
There are several components that can be factored
into the source of such behavior, and each of
these is different for each individual. This
study explores the relationship of both physical
aggression and self-destruction to three factors
prior criminal offenses, removal from treatment
by judge, and scores across months in treatment.
Physical aggression has been an important
element in previous research on those in
treatment and those in prison. In one such
study, Ireland (2004) found that those in prison
receiving therapy showed a significant
improvement in physical aggression than did those
who did not receive therapy. Aggressive behavior
against others and toward the self are due to
frustration, inability to express ideas, fear,
and anxiety (Federn, 1989), and for delinquents
the key impulse in dealing with these factors is
through aggression (Toch Adams, 2002).
Self-injury has also been studied in the midst of
treatment programs and in prisons. Just as
previous research has found increases in physical
aggression due to several factors, increases
among self-injury are found in similar
circumstances. Self-injuring behavior has been
found to be associated with more reports of
physical aggression towards other people and
objects (Matsumoto, et al., 2005). Hillbrand
(1993) suggests that self-injuring behavior is
higher in correctional than in non-correctional
psychiatric patients because prisoners use their
behavior to manipulate correctional officers.
Considering this previous research on physical
aggression and self-injuring behavior, the
purpose of this study was to examine their
relationship to prior criminal activity, removal
from treatment, and monthly scores in treatment.
It was expected that those having committed prior
criminal offenses and those being removed from
treatment by a judge (sent to juvenile prison)
would result in higher physical aggression and
self-injury scores than for those with no prior
criminal offenses and those who were kept in the
residential treatment. Those who were kept in
treatment were expected to show no increases in
physical aggression and self-injury across months
in treatment, while those removed from treatment
were expected to show increases in both behaviors
across monthly scores.

Discussion The purpose of this study was to
examine the relationship of physical aggression
and self-injury to prior criminal offenses,
removal from treatment, and monthly scores.
Previous research suggests that these two
behaviors are commonly found together, however
closer examination on adolescents may shed light
on different treatment methods. The overall
effect of prior offenses showed higher physical
aggression scores for those without prior
criminal offenses and showed no difference
between the groups for self-injury. Physical
aggression was found to increase overall when
adolescents were removed from treatment and sent
to prison, which shows that remaining in
treatment until completion helps the individual.
Removal or non-removal of treatment was
significantly related to prior offenses and
monthly scores in the 3-way interaction for
physical aggression. For those with no prior
criminal offenses who were kept in treatment
physical aggression scores where lower than those
removed from treatment for months 3 and 6,
however, for those with prior criminal offenses
who were kept in treatment physical aggression
scores were higher than those removed from
treatment for month 3. The results show
physical aggression differs when taking into
account other variables in relationship with
monthly scores and removal from treatment. There
were no score differences on successive months
for those kept in therapy, and for those removed
from treatment scores for month 3 were
significantly higher than month 1, but also
higher than month 6. Results show no significant
3-way relationship for the three variables in
regards to self-injury. There was also no
significant 2-ways or main effects involving
these variables. There is a possibility that
this was due to limitations of a clear and
comprehensive definition of self-injury and also
to the hidden nature of such behavior. There are
several different types of self-injury some of
which may not have been included in this study.
These results bring up a question as to why
physical aggression and self-injuring behavior
did not appear similarly in relationship to the
other three variables. Future research should
use these results as a pilot to investigate any
differences between these behaviors that
contradict previous research. These results show
there is a pattern on physical aggression and
this effect is irrelevant to self-injuring
behavior. This may be due to limitations in the
study and overall, there are several improvements
that might be made while considering future
research. An increase in the sample size is
needed to find more significant findings and
increase the number of self-injurers. Future
research should include all types of physically
aggressive and self-injuring behavior and
categorize different types in order to interpret
a more detailed perspective. Also there should
be a variety of different treatment methods used
to also increase the specificity of certain
options. It would also be helpful to examine
other characteristics such as socialization in
families and other deviant behaviors such as drug
and alcohol use, truancy, in-school suspensions,
theft, and verbal abuse. Toch and Adams (2002)
suggest that those who are aggressive in nature
tend to be unsocialized, and this lack of
socialization, such as through parenting, creates
defects in the mediating and expression of
impulses leading to reactions of inappropriate
behavior.

Methods Participants Participants were 43 youths
aged 9 to 17 with a mean age of 16.36. Each
participant had a history of behavioral disorders
and/or criminal activity and were placed into one
of four homes/treatments located in a large
Midwestern city by a family court. Each group
home housed 6-12 youths and was staffed 24 hours
a day with a supervisor and 3-5
counselors. Measures This study used archival
data collected from school, court, and treatment
records. The Daily Adjustment Indicator
(Burchard Schaefer, 1992) is a 25-item survey
that was used as a daily measure of data such as
physical aggression and self-injury for the first
six months of treatment in the group home.
Months 1, 3, and 6 were selected to show
successive score differences for the 6 month
period. Prior criminal offenses was divided into
yes (those who have committed one or more crimes)
and no (those who have no prior offenses).
Removal from treatment by judge was divided into
kept (those who remained in treatment for its
entirety) and removed (those who were kicked out
by judge and sent to prison). Procedures On a
daily basis participants received scholastic
tutoring and group therapy that focused on social
skills, interpersonal decision making, and anger
management. On a weekly basis each individual
participant received therapy focusing on similar
topics of the group therapy. The Daily
Adjustment Indicator scores were collected and
combined by the supervisor at the end of each
month.