Aggression in school and family contexts among youngsters with special needs: Qualitative and...

10
Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project Bojan Musil a, , Sara Tement a , Karin Bakracevic Vukman a , Ajda Sostaric b a Department of Psychology, Faculty of Arts, University of Maribor, Koroska 160, 2000 Maribor, Slovenia b CAAP Centre for Alternative and Autonomous Production, Valvasorjeva 42, 2000 Maribor, Slovenia abstract article info Article history: Received 17 December 2013 Received in revised form 2 June 2014 Accepted 3 June 2014 Available online 13 June 2014 Keywords: Aggression Community-based violence Youngsters Special needs TranSpace is an international project for empowering youngsters with disabilities to protect themselves from community-based violence. In the context of the project, an initial empirical study was carried out to assess the prevalence of aggression and victimization among these youngsters and to gain deeper insight into these phenomena. Participants in the project, aged from 11 to 21 years, came from six European countries (Austria, Bulgaria, Germany, Hungary, Italy and Spain). In a mixed-methods study, initial data were gathered using the Aggression Victimization Instrument, which is a composite of well-established and new measures and is used to assess aggressive acts in school and family contexts; further insights into emergent themes were obtained by conducting qualitative analyses using open-ended questions. Overall, youngsters reported a low incidence of aggressive acts; the prevalence is higher in the school context from the perspective of the victim. Females seem to be victims of relational aggression slightly more often than males; males are more frequently bullies than females. For early adolescents, the prevalence of aggressive acts is higher in the category of physical and ver- bal aggression in the school context. From the qualitative part (interviews), nine thematic frameworks appeared, with low self-esteem emerging as the most important issue, and one related to social context and coping strate- gies. From the research ndings, we can highlight that self-esteem is a central theme of any intervention concerning children and adolescents with disabilities who have been experiencing violence. In the process of empowerment, it is essential to develop an individual's social relations and appropriate coping strategies. © 2014 Elsevier Ltd. All rights reserved. 1. Introduction Aggression in the school and family among children and adolescents is an extensive phenomenon in Europe as well as in the United States. Findings of a national survey in the United States indicated that approx- imately 30% of the school-age population had experienced bullying as a perpetrator, victim or provocative victim (Nansel et al., 2001); the esti- mated prevalence of bullying involvement among adolescents from a large cross-national study from 40 countries, for instance, was around 26% (Craig et al., 2009). Bullying may take different forms (e.g., name calling, kicking or ignoring) and may range from being bullied to bully- ing others, both of which have been found to be associated with many different factors such as gender, age, family or neighbourhood socioeco- nomic status (e.g., Jansen et al., 2012). Since involvement in bullying is related to future detrimental outcomes in terms of physical health and psychiatric problems such as anxiety and depression (Jansen et al., 2012), continuous assessment of prevalence rates and detection of groups who are at special risk are crucial. There is also tentative support for the notion that victimization leads to additional problems over and above the pre-existing ones (Arseneault et al., 2006; Cluvera, Bowesc, & Gardnera, 2010). In fact, physical or psychological symptoms may also precede acts of bullying. A prospective cohort study, for instance, found that ill children, especial- ly in terms of psychological symptoms, are bullied more frequently (Fekkes, Pijpers, Fredriks, Vogels, & Verloove-Vanhorick, 2006). Such ndings also point to the complex nature of the relation between ag- gressive acts and psychological or physical health problems (i.e., special needs). Research into the prevalence of victimizations among young- sters with disabilities or special needs is thus crucial (e.g., Son, Parish, & Peterson, 2012). Following the US Maternal and Child Health Bureau (McPherson et al., 1998), children and adolescents with special needs are dened as those who have or are at increased risk for a chronic physical, devel- opmental, behavioral, or emotional condition and who also require health and related services of a type or amount beyond that required by children generally(p. 138). In the context of internal diversity of spe- cial needs, there are numerous studies addressing the prevalence of vio- lence against them. Monchy, Pijl, and Zandberg (2004) investigated social inclusion and bullying in a group of students with behavioral prob- lems, pervasive developmental disorders; ADHD, Tourette's syndrome, Children and Youth Services Review 44 (2014) 4655 Corresponding author. Tel.: +386 2 2293638, +386 31 634 704 (cell). E-mail address: [email protected] (B. Musil). http://dx.doi.org/10.1016/j.childyouth.2014.06.005 0190-7409/© 2014 Elsevier Ltd. All rights reserved. Contents lists available at ScienceDirect Children and Youth Services Review journal homepage: www.elsevier.com/locate/childyouth

Transcript of Aggression in school and family contexts among youngsters with special needs: Qualitative and...

Page 1: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

Children and Youth Services Review 44 (2014) 46–55

Contents lists available at ScienceDirect

Children and Youth Services Review

j ourna l homepage: www.e lsev ie r .com/ locate /ch i ldyouth

Aggression in school and family contexts among youngsters with specialneeds: Qualitative and quantitative evidence from the TranSpace project

Bojan Musil a,⁎, Sara Tement a, Karin Bakracevic Vukman a, Ajda Sostaric b

a Department of Psychology, Faculty of Arts, University of Maribor, Koroska 160, 2000 Maribor, Sloveniab CAAP — Centre for Alternative and Autonomous Production, Valvasorjeva 42, 2000 Maribor, Slovenia

⁎ Corresponding author. Tel.: +386 2 2293638, +386E-mail address: [email protected] (B. Musil).

http://dx.doi.org/10.1016/j.childyouth.2014.06.0050190-7409/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 17 December 2013Received in revised form 2 June 2014Accepted 3 June 2014Available online 13 June 2014

Keywords:AggressionCommunity-based violenceYoungstersSpecial needs

TranSpace is an international project for empowering youngsters with disabilities to protect themselves fromcommunity-based violence. In the context of the project, an initial empirical study was carried out to assessthe prevalence of aggression and victimization among these youngsters and to gain deeper insight into thesephenomena. Participants in the project, aged from 11 to 21 years, came from six European countries (Austria,Bulgaria, Germany, Hungary, Italy and Spain). In a mixed-methods study, initial data were gathered using theAggression Victimization Instrument, which is a composite of well-established and new measures and is usedto assess aggressive acts in school and family contexts; further insights into emergent themes were obtainedby conducting qualitative analyses using open-ended questions. Overall, youngsters reported a low incidenceof aggressive acts; the prevalence is higher in the school context from the perspective of the victim. Femalesseem to be victims of relational aggression slightly more often than males; males are more frequently bulliesthan females. For early adolescents, the prevalence of aggressive acts is higher in the category of physical and ver-bal aggression in the school context. From the qualitative part (interviews), nine thematic frameworks appeared,with low self-esteem emerging as the most important issue, and one related to social context and coping strate-gies. From the research findings, we can highlight that self-esteem is a central theme of any interventionconcerning children and adolescents with disabilities who have been experiencing violence. In the process ofempowerment, it is essential to develop an individual's social relations and appropriate coping strategies.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Aggression in the school and family among children and adolescentsis an extensive phenomenon in Europe as well as in the United States.Findings of a national survey in the United States indicated that approx-imately 30% of the school-age population had experienced bullying as aperpetrator, victim or provocative victim (Nansel et al., 2001); the esti-mated prevalence of bullying involvement among adolescents from alarge cross-national study from 40 countries, for instance, was around26% (Craig et al., 2009). Bullying may take different forms (e.g., namecalling, kicking or ignoring) and may range from being bullied to bully-ing others, both of which have been found to be associated with manydifferent factors such as gender, age, family or neighbourhood socioeco-nomic status (e.g., Jansen et al., 2012). Since involvement in bullying isrelated to future detrimental outcomes in terms of physical health andpsychiatric problems such as anxiety and depression (Jansen et al.,2012), continuous assessment of prevalence rates and detection ofgroups who are at special risk are crucial.

31 634 704 (cell).

There is also tentative support for the notion that victimizationleads to additional problems over and above the pre-existing ones(Arseneault et al., 2006; Cluvera, Bowesc, & Gardnera, 2010). In fact,physical or psychological symptoms may also precede acts of bullying.A prospective cohort study, for instance, found that ill children, especial-ly in terms of psychological symptoms, are bullied more frequently(Fekkes, Pijpers, Fredriks, Vogels, & Verloove-Vanhorick, 2006). Suchfindings also point to the complex nature of the relation between ag-gressive acts and psychological or physical health problems (i.e., specialneeds). Research into the prevalence of victimizations among young-sters with disabilities or special needs is thus crucial (e.g., Son, Parish,& Peterson, 2012).

Following the US Maternal and Child Health Bureau (McPhersonet al., 1998), children and adolescents with special needs are definedas thosewho “have or are at increased risk for a chronic physical, devel-opmental, behavioral, or emotional condition and who also requirehealth and related services of a type or amount beyond that requiredby children generally” (p. 138). In the context of internal diversity of spe-cial needs, there are numerous studies addressing the prevalence of vio-lence against them. Monchy, Pijl, and Zandberg (2004) investigatedsocial inclusion and bullying in a group of studentswith behavioral prob-lems, pervasive developmental disorders; ADHD, Tourette's syndrome,

Page 2: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

47B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

Asperger syndrome, reactive attachment disorder etc. Whitney, Smith,and Thompson (1994) included students with learning difficulties andphysical disability, as well as those with hearing and visual impairment.According toWhitney et al. (1994), just being noticeably different placesa person at risk for victimization.We could assume that, in the context ofaggressive behavior, “being noticeably different” is a more importantfactor that the type of disability.

Findings from previous research indicate that certain special needs,such as psychological disorders, may be related to a higher risk ofbeing bullied (Cluvera et al., 2010). Studies that involve students withdisabilities have shown victimization rates over 50% (Little, 2002;Singer, 2005), indicating that these students become the targets of vic-timization more often than do their non-disabled peers. Althoughmany studies have shown that students with disabilities are victimizedmore frequently than their able peers, results are inconsistent (Rose,Monda-Amaya, & Espelage, 2011). For example, Woods and Wolke(2004) found comparable victimization rates among students with orwithout disabilities. Nevertheless, themajority of studies on the victim-ization of students with special needs document increased levels ofverbal abuse, social exclusion, and physical aggression when comparedto other peers (Little, 2002;Marini, Fairbairn, & Zuber, 2001; Norwich &Kelly, 2004).

Alongwith the focus on children and adolescentswith special needs,another theme in the bullying literature requires special attention.Studies mostly focus either on bullying in the school context or on vic-timization in the family context. There is, however, ample evidencethat bullying in school is associated with a history of family violence(e.g., Espelage, Low, & De La Rue, 2012). In fact, those who experienceviolence in the family domain are more likely to be bullied or to bethe bully (e.g., Baldry, 2003). In order to gain a broad understanding ofaggression in school as well as in the family, one should combine bothcontexts in future studies.

The present research strives to follow three research challenges.First, it focuses on the prevalence of aggressive acts among childrenand adolescents with special needs in different European countriesand on making a thorough examination of age and gender differencesamong this group, since the latter two characteristics, in combinationwith special education, represent significant socio-demographic factorsin a multilevel analysis of maltreatment of children with disabilities(e.g., Algood, Hong, Gourdine, &Williams, 2011). In addition, it incorpo-rates aggression in both the school and family domains.

1.1. Theoretical background

A broad definition of aggression has been applied in order to encom-pass all possible forms of aggressive behavior among children and ado-lescents with special needs. Aggressive acts/school bullying constitute aphenomenon where a person is “exposed, repeatedly and over time, tonegative actions on the part of one or more other persons” and “when aperson intentionally inflicts injury or discomfort upon another person,through physical contact, through words or in other ways” (Olweus,1993, p. 9). Thus, in order to speak about aggressive acts, it is prolongedexposure, not single events, that is considered.

Aggressive acts may takemany different forms, for instance physicalaggression, verbal aggression, relational aggression/manipulation andcyber-bullying, all of which are relatively common in the school contextor among adolescents (e.g., Wang, Iannotti, & Nansel, 2009). Physicalaggression or harassment is characterized by direct physical violenceagainst others in terms of pushing, hitting or even kicking (e.g., Buss &Durkee, 1957; Solberg & Olweus, 2003). Verbal aggression can be de-fined as a negative attitude towards someone expressed directly by con-tent (e.g., saying hurtful things and calling someone mean names) andstyle of speech (e.g., yelling at someone or threatening him) (e.g., Buss& Durkee, 1957). Relational aggression, on the other hand, is moreindirect and includes purposeful manipulation and attempts to harmanother child’s or young adults’ friendships or feelings of inclusion by

a peer group (e.g., excluding schoolmates or other children from thepeer group, encouraging others not to be friends with schoolmates orother children or spreading lies; Crick & Grotpeter, 1995). Cyber-bullying is a rather new and increasingly common form of bullying,which can be defined as a form of “aggression that occurs throughpersonal computers (e.g., e-mail and instant messaging) or cell phones(e.g., text messaging)” (Wang et al., 2009, p. 369).

Since studies have shown that victims can also be bullies/offendersand vice versa (e.g., Solberg, Olweus, & Endersen, 2007), it is also veryimportant to take into account the perspective of the victim as well asthat of the bully/offender in the school context. It may be expected,however, that the prevalence of such bully victims will be greater inyounger children.

Since aggressive actsmay also occur in the family domain (includingnatural parents/grandparents/step-parents/adoptive parents/caregivers/aunts and uncles/or brothers, sisters or cousins), the fam-ily context should be examined in terms of aggressive acts. Two majorforms – psychological victimization and family neglect – are especiallysalient (Dunne et al., 2009; Zolotor et al., 2009). The former is character-ized by saying or doing things that make the child or adolescent feelembarrassed, ashamed or bad (e.g., parents screaming very loudly andaggressively). The latter encompasses a lack of appropriate care forthe child or adolescent as well as ignoring the child’s or adolescent’sneeds.

1.2. The TranSpace study

The present study was conducted in six European countries, withinthe consortium of TranSpace project partners, focusing on the empow-erment of children and adolescents with special needs to protect them-selves from community-based violence. Themain objective of the study,including quantitative and qualitative parts, was to assess the rate of ag-gressive acts among children and adolescents with special needs in var-ious contexts, to compare the participating countries as to prevalence,and to gain better understanding of the dynamic between aggressiveacts and victim responses by conducting structured interviews. Thesewere all necessary steps for further intervention planning in the agendaof the TranSpace project.

2. Method

2.1. Participants and procedure

The participants for this study were recruited through representa-tives of each project partner. First, agreement was obtained fromstudy participants (or their parents or legal representatives). The partic-ipants were informed that they could discontinue participation at anytime. Owing to the sensitive topic of the study, the participants hadthe possibility of skipping questionswith which theywere not comfort-able. All questionnaires were completed using a paper–pencil format ina face-to-face setting. More precisely, one contact person was presentduring questionnaire administration to help with possible difficultiesin understanding. All participants who indicated some level of aggres-sive acts (i.e., “a few times a month” at least 5 times) were invited toparticipate in an interview.

The participants were 204 children and adolescents indicating someform of special needs; they came from the six participating countries:Austria (n = 32; 46.9% females), Bulgaria (n = 35; 25.7% females),Hungary (n=45; 42.2% females), and Germany (n=20; 9.1% females).For Italy (n= 42) and Spain (n= 30), aswell as for several participantsfrom other countries, it was not possible to determine gender, age andother personal characteristics. Therefore, N varies between the personalcharacteristics and other (aggression) variables.

Age of the participants ranged from 11 to 21 years, (M= 15.69;SD = 2.67). 25.3% were aged 14 years or younger (i.e., early ado-lescence), 37.0% were aged between 15 and 18 years (i.e., middle

Page 3: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

48 B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

adolescence) and 14.2% were aged 19 years or older (i.e., lateadolescence).

The three most prevalent special needs among these participantswere as follows: attention deficit hyperactivity disorder (ADHD;24.7%), mood disorders, emotional problems (i.e., depression; 16.5%),learning difficulties (12.9%) and conduct disorders (i.e., violent andantisocial behavior, 11.8%).

It should be noted that some participating countries were unable tocarry out the questionnaire in its entirety. In Italy in particular, the diffi-culties were of a legal nature, since organisations would have had thelegal obligation to report to the police any detected abuse, and authori-sation is required by parents/legal guardians to ask questions of any na-ture tominors. Structured interviewswere not conducted among all theparticipating countries and not with all participants. Additionally, therewere concerns about anonymity. Several participants from other coun-tries declined to participate in the second part of the study or were noteligible, since those participants who reported some form of aggressiveacts were asked to take part in the qualitative part. From the total sam-ple, the percentage of participants who participated in the qualitativepart was 29.4%.

2.2. Measures

The aggression victimization instrument (AVI) was designed to as-sess the prevalence of aggressive acts or bullying against children andyoung adults with special needs and to detect those situations inwhich such acts occur. The instrument included a set of Likert-typeitems (Part I) as well as several open-ended questions (Part II). The in-strument in Part I addresses several forms of aggression in the schooland family context, as well as locus of control and social desirability aspotentially important correlates of aggression.

2.2.1. Part IAggressive acts in schools were adapted from the Olweus Bullying

Questionnaire (Olweus, 1996). Physical aggression was assessed usinga 7-item scale addressing acts such as pushing and kicking. A sampleitem is “How many times in the past 6 months have school mates orother children/youngsters pushed you around? Cronbach’s alpha forthis scale was .80. Verbal aggression was assessed using a 4-item mea-sure including aggressive acts such insults or yelling. A sample item is“How many times in the past 6 months have school mates or otherchildren/youngsters made fun of you because of nationality, religionor appearance?” Cronbach’s alpha for this scale was found to be .60. Re-lational aggression, which included aspects of social isolation or manip-ulative acts, was measured using a 7-item scale. A sample item is “Howmany times in the past 6 months school have mates or other children/youngsters encouraged others not to be friends with you?” Cronbach’salpha for this scale was .80. Cyber-bullying was measured with 3items addressing aggressive acts over the internet or via phone. A sam-ple item is “Howmany times in the past 6 months have school mates orother children/youngsters sentmeanmessages or inappropriate photosto you over the internet or via phone?” Cronbach’s alpha for this scalewas .55. The same items were used to address all forms of aggressionfrom the perspective of the victim and from the perspective of thebully/offender. The item stem for the bully items was as follows:“Howmany times in the past 6 months have you…?” Cronbach’s alphasfor all bully scales, except for cyber-bullying”, were (marginally) accept-able (Cronbach’s alpha (physical aggression) = .78; Cronbach’s alpha(verbal aggression) = .59; Cronbach’s alpha (relational aggression)= .65). Since cyber-bullying from the perspective of the bully had unac-ceptable reliability, it was excluded from further analyses. A possible ex-planation for the problematic Cronbach’s alpha could be the length ofthe scale (too few items) or low variance of cyber-bullying (i.e., insome countries Cronbach’s alpha could not be calculated since the var-iance was 0). Items for aggressive acts in the family domain wereadapted from the ISPCAN Child Abuse Screening Tool Children’s Home

Version (ICAST-CH) (Zolotor et al., 2009). Psychological victimizationwas measured using 7 items encompassing acts that could make thechild feel sad or hurt. A sample item is “How many times in the past6 months has someone in your family screamed at you very loudlyand aggressively?” Cronbach’s alpha for this scale was found to be .74.Neglect was assessed using a 6-item scale focusing on lack of care orconcern. A sample item is “How many times in the past 6 months didyou feel uncared for?” Cronbach’s alpha for this scale was found tobe .71.

In Part I, two control variables (i.e., locus of control, social desirabil-ity) which are likely to be linked to reports of aggressive acts wereassessed as well (Halloran, Doumas, John, & Margolin, 1999; Harris,1997). Locus of control refers “to the perceived causes of past events”(Connell, 1985, p. 1018) and can, among other effects, notably deter-mine one’s method of coping with an unfavorable event such as beingthe victim of an aggressive act. Based on this, locus of control is likelyto be interwoven with perceptions of victimization, as well as with thechild’s own aggressive acts. Social desirability is a tendency to give so-cially desirable answers and can frequently occur when data are collect-ed on sensitive or controversial topics (e.g., aggressive acts, alcohol ordrug usage). Thus, it may act as a response bias and should thereforebe controlled for. Locus of control, which can be defined as “an expec-tancy that reinforcement was under one’s own control (internal) ornot under one’s own control (external)” (Connell, 1985, p. 1018), wasassessed using theMultidimensional Measure of Children’s perceptionsof control (11 items; locus of control on general domain) (Connell,1985). The scale comprises three sources of control (subscales): inter-nal, external: powerful others, and external: unknown control.Cronbach’s alpha for internal control is .78 (from .62 to .78 in samplesfrom different countries), .50 for external control: powerful others(range from incalculable to .71), and .58 for external: unknown control(range from .09 to .69). Because of inadequate values of Cronbach’salpha in samples from different countries, we didn’t include externallocus (powerful others and unknown control) in the analysis.

The second questionnaire is a short form of the Crandall SocialDesirability Test for Children (CSDTC) (Carifio, 1994; 12 items), whichassesses “a young person’s tendency to give socially desirable (or ac-ceptable) answers to questions, rather than answers that express theyoung person’s true views, opinions, or feelings” (p. 74). The Cronbach’salpha is .73 (from .61 to .84 in samples from different countries).

All items addressing aggressive acts in school and family included a5-point Likert type response format ranging from “never” to “everyday”. A time frame of the past sixmonths is included in the response for-mat in order to reflect our definition of aggressive acts/bullying. For thescales addressing locus of control and social desirability, the original re-sponse format was preserved (i.e., 1 = not true at all; 4 = very true forlocus of control; 0 = yes; 1 = no for social desirability). We also addedthe response option “not applicable” to all items in case the child oryoung adult had problems understanding the item or did not identifyat all with the content of the item.

2.2.2. Part IIPart II of the instrument is an interview which focused on children

who actually identified themselves as victims of aggressive acts. It con-sists of two parts with 13 major open-ended questions. In the first part,the focus is on up to 3 different violence-related situations reported inthe questionnaire in order to establish the context of these violent epi-sodes, together with subjective responses of their victims. The secondpart of the interview comprises very diverse questions aimed at ad-dressing the issues of fear, discomfort, object relations, subjective realityand self-esteem. The aim of the interview is to exploremanifestations ofviolence in more detail, to gain an understanding of the contexts ofvarious violent episodes and the subjective responses of victims tothem, to touch upon the delicate arena of discomfort and thus tocheck its potential relation to acts of violence.

Page 4: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

49B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

In each of the questions below, an interviewer attempts to explorethe subject fully by asking additional sub-questions, adapted to a child’snarrative in an improvisational manner. Each of the questions isexplored in-depth, and the child or a young adult is encouraged toexplain the matter further.

2.3. Plan of analyses

Two approaches – a quantitative and qualitative – were used foranalyzing the data. The quantitative part included calculating the prev-alence of each aggression type in the school and family context in allcountries, examining the victim–bully relation and conducting cross-national and gender aswell as age comparison analyses. More precisely,MANOVA comparison analyses were calculated.

The plan of analysis for the qualitative part was more complex andfollowed the principles of interpretative phenomenological analysis(IPA; Smith, Flowers, & Larkin, 2009). The contextual nature of the inter-views allowed for cross-referential analysis, seeking to establish recur-ring themes of a violent nature to which the interviewed children andyoungsters are normally subjected, as well as the reasons for back-grounds to and their intimate emotional, cognitive and physical reac-tions to these violent episodes. The emphasis in this analysis wastherefore placed on subtracting the correlation between general pat-terns of violent incidents and the subjective realities of the childrenand youngsters that emerged in response to these incidents. Personalstories and their fragments were placed into generalized commontheme categories that were meant to encompass individual statementsabout (reaction to) violence into a wider thematic and contextualframework, thus attempting to understand the underlying intercon-nected dynamics of factors involved in the recurringmodels of violenceand/or subjective reactions of victims to this violence. In the light ofsuch analysis, several general themes emerged over and over again inthe answers given by children and youngsters; these may have beenarticulated differently in their individual expression and context, butnevertheless alluded to mutual thematic horizons.

3. Results and discussion

The following section includes results and discussion from thequantitative as well as the qualitative part of the study.

Table 1Descriptive statistics and pairwise comparisons (and ANOVAs/MANOVAs) between countries i

(1)Austria, M (SD)

(2)Germany, M (SD)

(3)Bulgar

VIOLENCE IN SCHOOLS(perspective of the victim)

F(16, 459) = 4.58, p b .05; Wilks’ Lambda = .64;

Physical aggression 1.28 (0.49) 1.44 (0.61) 1.55 (0Verbal aggression 1.57 (0.70) 1.56 (0.88) 1.95 (0Relational aggression/manipulation 2.00 (1.03) 1.63 (0.69) 1.73 (0Cyber-bullying 1.31 (0.64) 1.08 (0.19) 1.15 (0

Welch’s F(4, 67.43) = 14.29, p b .05Violence in schools: TOTAL 1.56 (0.62) 1.55 (0.66) 1.63 (0VIOLENCE IN SCHOOLS(perspective of the bully)

F(12, 400) = 5.79, p b .05; Wilks’ Lambda = .66;

Physical aggression 1.07 (0.16) 1.46 (0.59) 1.43 (0Verbal aggression 1.31 (0.66) 1.64 (0.65) 1.57 (0Relational aggression/manipulation 1.49 (0.56) 1.37 (0.54) 1.30 (0

Welch’s F(4, 63.67) = 16.38, p b .05Violence in schools: TOTAL 1.32 (0.49) 1.52 (0.60) 1.40 (0VIOLENCE IN THE FAMILY DOMAIN F(8, 288) = 5.47, p b .05; Wilks’ Lambda = .75; ηPsychological victimization 1.60 (0.70) 1.69 (0.61) 1.39 (0Neglect 1.70 (0.71) 1.44 (0.60) 1.28 (0

Welch’s F(4, 60.59) = 28.00, p b .05Violence in the family context: TOTAL 1.65 (0.63) 1.63 (0.61) 1.33 (0

Note. All variables in all countries, except for Italy, ranged from 1 to 5, with higher scores indicTherefore, direct comparisons between Italy and other countries were not computed.

3.1. Quantitative part of the study

Results from the quantitative part of the study include the preva-lence of each aggression type (in the family and school contexts) forthe full sample as well as for samples from each country separately pre-sented, a list of the most prevalent forms of aggression, overall genderand age differences as well as relations with some other measures(e.g., locus of control, social desirability).

Table 1 depicts descriptive statistics of aggressive acts for each coun-try separately. A MANOVA was computed to test the differences inaggressive acts in school and family context between countries. AnANOVA was computed to test differences in total score. The MANOVAresults, however, should be interpreted with caution, since the Box’stest was significant. Thus, the assumption of equality of covariance ma-tricesmay not hold. TheGames–Howell post hoc testwas applied to testfor pairwise comparisons between when the equality of variances wasnot given; and Bonferroni correction was applied to the subsequentpairwise comparisons. A similar procedure was performed in the analy-sis of aggressive acts between males and females (Table 2) and aggres-sive acts between age groups (Table 3).

Table 1 indicates that aggressive acts are generally not excessive,since the mean values rarely reach the value of 2, which shows that anaggressive act occurred a few times in the past 6 months or less. Alldomains are significantly positively intercorrelated: the correlation be-tween victim and bully perspectives in school is .56 (p b .01); the corre-lation between victimperspective in school and the family domain is .45(p b .01); and the correlation between the bully perspective in schooland the family domain is .42 (p b .01).

When looking at specific items of school violence from the perspec-tive of the victim, the formsmost frequentlymentionedwere: 1) talkingbehind someone’s back— relational aggression (M= 2.18; SD= 1.40);2) yelling and threats — verbal aggression (M = 2.00; SD = 1.34);3) spreading lies — relational aggression (M = 1.94; SD = 1.22);4) name calling — verbal aggression (M = 1.89; SD = 1.28) and5) damaging belongings— physical aggression (M= 1.83; SD= 1.16).When bullying others in the school context, the participants mostfrequently made the following responses: 1) ignoring — relational ag-gression (M = 1.77; SD = 1.13); 2) name calling — verbal aggression(M = 1.69; SD = 0.93) and 3) yelling and threats — verbal aggression(M = 1.61; SD = 0.86). In the family context, on the other hand, themost prevalent aggressive acts were as follows: 1) screaming veryloudly — psychological victimization (M = 2.17; SD = 1.24); 2) name

n aggressive acts in the school context.

ia, M (SD)(4)Hungary, M (SD)

(5)Spain, (SD)

(6)Italy, M (SD)

Pairwisecomparisons

η2 = .11

.54) 1.84 (0.80) 1.19 (0.22) 1.98 (0.65) 1, 2, 5 b 4

.89) 1.97 (1.14) 1.34 (0.42) 2.33 (1.15) 2, 5 b 4

.83) 1.96 (0.66) 1.20 (0.32) 2.26 (0.90) 5 b 1, 4; 2 b 4

.28) 1.21 (0.37) 1.06 (0.15) 1.10 (0.48)

.56) 1.81 (0.49) 1.20 (0.25) 2.12 (0.09) 5 b 3, 4η2 = .13

.48) 1.51 (0.59) 1.03 (0.09) 1.61 (0.71) 1, 5 b 4, 3

.69) 1.57 (0.46) 1.18 (0.27) 1.70 (0.81) 5 b 4

.40) 1.58 (0.41) 1.08 (0.20) 1.77 (0.86) 5 b 1, 4

.41) 1.55 (0.42) 1.08 (0.15) 1.66 (0.71) 5 b 3, 42 = .13.40) 1.67 (0.52) 1.02 (0.08) N.A. 5 b 1, 3, 4.40) 1.49 (0.58) 1.04 (0.15) N.A. 5 b 1, 4

.28) 1.58 (0.49) 1.03 (0.11) N.A. 5 b 1, 3, 4

ating greater levels of aggressive acts. In Italy a response format from 1 to 7 was applied.

Page 5: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

Table 2Descriptive statistics and pairwise comparisons (and t-tests, MANOVAs) between malesand females in aggressive acts in the school and family context.

MalesM (SD)

FemalesM (SD)

VIOLENCE IN SCHOOLS(perspective of the victim)

F(4, 115) = 2.58, p b .05; Wilks’Lambda = .92; η2 = .08

Physical aggression 1.60 (0.66) 1.59 (0.56)Verbal aggression 1.79 (0.80) 2.02 (0.97)Relational aggression/manipulation 1.77 (0.68) 2.13 (1.05)Cyber-bullying 1.16 (0.30) 1.32 (0.60)

t(121) = −1.28, p = .nsViolence in schools: TOTAL 1.64 (0.57) 1. 80 (0.60)VIOLENCE IN SCHOOLS(perspective of the bully)

F(3, 116) = 1.66, p = .ns; Wilks’Lambda = .96; η2 = .04

Physical aggression 1.47 (0.54) 1.25 (0.46)Verbal aggression 1.63 (0.70) 1.36 (0.40)Relational aggression/manipulation 1.52 (0.53) 1.39 (0.39)

t(121) = 2.88, p b .05Violence in schools: TOTAL 1.54 (0.53) 1.32 (0.30)VIOLENCE IN THE FAMILY DOMAIN F(2, 113) = 1.56, p = .ns; Wilks’

Lambda = .97; η2 = .03Psychological victimization 1.54 (0.46) 1.61 (0.68)Neglect 1.40 (0.59) 1.63 (0.59)

t(1210) = −1.55, p = .nsViolence in the family context: TOTAL 1.47 (0.44) 1.61 (0.58)

Note.MANOVAwas computed to test thedifferences in aggressive acts betweenmales andfemales. ANOVA was computed to test differences in total score. The MANOVA results,however, should be interpreted with caution, since the Box’s test was significant.

50 B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

calling, cursing — psychological victimization (M = 1.76; SD = 1.16)and 3) neglect in terms of making children or siblings feel unimportant(M= 1.72; SD= 1.16). Psychological victimization is a result of the be-haviors and acts of adults (parents or stepparents) but also of siblings.For instance, 49.4% of the participants reported that an adult hadscreamed at them loudly or aggressively, 24.7% reported that the actwas done by another child or adolescent, and 25.9% indicated thatboth adults and children/adolescents had screamed at them. In contrast,45% of the participants reported that their siblings called them names,31.7% reported that the act was done by adults and 23.3% indicatedthat both adults and children/adolescents called them names or saidmean things.

Generally, the prevalence of aggressive acts is higher in the schoolcontext from the perspective of the victim, followed by the family do-main and school context from the perspective of the bully. In the schoolcontext, verbal and relational aggression is emphasized, and in the

Table 3Descriptive statistics and pairwise comparisons (and ANOVAs/MANOVAs) between age groups

(1) Early adolescence, M (SD) (2)

VIOLENCE IN SCHOOLS (perspective of the victim) F(8, 230) = 4.44, p b .05; Wilks’ LamPhysical aggression 1.95 (0.69) 1.4Verbal aggression 2.11 (0.87) 1.7Relational aggression/manipulation 1.95 (0.90) 1.8Cyber-bullying 1.15 (0.30) 1.2

F(2, 121) = 3.04, p = .nsViolence in schools: TOTAL 1.87 (0.59) 1.5VIOLENCE IN SCHOOLS (perspective of the bully) F(6, 232) = 3.48, p b .05; Wilks’ LamPhysical aggression 1.59 (0.56) 1.2Verbal aggression 1.56 (0.53) 1.5Relational aggression/manipulation 1.44 (0.45) 1.5

Welch’s F(2, 53.72) = 0.53, p = .nsViolence in schools: TOTAL 1.52 (0.46) 1.4VIOLENCE IN THE FAMILY DOMAIN F(4, 226) = 1.88, p = .ns; Wilks’ LamPsychological victimization 1.45 (0.39) 1.6Neglect 1.33 (0.42) 1.5

Welch’s F(2, 51.00) = 3.31, p b .05Violence in the family context: TOTAL 1.39 (0.33) 1.5

Note. MANOVA was computed to test the differences in aggressive acts between special needhowever, should be interpreted with caution, since the Box’s test was significant.

family domain psychological victimization. It seems that more overtand direct forms of aggressive behavior are more frequent.

From the comparisons across countries, the sample from Spain isclearly the one with the lowest prevalence of each aggression type inboth school and family contexts. All other countries are more homoge-neous, with a slightly higher prevalence of aggressive acts in the schoolcontext in the sample from Hungary.

In addition to analysis of the prevalence of aggressive acts in schoolfrom different perspectives, we also explored the bully–victim relation.We split the sample into two victim and bully groups, where thosewithan average score higher than 2 (response category “a few times in thepast six months”) were labeled as bully or victim (others obtained alabel “not bully” or “not victim”). According to the prevalence rates,most participants (75.8%) fell into the category “Not victim, not bully”,15.5% were “Victim”, 1.9% “Bully”, and 6.8% were “Both victim andbully”.

Since previous studies indicate that gender may be an importantvariable to consider when examining aggression, we also conductedgender difference tests of aggression in the school and family contexts.Because of the rather small sample sizes in each partner country, theanalyses were based on the overall sample of participants. These resultsare depicted in Table 2.

From the perspective of the victim, there were overall significantgender differences. However, when carrying out post hoc comparisonsand applying the Bonferroni correction, the differences failed to reachsignificance. However, females seem to become victims of relational ag-gression slightly more often than males. From the perspective of thebully, a significant difference appears only in the total score. More pre-cisely, males seem to take the role of bully in general more frequentlythan do females. In terms of aggression in the family context, we againfound no significant gender differences.

It was further expected that age differencesmight be found in differ-ent forms of aggressive acts in the school and family context. Moreprecisely, we compared the early, middle and late adolescence groups(Table 3). Owing to sample size restrictions, differences between partic-ipating countries were again not included.

In the school context, the prevalence of aggressive acts is significant-ly higher in the group of early adolescents in the category of physicaland verbal aggression (from the perspective of the victim) and in thecategory of physical aggression (from the perspective of the bully). Itseems that, in the younger group, there are more overt, direct forms ofaggressive behavior in the school context than there are among olderage groups. However, this trend is reversed in the family context,

in aggressive acts in the school context.

Middle adolescence, M (SD) (3) Late adolescence, M (SD) Pairwise comparisons

bda = .75; η2 = .131 (0.46) 1.42 (0.61) 2, 3 b 15 (0.73) 1.66 (0.88) 3 b 14 (0.71) 1.91 (0.98)5 (0.53) 1.23 (0.33)

9 (0.49) 1.64 (0.71)bda = .84; η2 = .088 (0.44) 1.30 (0.56) 2, 3 b 10 (0.64) 1.55 (0.73)1 (0.47) 1.43 (0.59)

3 (0.44) 1.44 (0.59)bda = .93; η2 = .03

4 (0.62) 1.57 (0.56)1 (0.60) 1.73 (0.80)

7 (0.51) 1.67 (0.66)

s groups. ANOVA was computed to test differences in total score. The MANOVA results,

Page 6: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

Table 4Thematic frameworks and their incidence in the interviews.

Thematic frameworks Frequency(%)

Low self-esteem 8 (13.33)Everybody’s picking on me 12 (20)The rumor has it… 5 (8.33)Being a ghost 6 (10)Escapism 6 (10)The subversive power of humour 2 (3.33)Unchallenged physical violence 4 (6.67)The (self-)aggressor 19 (31.67)Home sweet home 15 (25)

Note. Number of interviews analyzed = 60.

51B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

where the younger group exhibits absolutely the lowest prevalence ofaggressive acts.

Finally, we address two additional concepts — locus of control andsocial desirability. Generally, an internal locus of control is significantlynegatively correlated with all aggressive acts in the family context (cor-relations range between − .18 and − .28). Respondents who perceivelife events as more under their own (internal) control report less ag-gressive behavior in the family context. If we speculate about causalitybased on these associations, better relations in the family can exert aninfluence on a more internally oriented perception of a person’s lifeevents and consequently of coping strategies. However, this speculationis limited only to the family context, since correlations between locus ofcontrol and aggressive acts in the school context are in mostly non-significant or marginal.

All aggressive acts except relational aggression in school from theperspective of the victim (r = .18; p = .06) are significantly positivelycorrelatedwith the social desirability scale (in the school context, corre-lations range between .19 and .30; in the family context, correlationsrange between .40 and .47). A person’s tendency to give socially desir-able (or acceptable) answers to questions is thus associated with ahigher prevalence of aggressive acts. This evidence suggests the possi-bility of a different perspective on the interpretation of our resultsconcerning the prevalence of aggressive acts. In that vein, respondentscould have reported inflated degrees of experience with aggressiveacts, as if these events were somehow normal in life, or as if it were ex-pected that they would have such experiences according to the contextof the questions in the instrument, all addressing aggressive behavior.

3.2. Qualitative part of the study

Given that the interviews in Part II were supposed to provide a rath-er more narrative background for further steps in the project (i.e., inter-vention planning),we attempted to identify several general themes thathad emerged repeatedly in the answers given by the children andyoungsters.

The analysis of the interviews took a twofold path: first of all by plac-ing (fragments of) individual narratives into various coherent thematicframeworks in order to demarcate a general referential arena ofviolence-related responses and episodes; and secondly, reporting indi-vidual examples of violent situations togetherwith subjective responsesof children and youngsters to such situations under each thematicframework in order to provide a more tangible set of data about theviolence erupting in individual realities. It would be false to claim thatthe interviews unanimously confirmed the themes outlined; sometimesit happened that a theme was invoked by merely two or three inter-viewees, such as the theme of the subversive power of humour. Howev-er, since humour as such isn't an episode or gesture of violence, but apotential response to it, with inherently embedded healing potential,this theme held the promise of a very complex and possibly veryrewarding landscape of role reversal and a highly subtle change inpower relations. It was therefore included as one of the themes; not asa theme of violence, though, but as a potentially extremely intriguingresponse to violence that does deserve further attention in futureintervention planning.

From the analysis of 60 interviews, nine thematic frameworksemerged. In Table 4 these frameworks are represented together withthe number of times they occurred in the interviews. In the followingtext each thematic framework is further described and accompaniedby actual examples from the interviews, where original answers of theyoungsters are embedded in contextually elaborated narratives of theinterviewees. The names of the individuals in these examples arefictitious.

3.2.1. Thematic framework 1: low self-esteemLow self-esteem seems to be the most prevalent and the most com-

plex problem that the interviewees report either openly or in a latent

manner and could be considered as the cornerstone lying silently atthe heart ofmost violence-related causes or consequences. The problemseems to be widely present with children who suffer from one or moreforms of peer violence and seems highly correlated with depressivemoods, anxiety attacks, shame and even self-injury.

The interviewees frequentlymention that an increase in self-esteemand feeling more comfortable with oneself would help them feel stron-ger, and they are most probably right in this assumption, which is whyany intervention should consistently strive to tackle this wide and com-plex problem in one way or another. In many cases low self-esteemseems to stand behind the inability to build up an adequate defensemechanism against violence. Victims of violent attackswill normally re-spond to these attacks with silence, a lowered head and escapism,whereas in certain cases, the anxiety induced by violence is so strongthat it evokes self-aggressive behavior. One victim of violence alsoreports acting out byhitting the surrounding objects after havingherselfbeen subjected to violence.

The key challenge in tackling the problemof low self-esteem and thesubsequent “victim behavior” is to produce a tool for these children tothink for themselves, articulate themselves, reflect upon themselvesand assert themselves. A genuine increase in self-esteem and a decreasein victim behavior may not come natural by to these youngsters, but itcan come as a result of gradually gaining the mental disposition of anactive agent, moving towards a more profound understanding ofoneself and others.

EXAMPLES. Mary (AT) admittedly has very low self-esteem, which seemsto be the root cause of her inability to fend off the aggressors in her life. Hadshe acquired a mechanism to help her increase her self-esteem, her abilityto cope with intrusions into her space would most likely increase as well.

Gary (BG) admits to being insecure and sees it as one of his negativefeatures, together with being depressed sometimes.

3.2.2. Thematic framework 2: everybody’s picking on meThis theme is extremely common and highly important because it

introduces the factor of group dynamics and its function in violencepro-duction against an isolated victim. The numerous reports of a group ofadolescents bullying, molesting or ridiculing one youngster indicatethat this is a major theme to be tackled in possible intervention.

Such continuous group aggression will normally combine varioustypes of violence, ranging from being called names, being laughed at,having things hidden away, being shouted at, being prevented fromparticipating in group games or being excluded from spaces, and alsobeing beaten or otherwise physically molested. The main point to beaddressed here is the helplessness of an individual in the face of facelesscollective aggression; how can one protect oneself against a mass ofmutually encouraging individuals, who are being socially and otherwiseprotected by the anonymity of a large group?

EXAMPLES. Anna (AT) suffered some rather harsh aspects of groupmock-ery, having been called names in a sexual manner, having been laughed at,

Page 7: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

52 B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

having been called silly and dumb, having been gossiped about behind herback, even by her supposed friends.

Loretta (HG) seems to be having a hard time with her schoolmates; shereports several incidents that seem to have happened at school. Some class-mates once dragged down her pants, and then there was another incidentwith having her school bag hidden away and, finally, she says others aregossiping about her behind her back. She says she doesn’t feel good in theclassroom, because there are too many intrigues and quarrels going onthere.

3.2.3. Thematic framework 3: rumor has it…Another major theme perceived as a powerful source of violent inci-

dents amongst children and adolescents is the intentional spreading ofrumors. All of the interviewees claim that such rumors aren‘t true andthat they feel hurt because of the lies about them spread around, sincethey paint an entirely erroneous and highly negative picture of who orwhat a person is. A rumor may not exactly constitute a violent incidentcompared with a heavy punch in one's face, but rumors may neverthe-less have an immensely destructive influence on one's life and can inthis respect be considered as clearly violent. Even though most ofthe victims tried confronting the rumor-spreading peers, the rumorsdidn't come to an end, but kept on living their own life, despite theconfrontation.

EXAMPLES. When Leo (AT) wasmoving from one place to another, a falserumor started circulating that he had stolen or broken something and thatthis was the reason for his going away. This false rumor made him veryupset, even to the point of getting a new phone number in order tocompletely break contact with everyone. What seems positive about Leo’sattitude towards the problem is that he tried confronting the people whowere spreading the rumor, but they didn’t show up, and he now retrospec-tively sees them as cowards for that reason.

Loretta (HG) complains about her classmates gossiping behind herback. She feels terrible at school for that reason, since she sees herself as atarget of “intrigues and quarrels”.

3.2.4. Thematic framework 4: being a ghostAnother theme of violence that seems to arise repeatedly inmultiple

interviews is intentional, active and harsh social exclusion: several ado-lescents report being completely excluded from social groups. Otherswalk right through them as if they were ghosts. When the victims tryto approach their peers and enter their peer groups, they are pushedaway intentionally and with militant determination by the groups'members. Normally being ignored is accompanied by other, mostlyharsher forms of violence. Social exclusion thus generally works as anadditional ostracizing mechanism that accompanies other acts of vio-lence and reconfirms the status of the victim as the helpless, isolatedsubject.

EXAMPLES. In Greta's (AT) case, acts of psychological and physicalaggression (by having things thrown at her or hidden from her) combinewith acts of social exclusion, such as being chased away every time shemakes an attempt to enter a group of peers.

Joanne (AT) seems to lack a true place of her own, a space she wouldclearly inhabit and speak from. She complains about not being includedin peer groups and not being listened to when she tries to speak, thus resid-ing in an invisible space, which does not allow her to produce sounds orimages. Furthermore, Joanne's invisibility extends to her family life, whereshe is not invited to take part in family decisions and has to share hermom with four younger siblings, which evokes feelings of guilt withwhich she has to deal subsequently.

3.2.5. Thematic framework 5: escapismEscapism is an important violence-related theme, since it stretches

well beyond violent incidents and thus functions as a copingmechanism

for life in general rather than a specifically narrow defense mechanismagainst violence in particular. Running off, shutting out the world orpressing a mental “off” button is a reaction that frequently emerges inadolescents in reaction to violence. The escapist tendencies extend wellbeyond the violent situation and penetrate the everyday functioninglife of a victim (for example, by the latter not going to work or to school,etc.). Escapism should therefore be considered as much more than justmentally shutting oneself out of a violent scene: it is a mode of being,and as such it functions as a crippling defense against reality, againstletting life happen, against feeling life as it happens and against lettingoneself build something meaningful by insisting and persisting throughthe ups and downs of reality.

EXAMPLES. Anna (AT) has escapist tendencies provoked by the aggressiveacts to which she's been subjected, meaning that she retreats from a violentsituation or simply doesn't show up for work or school and runs away. Aconfrontational mechanism instead of a defensive one would be crucialfor Anna to have meaningful continuity in her life.

Patricia (ES) admits that sometimes she desires to be far, far away,especially when she cannot find the way. Apparently this sometimeshappens at work.

3.2.6. Thematic framework 6: the subversive power of humourThis theme is different from previous ones, since it doesn't involve a

particular type of violent incident, but suggests a very lucid, plausibleand highly creative solution to such violence. For this reason, it seemsan important theme to include, even though only three intervieweeshave mentioned it. Each of them in their own manner suggested thatthey would like to respond to violent episodes with humour instead ofsaying nothing. It seems an important suggestion, because humourhas a strong link with subversive power in the carnivalesque sense,namely by turning the existing power relations upside down andthus dismantling the dominant social order. Applying humour as adefensive shield against aggression therefore aims not at keepingthe victim unharmed, but goes much further by penetrating the at-tacker‘s space and changing its structure. A victim becomes an activeagent in responding to violence, thus making that crucial step frombeing an object (of aggression) to being a subject (of action againstaggression).

EXAMPLES. Anna (AT) would like to learn how to cope with mockery andother acts of group peer violence not merely by ignoring such attacks, butbecoming an active agent in responding to them, thus making that crucialstep from being an object (of harassment) to being a subject (of actionwith regard to harassment). She specifically mentions humour or talkingback in a funnyway as the most desired manner of response, but apparent-ly she lacks the concentration to produce jokes at the given moment.

Joanne (AT) says that she would like to respond to violence with hu-mour, but when she's in the middle of the violent whirlwind, she cannotthink of anything funny to say. Afterwards, sentences or words come toher mind.

3.2.7. Thematic framework 7: unchallenged physical violenceThere have been a couple of reports of serious forms of physical vio-

lence. The emphasis in any intervention should be placed on trainingthe group‘s sensibility towards the intolerable nature of physical vio-lence and the consequent responsibility of each individual to intervenein acts of such violence to try and stop them.

EXAMPLES. The level of physical aggression to which Greta (AT) is sub-jected seems to have reached horrific dimensions: a wooden billet, rustynails, bags, books, chairs and water bottles have been thrown at her. Thistorture reached the point of having a police officer at school, who spokewith the perpetrators, which made things better.

Andrea (HG) states that –when she was outside – someone spat at her,and apparently the incident was repeated on a weekly basis.

Page 8: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

53B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

3.2.8. Thematic framework 8: the (self-)aggressor?Even though the majority of adolescents in the interviews are vic-

tims of violence, a surprisingly high number of them indicate that theyhave certain aggressive tendencies of their own, especially in terms oflosing control quickly. There are a very high number of cases where vi-olence has been applied as a protective mechanism against attacks andshould therefore be seen as an act of legitimate self-defense. The largenumber of aggressive tendencies reported by girls stops at shouting,getting angry at people for no reason or being impatient; boys alsospeak of being physically violent in some cases, but this violence has fre-quently been provoked by a violent incident. There is a fairly high occur-rence of auto-aggressive behavior, connected to low self-esteem andguilt issues. In some cases, auto-aggressive behavior is used as an ag-gressive manipulation strategy for adolescents to get what they wantfrom their parent. Some adolescents also mention hitting other objectsas a way of transferring the aggression from inside to outside.

The problem ofmaintaining control over an aggressive situation andkeeping one's head straight when in the vortex of intense interaction issomething that should be given a fair amount of attention, especially interms of devising amechanism for creating of a tranquilmental haven, aspace that does not allow poorly reflected aggression in terms of “actingout” to either enter or exit.

EXAMPLES. Even though Greta (AT) is a victim of serious physical andmental violence, she also manifests certain aggressive tendencies herself.She admits to becoming aggressive rather quickly and thinks it’s a negativepersonal feature. She reports that shouting happens in her home and addsthat she makes “a total fool out of them”.

Andy (BG) admits having hit a boy and injured his head very badly, be-cause this boy called him and his mother names. Andy does hint at havingcertain potentially aggressive tendencies, such as saying that he's notalways in a good mood, especially when people aren't nice to him andthat he sometimes behaves in a naughty way.

3.2.9. Thematic framework 9: home sweet homeIn general, family violence remains limited to shouting and scream-

ing at home, which seems to have to do with educational approaches.There are very few reports of physical violence accompanying theshouting. In some cases, adolescents express a desire for greater inde-pendence from their parents, who are naturally even more protectiveof their children than they would be if they lacked special needs.

EXAMPLES. Loretta (HG) says that, when her parents get upset over herlaziness, they start shouting at her, and at times she gets slapped in the face.

Martha (AT) sometimes feels uncomfortable at home, since her father isa choleric person, who is offended quite easily and gets crazy. Then, whenthings get tense, they all start screaming at each other.

4. Conclusions

What are the major conclusions from the TranSpace study onaggression victimization among children and adolescents with specialneeds?

From the quantitative part of the investigation, we can highlight thefollowing points:

i. Aggressive acts are generally not excessive — they occurred a fewtimes in the past 6 months or less. The prevalence of aggressiveacts is higher in the school context from the perspective of the vic-tim. Generally, more overt and direct forms of aggressive behaviourare more frequent (verbal and partially relational aggression inthe school context and psychological victimization in the familycontext).

ii. Considering the gender differences, females seems to be victims ofrelational aggression slightly more often than males; males morefrequently become bullies than do females.

iii. In the school context, the prevalence of aggressive acts is higher inthe group of early adolescents in the category of physical and verbalaggression. Thus, because of certain developmental characteristics,this age-group needs special attention.

These conclusions correspond with findings from previous researchin the context of bullying and victimization.Wang et al. (2009) conduct-ed a nationally representative US study, where the most prevalentforms of bullying and victimization were verbal and relational; andboys were more often involved in physical or verbal bullying, whilegirls were more often involved in relational bullying. According toother studies (e.g., Crick & Grotpeter, 1995; Monks et al., 2009; Smithet al., 1999), physical aggression is more common among boys, whilegirls exhibit more indirect and relational aggression. Regarding theschool context, studies suggest that aggression is more direct (e.g., tak-ing the form of physical bullying) in the early stages of development,becoming more indirect with age (e.g. taking the form of relational bul-lying) (Monks, Smith, & Swettenham, 2005; Monks et al., 2009; Smithet al., 1999). According to Ortega and Lera (2000), school bullyingincreases among children aged 12–14 years and decreases whenstudents get older. Despite the fact that our early adolescent groupexhibited the lowest prevalence of aggression in the family domain,violence in the school context was significantly correlated with vio-lence in the family, which is in line with other research studying re-lations between peer victimization and family violence (Espelageet al., 2012).

Next, analysis of the control variable of social desirability confirmsour expectations that a person’s tendency to give socially desirable (oracceptable) answers to questions is linked to reports of aggressiveacts. However, positive relations are contrary to thefindings fromprevi-ous research, where social desirability was negatively related to aggres-sion scales (Harris, 1997). One possible explanation could be attributedto the specific research context (youngsters with special needs directedto indicate and report potentially aggressive experiences), whichmightsubtly influence respondents to report inflated degrees of experiencewith aggressive acts, as if these events were somehow normal in life,or as if it were expected that they would have had such experiences ac-cording to the context of the questions in the instrument, all of whichaddressed aggressive behavior. Additionally, in the context of researchinto bullying among adolescents with autism spectrum disorder, VanRoekel, Scholte, and Didden (2010) report that adolescents who scoredhigh on teacher- and self-reported victimization were more likely tomisinterpret non-bullying situations as bullying.

The qualitative part of the study contributes to a deeper insight intothe subjective realities of youngsters experiencing violence. From theanalysis of interviews, nine thematic frameworks emerged, which canbe further organised into the interrelated metathemes, represented inFig. 1.

The central, but most indirectly accessible is the subjective (or psy-chological) context represented by low self-esteem (thematic frame-work 1), and the most important aim of intervention is to increase it.In this subjective equation, social context indirectly (thematic frame-work 9) and social relations directly (thematic frameworks 2, 3, 4, 7) in-fluence the subjective context (self-esteem), consequently producingstrategies to copewith the social and to defend the subjective (thematicframeworks 5, 6, 8). Once subjective context and coping strategies areestablished, they in turn influence social context.

The highlighted metathemes and thematic frameworks from the in-terviews are consistent with previous research. Son et al. (2014) sum-marized findings from a range of studies and related victimization ofchildren to depression, low self-esteem, physical health problems, sub-stance abuse, school absences and avoidance, self-harm, and suicidalideation. In a qualitative analysis of Dutch children with dyslexia andtheir reaction to school bullying, Singer (2005) emphasized low self-esteem, mostly accompanied by problems in relations with peers andfeelings of being different. Victimized adolescents with ADHD report

Page 9: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

Fig. 1. Subjective reality of youngsters experiencing violence.

54 B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

social exclusion (Shea & Wiener, 2003) and lower levels of social sup-port (Timmermanis & Wiener, 2011).

In the context of the TranSpace project, it is crucial to emphasize theconcept of Otherness, or, according to Annerback, Sahlqvist, andWingren (2014), the category of children who are “different”. This in away represents the final thematic framework that encompasses the en-tire range of violence-related themes in connection with mental healthand special needs and thus has a conceptual status exceeding that of theseparate themes listed above. Otherness is something closely related tomental health difficulties or learning disabilities and should be ad-dressed as a separate theme. Anyonewhodoesn‘tfit the picture of dom-inant representations is the “Other”, especially people with mentalhealth problems and learning disabilities. It is therefore crucial forthem to learn how to experience Otherness inwhatever form and incar-nation in order to learn to incorporate it into their own emotional andcognitive realities and thus to become skilled at voicing it as a legitimatesubject position that deserves its rightful space.

In general, the present study has several strengths, including amixed-method design applied in several European countries, a broadconceptualization of aggression and a focus on a rather poorly examinedgroup of youngsters who may be particularly susceptible to aggressionvictimization (i.e., youngsters with special needs). However, the studyis not without limitations. First, the sample sizes in each of the partici-pating countries were rather small. Therefore, inferences drawn fromcross-national comparisonsmay be limited. Results based on the overallsample, however, are more substantial. Future studies should considerincluding broader samples to compare aggression prevalence estimatesacross countries, aswell as comparingdifferent groups (e.g., age groups)within and between countries. Second, some of the scales in the quanti-tative part had lower rates of reliability (i.e., internal consistency) thanwould normally be expected. Since only well-established scales wereused, such a finding could indicate that the psychometric properties ofthe questionnaires are contingent on sample specifics. One promisingfuture research avenue could be establishing the psychometric charac-teristics of instruments in distinct contexts (e.g., youngsterswith specialneeds). Finally, our study included a very broad definition of specialneeds. This may be regarded as a strength, allowing for better generali-zation of the findings. On the other hand, the distinctive features of thedifferent special needs groups with respect to aggression victimizationmay remain unclear.

From a practical viewpoint, the present study provides valuableinsights into bullying/aggression prevention and intervention pro-grams. Based on our results, intervention and prevention programsshould encompass country specifics and target the most salientbullying/aggression aspect in each country. Second, gender may be acrucial variable to consider, especially when working with bullies.Third, an important target group for intervention is the early adoles-cents. Based on the results of the qualitative part of the study, programssuch as social and interpersonal skills training should focus on the vic-tims’ self-esteem. More precisely, to increase levels of self-esteem, orin other words, to ensure that the individual acquires themental dispo-sition of an active agent, any intervention should focus on enhancing aperson’s social relations and at the same time developing appropriatecoping strategies. In that vein, talking about past violent episodes(via AVI) is a subtle (projective) means of affecting someone’s subjec-tive (psychological) reality.

In sum, our findings highlight several country, age and gender differ-ences in aggressive acts among a sample of youngsters with specialneeds. By using a mixed-method design, the study provides a deeperunderstanding of the phenomenon and, in turn, benefits the literatureby giving directions for future intervention and prevention programs.

Acknowledgements

The Research in TranSpace project was funded with financial sup-port from the Daphne Programme of the European Union. The studywas carried out with the support of the University of Maribor(Slovenia); Counselling Center for Children, Adolescents and ParentsMaribor (Slovenia); Fundación INTRAS (Spain); Mental Health Center(Bulgaria); CJD Berufsbildungswerkes Frechen (Germany); Pro MenteOberösterreich (Austria); Borgorete Cooperativa sociale (Italy); andthe Way Out Psychiatric Self Help Association (Hungary).

References

Algood, C. L., Hong, J. S., Gourdine, R. M., &Williams, A.B. (2011). Maltreatment of childrenwith developmental disabilities: An ecological systems analysis. Children and YouthServices Review, 33, 1142–1148.

Annerback, E. -M., Sahlqvist, L., & Wingren, G. (2014). A cross-sectional study ofvictimisation of bullying among schoolchildren in Sweden: Background factors andself-reported health complaints. Scandinavian Journal of Public Health, 42, 270–277.

Page 10: Aggression in school and family contexts among youngsters with special needs: Qualitative and quantitative evidence from the TranSpace project

55B. Musil et al. / Children and Youth Services Review 44 (2014) 46–55

Arseneault, L., Walsh, E., Trzesniewski, K., Newcombe, R., Caspi, A., & Moffitt, T. E. (2006).Bullying victimization uniquely contributes to adjustment problems in youngchildren: A nationally representative cohort study. Pediatrics, 118, 130–138.

Baldry, A.C. (2003). Bullying in schools and exposure to domestic violence. Child Abuse &Neglect, 27, 713–732.

Buss, A. H., & Durkee, A. (1957). An inventory for assessing different kinds of hostility.Journal of Consulting Psychology, 21, 343–349.

Carifio, J. (1994). Sensitive data and students' tendencies to give socially desirableresponses. Journal of Alcohol and Drug Education, 39, 74–84.

Cluvera, L., Bowesc, L., & Gardnera, F. (2010). Risk and protective factors for bullyingvictimization among AIDS-affected and vulnerable children in South Africa. ChildAbuse & Neglect, 34, 793–803.

Connell, J. P. (1985). A new multidimensional measure of children's perceptions ofcontrol. Child Development, 56(4), 1018–1041.

Craig, W., Harel-Fisch, Y., Fogel-Grinvald, H., Dostaler, S., Hetland, J., Simons-Morton, B.,et al. (2009). A cross-national profile of bullying and victimization among adolescentsin 40 countries. International Journal of Public Health, 54(Suppl. 2), 216–224.

Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-psychological adjustment. Child Development, 66, 710–722.

Dunne, M. P., Zolotor, A. J., Runyan, D. K., Andreva-Miller, I., Choo, Y. W., Gerbaka, B., et al.(2009). ISPCAN Child Abuse Screening Tools Retrospective version (ICAST-R): Delphistudy and field testing in seven countries. Child Abuse & Neglect, 33, 815–825.

Espelage, D. L., Low, S., & De La Rue, L. (2012). Relations between peer victimization sub-types, family violence, and psychological outcomes during early adolescence.Psychology of Violence, 2, 313–324.

Fekkes, M., Pijpers, F. I., Fredriks, A.M., Vogels, T., & Verloove-Vanhorick, S. P. (2006). Do bul-lied children get ill, or do ill children get bullied? A prospective cohort study on the rela-tionship between bullying and health-related symptoms. Pediatrics, 117, 1568–1574.

Halloran, E. C., Doumas, D.M., John, R. S., & Margolin, G. (1999). The relationship betweenaggression in children and locus of control beliefs. The Journal of Genetic Psychology,160, 5–21.

Harris, J. A. (1997). A further evaluation of the aggression questionnaire: Issues of validityand reliability. Behaviour Research and Therapy, 35, 1047–1053.

Jansen, P. W., Verlinden, M., Dommisse-van Berkel, A., Mieloo, C., van der Ende, J.,Veenstra, R., et al. (2012). Prevalence of bullying and victimization among childrenin early elementary school: Do family and school neighbourhood socioeconomicstatus matter? BMC Public Health, 12, 1–10.

Little, L. (2002). Middle-class mothers’ perceptions of peer and sibling victimizationamong children with Asperger’s syndrome and nonverbal learning disorders. Issuesin Comprehensive Pediatric Nursing, 25, 43–57.

Marini, Z. A., Fairbairn, L., & Zuber, R. (2001). Peer harassment in individuals with devel-opmental disabilities: Towards the development of a multi-dimensional bullyingidentification model. Developmental Disabilities Bulletin, 29, 170–195.

McPherson, M., Arango, P., Fox, H., Lauver, C., McManus, M., Newacheck, P. W., et al.(1998). A new definition of children with special health care needs. Pediatrics, 102,137–140.

Monchy, M.D., Pijl, S. J., & Zandberg, T. (2004). Discrepancies in judging social inclusionand bullying of pupils with behaviour problems. European Journal of Special NeedsEducation, 19, 317–330.

Monks, C. P., Smith, P. K., Naylor, P., Barter, C., Ireland, J. L., & Coyne, I. (2009). Bullying indifferent contexts: Commonalities, differences, and the role of theory. Aggression andViolent Behavior, 14, 146–156.

Monks, C. P., Smith, P. K., & Swettenham, J. (2005). The psychological correlates of peervictimization in schools: Social cognitive skills, executive functions and attachmentprofiles. Aggressive Behavior, 31, 571–588.

Nansel, T. R., Overpeck, M., Pilla, R. S., Ruan, W. J., Simons-Morton, B., & Scheidt, P. (2001).Bullying behaviours among US youth: Prevalence and association with psychosocialadjustment. Journal of the American Medical Association, 285, 2094–2100.

Norwich, B., & Kelly, N. (2004). Pupils’ views of inclusion: Moderate learning difficultiesand bullying in mainstream and special schools. British Educational Research Journal,30, 43–65.

Olweus, D. (1993). Bullying at school: What we know and what we can do. Oxford:Blackwell.

Olweus, D. (1996). The Revised Olweus Bully/Victim Questionnaire. Mimeo. Bergen,Norway: Research Centre for Health Promotion, University of Bergen.

Ortega, R., & Lera, M. -J. (2000). The Seville anti-bullying in school project. AggressiveBehavior, 26, 113–123.

Rose, C. A., Monda-Amaya, L. E., & Espelage, D. L. (2011). Bullying perpetration and victim-ization in special education: A review of the literature. Remedial and Special Education,32, 114–130.

Shea, B., & Wiener, J. (2003). Social exile: The cycle of peer victmization for boys withADHD. Canadian Journal of School Psychology, 18, 55–90.

Singer, E. (2005). The strategies adopted by Dutch children with dyslexia to maintaintheir self-esteem when teased at school. Journal of Learning Disabilities, 38, 411–423.

Smith, J. A., Flowers, P., & Larkin, M. (2009). Interpretative phenomenological analysis:Theory, method and research. London: Sage.

Smith, P. K., Morita, Y., Junger-Tas, J., Olweus, D., Catalano, R., & Slee, P. (1999). The natureof school bullying: A cross-national perspective. London & New York: Routledge.

Solberg, M., & Olweus, D. (2003). Prevalence estimation of school bullying with theOlweus Bully/Victim Questionnaire. Aggressive Behavior, 29, 239–268.

Solberg, M. E., Olweus, D., & Endersen, I. M. (2007). Bullies and victims at school: Are theythe same pupils? British Journal of Educational Psychology, 77, 441–464.

Son, E., Parish, S. L., & Peterson, N. A. (2012). National prevalence of peer victimizationamong young children with disabilities in the United States. Children and YouthServices Review, 34, 1540–1545.

Son, E., Peterson, N. A., Pottick, K. J., Zippay, A., Parish, S. L., & Lohrmann, S. (2014). PeerVictimization among young children with disabilities: Early risk and protectivefactors. Exceptional Children, 80, 368–384.

Timmermanis, V., & Wiener, J. (2011). Social correlates of bullying in adolescents withattention-deficit/hyperactivity disorder. Canadian Journal of School Psychology, 26,301–318.

Van Roekel, E., Scholte, R. H. J., & Didden, R. (2010). Bullying among adolescents withautism spectrum disorders: Prevalence and perception. Journal of Autism andDevelopmental Disorders, 40, 63–73.

Wang, J., Iannotti, R. J., & Nansel, T. R. (2009). School bullying among US adolescents:Physical, verbal, relational and cyber. Journal of Adolescent Health, 45, 368–375.

Whitney, I., Smith, P. K., & Thompson, D. (1994). Bullying and children with specialeducational needs. In P. K. Smith, & S. Sharp (Eds.), School bullying: Insights andperspectives (pp. 213–240). London: Routledge.

Woods, S., & Wolke, D. (2004). Direct and relational bullying among primary schoolchildren and academic achievement. Journal of School Psychology, 42, 135–155.

Zolotor, A. J., Runyan, D. K., Dunne, M. P., Jain, D., Péturs, H. R., Ramirez, C., et al. (2009).ISPCAN Child Abuse Screening Tool Children’s Version (ICAST-C): Instrumentdevelopment and multi-national pilot testing. Child Abuse & Neglect, 33, 833–841.