ArticlePDF Available

Resilience to bullying victimization: The role of individual, family and peer characteristics

Authors:

Abstract and Figures

Little research attention has been paid to bullied students who function better than expected and are therefore defined as "resilient". The present longitudinal study aimed to identify individual, family and peer factors that predict fewer than expected levels of depression and delinquency following experiences of bullying victimization. The sample consisted 3,136 adolescents. Self-report data were used to measure bullying victimization at age 13 and 14 and depression and delinquency at age 14. We examined the effects of gender, self-esteem, social alienation, parental conflict, sibling victimization and number of close friends on levels of emotional and behavioral resilience following bullying victimization. The resilience measures were derived by regressing depression and delinquency scores at age 14 on levels of bullying victimization at age 13 and 14, respectively. The adolescents who reported low depression despite frequently experiencing bullying tended to be male, had higher self-esteem, were feeling less socially alienated, were experiencing low levels of conflict with parents and were not victimized by siblings. On the other hand, the adolescents who reported low delinquency despite frequently experiencing bullying tended to be female, had higher self-esteem, were experiencing low levels of conflict with parents, were not victimized by siblings and had less close friends. Relationships with parents and siblings continue to play some role in promoting emotional and behavioral adjustment among victims of bullying and, therefore, interventions are more likely to be successful if they target both the psychosocial skills of adolescents and their relationships with their family.
Content may be subject to copyright.
Child
Abuse
&
Neglect
37
(2013)
997–1006
Contents
lists
available
at
ScienceDirect
Child
Abuse
&
Neglect
Resilience
to
bullying
victimization:
The
role
of
individual,
family
and
peer
characteristics
Maria
Sapounaa,,
Dieter
Wolkeb,c
aSchool
of
Social
Sciences,
University
of
the
West
of
Scotland,
ML3
0JB,
UK
bDepartment
of
Psychology,
University
of
Warwick,
CV4
7AL,
UK
cDivision
of
Mental
Health
and
Wellbeing,
Warwick
Medical
School,
University
of
Warwick,
CV4
7AL,
UK
a
r
t
i
c
l
e
i
n
f
o
Article
history:
Received
9
March
2013
Received
in
revised
form
21
May
2013
Accepted
25
May
2013
Available
online
1
July
2013
Keywords:
Bullying
Victimization
Resilience
Parents
Peers
Individual
differences
a
b
s
t
r
a
c
t
Little
research
attention
has
been
paid
to
bullied
students
who
function
better
than
expected
and
are
therefore
defined
as
“resilient”.
The
present
longitudinal
study
aimed
to
identify
individual,
family
and
peer
factors
that
predict
fewer
than
expected
levels
of
depression
and
delinquency
following
experiences
of
bullying
victimization.
The
sample
consisted
3,136
adolescents.
Self-report
data
were
used
to
measure
bullying
victimization
at
age
13
and
14
and
depression
and
delinquency
at
age
14.
We
examined
the
effects
of
gender,
self-esteem,
social
alienation,
parental
conflict,
sibling
victimization
and
number
of
close
friends
on
levels
of
emotional
and
behavioral
resilience
following
bullying
victim-
ization.
The
resilience
measures
were
derived
by
regressing
depression
and
delinquency
scores
at
age
14
on
levels
of
bullying
victimization
at
age
13
and
14,
respectively.
The
adolescents
who
reported
low
depression
despite
frequently
experiencing
bullying
tended
to
be
male,
had
higher
self-esteem,
were
feeling
less
socially
alienated,
were
experienc-
ing
low
levels
of
conflict
with
parents
and
were
not
victimized
by
siblings.
On
the
other
hand,
the
adolescents
who
reported
low
delinquency
despite
frequently
experiencing
bul-
lying
tended
to
be
female,
had
higher
self-esteem,
were
experiencing
low
levels
of
conflict
with
parents,
were
not
victimized
by
siblings
and
had
less
close
friends.
Relationships
with
parents
and
siblings
continue
to
play
some
role
in
promoting
emotional
and
behavioral
adjustment
among
victims
of
bullying
and,
therefore,
interventions
are
more
likely
to
be
successful
if
they
target
both
the
psychosocial
skills
of
adolescents
and
their
relationships
with
their
family.
©
2013
Elsevier
Ltd.
All
rights
reserved.
Introduction
Bullying
is
a
form
of
aggressive
behavior
that
is
repeated
over
time
against
a
person
who
feels
powerless
to
defend
him
or
herself
(Monks,
Smith,
Naylor,
Barter,
Ireland,
&
Coyne,
2009).
It
can
take
many
forms
such
as
hitting,
name
calling,
social
exclusion,
spreading
nasty
rumors
and
sending
insulting
messages
by
phone.
A
recent
comparison
of
bullying
prevalence
across
40
countries
revealed
that,
on
average,
26%
of
adolescents
are
involved
in
bullying:
12.6%
as
victims,
10.7%
as
bullies
and
3.6%
as
bully
victims
(Craig
et
al.,
2009).
Bullying
increases
steadily
in
primary
school,
peaks
during
the
first
years
of
sec-
ondary
school
(ages
12–14)
as
students
re-negotiate
their
position
in
the
new
peer
group
and
tapers
off
in
late
adolescence.
The
consequences
of
bullying
can
be
severe
and
long-lasting,
including
low
self-esteem,
depression,
academic
failure,
con-
duct
problems,
psychosis
and
increased
risk
of
suicide
(Arseneault,
Bowes,
&
Shakoor,
2009;
Barker,
Arseneault,
Brendgen,
Fontaine,
&
Maughan,
2008;
Brunstein-Klomek,
Sourander,
&
Gould,
2010;
Schreier
et
al.,
2009).
Corresponding
author
address:
School
of
Social
Sciences,
University
of
the
West
of
Scotland,
Almada
Street,
Hamilton
ML3
0JB,
UK.
0145-2134/$
see
front
matter
©
2013
Elsevier
Ltd.
All
rights
reserved.
http://dx.doi.org/10.1016/j.chiabu.2013.05.009
998
M.
Sapouna,
D.
Wolke
/
Child
Abuse
&
Neglect
37
(2013)
997–1006
Although
bullied
students
are
clearly
at
risk
of
the
problems
mentioned
above,
not
all
of
them
will
experience
such
difficulties.
Those
individuals
who
show
positive
developmental
outcomes
despite
facing
stressors
such
as
bullying
are
referred
to
as
“resilient”
(Rutter,
2006).
Resilience
has
been
neglected
in
bullying
research
(Rothon,
Head,
Klineberg,
&
Stansfeld,
2011),
and,
as
a
result,
it
is
not
currently
known
how
some
bullied
students
manage
to
bounce
back
and
function
well
over
time
despite
their
negative
experience.
Studies
that
have
investigated
resilience
to
child
maltreatment
find
that
12–22%
of
children
or
adults
who
were
abused
as
children
manifest
better
outcomes
than
expected
given
their
experiences
of
abuse
(Jaffee,
Caspi,
Moffitt,
Polo-Tomás,
&
Taylor,
2007).
However,
even
in
this
field,
longitudinal
studies
are
relatively
few
and
often
limited
to
small
samples
(Cicchetti,
2010;
Werner,
2013).
Identifying
the
factors
that
promote
positive
outcomes
in
young
people
who
have
experienced
negative
events
such
as
bullying
could
steer
the
development
of
successful
interventions
for
victims.
The
present
study
attempts
to
address
this
by
investigating
individual,
family
and
peer
predictors
of
resilience
to
bullying
using
a
large
cohort
of
adolescents
in
Scotland.
Defining
resilience
Although
definitions
of
resilience
vary
among
studies,
a
consensus
view
is
emerging
that
resilient
individuals
are
those
who
manifest
positive
outcomes
over
time
despite
facing
significant
adversities
(Luthar,
Cicchetti,
&
Becker,
2000).
In
ado-
lescence,
not
being
depressed
is
an
indicator
of
emotional
adjustment,
performing
well
at
school
is
an
indicator
of
academic
adjustment
and
not
being
delinquent
is
an
indicator
of
behavioral
adjustment
(Jaffee
et
al.,
2007;
Luthar
et
al.,
2000).
Most
authors
are
also
in
agreement
that
resilience
is
not
a
personality
trait
but
rather
a
capacity
that
develops
over
time
in
the
context
of
positive
relationships
with
family
members
and
peers
(Egeland,
Carlson,
&
Sroufe,
1993;
Luthar,
2003;
Rutter,
1999).
Garmezy
(1985)
was
one
of
the
first
researchers
in
this
field
to
recognize
the
importance
of
positive
relationships
within
and
outside
the
family
in
fostering
resilience.
He
distinguished
between
three
basic
sources
of
protection:
individ-
ual
characteristics
(including
high
self-esteem
and
autonomy),
family
environment
(defined
as
positive
relationships
with
parents
characterized
by
warmth,
harmony
and
absence
of
neglect
and
conflict)
and
community
(including
positive
peer
interactions
characterized
by
trust,
support
and
absence
of
conflict,
quality
neighborhoods
and
schools).
Garmezy’s
the-
oretical
framework
influenced
much
subsequent
theorizing
and
research
on
resilience
(Luthar
et
al.,
2000).
For
example,
life-course
theories
of
resilience
also
place
emphasis
on
positive
relationships
with
family
members
and
peers
as
determi-
nants
of
resilience
(Rutter,
1999).
According
to
this
perspective,
relationships
within
and
outside
the
family
serve
either
to
increase
or
decrease
the
risk
of
negative
outcomes
following
adversity
depending
on
their
quality.
Rutter
(1999)
has
argued
that
negative
relationships
with
family
and
peers
may
be
genetically
mediated
in
the
sense
that
the
child’s
characteristics
and
behavior
shapes
the
type
of
relationship
they
develop
with
other
people.
Resilience
against
bullying
Although
central
in
Garmezy’s
and
others’
theoretical
frameworks,
family
and
peer
predictors
of
positive
adjustment
among
victims
of
bullying
have
been
largely
overlooked
in
the
literature.
Rather,
the
literature
has
primarily
explored
how
individual
characteristics
influence
the
outcomes
that
bullied
children
and
young
people
will
experience.
For
example,
research
has
identified
that
cognitive
interpretations
of
events
(e.g.
how
great
the
threat
is
perceived
to
be)
partially
medi-
ate
the
extent
to
which
bullied
children
will
report
feeling
lonely
(Catterson
&
Hunter,
2010).
However,
other
individual
characteristics
remain
under-investigated.
For
example,
although
there
is
evidence
that
adolescents
who
are
resilient
to
sexual
abuse
are
characterized
by
high
levels
of
self-esteem
(i.e.
have
a
positive
view
of
themselves;
Turner,
Finkelhor,
&
Ormrod,
2010),
the
possible
mediating
role
of
self-esteem
in
the
association
between
bullying
and
positive
adjustment
remains
under-investigated.
It
is
also
not
clear
how
gender
may
affect
adolescents’
adjustment
to
bullying.
Some
studies
suggest
that
females
are
more
vulnerable
to
the
effects
of
bullying
than
males
(Barker
et
al.,
2008;
Klomek
et
al.,
2009;
Perren,
Dooley,
Shaw,
&
Cross,
2010)
while
others
have
found
that
bullying
affects
both
genders
equally
(Bakker,
Ormel,
Verhulst,
&
Oldehinkel,
2010;
Turner,
Exum,
Brame,
&
Holt,
2013).
Only
recently
have
studies
begun
to
examine
how
family
relationships
influence
longer-term
developmental
outcomes
for
bullied
students.
Bowes,
Maughan,
Caspi,
Moffitt
and
Arseneault
(2010)
investigated
predictors
of
positive
adjustment
following
experiences
of
bullying
victimization
in
primary
school
in
a
sample
of
1,116
pairs
of
twins
aged
10–12.
The
study
found
that
maternal
warmth,
sibling
warmth
and
a
positive
atmosphere
at
home
were
associated
with
fewer
than
expected
emotional
and
behavioral
problems
such
as
depression
and
aggression
over
a
two-year
period
following
bullying
victimiza-
tion.
Maternal
warmth,
in
particular,
exerted
a
protective
effect
independent
of
genetic
factors
such
that
the
bullied
twin
who
received
more
maternal
warmth
had
fewer
behavioral
problems
than
the
twin
who
received
less
maternal
warmth.
This
study
provides
strong
evidence
of
an
environmental
effect
of
families
in
protecting
children
aged
10–12
years
from
negative
outcomes
associated
with
being
bullied.
However,
it
is
not
known
whether
family
relationships
are
equally
important
for
older
age
groups
that
tend
to
spend
less
time
at
home.
The
importance
of
positive
relationships
with
siblings
has
been
further
highlighted
in
a
recent
review
of
sibling
bullying
(Wolke
&
Skew,
2012).
Although
based
on
a
small
number
of
cross-sectional
studies,
the
review
suggests
that
those
bullied
at
home
and
at
school
show
increased
odds
of
emotional
and
behavioral
problems
compared
to
those
victimized
in
only
one
context
or
not
at
all.
This
is
not
surprising
given
emerging
evidence
that
sibling
relationships
have
a
significant
bearing
on
a
range
of
developmental
outcomes
including
self-esteem
and
social
competence
in
peer
relationships.
Adolescents
who
report
M.
Sapouna,
D.
Wolke
/
Child
Abuse
&
Neglect
37
(2013)
997–1006
999
positive
relationships
with
siblings
tend
to
have
higher
self-esteem
and
lower
levels
of
anxiety
(Campione-Barr,
Bassett-
Greer,
&
Kruse,
2012).
Moreover,
longitudinal
research
suggests
that
positive
relationships
with
siblings
characterized
by
warmth
and
absence
of
hostility
and
aggression
can
provide
a
buffer
against
maladjustment
following
stressful
life
events
(Soli,
McHale,
&
Feinberg,
2009).
Peer
relationships
may
also
play
a
role
in
promoting
resilience
to
bullying.
For
example,
bullied
adolescents
who
report
high
levels
of
support
from
peers
are
more
likely
to
maintain
appropriate
academic
achievement
for
their
age
group
compared
to
those
with
low
peer
support
(Rothon
et
al.,
2011;
Wang,
Iannotti,
&
Luk,
2011).
However,
in
other
domains
of
adjustment
results
are
mixed.
According
to
some
studies,
high
levels
of
peer
support
moderate
the
effect
of
bullying
victimization
on
adolescents’
emotional
adjustment
(Flaspohler,
Elfstrom,
Vanderzee,
Sink,
&
Birchmeier,
2009;
Stadler,
Feifel,
Rohrmann,
Vermeiren,
&
Poustka,
2010;
Yeung
&
Leadbeater,
2010).
Conversely,
other
studies
found
that
support
from
friends
alone
cannot
mitigate
against
the
strong
negative
effect
that
bullying
has
on
adolescent
emotional
adjustment
(Pouwelse,
Bolman,
Lodewijkx,
&
Spaa,
2011;
Rothon
et
al.,
2011).
There
is
a
suggestion
that
relationships
with
peers
gain
increasing
importance
in
adolescence
as
buffers
of
stress
compared
to
relationships
with
parents
(La
Greca
&
Harrison,
2005;
Stadler
et
al.,
2010),
however
further
research
with
larger
samples
is
needed.
The
present
study
The
present
short-term
longitudinal
study
aims
to
shed
some
light
on
the
process
of
resilience
to
bullying
by
identifying
individual,
family
and
peer
factors
that
predict
fewer
than
expected
levels
of
depression
and
delinquency
following
expe-
riences
of
bullying
victimization.
In
particular,
we
were
interested
in
examining
the
extent
to
which
positive
qualities
of
relationships
with
parents,
siblings
and
peers
predict
resilience
over
and
above
individual
characteristics.
We
also
sought
to
examine
whether
it
is
relationships
with
family
members
or
peer
relationships
that
provide
the
strongest
protection
against
bullying
during
early
adolescence
as
youths
make
their
transition
from
predominantly
adult-
to
peer-centered
relationships.
Methods
Sample
The
sample
for
this
study
was
drawn
from
a
cohort
of
4,597
individuals
participating
in
the
Edinburgh
Study
of
Youth
Transitions
and
Crime,
a
prospective
longitudinal
study
of
pathways
in
and
out
of
youth
offending
and
victimization
man-
aged
by
the
University
of
Edinburgh
(McAra
&
McVie,
2007;
Smith
&
McVie,
2003;
http://www.law.ed.ac.uk/cls/esytc/).
Participants
were
recruited
at
age
12
and
followed
annually
up
to
age
17.
At
initial
recruitment
in
autumn
1998,
the
sample
consisted
of
92%
of
the
total
population
of
young
people
who
were
enrolled
at
secondary
school
in
the
City
of
Edinburgh.
Only
data
from
sweeps
1
(age
12)
to
4
(age
15)
have
been
made
publicly
available
via
the
UK
Data
Archive
for
use
by
researchers.
For
the
purposes
of
this
study,
self-report
data
from
the
first
three
waves
were
analyzed
(ages
12,
13
and
14).
We
were
not
able
to
use
all
four
waves
of
data
because
there
was
no
measure
of
depression
at
wave
4
(age
15).
The
final
sample
consisted
of
3,136
adolescents
for
whom
data
were
available
on
all
the
variables
needed
for
this
analysis.
A
total
of
48.5%
of
our
final
sample
were
males
and
94.9%
were
of
White
ethnic
background.
Procedures
Parental
and
child
consent
was
obtained
for
all
participants
in
the
study.
Self-report
data
was
obtained
via
question-
naires
administered
at
schools
by
trained
researchers.
Follow-up
visits
were
arranged
to
capture
absent
students’
views.
All
participants
were
assured
of
confidentiality
prior
to
completing
the
questionnaire.
Measures
A
summary
of
the
variables
used
in
the
study
is
provided
in
Table
1.
Table
1
Summary
of
variables
used
in
the
study.
Time
1
(age
12)
Time
2
(age
13)
Time
3
(age
14)
Bullying
X
X
Depression
X
Delinquency
X
Self-esteem
X
Social
alienation
X
Parental
conflict
X
X
Sibling
victimization
X
Number
of
close
friends
X
1000
M.
Sapouna,
D.
Wolke
/
Child
Abuse
&
Neglect
37
(2013)
997–1006
Bullying
at
age
13
and
age
14
(waves
2
and
3)
was
a
composite
score
of
responses
to
the
following
four
items
adapted
from
Olweus
(1993):
being
ignored/left
out,
being
‘slagged
off’/called
names,
being
threatened
and
being
attacked
(0
=
hardly
ever/never,
1
=
less
than
once
a
week,
2
=
at
least
once
a
week,
or
3
=
most
days).
Scores
ranged
from
0
to
12,
representing
the
frequency
of
bullying
victimization
in
the
last
year
(at
13:
M
=
1.48,
SD
=
2.36;
at
14:
M
=
1.39,
SD
=
2.20,
˛
=
.81).
Depression
was
assessed
at
age
14
(wave
3)
using
6
self-reported
items
that
asked
how
often
in
the
last
month
the
participant
felt
too
tired
to
do
things,
had
trouble
sleeping,
felt
unhappy,
sad
or
depressed,
felt
hopeless
about
the
future,
felt
nervous
or
tense,
and
had
worried
too
much
about
things
(0
=
hardly
ever/never,
1
=
less
than
once
a
week,
2
=
at
least
once
per
week,
or
3
=
most
days).
This
measure
was
adapted
from
the
West
of
Scotland
11
to
16
Study
of
Teenage
Health
(Sweeting,
Young,
West,
&
Der,
2006)
and
originally
validated
by
Kandel
and
Davies
(1982).
A
composite
score
was
created
to
denote
the
frequency
of
depression
symptoms
in
the
last
month
(range
0–18,
M
=
7.03,
SD
=
4.32,
˛
=
.82).
Delinquency,
defined
as
criminal
offenses
committed
by
young
people,
was
assessed
at
age
14
(wave
3)
using
14
self-
reported
items
that
asked
whether
the
participant
had
committed
any
of
the
following
14
offenses
in
the
last
year:
not
paying
the
correct
fare,
being
noisy
or
cheeky
in
public,
stealing
something
from
a
shop,
riding
in
a
stolen
vehicle,
stealing
something
from
the
school,
carrying
a
weapon
or
knife,
breaking
into
a
vehicle,
using
force/weapon/threats
to
rob
someone,
vandalizing
property,
breaking
into
a
house,
stealing
something
from
home,
writing
or
spraying
graffiti,
setting
fire
to
something
and
hitting,
punching
or
kicking
someone.
Responses
were
coded
1
for
yes
and
0
for
no
and
summed
across
the
items
to
provide
a
measure
of
delinquency
(range
0–14,
M
=
2.97,
SD
=
2.86,
˛
=
.81).
Outcome
measures.
Two
domains
of
positive
adjustment
following
experiences
of
bullying
were
investigated
for
the
purposes
of
this
study.
First,
a
measure
of
“emotional
resilience”
to
bullying
was
derived
using
the
methodology
adopted
by
Bowes
et
al.
(2010).
Emotionally
resilient
participants
were
considered
those
who
reported
fewer
than
expected
levels
of
depression
given
the
frequency
of
bullying
they
experienced
at
ages
13
and
14.
The
emotional
resilience
measure
was
derived
by
regressing
depression
scores
at
age
14
on
levels
of
bullying
victimization
at
ages
13
and
14.
Residual
scores
were
saved
and
reverse-coded
(14.58
to
12.17)
so
that
positive
residual
scores
indicated
participants
with
fewer
than
expected
levels
of
depression
at
age
14
given
the
frequency
of
bullying
experienced
at
ages
13
and
14
(i.e.
emotionally
resilient
participants).
Second,
a
measure
of
“behavioral
resilience”
to
bullying
was
derived
using
the
same
methodology
above.
Behaviorally
resilient
participants
were
considered
those
who
reported
fewer
than
expected
levels
of
delinquency
given
the
frequency
of
bullying
they
experienced
at
ages
13
and
14.
The
behavioral
resilience
measure
was
derived
by
regressing
delinquency
scores
at
age
14
on
levels
of
bullying
victimization
at
ages
13
and
14.
Residual
scores
were
saved
and
reverse-coded
(4.41
to
11.24)
so
that
positive
residual
scores
indicated
participants
with
fewer
than
expected
levels
of
delinquency
at
age
14
given
the
frequency
of
bullying
experienced
at
ages
13
and
14
(i.e.
behaviorally
resilient
participants).
Testing
the
validity
of
the
resilience
measures.
There
was
no
correlation
between
our
measures
of
resilience
and
the
frequency
of
bullying
victimization
at
ages
13
and
14
(r
=
.00,
p
=
1.00)
which
proves
that
our
measures
were
not
purely
a
function
of
the
low
frequency
of
bullying
victimization
(we
would
like
to
thank
an
anonymous
reviewer
for
pointing
this
out
to
us;
see
Table
3
below).
To
further
test
the
validity
of
our
measures,
we
examined
how
strongly
they
correlated
to
other
develop-
mental
outcomes
that
were
measured
in
the
Edinburgh
Study
of
Youth
Transitions
and
Crime.
We
found
that
emotional
and
behavioral
resilience
was
negatively
correlated
to
self-reported
alcohol
(r
=
.16,
p
<
.01
for
emotional
and
r
=
.41,
p
<
.01
for
behavioral)
and
drug
use
(r
=
.03,
p
=
ns
for
emotional
and
r
=
.28,
p
<
.01
for
behavioral)
and
teacher-reported
truancy
(r
=
.05,
p
<
.05
for
emotional
and
r
=
.09,
p
<
.01
for
behavioral).
Furthermore,
behaviorally
resilient
children
were
less
likely
to
be
referred
to
the
Children’s
Reporter
(t(3,134)
=
4.96,
p
<
.001)
which
is
the
body
responsible
for
judging
cases
of
adolescent
delinquency
in
Scotland.
Individual,
family
and
peer
predictors.
Individual
factors
considered
were
Gender
coded
1
for
males
and
0
for
females.
Social
alienation
at
age
12
(sweep
1),
defined
as
the
experience
that
one
is
not
an
integral
part
of
a
group,
was
the
composite
index
of
6
self-reported
items
rated
on
a
5-point
Likert
type
scale
ranging
from
0
(disagree
a
lot)
to
4
(agree
a
lot).
An
example
item
was:
Some
people
are
against
me
for
no
good
reason;
range
0–24,
M
=
9.42,
SD
=
6.23,
˛
=
.85.
The
scale
was
a
modified
version
of
the
Alienation
scale
of
the
MPQ
(Tellegen,
1982)
that
taps
negative
emotionality.
Being
estranged
from
peers
has
been
found
to
increase
the
risk
of
peer
victimization
and
depression
(Cook,
Williams,
Guerra,
Kim,
&
Sadek,
2010;
Gooren,
van
Lier
Stegge,
Terwogt,
&
Koot,
2011,
Ostrov,
2008).
Self-esteem,
defined
as
a
positive
evaluation
of
self,
was
assessed
at
age
12
(sweep
1)
using
a
modified
version
of
the
Rosenberg
self-esteem
scale
(Rosenberg,
1965)
consisting
of
6
self-reported
items
rated
on
a
5-point
Likert
type
scale
ranging
from
0
(disagree
a
lot)
to
4
(agree
a
lot).
An
example
item
was:
I
like
myself;
range
0–24,
M
=
15.32,
SD
=
4.39,
˛
=
.72.
To
develop
a
proxy
indicator
of
positive
relationships
with
parents
we
conducted
a
principal
component
analysis
of
the
following
items:
(1)
composite
index
of
6
self-reported
items
measuring
the
frequency
of
arguments
with
parents
at
age
12
(sweep
1;
e.g.
“argue
with
parents
about
homework”,
“argue
with
parents
about
friends”)
(from
0
=
hardly
ever/never
to
4
=
most
days;
range
0–18,
M
=
5.64,
SD
=
4.04,
˛
=
.75),
(2)
composite
index
of
6
self-reported
items
measuring
the
frequency
of
arguments
with
parents
at
age
13
(sweep
2;
e.g.
“argue
with
parents
about
homework”,
“argue
with
parents
about
friends”)
(from
0
=
hardly
ever/never
to
4
=
most
days;
range
0–18,
M
=
5.05,
SD
=
4.02,
˛
=
.77)
and
3
composite
index
of
5
self-reported
items
measuring
the
frequency
of
punitive
behaviors
exhibited
by
parents
at
age
13
(sweep
2;
e.g.
“parents
tell
you
off
or
give
you
a
row”,
“parents
stop
you
going
out”;
from
0
=
hardly
ever/never
to
4
=
most
days;
range
0–15,
M
=
2.91,
SD
=
2.38,
˛
=
.66).
M.
Sapouna,
D.
Wolke
/
Child
Abuse
&
Neglect
37
(2013)
997–1006
1001
We
extracted
one
principal
component
that
explained
59.4%
of
the
variance:
the
composite
scores
of
arguments
with
parents
at
sweep
1
(.73)
and
sweep
2
(.84),
and
punitive
parenting
at
sweep
2
(.73)
all
loaded
high
on
the
factor
“parental
conflict”
defined
as
frequent
arguments
with
parents.
Factor
scores
were
saved
as
variables
through
the
use
of
the
Anderson–Rubin
method,
and
the
“parental
conflict”
factor
was
reverse-coded
so
that
positive
scores
indicated
low
levels
of
conflict
with
parents
(range
1–7.15,
M
=
5.50,
SD
=
.99).
Sibling
victimization
was
measured
at
age
13
(sweep
2)
by
asking
respondents
how
often
their
siblings
had
threatened
to
hurt
them,
had
hurt
them
by
hitting
or
had
hurt
them
using
a
weapon.
Responses
were
rated
on
a
4-point
Likert-type
scale
ranging
from
0
(hardly
ever/never)
to
3
(most
days).
We
created
a
composite
score,
ranging
from
0
to
9,
which
represented
the
frequency
of
victimization
by
siblings
in
the
past
year
(M
=
2.02,
SD
=
2.38,
˛
=
.72).
We
reverse-coded
scores
so
that
positive
values
indicated
low
levels
of
sibling
victimization
(range
0–9,
M
=
6.98,
SD
=
2.38).
The
adolescents’
size
of
peer
group
was
assessed
by
means
of
a
single
time
measuring
the
number
of
close
friends
the
participants
reported
having
at
age
12
(sweep
1;
“How
many
close
friends
do
you
have?”).
Responses
ranged
from
0
(none)
to
3
(between
six
and
ten;
Mdn
=
1).
Statistical
analyses
Statistical
analyses
were
conducted
using
SPSS
18.0.
First,
descriptive
statistics
and
correlations
among
all
the
variables
used
in
the
study
were
examined.
Second,
we
tested
for
significant
associations
between
the
predictors
(independent
vari-
ables:
gender,
social
alienation,
self-esteem,
parental
conflict,
sibling
victimization
and
number
of
close
friends)
and
the
two
dependent
outcomes
of
emotional
and
behavioral
resilience
to
bullying
using
a
hierarchical
regression
model,
one
for
each
of
the
two
outcomes.
Individual
variables
(i.e.
gender,
social
alienation,
self-esteem)
were
entered
at
step
1
and
family
and
peer-related
variables
(i.e.
parental
conflict,
sibling
victimization
and
number
of
close
friends)
were
entered
at
step
2
to
establish
the
extent
to
which
resilience
to
bullying
can
be
predicted
by
family
and
peer-related
characteristics
over
and
beyond
individual
characteristics
of
adolescents
who
were
entered
in
the
model
first.
For
these
regression
analyses,
depend-
ent
and
independent
variables
(with
the
exception
of
gender)
were
transformed
into
standardized
z-scores.
The
resulting
standardized
coefficients
(betas)
measured
standard
deviation
change
in
the
dependent
variable
per
standard
deviation
increase
in
the
predictor
variables.
Results
Descriptive
statistics,
by
gender,
for
the
study
population
are
provided
in
Table
2.
Significant
gender
differences
emerged
for
depression,
delinquency,
emotional
and
behavioral
resilience,
self-esteem,
sibling
victimization
and
parental
con-
flict.
More
specifically,
girls
reported
significantly
more
depression
(t(3,134)
=
12.85,
p
<
.001),
less
conflict
with
parents
(t(3,134)
=
5.40,
p
<
.001)
and
higher
behavioral
resilience
(t(3,134)
=
7.21,
p
<
.001)
whereas
boys
reported
higher
self-
esteem
(t(3,134)
=
12.90,
p
<
.001),
more
delinquency
(t(3,134)
=
7.17,
p
<
.001),
less
sibling
victimization
(t(3,134)
=
3.23,
p
<
.01)
and
higher
emotional
resilience
(t(3,134)
=
13.93,
p
<
.001).
The
correlations
table
shows
that
being
bullied
at
age
13
was
associated
with
higher
levels
of
depression
(r
=
.28,
p
<
.01)
and
delinquency
(r
=
.07,
p
<
.01)
one
year
later.
Bullying
at
age
13
was
also
associated
with
experiencing
more
bullying
victimization
at
age
14,
having
less
self-esteem,
feeling
more
socially
alienated,
experiencing
more
conflict
with
parents,
experiencing
more
sibling
victimization
and
having
less
close
friends
(see
Table
3).
Table
2
Descriptive
statistics,
by
gender,
for
the
study
population
(N
=
3,136).
Boys
(N
=
1,521)
Girls
(N
=
1,615)
Range
M
(SD)
Range
M
(SD)
Bullying
victimization
at
age
13
0–12
1.51
(2.49)
0–12
1.46
(2.24)
Bullying
victimization
at
age
14
0–12
1.37
(2.28)
0–12
1.41
(2.11)
Depression
0–18
6.04
(4.03)
0–18
7.97
(4.37)
Delinquency
0–14
3.34
(3.01)
0–13
2.61
(2.65)
Emotional
resilience
14.58
to
12.17
1.42
(3.68)
14.58
to
8.80
3.34
(4.04)
Behavioral
resilience
4.41
to
11.10
6.45
(3.01)
3.41
to
11.24
7.18
(2.63)
Self-esteem
0–24
16.34
(4.13)
0–24
14.37
(4.42)
Social
alienation
0–24
9.58
(6.32)
0–24
9.27
(6.15)
Low
parental
conflict
1–7.15
5.41
(1.02)
1.15–7.15
5.60
(0.95)
Low
sibling
victimization
0–9
7.12
(2.38)
0–9
6.85
(2.37)
Range
Mdn
Range
Mdn
Number
of
friends
0–3
1
0–3
1
1002
M.
Sapouna,
D.
Wolke
/
Child
Abuse
&
Neglect
37
(2013)
997–1006
Table
3
Correlations
between
the
study
variables
(N
=
3,136).
Correlations
matrix
Variables
1
2
3
4
5
6
7
8
9
10
11
1
Bullying
victimization
at
age
13
1.00
.45** .28** .07** .00
.00
.14** .38** .21** .18** .08**
2
Bullying
victimization
at
age
14
1.00
.36** .09** .00
.00
.11** .29** .17** .13** .04*
3
Depression
1.00
.18** .92** .15** .27** .25** .24** .19** .05**
4
Delinquency
1.00
.16** .10** .07** .02** .32** .14** .11**
5
Emotional
resilience 1.00 .16** .24** .12** .17** .14** .03
6
Behavioral
resilience 1.00 .05** .06** .30** .13** .12**
7
Self-esteem 1.00
.29** .23** .14** .07*
8
Social
alienation
1.00
.27** .16** .13**
9
Low
parental
conflict
1.00
.31** .03
10
Low
sibling
victimization
1.00
.01
11
Number
of
close
friends
1.00
*p
<
.05.
** p
<
.01.
Table
4
Summary
of
hierarchical
regression
models
predicting
emotional
resilience
to
bullying
victimization
(N
=
3,136).
Predictor
variable
ˇ
R2R2F
change
Step
1
0.10
112.96
Male
.21***
Self-esteem
.17***
Social
alienation .08***
Step
2
0.12
0.02
27.22***
Low
parental
conflict
.13***
Low
sibling
victimization
.06**
Number
of
close
friends
.00
*p
<
.05.
** p
<
.01.
*** p
<
.001.
Predictors
of
emotional
resilience
to
bullying
The
two-step
model
of
the
hierarchical
regression
is
presented
in
Table
4
and
shows
that
individual
variables
alone
accounted
for
10%
of
the
total
variance
of
emotional
resilience
to
bullying.
Being
male
(ˇ
=
.21,
p
<
.001),
having
high
self-
esteem
(ˇ
=
.17,
p
<
.001)
and
feeling
less
socially
alienated
(ˇ
=
.08,
p
<
.001)
significantly
predicted
emotional
resilience
to
bullying
victimization.
The
inclusion
of
environmental
variables
in
step
2
of
the
model
further
only
slightly
increased
its
predictive
power
(R2=
.02,
p
<
.001).
Only
low
levels
of
family
discord
(ˇ
=
.13,
p
<
.001)
and
sibling
victimization
(ˇ
=
.06,
p
<
.01)
were
statistically
significant
predictors
of
emotional
resilience
to
bullying.
The
full
model
accounted
for
12%
of
the
variance
in
emotional
resilience
to
bullying
victimization.
Predictors
of
behavioral
resilience
to
bullying
The
two-step
model
of
the
hierarchical
regression
is
presented
in
Table
5
and
shows
that
individual
variables
alone
accounted
for
2%
of
the
total
variance
of
behavioral
resilience
to
bullying.
Being
female
(ˇ
=
.14,
p
<
.001)
and
having
Table
5
Summary
of
hierarchical
regression
models
predicting
behavioral
resilience
to
bullying
victimization
(N
=
3,136).
Predictor
variable
ˇ
R2R2F
change
Step
1
0.02
25.66
Male
.14***
Self-esteem
.08***
Social
alienation
.03
Step
2
0.116
0.09
108.59***
Low
parental
conflict
.28***
Low
sibling
victimization
.05*
Number
of
close
friends
.11***
*p
<
.05.
**p
<
.01.
*** p
<
.001.
M.
Sapouna,
D.
Wolke
/
Child
Abuse
&
Neglect
37
(2013)
997–1006
1003
high
self-esteem
(ˇ
=
.08,
p
<
.001)
significantly
predicted
behavioral
resilience
to
bullying
victimization.
The
inclusion
of
environmental
variables
in
step
2
of
the
model
further
increased
its
predictive
power
(R2=
.09,
p
<
.001).
Low
levels
of
family
discord
(ˇ
=
.28,
p
<
.001),
low
levels
of
sibling
victimization
(ˇ
=
.05,
p
<
.05)
and
less
close
friends
(ˇ
=
.11,
p
<
.001)
were
statistically
significant
predictors
of
emotional
resilience
to
bullying.
The
full
model
accounted
for
12%
of
the
variance
in
behavioral
resilience
to
bullying
victimization.
Discussion
This
study
investigated
positive
outcomes
following
experiences
of
bullying
victimization
in
early
adolescence.
Using
prospective
data
from
a
large
cohort
of
adolescents,
we
found,
similar
to
other
studies
(Klomek
et
al.,
2009;
Ttofi,
Farrington,
Lösel,
&
Loeber,
2011;
Winsper,
Lereya,
Zanarini,
&
Wolke,
2012),
that
frequent
bullying
victimization
is
typically
associated
with
higher
levels
of
depression
and
delinquency.
Despite
experiencing
frequent
bullying,
however,
some
adolescents
were
“resilient”
in
that
they
showed
fewer
than
expected
levels
of
depression
and
delinquency
over
time.
The
adolescents
who
reported
low
depression
despite
frequently
experiencing
bullying
tended
to
be
male,
had
higher
self-esteem,
were
feeling
less
socially
alienated,
were
experiencing
low
levels
of
conflict
with
parents
and
were
not
victimized
by
siblings
compared
to
those
who
reported
more
depression.
On
the
other
hand,
the
adolescents
who
reported
low
delinquency
despite
frequently
experiencing
bullying
tended
to
be
female,
had
higher
self-esteem,
were
experiencing
low
levels
of
conflict
with
parents,
were
not
victimized
by
siblings
and
had
less
close
friends
compared
to
those
who
reported
more
delinquency.
The
resilient
adolescents
in
our
study
were
also
less
likely
to
report
frequent
alcohol
and
drug
use,
be
truant
from
school
and
be
referred
to
the
Children’s
Reporter,
pointing
to
the
validity
of
our
measures
of
resilience.
In
terms
of
individual
characteristics
that
predict
positive
adjustment
among
bullied
adolescents,
males
were
more
likely
to
report
lower
than
expected
levels
of
depression
following
bullying.
On
the
other
hand,
females
were
more
likely
to
report
lower
than
expected
levels
of
delinquency
following
bullying.
We
cannot
rule
out
the
possibility
that
these
gender
differences
reflect
differences
in
the
prevalence
of
depression
and
delinquency
between
females
and
males
rather
than
“true”
gender
differences
in
resilience.
Most
previous
studies
have
shown
females
to
be
at
greater
risk
of
depression
following
experiences
of
bullying
victimization
(Barker
et
al.,
2008;
Gower
&
Borowsky,
2013;
Klomek
et
al.,
2009;
Perren
et
al.,
2010)
and
epidemiological
data
that
point
to
a
female
preponderance
in
prevalence
of
depression
over
the
life-course
(Velde,
Bracke,
&
Levecque,
2010).
Previous
studies
have
also
found
that
girls
are
more
likely
to
be
subject
to
relational
forms
of
victimization
that
have
been
shown
to
be
more
hurtful
and
more
difficult
to
escape
from
(Smith,
Rose,
&
Schwartz-Mette,
2009;
Wolke,
Woods,
&
Samara,
2009).
Gender
differences
in
the
prevalence
of
delinquency
following
experiences
of
bullying
victimization
are
less
clear.
Some
studies
have
found
that
bullied
boys
are
more
likely
to
get
involved
in
delinquent
behaviors
compared
to
bullied
girls
(Sullivan,
Farrell,
&
Kliewer,
2006)
while
others
have
found
that
bullied
boys
and
girls
are
equally
likely
to
subsequently
engage
in
delinquency
(Barker
et
al.,
2008;
Cullen,
Unnever,
Hartman,
Turner,
&
Agnew,
2008).
Findings
further
indicated
that
internal
resources
such
as
self-esteem
are
influential
in
successfully
overcoming
victim-
ization
experiences
as
previously
found
with
samples
of
sexually
victimized
and
maltreated
youth
(Turner
et
al.,
2010).
Furthermore,
our
study
found
that
negative
emotionality
is
negatively
associated
to
emotional
resilience
to
bullying
sup-
porting
earlier
findings
in
the
literature
that
have
linked
negative
emotionality
with
greater
risk
of
depression
(Doane
et
al.,
2011).
This
study
also
examined
the
role
of
family
and
peer
characteristics
in
promoting
positive
outcomes
for
bullied
students
in
adolescence.
Results
indicated
that
adolescents
who
experienced
no
or
low
levels
of
conflict
with
their
parents
were
more
likely
to
report
low
levels
of
depression
and
delinquency
following
bullying
victimization.
These
results
are
similar
to
another
study
that
examined
resilience
to
bullying
among
children
(Bowes
et
al.,
2010).
This
finding
adds
to
the
previous
literature
by
showing
that
parents
maintain
a
buffering
role
against
victimization
experiences
in
early
adolescence,
contrary
to
what
some
recent
research
has
suggested
(Reavis,
Keane,
&
Calkins,
2010).
The
mechanisms
by
which
relationships
with
parents
protect
against
maladjustment
of
bullied
students
are
not
well-understood
but
there
is
an
indication
from
some
studies
that
support
from
within
and
outside
the
family
promotes
self-esteem
and
induces
positive
feelings
about
one’s
relationships
with
others,
thereby
increasing
an
individual’s
perception
that
he
or
she
can
cope
effectively
with
negative
experiences
such
as
bullying
(Burton,
Stice,
&
Seeley,
2004;
Ueno,
2005).
It
has
further
been
suggested
that
a
positive
family
environment
may
serve
a
stress-relieving
function,
enabling
adolescents
to
cope
more
effectively
with
the
emotional
sequelae
of
bullying
(Bowes
et
al.,
2010).
In
addition,
this
study
confirmed
that
a
victimization-free
sibling
environment
can
make
a
contribution,
albeit
small,
to
developmental
outcomes
for
bullied
adolescents
over
and
above
the
effects
of
parental
relationships
as
found
in
previous
studies
of
younger
samples
(Bowes
et
al.,
2010).
This
points
to
the
need
to
give
siblings
some
role
in
interventions
aimed
to
improve
health
outcomes
for
adolescents.
In
the
present
study,
having
less
close
friends
predicted
lower
than
expected
levels
of
delinquency
following
bullying
victimization.
This
can
be
explained
from
a
routine
activities
perspective
according
to
which
young
people
who
spend
more
time
socializing
with
peers
are
at
an
increased
risk
of
crime
involvement
(Osgood,
Wilson,
O’Malley,
Bachman,
&
Johnston,
1996).
The
reported
lack
of
an
association
between
the
number
of
close
friends
and
levels</