Content uploaded by James O'Higgins Norman
Author content
All content in this area was uploaded by James O'Higgins Norman on Nov 17, 2021
Content may be subject to copyright.
British Journal of Educational Psychology (2019)
©2019 The British Psychological Society
www.wileyonlinelibrary.com
Bullying by siblings and peers: Poly-setting
victimization and the association with problem
behaviours and depression
Mair
ead Foody
1
*, Muthanna Samara
2
and James O’Higgins
Norman
1
1
Anti-Bullying Research and Resource Centre, Dublin City University, Ireland
2
Department of Psychology, Kingston University London, UK
Background. In recent times, sibling bullying has emerged of interest to researchers
concerned with the emotional and behavioural implications for victimization regardless of
type and setting.
Aims. This research attempts to extend current knowledge on both peer and sibling
bullying and to determine the effects of poly-setting victimization. This paper is concerned
with the following objectives: (1) determining the current rate of bullying and victimization
among siblings and peers in a large sample of adolescents; (2) investigating the relationship
between sibling and peer bullying and depression and behaviour; (3) highlighting the carry-
over effects of bullying from one setting to another; and (4) determining the overall
association of poly-setting victimization with depression and behaviour.
Sample and methods. Over 2,000 adolescents aged between 12 and 15 years
participated in an online survey.
Results. Results found lower rates of sibling bullying compared to international studies.
Sibling victims of bullying were at increased risk of becoming peer victims. Poor friendship
quality, disliking school, along with peer and sibling bullying involvement predicted scores
in the clinical range for outcome measures of internalizing and externalizing problem.
Conclusions. The current study has clinical and educational implications for working
with all important stakeholders (i.e., schools, parents, siblings) to reduce bullying and
improve mental health.
The burgeoning field of peer bullying research has provided much-needed insights into
child and adolescent development, such as the association with childhood adversity and
psychopathology in later life (Arseneault et al., 2011). Regardless of the methods used,
research has demonstrated a significant link between bullying experiences and social,
behavioural, and psychological problems. For peer bullying, exposure to such incidences
has been associated with anxiety, depression, psychosis, lower self-esteem, borderline
personality disorder, and even suicide across all age groups (Fisher et al., 2013; Kelleher
et al., 2013; Patchin & Hinduja, 2010; Winsper, Hall, Strauss, & Wolke, 2017; Wolke,
Copeland, Angold, & Costello, 2013). Other factors such as lower academic achievement
*Correspondence should be addressed to Mair
ead Foody, Anti-Bullying Research and Resource Centre, Dublin City University,
Drumcondra, Dublin D09 DY00, Ireland (email: Mairead.foody@dcu.ie).
DOI:10.1111/bjep.12311
1
and early school leaving are demonstrated outcomes of bullying (Cornell, Huang, Grego ry,
& Xitao, 2013; Hammig & Jozkowski, 2013), and the impact has been shown to be long-
lasting (Takizawa, Maughan, & Arseneault, 2014).
While the literature is abundant with studies on the impact of peer bullying, bullying
between brothers and sisters is less researched. Sibling bullying is the term used to refer to
bullying behaviour when occurring between siblings of any age and gender. For many,
aggression between siblings is considered a normal part of childhood and individuals are
generally more complacent about sibling violence compared to peer violence (Reese-
Weber, 2008). One systematic review in 2015 found only 19 studies that specifically
investigated risk factors, relationship to peer bullying, and/or the mental health
consequences of sibling bullying (Wolke, Tippett, & Dantchev, 2015). The same review
found the rate of sibling bullying to be generally higher than peer bullying (e.g., 10–40%
compared to 2–20%, respectively), with similar results documented elsewhere in the
literature (Hoffman, Kiecolt, & Edwards, 2005). It is astonishing that this research area has
not developed more, especially when one considers that children spend more time with
siblings than parents by the time they reach middle childhood (Faith, Elledge, Newgent, &
Cavell, 2015). Despite this, studies have recognized sibling bullying as one of the most
prevalent, long-lasting, and damaging experiences for young people (e.g., Bowes, Wolke,
Joinson, Lereya, & Lewis, 2014; Dantchev, Zammit, & Wolke, 2018; Hardy, 2001; Khan &
Cooke, 2013).
For the most part, sibling bullying is defined in the same manner that peer bullying is,
only that it applies to the unique relationship between siblings of any age or gender. As
such, there are particular elements specific to sibling bullying which separate it from
aggression, harassment, and fighting. These include repetition, intent to hurt, negative
outcomes, and a power hierarchy (Olweus, 1991). In general, research suggests that
individuals involved in sibling bullying are more likely than those not involved to report
mental health problems (Tucker, Finkelhor, Turner, & Shattuck, 2013), abuse substances
(Button & Gealt, 2010), and engage in anti-social behaviour towards peers (Ensor, Marks,
Jacobs, & Hughes, 2010; Menesini, Camodeca, & Nocentini, 2010). As the number of
siblings increases, there is also an increase in the likelihood of both bullying perpetration
and victimization (Bowes et al., 2014; Tippett & Wolke, 2015). In addition, the outcomes
for those involved in both sibling and peer bullying can be particularly damaging
(Dantchev et al., 2018; Duncan, 1999). Furthermore, the relationship between sibling
and peer bullying appears to have the potential to predict, in that poor sibling
relationships may influence and even determine poor peer relationships (e.g., Defoe
et al., 2013; Johnson et al., 2014).
While there have been some international investigations of sibling bullying elsewhere
(e.g., Tanrikulu & Campbell, 2015; Wolke & Samara, 2004), this is the first large-scale study
to investigate this issue with adolescents in the Republic of Ireland. For the most part,
current research has established a link between sibling bullying and mental health
problems (particularly to depression; Bar-Zomer & Brunstein Klomek, 2018) and on the
role of social support in preventing internalizing problems (e.g., Coyle, Demaray, Malecki,
Tennant, & Klossing, 2017). As such, this study aims to build on previous international
research on the implications of sibling bullying, while investigating this issue with an Irish
population.
We were specifically concerned with the following research questions: What is the
current rate of bullying and victimization among siblings and does it differ for specific
types of bullying behaviour including physical, verbal, and relational bullying? What are
the carry-over effects between sibling and peer bullying? What is the relationship between
2Mair
ead Foody et al.
involvement in sibling bullying and internalizing and externalizing problems? What is the
relationship of poly-setting bullying and/or victimization (i.e., sibling victimization and
peer victimization) and internalizing and externalizing problems? And finally, is
involvement in sibling and/or peer bullying associated with levels of internalizing and
externalizing problems in a cross-sectional sample?
Method
This study involved a cross-sectional analysis of sibling and peer bullying involvement in
teenagers aged 12–15 years, M(SD): 13.5(1), and attending 1st to 3rd year in post-primary
schools across Ireland. All post-primary schools in the country (N=811) were initially
contacted by email asking them to take part in the study. A reminder email was sent
2 weeks later. Thirty-two schools initially responded and agreed to participate, two of
which later declined, leaving 30 participating schools representing 3.7% of the entire post-
primary population. Originally, 2,606 students were recruited from these schools;
however, 196 students declined to participate by selecting the relevant option when
presented with the survey. This resulted in a final sample of 2,410 junior cycle post-
primary students, 1,018 males (43.2%) and 1,338 (56.8%) females. Out of the final sample,
2,144 participants (87.7%) were Irish (see Table 1). The number of participants who
reported having at least one sibling was 2,247 (93%). The mean number of siblings was
2.3.
Ethical issues
This study received ethical approval from the first author’s university ethics review board.
Principals were contacted initially, and written information about the study was provided
by email. Once consent was obtained at this level, parental information and consent forms
were provided to parents by the participating schools. The survey was delivered online
and took place during one class sitting. A unique link to complete the survey was given to
each participating school and the principal shared it with the teachers of the respective
Table 1. Gender, year group, and nationality of the sample
N% of sample
Gender
Male 1,018 42.2
Female 1,338 55.5
Missing
a
54 2.3
Post-primary school year
First 1,257 52.2
Second 1,024 42.5
Third 21 0.9
Missing
a
108 4.4
Nationality
Irish 2,144 87.7
Non-Irish 274 11.4
Missing
a
22 0.9
Note.
a
Questions did not include mandatory answers, so some questions left unanswered and labelled as
missing.
Sibling and peer bullying in Ireland 3
class groups. Before the students were presented with the survey, they were first given a
plain language statement about the research. Students were also informed that they do not
have to complete the survey and were free to stop participating at any time. Students had
to actively select a response saying they gave their consent before they were able to access
the survey. Responses were completely anonymous at both the pupil and school level.
Data collection took place between March and May 2017.
Survey instruments
Demographic variables
Participants were first asked to provide information about their gender (male/female),
year group (first, second, or third year), nationality (stating it specifically), and the number
of siblings they had.
Bullying questionnaires
Participants were presented with the following definition of bullying based on the Olweus
Bullying Questionnaire (OBQ, 1996):
We say a student is being bullied when another student, or several other students: (1) say
mean and hurtful things or make fun of him/her or call him/her mean and hurtful names; (2)
completely ignore or exclude him/her from their group of friends or leave him/her out of
things on purpose; (3) hit, kick, push, shove around or lock him/her inside a room; (4) tell lies
or spread false rumours about him/her or send mean notes and try to make other students
dislike him/her; (5) and any other hurtful things like that. When we talk about bullying, these
things happen repeatedly, and it is difficult for the student being bullied to defend himself or
herself. We also call it bullying, when a student is teased repeatedly in a mean and hurtful way.
But we don’t call it bullying when the teasing is done in a friendly and playful way. It is not
bullying when two students of about equal strength or power argue or fight.
Peer bullying instrument. Involvement in peer bullying was determined using a
modified version of the OBQ. This included three questions relating to physical bullying
(e.g., I was hit, kicked, pushed, shoved around, or locked indoors); three questions
relating to verbal bullying (e.g., I was called mean names, was made fun of, or teased in a
hurtful way); and two relating to relational bullying (e.g., other students let me out of
things on purpose, excluded me from their group of friends or completely ignored me).
Response options were as follows: ‘I haven’t been bullied in school (0), it has only
happened once or twice (1), 2 or 3 times a month (2), about once a week (3), and several
times a week (4)’. Responses were coded as not involved (0 and 1) and involved (2, 3,4)
and then all questions relating to specific bullying type were combined to give an overall
variable for involvement in three types of bullying: physical, verbal, and relational. Similar
questions and coding were asked in relation to peer bullying perpetration. We categorized
the variable into pure bullies (frequently involved in bullying others (2, 3, 4) but never or
rarely victimized 0, 1), pure victims (frequently involved in victimization (2, 3, 4) but
never or rarely bullied others 0, 1), bully–victims (frequently involved in bullying others
and victimization 2, 3, 4), and neutrals (never or rarely bullied and victimized 0, 1). This
instrument had good internal consistency in the current study (Cronbach alpha
coefficient was .78 for victimization and .9 for bullying perpetration).
4Mair
ead Foody et al.
Happiness with school. Participants were asked a question relating to how much they
liked school: ‘How much do you like school?’ Participants chose the answer that applied
to them the most from the following response options: I dislike school very much; I dislike
school; I neither like nor dislike school; I like school; and I like school very much.
Sibling bullying. Involvement in sibling bullying was measured with a 7-item scale
based on a modified version of the OBQ used in a previous study (Wolke & Samara, 2004).
Participants were given the following instruction before being presented with the items:
During the last 3 months, please say if any of these things have been done to you by a
brother or sister on purpose. Two questions related to physical bullying (e.g., I was hit,
kicked, pushed, or threatened); one item measured verbal bullying (e.g., I was called bad
or nasty names); and two measured relational bullying (e.g., I was tricked in nasty way).
Response options were as follows: never (0), it has only happened once or twice (1), 2 or 3
times a month (2), about once a week (3), and several times a week (4). Responses were
coded as not involved (0 and 1) and involved (2, 3,4) and then all questions relating to
specific bullying type were combined to give an overall variable for involvement in three
types of bullying: physical, verbal, and relational. Similar questions and coding were asked
in relation to sibling bullying perpetration (e.g., during the last 3 months, please say if you
have done any of the following things to a brother or sister on purpose). This instrument
had good internal consistency in the current study (Cronbach alpha coeffic ient was .87 for
sibling victimization and .89 for sibling bullying). The same classification was done for
sibling bullying to pure bullies, pure victims, bully–victims, and neutrals.
Behaviour questionnaire. The Strengths and Difficulties Questionnaire (SDQ) (Good-
man, 1997, 2001) was administered to all pupils (www.sdqinfo.com). It includes five
subscales which relate to conduct (e.g., I get very angry), emotional (e.g., I worry a lot),
peer (e.g., I am usually on my own), and hyperactivity problems (e.g., I am easily
distracted), as well as prosocial behaviour (e.g., I try to be nice to other people). Higher
scores indicate higher levels of each category. Information about how the SDQ was coded
and analysed is provided in the ‘Statistical Analysis’ section. This instrument had
satisfactory internal consistency in the current study for the total difficulties scale
(Cronbach alpha coefficient was .77); emotional subscale (Cronbach alpha coefficient
was .77); conduct problems (Cronbach alpha coefficient was .59); hyperactivity
(Cronbach alpha coefficient was .6); and prosocial subscales (Cronbach alpha coefficient
was .79). The reliability for the peer problems subscale (Cronbach alpha was .26) was
considered too low to be included in further analysis but the conduct problems and
hyperactivity subscales were retained.
Moods and feelings questionnaire. The Moods and Feelings Questionnaire short
version (MFQ, Angold et al., 1995; Messer et al., 1995) was used to determine how
participants were feeling in the past 2 weeks. Previous studies have found it to be a
reliable and valid measure of depression in populations aged 6–17 years (e.g., Jeffreys
et al., 2016). Answer options included: not true (0), sometimes true (1), and true (2). A
higher overall score indicates higher depression. This instrument had good internal
consistency in the current study (Cronbach alpha coefficient was .93).
Sibling and peer bullying in Ireland 5
Cambridge hormones and moods friendship quality questionnaire. A modified
version of this scale was included to investigate the quality of the friendships the
participants reported having with their peers (Goodyer, Wright, & Altham, 1989, 1990). It
contained five questions: (1) Are you happy with the number of friends you have? (2) Do
your friends know what makes you happy or sad? (3) How often do you see your friends
outside of school? (4) Do you talk to your friends about problems? And (5) overall, are you
happy with your friends? Response options ranged from simple YES/NO answers (e.g.,
Q2) to Likert-type answers where a higher number for the coded response represented
poorer friendship quality (e.g., Q5: very happy (1); quite happy (2); quite unhappy (3);
and unhappy (4)). This instrument had Cronbach alpha coefficient of .6.
Statistical analysis
The statistical software package SPSS version 24, (https://www.ibm.com/analytics/spss-
statistics-software) was used to conduct all analysis. Chi-square test for independence was
used to determine whether there were gender differences in involvement in the various
types and roles of sibling bullying and victimization, as well as to determine the carry-over
effects with peer bullying. To account for the nested nature of the data, we used the
complex samples module taking into account schools as a cluster-level factor and
conducted general linear model ANOVAs. This is to explore the relationship between
involvement in different types of sibling bullying and behaviour and depression levels
(measured by the MFQ, SDQ, and SDQ subscales). We used the same method to
investigate the association with depression and behaviour when individuals were
involved in bullying with siblings, peers, or both (i.e., poly-setting involvement).
Bonferroni corrections with the significance value set to .01 were used.
We also conducted a two-level logistic regression analysis to investigate which
variables predicted higher scores on the MFQ, SDQ, and its subscales. The first level
involved entering the variable ‘school’ only while the second level included the other
independent variables. Scores on the SDQ were re-coded and split for the borderline and
clinical range (≥80th percentile) versus normal range (<80th percentile) as dependent
variables. Using the 80th percentile as a cut-off point for the borderline and clinical range
is standard practice and has been demonstrated as having concordance with DSM-IV
diagnosis (He, Burstein, Schmitz, & Merikangas, 2013). The independent variables
included friendship quality, happiness with school, age, gender, number of siblings,
nationality, and involvement in peer or sibling bullying and/or victimization. Multiple
regression was used to investigate the role of the same variables in predicting scores on the
MFQ (as this involved a continuous variable).
Considering the nature of the student data reported (i.e., nested within schools),
hierarchical linear modelling was also conducted to determine the role of school in the
prevalence of bullying where bullying and victimization by peers and siblings were
considered as outcomes. For this analysis, four new continuous variables were created
for: (1) sibling victimization, (2) sibling bullying, (3) peer victimization, and (4) peer
bullying, where a higher number indicated more of each. We then used HLM to generate a
random-intercept model first with subsequent fixed predictors. A model containing only
school and these four new continuous variables as outcomes was first generated to
determine whether school was significant. A second model was built with fixed predictor
variables (e.g., gender, age). School was added as a random factor while the other
predictors (e.g., age, gender, nationality, friendship) as fixed model factors. Interclass
correlation coefficients (ICCs) were also calculated for each model. The ICC is the
6Mair
ead Foody et al.
proportion of variance in the outcome variable that is explained by the grouping structure
of the hierarchal model. The ICC is used to determine whether there is a significant
clustering of observations within higher level units or in this case, at the school level.
Results
Sibling victimization and bullying
Involvement in sibling and peer bullying was categorized into four groups: bully, victim,
bully–victim (both a victim and a bully), and neutrals (no involvement in bullying) for each
bullying type (see Table 2). When asked who they were bullied by, 25.9% said older
brother, 21.2% said older sister, 25.2% said younger brother, 24.9% said younger sister,
and 2.8% said a mixture of ages/siblings.
Chi-square analysis revealed a significant effect for gender and overall involvement in
sibling bullying (not split for specific types), v
2
(3, n=2,026) =14.4, p<.05, φ=.083;
physical sibling bullying, v
2
(3, n=2,030) =11.2, p<.05, φ=.074; verbal sibling
bullying, v
2
(3, n=2,040) =19.9, p<.001, φ=.099, but not relational sibling bullying
(p>.05). There were more female victims and bully–victims across all types of sibling
bullying involvement (see Table 2).
There was a significant effect for gender and overall involvement in peer bullying, v
2
(3, n=2,246) =11.1, p<.05, φ=.07, as well as in each of the types of peer bullying:
physical bullying, v
2
(3, n=2,256) =10.6, p<.05, φ=.068; verbal, v
2
(3,
n=2,255) =13.2, p<.005, φ=.076; and relational, v
2
(3, n=2,251) =29.1,
p<.001, φ=.114. More males were physical peer bullies, victims, and bully–victims,
while females were more likely to be verbal and relational peer victims compared to males.
Overlap between sibling and peer bullying
Chi-square analysis was used to determine whether there were any carry-over effects
between the subgroups of peer and sibling bullying. Or to put it another way, we wanted
to determine the odds of individuals being involved in peer bullying (across the different
groups) while taking account of their role in sibling bullying. There was a significant carry-
over effect for all peer and sibling bullying subgroups’ relationships (all ps<.001). Odds
ratios demonstrated the relative risk of involvement in sibling bullying/victimization and
peer bullying/victimization (see Table 3). This indicates that all sibling subgroups (bullies,
victims, and bully–victims) were more likely to be peer bullies, victims, or bully–victims
compared to neutrals. The strongest carry-over effect was for the bully–victim subgroup
(sibling bully–victim to peer bully–victim; OR: 12.2) followed by those who are sibling
victims and also peer bully–victims (OR: 7.34). The weakest carry-over effect, although
significant, was from sibling bully to peer victim (OR: 1.98; see Table 3).
Sibling bullying, behaviour, and depression
We conducted general linear model ANOVAs with Bonferroni corrections to determine
the role of sibling bullying involvement on internalizing and externalizing problems (SDQ
and MFQ). Due to the fact that there were several ANOVAs conducted, and an increased
chance of a type 1 error, we set the significance level of .01 (instead of .05). In addition, as
the data were nested among schools, we used the complex samples module in SPSS to
conduct the general linear models and conduct means comparisons. A new plan file was
Sibling and peer bullying in Ireland 7
Table 2. Frequency and prevalence for sibling and peer bullying type, bullying role, and gender
a
Sibling bullying Peer bullying
Physical
N/2,030 (%)
Verbal
N/2,040 (%)
Relational
N/2,039 (%)
Overall
N/2,026 (%)
Physical
N=2,258 (%)
Verbal
N=2,255 (%)
Relational
N=2,251 (%)
Overall
N=2,246 (%)
Bullies 55/2,030 (2.71) 66/2,040 (3.24) 31/2,039 (1.52) 64/2,026 (3.16) 14/2,258 (0.62) 25/2,255 (1.11) 13/2,251 (0.58) 26/2,246 (1.16)
Males 28/55 (50.9) 34/66 (51.5) 15/31 (48.4) 34/854 (3.98) 9/14 (64.3) 8/25 (32) 11/13 (84.6) 13/26 (50)
Females 27/55 (49.1) 32/66 (48.5) 16/31 (51.6) 30/1172 (2.56) 5/14 (35.7) 17/25 (68) 2/13 (15.4) 13/26 (50)
Victims 224/2,030 (11) 208/2,040 (10.2) 157/2,039 (7.70) 268/2,026 (13.2) 100/2,258 (4.43) 175/2,255 (7.76) 261/2,251 (11.6) 330/2,246 (14.7)
Males 92/224 (41.1) 66/208 (31.7) 50/157 (31.8) 97/854 (11.4) 51/100 (51) 73/175 (41.7) 77/261 (29.5) 116/330 (35.2)
Females 132/224 (58.9) 142/208 (68.3) 107/157 (68.2) 171/1,172 (14.6) 49/100 (49) 102/175 (58.3) 184/261 (70.4) 214/330 (64.8)
Bully–Victims 213/2,030 (10.5) 228/2,040 (11.2) 68/2,039 (3.34) 313/2,026 (15.4) 9/2,258 (0.40) 23/2,255 (1.02) 9/2,251 (0.40) 38/2,246 (1.69)
Males 69/213 (32.4) 78/228 (34.2) 30/68 (44.1) 113/854 (13.2) 7/9 (77.8) 18/23 (78.3) 4/9 (44.4) 21/38 (55.3)
Females 144/213 (67.6) 150/228 (65.8) 38/68 (55.9) 200/1,172 (17.1) 2/9 (22.2) 5/23 (21.7) 5/9 (55.6) 17/38 (44.7)
Neutrals 1,538/2,030 (75.8) 1,538/2,040 (75.4) 1,783/2,039 (87.4) 1,381/2,026 (68.2) 2,135/2,258 (94.6) 2,032/2,255 (90.1) 1,968/2,251 (87.4) 1,852/2,246 (82.5)
Males 668/1,538 (43.4) 678/1,538 (44.1) 762/1,783 (42.7) 610/854 (71.4) 894/2,135 (41.9) 862/2,032 (42.4) 864/1,968 (43.9) 804/1,852 (43.4)
Females 870/1,538 (56.6) 860/1,538 (55.9) 1,021/1783 (57.3) 771/1,172 (65.8) 1,241/2135 (58.1) 1,170/2,032 (57.6) 1,104/1,968 (56.1) 1,048/1,852 (56.6)
Note.
a
Answers were not mandatory so frequencies or percentages do not add to total sample number.
8Mair
ead Foody et al.
first established with ‘school’ inputted at stage one as a ‘cluster’. The results found
significant differences between sibling physical bullying involvement on the MFQ, (Wald
F(3, 23) =40.1, p<.001], total difficulties, Wald F(3, 23) =34.4, p<.001, emotional
problems, Wald F(3, 23) =15.2, p<.001, conduct problems, Wald F(3,23) =30.7,
p<.001, hyperactivity, Wald F(3, 23) =9.93, p<.001, and prosocial behaviour, Wald F
(3, 23) =14.8, p<.001, scales (see Table 4).
There was a significant difference between involvement in relational sibling bullying
subgroups on the MFQ, Wald F(3, 23) =42.7, p<.001, total difficulties, Wald F
(3,23) =34.2, p<.001, emotional problems, Wald F(3, 23) =15.2, p<.001, conduct
problems, Wald F(3, 23) =37.5, p<.001, hyperactivity, Wald F(3, 23) =9.12, p<.01,
and prosocial behaviour, Wald F(3, 23 =14.8, p<.001, scales.
In addition, there was a significant difference for verbal sibling bullying subgroups on
the MFQ, Wald F(3, 23) =68.8, p<.001, total difficulties, Wald F(3, 23) =28.9,
p<.001, emotional problems, Wald F(3, 23) =18.1, p<.001, conduct problems, Wald F
(3, 23) =30.9, p<.01, hyperactivity, (Wald F(3, 23) =9.68, p<.001, and prosocial
behaviour, Wald F(3, 23) =5.01, p<.01, scales. Post hoc comparisons using Bonferroni
indicated a range of significant differences between victims, bullies, bully–victims when
compared to neutrals (see Table 4).
Involvement in multiple settings and association with depression and behaviour
Using the same procedure as above, we also investigated overall involvement, regardless
of bullying type (e.g., physical, relational, or verbal), and association with the SDQ, SDQ
subscales, and MFQ. Participants were split into four groups: neutrals, sibling only, peer
only, and poly for victimization, perpetration, and bully–victims (see Table S1).
Victimization
General linear model analyses revealed significant differences between the four groups on
the MFQ, F(3, 23) =157, p<.001, total, F(3, 23) =20.7, p<.001, emotional, F(3,
Table 3. Carry-over effect from sibling bullying and peer bullying compared to no carry-over effect
Peer
Bully (%)
Carry-over
effect from
sibling to peer
bullying Peer
Bully–Victim
(%)
No carry-over
effect from
sibling bullying
to peer bullying
OR CI Sig (p)
Peer
victim (%)
No carry-over
effect
Sibling bully 49.3 18.4 4.32 2.66–7.02 <.001
29.3 17.3 1.98 1.52–2.57 <.001
57.1 17.1 6.47 3.47–12.1 <.001
Sibling victim 60 28.7 3.73 2.29–6.07 <.001
46.6 26.2 2.44 1.93–3.09 <.001
72.1 26.0 7.34 3.74–14.4 <.001
Sibling
bully–victim
56.9 18.2 5.93 3.36–10.5 <.001
33.1 16.8 2.44 1.82–3.27 <.001
71 16.6 12.2 5.57–26.9 <.001
Note.CI=confidence interval; OR =odds ratio.
Sibling and peer bullying in Ireland 9
23) =25.4, p<.001, conduct, F(3, 23) =13.2, p<.001, and hyperactivity, F(3,
23) =6.01, p<.01, subscales. The significance level was set to .01 or .001, and all
differences from neutrals are noted in Table S1. Extra-group comparisons indicated that
poly-victim scores were significantly higher than sibling victims on the MFQ (p<.001).
Bullying perpetration
There were significant differences for perpetration between the four groups on the MFQ,
F(3, 23) =22.4, p<.001, total difficulties, F(3, 23) =29.3, p<.001, emotional, F(3,
23) =5.30, p<.01, conduct, F(3, 23) =34.2, p<.001, hyperactivity, F(3, 23) =8.05,
p<.001, and prosocial behaviour scale, F(3, 23) =4.93, p<.01. Extra-group compar-
isons indicated that peer bullies had significantly less conduct problems compared to
sibling and poly-bullies (p<.001 and p<.01, respectively). Poly-bullies demonstrated
significantly higher depression scores than sibling bullies (p<.001). Poly-bullies showed
significantly higher scores than peer bullies on the total difficulties scale (p<.001; see
Table S1).
Bully–victims
General linear model analyses revealed significant differences for the groups on the MFQ,
F(3, 23) =85.2, p<.001, total difficulties, F(3, 23) =19.5, p<.001, emotional
Table 4. Means (M) and standard error (SE) for the SDQ total difficulties, SDQ subscales, and
depression for sibling bullying subgroups and types
Variable
Neutral Victim Bully Bully–Victim
M(SE)M(SE)M(SE)M(SE)
Physical
Total difficulties 13.3 (0.28) 15.7 (0.62)** 15.8 (1.08) 16.8 (0.34)***
Emotional problems 3.82 (0.15) 4.78 (0.29)** 3.76 (0.42) 4.63 (0.15)***
Conduct problems 2.32 (0.10) 2.79 (0.15) 3.48 (0.36) 3.65 (0.16)***
Hyperactivity 4.46 (0.09) 4.90 (0.23) 5.36 (0.30) 5.36 (0.15)***
Prosocial behaviour 7.45 (0.14) 7.43 (0.18) 6.5 (0.36) 6.9 (0.18)
MFQ 4.17 (0.20) 6.44 (0.50)*** 6.06 (0.72) 6.34 (0.3)***
Relational
Total difficulties 13.5 (0.26) 17.2 (0.77)*** 17 (1.03)*** 18.4 (0.97)***
Emotional problems 3.85 (0.14) 5.28 (0.34)*** 4.50 (0.55) 4.86 (0.37)
Conduct problems 2.37 (0.10) 3.26 (0.24)** 4.47 (0.32)*** 4.33 (0.38)***
Hyperactivity 4.52 (0.08) 5.17 (0.20)** 5.69 (0.34)** 5.54 (0.31)**
Prosocial behaviour 7.43 (0.13) 7.44 (0.25) 5.25 (0.37)*** 6.66 (0.32)**
MFQ 4.31 (0.20) 7.28 (0.47)*** 7.63 (0.69)*** 6.70 (0.64)**
Verbal
Total difficulties 13.3 (0.28) 16.0 (0.56)*** 15.3 (0.95) 17 (0.42)***
Emotional problems 3.77 (0.15) 5.05 (0.25)*** 3.44 (0.47) 4.89 (0.19)***
Conduct problems 2.32 (0.11) 3.54 (0.34) 2.83 (0.13) 3.65 (0.19)**
Hyperactivity 4.47 (0.08) 4.99 (0.20) 5.33 (0.23)** 5.24 (0.15)***
Prosocial behaviour 7.42 (0.14) 7.50 (0.16) 6.18 (0.33)** 7.10 (0.18)
MFQ 4.02 (0.21) 6.8 (0.35)*** 5.59 (0.57) 7.19 (0.35)***
Notes.**Significant difference compared to neutrals (p<.01).
***Significant difference compared to neutrals (p<.001).
10 Mair
ead Foody et al.
Table 5. Two-level regression models at level 1 (school level only) and level 2 (all predictors) on the
SDQ total difficulties, SDQ subscales, and depression: unstandardized coefficients
Variable BSEbWald df
95% confidence
interval for B
Sig. (p)Lower Upper
Total difficulties
Level 1
School 0.024 0.006 1.03 17.6 1 1.01 1.04 <.001
Level 2
a
School 0.023 0.006 1.02 14.3 1 1.01 1.02 <.001
Friendship quality (poorer) 0.060 0.020 1.06 8.88 1 1.06 1.02 <.01
Happiness with school
(/dislike)
0.33 0.052 0.72 40.1 1 0.65 0.80 <.001
Peer victim 1.01 0.14 2.74 51.6 1 2.74 2.08 <.001
Peer bully–victim 0.89 0.39 2.43 5.11 1 2.43 1.13 <.05
Sibling victim 0.50 0.15 1.65 10.7 1 1.65 1.22 <.001
Sibling bully–victim 0.38 0.15 1.46 6.71 1 1.46 1.10 <.01
Emotional
Level 1
School 0.005 0.005 1.01 0.84 1 0.99 1.02 >.05
Level 2
Friendship quality (poorer) 0.080 0.020 1.08 16.4 1 1.04 1.13 <.001
Gender (Female) 0.94 0.11 2.56 71 1 2.06 3.18 <.001
How do you like school
(/less)
0.16 0.051 0.86 9.31 1 0.78 0.95 <.01
Peer victim 0.10 0.14 2.71 50.5 1 2.06 3.56 <.001
Peer bully–victim 0.82 0.39 2.26 4.32 1 1.05 4.89 <.05
Sibling victim 0.52 0.15 1.67 12 1 1.25 2.24 <.001
Conduct problems
Level 1
School 0.010 0.005 1.01 3.54 1 1.00 1.02 >.05
Level 2
Gender (male) 0.502 0.11 0.61 22.4 1 0.49 0.75 <.001
How old are you (/younger) 0.18 0.068 0.84 6.9 1 0.73 0.96 <.01
How much do you like school
(/less)
0.27 0.051 0.76 29.4 1 0.69 0.84 <.001
Peer victim 0.53 0.142 1.70 14 1 1.29 2.24 <.001
Peer bully 1.5 0.46 4.48 10.8 1 1.83 11 <.01
Peer bully–victim 1.30 0.41 3.66 10 1 1.64 8.17 <.01
Sibling bully 0.56 0.27 1.75 4.12 1 1.02 2.30 <.05
Sibling bully–victim 0.77 0.14 2.17 31.5 1 1.65 2.84 <.001
Hyperactivity
Level 1
School 0.017 0.005 1.02 10.5 1 1.01 1.03 <.001
Level 2
School 0.016 0.005 1.02 8.97 1 1.01 1.03 <.01
How much do you like school
(/less)
0.33 0.048 0.72 46.6 1 0.66 0.79 <.001
Peer victim 0.38 0.14 1.46 7.75 1 1.12 1.91 <.01
Continued
Sibling and peer bullying in Ireland 11
problems, F(3, 23) =11.3, p<.001, conduct, F(3, 23) =25.2, p<.001, hyperactivity, F
(3, 23) =8.73, p<.01, and prosocial behaviour scale, F(3, 23) =4.55, p<.01. Extra-
group comparisons indicated sibling bully–victims were significantly more prosocial
compared to poly-bully–victims (p<.01), while poly-bully–victims had significantly
higher scores on the MFQ compared to sibling bully–victims (p<.01; see Table S1).
Predicting externalizing and internalizing problems
Taking into account the clustered nature of the data (within schools), two-level regression
analyses were conducted to test the predictors of being in the borderline/clinical range for
total difficulties, emotional problems, conduct problems, hyperactivity, and prosocial
behaviour. The first level tested differences at the school level, while the second level in
Table 5. (Continued)
Variable BSEbWald df
95% confidence
interval for B
Sig. (p)Lower Upper
Peer bully–victim 0.97 0.41 2.64 5.71 1 1.19 5.85 <.05
Sibling victim 0.33 0.14 1.39 5.24 1 1.05 1.83 <.05
Sibling bully 0.85 0.27 2.34 9.84 1 1.38 3.97 <.01
Prosocial
Level 1
School 0.002 0.007 1.0 0.13 1 0.99 1.01 >.05
Level 2
Nationality (non-Irish) 0.45 0.19 1.57 5.87 1 1.09 2.27 <.05
Friendship quality (poorer) 0.057 0.023 1.06 6.06 1 1.01 1.11 <.05
Gender (male) 0.77 0.13 0.46 36.1 1 0.36 0.59 <.001
How much do you like school
(/less)
0.43 0.060 0.65 50.7 1 0.58 0.73 <.001
Peer bully 1.00 0.46 2.73 5.85 1 1.12 6.68 <.05
Sibling bully 0.71 0.29 2.04 5.91 1 1.15 3.62 <.05
Depression
Variable BSEbt
95% confidence
interval for B
Sig. (p)
Lower
bond
Upper
bond
Level 1
School 0.019 0.013 0.040 1.4 0.045 0.007 >.05
Level 2
Gender (Female) 1.83 0.24 0.21 7.69 1.36 2.30 <.001
How much do you like school
(-/less)
0.51 0.12 0.12 4.30 0.74 0.27 <.001
Friendship quality (higher) 0.36 0.05 0.19 7.22 0.26 0.46 <.001
Peer victim 3.01 1.02 0.08 2.95 1.01 5.02 <.01
Sibling bully 0.73 .15 0.13 4.76 0.43 1.03 <.001
Note.
a
Only significant pvalues reported for level 2.
12 Mair
ead Foody et al.
each model involved the addition of the predictor variables. The following variables were
entered into the model at the second stage: friendship quality, nationality (Irish or non-
Irish), happiness with school, age, gender, number of siblings, and involvement in peer or
sibling bullying (see Table 5). The school variable was significant at level one for total
difficulties, v
2
(1, 1943) =18, p<.001, while the full model (with the predictors school,
poor friendship quality, disliking school, being a peer victim, peer bully–victim, sibling
victim, and a sibling bully–victim) was also significant, v
2
(13, 1943) =194, p>.001, see
Table 5. School was not significant at level 1 for the emotional subscale but the full model
with six significant variables was significant, v
2
(13, 1943) =211, p<.001, see Table 5.
Similar results were found for conduct problems where level 1 predictor school variable
was not significant but the full model with eight significant variables was significant, v
2
(13, 1943) =155, p<.001. School was a significant predictor at level 1 for hyperactivity,
v
2
(1, 1943) =10.7, p<.001, as was the full model with six significant variables, v
2
(13,
1943) =105, p<.001. School was not significant at level 1 for prosocial behaviour, but
the full model with six significant variables was significant, v
2
(13, 1943) =146] (all
ps<.001).
Multiple regression analysis (DV: total MFQ score) was employed to determine the role
of friendship quality, happiness with school, age, gender, number of siblings, involvement
in peer or sibling bullying, and nationality on depression levels. Five significant
contributing factors were significant (see Table 5). The final total model for depression
was significant (R
2
=.12, adjusted =.16; p<.001).
In terms of peer bullying, results from the HLM found that the first model indicated that
school was significant for peer victimization (p<.05), and not for peer bullying
perpetration (p>.05). The second model indicated that school, poor friendship quality,
disliking school, and being of non-Irish nationality were significant predictor variables for
peer victimization, while only negative friendship quality, disliking school, and non-Irish
nationality were significant predictors of peer bullying (see Table S2). For sibling
victimization, the first model with school only was not significant (p>.05) while the
model for sibling perpetration was significant (p<.05). Being female, disliking school,
and having poor friendship quality were significant predictors for sibling victimization (all
ps<.01). Being female, having higher number of siblings and disliking school were
significant predictors for being a sibling bully (all ps<.05).
Discussion
This study aimed to investigate bullying between siblings in Ireland and to provide a
detailed analysis of prevalence for types of sibling and peer bullying (physical, verbal, and
relational). In addition, it documents the carry-over effect of bullying involvement from
one setting to another (i.e., sibling to peer) and demonstrates the mental health and
behaviour implications of poly-involvement for all of the roles (victim, bully, and bully–
victim). The majority of this sample (93.5%) reported having at least one sibling, and the
average number of siblings was 2.3. The central statistics office in Ireland reported an
average number of children per family as 1.38 in 2016, although it has been reported to be
on the rise in recent years. This is considered one of the highest rates in Europe (European
Commission, 2015). These differences may arise from the lack of a definition of sibling
bullying provided in the current study. For example, the central statistics office refers to
the number of ‘children per family’ (either for a couple or a single parent). Our study may
suggest a higher number of children per family because participants were simply asked to
Sibling and peer bullying in Ireland 13
report the number of brothers and sisters they had. No definition of ‘sibling’ was provided,
and so, participants could have reported on all brothers or sisters that they had (including
step-siblings, half-siblings, or foster siblings), as opposed to those connected to only one
family unit (one or two parents). Future research needs to take this into account and
ensure that all participants have a clear understanding of what the question refers to when
using the terms sibling, half-sibling, step-sibling etc.
For sibling bullying, 13.2% reported being sibling victims, 3.2% reported being sibling
bullies, and 15.4% reported being bully–victims. These prevalence rates are generally
much lower than international comparisons. For example, one study in the United
Kingdom reported rates as high as 45% for victims of sibling aggression (Tippett & Wolke,
2015). It is likely that the strict definition of bullying prohibited individuals from reporting
sibling aggression or more general negative interactions with siblings. By providing a
definition of bullying to the participants, we attempted to clearly articulate what this
meant so that students did not confuse it with aggression or even fighting. Of course, our
findings are limited in that we cannot say with absolute certainty that this is how
participants interpreted the concept. Much like the peer bullying literature, the sibling
bullying literature may struggle with definitional issues and with ensuring that all
participants understand the concept in a similar manner.
For peer bullying, 14.7% reported being victims, 1.2% reported being peer bullies, and
1.7% reported being bully–victims. The rate of peer victimization is similar to that
reported in a recent meta-analysis of all the studies published in Ireland on bullying in the
last 20 years (Foody, Samara, & O’Higgins Norman, 2017). However, for peer bullying
perpetration, the rate appears much lower than the figure of 6.9% generated by the same
meta-analysis. It is difficult to determine why this may be the case. One explanation is that
the current study used a strict definition of bullying with regard to repetition and
considered the first two Likert answer options (i.e., I haven’t been bullied in school and It
has only happened once or twice) as neutrals (no bullying involvement). Previous
literature has highlighted that the coding of answers can greatly modify bullying rates from
one study to the next (Foody et al., 2017).
In terms of gender, females were more likely to be sibling victims and sibling bully–
victims for physical and verbal bullying when compared to males, while gender
differences were minimal for sibling bullies with regard to all types of bullying
perpetration. There were obvious gender differences for involvement in peer bullying
(see Table 2). These results are generally in keeping with international comparisons in the
peer bullying literature which demonstrate that males are more likely to be physical
bullies (e.g., Silva, Pereira, Mendoncßa, Nunes, & de Oliveira, 2013) and females report
more relational victimization compared to males (Wang, Iannotti, & Nansel, 2009).
However, the results from the sibling bullying data show that gender does not provide the
same indicator of bullying involvement between brothers and sisters.
The carry-over analysis gave a clearer picture of the overlap between sibling and peer
bullying. The strongest carry-over effect was for the bully–victim subgroup (sibling bully–
victim to peer bully–victim) followed by the sibling victim to peer bully–victim. The
weakest carry-over effect, although significant, was from sibling bully to peer victim. This
is perhaps the most sobering result of this study and it clearly outlines the risk for further
bullying experiences when there is previous exposure at home. It highlights the
importance of parenting and the family unit in the prevention of bullying involvement
(Bar-Zomer & Brunstein Klomek, 2018; Lereya et al., 2013). Furthermore, it suggests that
the effectiveness of anti-bullying programmes at the school level may have limited
effectiveness if not considered within the wider community context.
14 Mair
ead Foody et al.
Interestingly, the prevalence of bully–victims varied a lot from sibling to peer bullying.
The bully–victim group in the peer bullying is usually a small group but it seems that this is
not the case in sibling bullying, where they represented the largest percentage. One
explanation is that siblings exchange roles more regularly as they strive to gain more out of
bullying including resources such as parental attention, affection, love, and other material
gains. In addition, the power imbalance changes more regularly among siblings (from
bully to victims and vice versa) due to the interference of parents and othe r siblings, which
could potentially play less of a role in peer interaction. There are many other
developmental, psychological, and logistical factors which may account for a changing
imbalance of power between siblings such as age, conflict, time together, and/or sibling
spacing and location within the family unit. For example, some studies have suggested
that sibling relationships become more egalitarian with age (e.g., Furman & Buhrmester,
1992), while others have suggested that firstborns maintain higher levels of control in
family relationships than young siblings (Tucker, Updegradd & Baril, 2010).
Sibling bullying subgroups reported significantly more problems compared to neutrals
in terms of depression and behavioural problems. For example, sibling relational bullies
showed significantly more total difficulties, conduct, and hyperactivity problems and less
prosocial behaviour compared to neutrals. In addition, victims and bully–victims differed
significantly from neutrals on measures of emotional problems, hyperactivity, and
depression (across all types of sibling bullying). This study adds to the growing literature
demonstrating the negative implications of sibling bullying, specifically in terms of the
link between sibling bullying involvement and depression. For example, Bowes et al.
(2014) found that children who were frequently bullied by a sibling were twice as likely to
show symptoms of depression and self-harm in adulthood.
This is one of the few studies to use the SDQ with sibling bullying subgroups and to
highlight the specific behavioural difficulties for these groups using the subscales. By so
doing, we were able to look at the behavioural difficulties associated with sibling and peer
bullying involvement in more detail. Poly-setting victimization was associated with
significantly more behaviour difficulties, depression, and emotional problems compared
to neutrals, highlighting this issue as a serious concern for our adolescents. Similar
negative outcomes were demonstrated for poly-setting bully–victims, raising concern
over individuals who have problematic relationships with both siblings and peers.
Interestingly, the individuals who were only peer bullies displayed few significant
differences to neutrals in terms of internalizing and externalizing problems thus
suggesting that problematic sibling relationships may in fact be key to negative outcomes.
Sibling relationships need to be investigated in future research as potential predictors of
poly-bullying and victimizations. It is possible that interventions designed to reduce
maladaptive social behaviour such as empathy and perspective-taking interventions could
be helpful in this regard when implemented at home and at school (Foody & Samara,
2018).
For the most part, these conclusions are based on the SDQ being a valid instrument to
measure internalizing and externalizing problems. However, we found extremely low
reliability for the peer problems subscale in the current sample and decided to exclude it
from the analysis. Two other subscales (conduct problems and hyperactivity), as well as
the friendship questionnaire, had less than optimal internal reliability but we decided that
they could be included in the analysis. We determined that the Cronbach alpha level was
appropriate for this research study, even though a higher score is recommended for
measures to be used in applied settings (Nunnally, 1973). With this in mind, and the fact
that the Cronbach alpha level for the subscales that were used in this study went from .60
Sibling and peer bullying in Ireland 15
to .77, it is important to note that some recent literature has suggested that it may not be
the most ideal screening tool for such difficulties in young population (Garrido et al.,
2018). Future research will need to be conducted to determine the appropriateness of the
SDQ for research purposes, and these results will need to be considered in the light of this
new research agenda.
Regression analysis found that school, happiness with school, poor friendship quality,
being a peer victim, peer bully–victim, sibling victim, and sibling bully–victim were all
predictors of being in the borderline-clinical range on the total difficulties scale. For the
most part, being a victim of peer and sibling bullying and/or bully–victims were important
predictors of overall behaviour problems and depression. In some cases (e.g., with
emotional problems and depression) gender, friendship quality, and not liking school
were also important predictors of outcomes when combined with bullying involvement.
The predictor variables for significantly more conduct problems in the borderline-clinical
range were unique in that they included being male, being younger in age, and being
involved in both sibling and peer bullying (as victim, bully, and bully–victim).
Some limitations need to be mentioned when considering the current results. Firstly,
while we report on a substantial sample size, it is by no means representative of the post-
primary population in Ireland. Only 3.7% of all schools in the Republic of Ireland agreed to
participate and many students from these schools did not complete the survey. In an
attempt to ensure the feeling of anonymity for the schools, we did not collect information
regarding school type, ethos, size, gender, or location. It is important to note that our
results should be considered in the light of this.
A second limitation of this study is that it utilized a self-report tool to determine
levels of bullying. While this is useful to allow us to compare prevalence rates with
other countries, it is limited for many reasons. This method of data collection means
our participants have to consider their answers subjectively, rather than objectively.
Third, the results presented only demonstrated a cross-sectional relationship between
bullying and psychological outcomes. Future research utilizing longitudinal research
methods would give a clearer picture of how these variables interact. Fourth, the
definition provided to participants referred to peer bullying (and was inherently
school-focused) and no definition of sibling bullying was provided. A clearer outline
of what sibling bullying entails (i.e., intention, repetition, power hierarchy) needs to
be included in future research. Finally, the standard OBQ asks students about bullying
experiences in the current school year. The sibling bullying questionnaire asked about
the previous 3 months. Given that data collection took place between the period of
March–May, the timeframe for reporting peer bullying was greater than that for
sibling bullying. While the negative association of involvement in either sibling or
peer bullying would not be affected by this, it is important to consider the prevalence
rates in the light of this.
Despite the limitations, the current research extends knowledge on sibling bullying
and adds to the poly-victimization literature. It demonstrates novel findings in terms of the
role of school, friendship quality as well as happiness in school as important predictors of
behaviour problems within the clinical range (when combined with sibling and peer
bullying involvement). The study also took into account the cluster nature of the data and
included schools as a cluster variable. These findings are important for both mental health
and school-based anti-bullying interventions and demonstrate that promoting positive
siblingship, friendships, parental and school involvement are important factors in bullying
prevention (Samara & Smith, 2008).
16 Mair
ead Foody et al.
Acknowledgements
We would like to thank the Irish Research Council Government of Ireland Postdoctoral
Fellowship, the Marie Skłodowska-Curie Actions COFUND Collaborative Research Fellowships
for a Responsive and Innovative Europe (CAROLINE), and the Qatar National Research Fund
(QNRF), a member of Qatar Foundation Doha, Qatar, National Priority Research Programs
(NPRP) under Grant (NPRP 5-1134-3-240), for their support.
The authors would also like to thank all the schools, teachers, and children for their time
and willingness to participate in this research study.
Funding
This research was funded by the Irish Research Council Government of Ireland
Postdoctoral Fellowship. The first author is funded by the Marie Skłodowska-Curie
Actions COFUND Collaborative Research Fellowships for a Responsive and Innovative
Europe (CAROLINE). This work was also supported by the Qatar National Research Fund
(QNRF), a member of Qatar Foundation Doha, Qatar, National Priority Research Programs
(NPRP) under Grant (NPRP 5-1134-3-240) funded to Professor Muthanna Samara.
References
Angold, A., Costello, E. J., Messer, S. C., Pickles, A., Winder, F., & Silver, D. (1995). The development
of a short questionnaire for use in epidemiological studies of depression in children and
adolescents. International Journal of Methods in Psychiatric Research,5, 237–249.
Arseneault, L. P., Cannon, M. M. D. P., Fisher, H. L. P., Polanczyk, G. M. D. P., Moffitt, T. E. P., & Caspi,
A. P. (2011). Childhood trauma and children’s emerging psychotic symptoms: A genetically
sensitive longitudinal cohort study. The American Journal of Psychiatry,168(1), 65–72.
https://doi.org/10.1176/appi.ajp.2010.10040567
Bar-Zomer, J., & Brunstein Klomek, A. (2018). Attachment to parents as a moderator in the
association between sibling bullying and depression or suicidal ideation among children and
adolescents. Frontiers in Psychiatry,9,1–9. https://doi.org/10.3389/fpsyt.2018.00072
Bowes, L., Wolke, D., Joinson, C., Lereya, S. T., & Lewis, G. (2014). Sibling bullying and risk of
depression, anxiety, and self-harm: A prospective cohort study. Pediatrics,134(4), e1032–
e1039. https://doi.org/10.1542/peds.2014-0832
Button, D. M., & Gealt, R. (2010). High risk behaviors among victims of sibling violence. Journal of
Family Violence,25, 131–140. https://doi.org/10.1007/s10896-009-9276-x
Cornell, D., Huang, F., Gregory, A., & Xitao, F. (2013). Perceived prevalence of teasing and bullying
predicts high school dropout rates. Journal of Educational Psychology,105(1), 138–149.
https://doi.org/10.1037/a0030416
Coyle, S., Demaray, M. K., Malecki, C. K., Tennant, J. E., & Klossing, J. (2017). The associations
among sibling and peer-bullying, social support and internalizing behaviors. Child & Youth Care
Forum,46, 895–922. https://doi.org/10.1007/s10566-017-9412-3
Dantchev, S., Zammit, S., & Wolke, D. (2018). Sibling bullying in middle childhood and psychotic
disorder at 18 years: A prospective cohort study. Psychological Medicine,48, 2321–2328.
https://doi.org/10.1017/s0033291717003841
Defoe, I. N., Keijsers, L., Hawk, S. T., Branje, S., Dubas, S. J., Buist, K., ...Meeus, W. (2013). Siblings
versus parents and friends: Longitudinal linkages to adolescent externalizing problems. Journal
of Child Psychology and Psychiatry,54(8), 881–889. https://doi.org/10.1111/jcpp.12049
Duncan, R. (1999). Peer and sibling aggression: An investigation of intra-and extra-familial
bullying. Journal of Interpersonal Violence,14,871–886. https://doi.org/10.1177/
088626099014008005
Sibling and peer bullying in Ireland 17
Ensor, R., Marks, A., Jacobs, L., & Hughes, C. (2010). Trajectories of antisocial behaviour towards
siblings predict antisocial behaviour towards peers. The Journal of Child Psychology and
Psychiatry,51(11), 1208–1216. https://doi.org/10.1111/j.1469-7610.2010.02276.x
European Commission (2015). Eurostat demography Report. Luxemburg, Luxemburg: European
Union.
Faith, M. A., Elledge, L. C., Newgent, R. A., & Cavell, T. A. (2015). Conflict and dominance between
siblings as predictors of children’s peer victimization. Journal of Child Family Studies,24,
3623–3635. https://doi.org/10.1037/a0020313
Fisher, H. L., Schreier, A., Zammit, S., Maughan, B., Munaf
o, M. R., Lewis, G., & Wolke, D. (2013).
Pathways between childhood victimization and psychosis-like symptoms in the ALSPAC birth
cohort. Schizophrenia Bulletin,39, 1045–1055. https://doi.org/10.1093/schbul/sbs088
Foody, M., & Samara, M. (2018). Considering mindfulness techniques in school-based anti-bullying
programmes. Journal of New Approaches in Educational Research,7(1), 3–9. https://doi.org/
10.7821/naer.2018.1.253
Foody, M., Samara, M., & O’Higgins Norman, J. (2017). A meta-analysis of bullying and cyberbullying
studies on the island of Ireland. The British Journal of Educational Psychology,87, 535–557.
https://doi.org/10.1111/bjep.12163
Furman, W., & Buhrmester, D. (1992). Age and sex differences in perceptions of networks of
personal relationships. Child Devlopment,63(1), 103–115.
Garrido, L. E., Barrada, J. R., Aguasvivas, J. A., Mart
ınez-Molina, A., Arias, V. B., Golino, H. F., ...Rojo-
Moreno, L. (2018). Is small still beautiful for the Strengths and Difficulties Questionnaire? Novel
findings using exploratory structural equation modelling Assessment. https://doi.org/10.1177/
1073191118780461
Goodman, R. (1997). The Strength and Difficulties Questionnaire: A research note. Journal of
Child Psychology and Psychiatry,38,581–586. https://doi.org/10.1111/j.1469-7610.1997.
tb01545.x
Goodman, R. (2001). Psychometric properties of the Strength and Difficulties Questionnaire.
Journal of the American Academy of Child and Adolescent Psychiatry,40, 1337–1345.
https://doi.org/10.1097/00004583-200111000-00015
Goodyer, I. M., Wright, C., & Altham, P. M. E. (1989). Recent friendships in anxious and depressed
school age children. Psychological Medicine,19(1), 165–174.
Goodyer, I., Wright, C., & Altham, P. (1990). The friendships and recent life events of anxious and
depressed school-age children. The British Journal of Psychiatry,156(5), 689–698.
Hammig, B., & Jozkowski, K. (2013). Academic achievement, violent victimization, and bullying
among U.S. high school students. Journal of Interpersonal Violence,28, 1424–1436. https://
doi.org/10.1177/0886260512468247
Hardy, M. S. (2001). Physical aggression and sexual behavior among siblings: A retrospective study.
Journal of Family Violence,16(3), 255–267. https://doi.org/10.1023/A:1011186215874
He, J.-P., Burstein, M., Schmitz, A., & Merikangas, K. R. (2013). The Strengths and Difficulties
Questionnaire (SDQ): the factor structure and scale validation in U.S. adolescents. Journal
of Abnormal Child Psychology,41, 583–595. https://doi.org/10.1007/s10802-012-9696-6
23183936
Hoffman, K. L., Kiecolt, J., & Edwards, J. N. (2005). Physical violence between siblings: A theoretical
and empirical analysis. Journal of Family Issues,26, 1103–1130. https://doi.org/10.1177/
0192513X05277809
Jeffreys, M., Rozenman, M., Gonzalez, A., Warnick, E. M., Dauser, C., Scahill, L., ... Weersing, R.
(2016). Factor structure of the parent-report Mood and Feelings Questionnaire (MFQ) in an
outpatient mental health sample. Journal of Abnormal Child Psychology,44, 111–120.
https://doi.org/10.1007/s10802-015-0108-6
Johnson, R. M., Duncan, D. T., Rothman, E. F., Gilreath, T. D., Hemenway, D., Molnar, B. E., & Azrael,
D. (2014). Fighting with siblings and with peers among urban high school students. Journal of
Interpersonal Violence,30, 2221–2237. https://doi.org/10.1177/0886260514552440
18 Mair
ead Foody et al.
Kelleher, I. M. D. P., Keeley, H. M. D., Corcoran, P. P., Ramsay, H. M. D., Wasserman, C. M. A., Carli, V.
M. D. P., ...Cannon, M. M. D. P. (2013). Childhood trauma and psychosis in a prospective cohort
study: Cause, effect, and directionality. The American Journal of Psychiatry,170(7), 734–741.
https://doi.org/10.1176/appi.ajp.2012.12091169
Khan, R., & Cooke, D. J. (2013). Measurement of sibling violence: A two-factor model of severity.
Criminal Justice and Behavior,40(1), 2639. https://doi.org/10.1177/0093854812463349
Lereya, S., Samara, M., & Wolke, D. (2013). Parenting behavior and the risk of becoming a victim and
a bully/victim: A meta-analysis study. Child Abuse & Neglect,37(12), 1091–1108.
Menesini, E., Camodeca, M., & Nocentini, A. (2010). Bullying among siblings: The role of personality
and relational variables. British Journal of Developmental Psychology,28, 921–939. https://
doi.org/10.1348/026151009X479402
Messer, S. C., Angold, A., Costello, E. J., Loeber, R., Van Kammen, W., & Stouthamer-Loeber, M.
(1995). Development of a short questionnaire for use in epidemiological studies of depression in
children and adolescents: Factor composition and structure across development. International
Journal of Methods in Psychiatric Research,5, 251–262.
Nunnally, J. C. (1973). Psychometric theory (2nd ed.). New York, NY: McGraw-Hill.
Olweus, D. (1991). Bully/victim problems among schoolchildren: Basic facts and effects of a school
based intervention program. In D. J. Pepler & K. H. Rubin (Eds.), The development and
treatment of childhood aggression (pp. 411–448). Hillsdale, MI: Erlbaum.
Olweus, D. (1996). The Revised Olweus Bullying Questionnaire [Mimeo]. Bergen, Norway:
Research Center for Health Promotion (HEMIL): University of Bergen, N-5015 Bergen, Norway.
Patchin, J. W., & Hinduja, S. (2010). Cyberbullying and self-esteem. Journal of School Health,80,
614–621. https://doi.org/10.1111/j.1746-1561.2010.00548.x
Reese-Weber, M. (2008). A new experimental method assessing attitudes toward adolescent dating
and sibling violence using observations of violent interactions. Journal of Adolescence,31, 857–
876. https://doi.org/10.1016/j.adolescence.2007.11.002
Samara, M., & Smith, P. K. (2008). How schools tackle bullying, and the use of whole school policies:
changes over the last decade. Educational Psychology,28(6), 663–676. https://doi.org/10.
1080/01443410802191910
Silva, M. A. I., Pereira, B., Mendoncßa, D., Nunes, B., & de Oliveira, W. A. (2013). The involvement of
girls and boys with bullying: An analysis of gender differences. International Journal of
Environmental Research and Public Health,10, 6820–6831. https://doi.org/10.3390/ije
rph10126820
Takizawa, R., Maughan, B., & Arseneault, L. (2014). Adult health outcomes of childhood bullying
victimization: Evidence from five-decade longitudinal British birth cohort. The American
Journal of Psychiatry,171(7), 777–784. https://doi.org/10.1176/appi.ajp.2014.13101401
Tanrikulu, I., & Campbell, M. A. (2015). Sibling bullying perpetration: Associations with gender,
grade, peer perpetration, trait anger and moral disengagement. Journal of Interpersonal
Violence,30, 1010–1024. https://doi.org/10.1177/0886260514539763
Tippett, N., & Wolke, D. (2015). Aggression between siblings: Association with the home
environment and peer bullying. Aggressive Behavior,41(1), 14–24. https://doi.org/10.1002/
ab.21557
Tucker, C. J., Finkelhor, D., Turner, H., & Shattuck, A. (2013). Association of sibling aggression with
child and adolescent mental health. Pediatrics,132(1), 79–84. https://doi.org/10.1542/peds.
2012-3801
Tucker, C. J., Updegraff, K., & Baril, M. (2010). Who’s the boss? Patterns of control in adolescents’
sibling relationships. Family Relations,59, 520–532. https://doi.org/10.111/j.1741-3729.2010.
00620.x.
Wang, J., Iannotti, R. J., & Nansel, T. R. (2009). School bullying among adolescents in the United
States: Physical, verbal, relational and cyber. Journal of Adolescent Health,45(4), 368–375.
https://doi.org/10.1016/j.jadohealth.2009.03.021
Winsper, C., Hall, J., Strauss, V. Y., & Wolke, D. (2017). Aetiological pathways to borderline
personality disorder symptoms in early adolescence: Childhood dysregulated behaviour,
Sibling and peer bullying in Ireland 19
maladaptive parentings and bully victimisation. Borderline Personality Disorder and Emotion
Dysregulation,4,1–10. https://doi.org/10.1186/s40479-017-0060-x
Wolke, D., Copeland, W., Angold, A., & Costello, E. J. (2013). Impact of bullying in childhood on
adult health, wealth, crime and social outcomes. Psychological Science,24, 1958–1970.
https://doi.org/10.1177/0956797613481608
Wolke, D., & Samara, M. (2004). Bullied by siblings: association with peer victimisation and
behaviour problems in Israeli lower secondary school children. Journal of Child Psychology
and Psychiatry,45, 1015–1029. https://doi.org/10.1111/j.1469-7610.2004.t01-1-00293.x
Wolke, D., Tippett, N., & Dantchev, S. (2015). Bullying in the family: Sibling bullying. The Lancet
Psychiatry,2(10), 917–929. https://doi.org/10.1016/S2215-0366(15)00262-X
Received 9 March 2018; revised version received 9 July 2019
Supporting Information
The following supporting information may be found in the online edition of the article:
Table S1. Mean scores and standard errors of SDQ, SDQ sub-scales and MFQ by overall
bullying involvement in various settings: sibling, peer and poly.
Table S2. Regression and HLM regression models 1 (school level only) and model 2 (all
predictors) for peer and sibling bullying: Unstandardized coefficients.
`
20 Mair
ead Foody et al.
