Abstract
This study compares staff reports of bullying among institutionalized youth with residents’ own self-reported prevalence of bullying and victimization collected in the previous study (the Self-Report Study on Bullying in Croatian Residential Care [SSBCRC]) and staff reports of reduction strategies are compared with evidence-based proposed policy solutions arising from residents’ reports. The study also compares reduction strategies used by staff with evidence-based proposed policy solutions arising from residents’ reports arising from the SSBCRC. A total of 140 staff from 20 Croatian youth facilities completed an anonymous questionnaire. The results revealed that staff estimates of the prevalence of bullying and victimization were significantly lower than resident reports. Staff were better aware of the prevalence of certain types of bullying, but they held stereotypical views of bullies and victims and had difficulties in recognizing the true times and places of bullying. Staff described their anti-bullying policies as being predominantly reactive, rather than proactive and evidence-based. It is concluded that more effort needs to be made to change the current anti-bullying policies used by staff.
Introduction
Research on bullying in Croatia is still limited in scope, with the existing studies on bullying in Croatian schools being based on variable research designs and methods used. Consequently, the prevalence of bullying and victimization varies greatly across Croatian school-based studies. For instance, Rajhvajn, Bulat, and Ajduković (2012) used a self-report checklist based on a number of items indicative of bullying and victimization and found that 37.8% of second graders from five Croatian high schools were victimized by their peers on a weekly basis. Using a similar questionnaire indicative of bullying and victimization, Buljan Flander, Durman Marijanović, and Ćorić Špoljar (2007) found that 27.0% of fourth to eighth graders from 25 Croatian primary schools were victimized by their peers on a daily basis.
The United Nations Children’s Fund (UNICEF) Office for Croatia used the Olweus bully/victim questionnaire in 2004 with a sample of 11- to 14-year-old children from 84 Croatian primary schools (Pregrad, 2010). Given that the Olweus questionnaire includes the term “bullying” that has been translated to Croatian as “peer violence,” presumably because of the participants’ inaccurate interpretations of the term “bullying,” UNICEF found that the prevalence of victimization occurring 2 to 3 times a month or more often was 10.4%. The UNICEF Office for Croatia also reported that children were most frequently bullied by indirect forms of violence, such as being called names or made fun of in a hurtful way, and that bullying usually occurred in the school hallways (Pregrad, 2010).
Using a behavioral checklist, the first large-scale Self-Report Study on Bullying in Croatian Residential Care (SSBCRC; Sekol & Farrington, 2009, 2010, 2016a, 2016b) demonstrated that bullying in Croatian residential care is a serious and prevalent problem, with almost three quarters of residents being involved in bullying at least 2 to 3 times a month. Based on self-reported background and psychological correlates of bullying and victimization, as well as on residents’ qualitative accounts of residential peer cultures, the SSBCRC has offered valuable evidence-based policy recommendations. However, no residential care policy can be effectively delivered without staff being aware of the prevalence, times, places, causes, and consequences of bullying in their facilities.
This article, therefore, has two main aims: first, to assess how residential care staff view the nature and prevalence of bullying in their facilities, and compare this with resident reports collected in the previous SSBCRC; second, to assess what strategies staff use to combat bullying, and compare these with evidence-based proposed policy solutions arising from the SSBCRC. These questions have never before been addressed in residential care facilities worldwide. However, there is some relevant research comparing teacher and student reports that will be reviewed in this introduction first. There will then be a review of the existing residential care bullying literature conducted previously outside Croatia. Finally, the SSBCRC results will be summarized, as they serve as a baseline to which staff data are compared.
Teacher Reports of School Bullying
Although still relatively under-researched compared with other topics related to bullying in schools, the comparison between self-reports and teacher-reports of school bullying problems has recently received increased research attention. The results of this research have been relatively clear-cut, suggesting that teachers tend to underreport problems of bullying in their schools. For instance, there is evidence that, compared with student self-reports, teachers underreport the prevalence of bullying (Bradshaw, Sawyer, & O’Brennan, 2007; Mishna, Scarcello, Pepler, & Wiener, 2005; Pervin & Turner, 1994), with the convergence between teachers’ and students’ reports being only low to moderate (e.g., Beran & Stewart, 2008; Cornell & Bandyopadhyay, 2010).
There is also evidence that teachers are more likely to view overt types of aggression (e.g., physical aggression or verbal threats) as bullying, but that they often do not perceive intimidating looks, gossiping, or name calling as bullying (Boulton, 1997). In cases in which teachers do consider verbal aggression as bullying, they tend to view it as less serious than physical bullying (Craig, Henderson, & Murphy, 2000; Smith et al., 2002). Overall, teachers tend to view indirect and relational bullying as less serious than direct bullying, are less likely to recognize indirect bullying when it happens, and are less likely to intervene in indirect bullying situations (Mishna et al., 2005). However, both indirect and relational bullying, although more difficult for teachers to detect because of their covert nature, do constitute bullying and have serious consequences for victims if done repeatedly.
Teachers also seem to hold rather stereotypical views about characteristics of bullies and victims in schools. For instance, Nicolaides, Toda, and Smith (2002) found that preservice teachers believed that bullies had low self-esteem and poor social skills, although this has not been empirically confirmed (e.g., Johnson & Lewis, 1999; Rigby & Slee, 1991, 1993; Salmivalli, Kaukiainen, Kaistaniemi, & Lagerspetz, 1999; Salmon, James, & Smith, 1998). Similarly, in their qualitative study, Mishna et al. (2005) found that teachers believed that victims were poorly adjusted and unassertive, although these were not the actual characteristics of the victims included in their study. Finally, there is evidence that teachers might identify bullying among boys more quickly than bullying conducted by girls (Peters, 2012). Probably because of all the above reasons, Nicolaides et al. (2002) found that preservice teachers wanted to learn more about bullying, especially how to talk to bullies and victims and how to develop a whole-school policy on bullying.
Residential Care Bullying Research
Nature, prevalence, and correlates of bullying in residential care outside of Croatia
Although still limited in scope, research on bullying and victimization among institutionalized children and young people has been increasing over the last decade. As such, this research has mainly focused on the nature and prevalence of bullying, with some attempts to also establish basic correlates and predictors of bullying and victimization in care. In terms of the nature and prevalence of bullying, the existing research outside of Croatia has demonstrated that the prevalence of bullying and victimization among institutionalized youth appears to be considerably higher than reported among children in schools. For instance, a recent systematic review that included 80 studies on school bullying found mean prevalence rates of 35% for bullying perpetration and 36% for victimization (Modecki, Minchin, Harbaugh, Guerra, & Runions, 2014). On the contrary, Barter, Renold, Berridge, and Cawson (2004) found that almost all of 71 interviewed residents from 14 English children’s homes had experienced some form of peer violence. Similarly, in their self-report study of 1,324 Jewish and Arab institutionalized adolescents aged 11 to 19 years from 32 residential care facilities, Attar-Schwartz and Khoury-Kassabri (2015; Khoury-Kassabri & Attar-Schwartz, 2014) found that 73% of residents were verbally victimized by their peers at least once in the previous month, while 62% and 56% of residents were victimized indirectly and physically, respectively. Residents who were prone to verbal, indirect, and physical victimization were younger; had higher levels of adjustment difficulties; and were experiencing higher levels of maltreatment by residential care staff. Indirect victimization was found to be more prevalent among girls, while physical victimization was more prevalent among boys. Indirect and physical victimization were also more prevalent among residents with low perceptions of their social self-efficacy, while verbal victimization was more prevalent in Jewish residents as well as in facilities with higher numbers of vulnerable youth. Wright (2016) compared the rates of bullying and victimization among 50 male adolescents in residential care to the rates of bullying and victimization among 50 male adolescents in public schools. The findings revealed that adolescents from residential care reported significantly higher levels of bullying and victimization than their control counterparts. 1 Similarly, compared with a control group, bullying and victimization in a residential care group were more strongly predicted by low levels of attachment to peers and school and having been subjected to permissive parenting styles.
The SSBCRC
Following highly publicized incidents of bullying in Croatian residential care, Sekol and Farrington (2009, 2010, 2016a, 2016b) conducted the above-mentioned SSBCRC. The SSBCRC was the first large-scale study that established quantifiable estimates of bullying and victimization in residential care globally. 2 The study was based on self-reports collected from 601 residents of Croatian children’s homes and correctional homes which revealed that 70.0% of residents were involved in bullying 2 to 3 times a month or more often, either as bullies or as victims. The majority of bullies in the SSBCRC were also victims (and vice versa), and bullying, which often took indirect forms, usually occurred in bedrooms during the night.
Other results of the SSBCRC demonstrated that: (a) both male and female bullies tend to be careless, neurotic, disagreeable, likely to bully others in school, and likely to hold attitudes approving of bullying, while male bullies also tend to have low affective empathy and high extraversion as well as a history of bullying during their earlier placements (Sekol & Farrington, 2016b); (b) both male and female victims tend to be neurotic, lacking in self-esteem, and believing that bullying was just part of life in residential care, with female victims also being disagreeable and not conscientious and male victims being young, having a history of victimization during their previous placement, in school, and at the beginning of their current placements (Sekol & Farrington, 2016a); and (c) residential peer cultures and hierarchies, a poor relationship between residents and staff, stigmatization, and deprivations of material goods/services typical of institutional life all contribute to bullying and peer violence among residents (Sekol, 2013).
Based on the above-described self-report results collected in the SSBCRC, a number of evidence-based policy recommendations have been proposed to reduce bullying in Croatian care and correctional facilities (for details see Sekol, 2013, Sekol & Farrington, 2009, 2010, 2016a, 2016b). For the ease of later interpretation and in line with the “whole school approach,” according to which school anti-bullying policies should be multifaceted and target individual pupils, classes, and schools as institutions, this article summarizes the proposed evidence-based policy recommendations by dividing them into four levels: L1 = Level 1 (situational strategies); L2 = Level 2 (strategies on an individual level); L3 = Level 3 (strategies on a residential group level); and L4 = Level 4 (strategies on an institutional level). 3
First (L1), supervision of residents needs to be increased at times and places of high risk for bullying, and special attention needs to be paid to recognizing indirect forms of bullying. Staff should be particularly observant of events occurring in bedrooms during the night, but public communal areas such as living rooms, yards, and corridors should also be adequately supervised. Residential units should be as homogeneous as possible and house residents of the same age and reasons for admission (Sekol & Farrington, 2009). They should also aim to admit residents at approximately the same time to prevent those who are knowledgeable about the current group dynamics from exploiting newcomers. Similarly, residents who are in care for the first time should not be housed together with residents who have a long history of being in care. Efforts also should be made not to house residents who have a history of bullying during previous placements and/or in school together with residents who have a history of victimization. Similarly, residents who are new to the facility and who have a history of victimization in other settings should be intensively protected (Sekol & Farrington, 2016a, 2016b).
Second (L2), both male and female bullies and victims may benefit from cognitive-behavioral programs aimed at changing attitudes approving of bullying (Sekol & Farrington, 2016a, 2016b). Bullies of both genders may also benefit from techniques to control their carelessness, lack of self-control, and impulsive tendencies as well as from techniques that would improve their cooperation with others. The STOP-THINK-DO technique and cooperation skills training may be especially suitable for achieving these aims. Empathy enhancement may also be useful for changing the behavior of male bullies (Sekol & Farrington, 2016b). Finally, while victims may also benefit from programs such as the STOP-THINK-DO technique, programs aiming at improving their self-esteem, namely social skills and assertiveness training, seem to be the most important interventions targeting the behavior of victims (Sekol & Farrington, 2016a). Overall, staff need to be able to understand residents’ behavior and identify the causes of bullying (Anglin, 2002).
Third (L3), as deficits in perceived peer support were related to bullying and victimization, it is essential that staff should encourage residents to establish and maintain warm and supportive relationships with each other. This can be achieved by various forms of group work including group discussions or workshops, emphasizing similarities between residents (e.g., their similar life experiences), rather than differences between them. Group work should also be used to make residents aware that they have the power to create and promote a residential environment where bullying is not tolerated. Establishing a set of clear group rules about bullying and making a poster describing those rules could serve as a good starting point in such attempts (Sekol, 2013).
Fourth (L4), a positive relationship between residents and staff needs to be encouraged. Staff need to be consistently proactive and positive toward residents, treat them with respect, and consistently communicate a nonviolent philosophy (Sekol, 2013). Residents need to feel that they are all treated by the same manner and that rules and decisions made about them are transparent and fair.
Although research on bullying in residential care has been increasing recently, no research has examined how residential care staff view the problem of bullying and the most effective ways to tackle it. To date, only Connell and Farrington (1997) have investigated staff estimates of bullying and victimization in Young Offender Institutions, while Sekol and Farrington (2013) investigated staff estimates of bullying and victimization in residential care. However, the main aim of these two studies was to validate a bullying questionnaire, not to gain a broader understanding of staff views on bullying in their facilities. For instance, Sekol and Farrington’s (2013) study was nonanonymous and 12 staff members assessed 35 residents who they knew very well on a range of items indicative of being a bully or victim. Although the results demonstrated good general agreement between staff reports and residents’ self-reports, it was concluded that self-reports provide the most reliable and valid data about bullying among institutionalized youth.
An effective anti-bullying strategy can only be implemented if staff are well aware of the nature, prevalence, causes, and consequences of bullying in their facilities. The main aim of the present study is to assess how the staff view the problem of bullying in their facilities in broader terms, namely when they are asked to estimate overall bullying problems in their facilities, without having particular residents in mind. The specific aims of the present study are (a) to assess staff understanding of the nature and prevalence of bullying in their facilities, including some basic characteristics of bullies and victims; (b) to examine current staff policies regarding bullying; (c) to compare staff reports with residents’ own self-report data collected in the SSBCRC; and (d) to compare anti-bullying policies currently used by staff with evidence-based proposed policy solutions arising from the SSBCRC.
Method
Sample
Staff were sampled from 22 Croatian residential care facilities for children and youth during the data collection in the SSBCRC. All staff who were present in the facilities at the time of the SSBCRC data collection were asked to participate in the study. Additional questionnaires and envelopes were also left with the head of each facility for other members of staff to fill in. The completed questionnaires were either returned in sealed envelopes directly to the researcher during the data collection or posted to her home address. Staff in one correctional facility refused to participate in the study. In another correctional facility, only three staff members completed the questionnaire, one of which was returned incomplete. Those three questionnaires were thus excluded from the sample.
In total, 151 questionnaires from the remaining 20 facilities were returned. Of these, 11 questionnaires were either blank or incomplete. The final sample, therefore, comprised 140 members of staff from 20 Croatian residential care facilities. Of 140 staff members, 16 (11.4%) were male and 124 (88.6%) were female. On average, they had spent 12.6 years working in their current facility (SD = 9.9). Most of the staff were working directly with residents on a daily basis as members of the “treatment division.”
Of 20 facilities included in the sample, 10 were children’s homes, five were community residential homes, three were state residential homes, and two were correctional institutions. Community residential homes, state residential homes, and correctional institutions all accommodate young people who manifest antisocial or delinquent behavior, while children’s homes formally accommodate only children and young people without behavioral problems aged from 7 to 21 who either have no parent who is responsible for them or who have been neglected or abused in their biological or foster families.
Procedure
The proposal for the study was submitted to the Croatian Ministry of Health and Social Care which granted the ethical approval for the study. Each institution that participated in the study was then sent the Ministry’s approval, together with a detailed description of the study, its instruments, and procedures. All 22 facilities confirmed their approval of the study.
The data were collected as part of the SSBCRC, which also collected self-report data on bullying from 601 residents from 22 facilities, and which is described both in the introduction to this article and elsewhere (for details, see Sekol & Farrington, 2009). Before collecting the data from residents, the researcher approached all staff who were present in the facilities at the time of the study and asked them to participate in the study. The goals of the study were verbally explained to all members of staff who were present at the facility and the anonymity of the research was stressed. Unmarked envelopes were provided with each questionnaire. Staff were told to complete their questionnaires alone at any time and to return their completed questionnaires in a sealed envelope to the researcher on the next day.
Originally, the first part of the questionnaire also asked staff about their age and role in the facility. However, in the first facility, staff expressed concerns about anonymity because they believed that revealing their age and professional role, together with their work experience and gender, would make it easy for the head of the facility to identify them if that person obtained the questionnaires. Therefore, no data about age and professional roles was collected. Although some staff members returned their completed questionnaires directly to the researcher on the next day, others insisted on posting their questionnaires to the researcher’s home address, again because of concerns about privacy. In each facility, the researcher asked the head of the facility or other staff members to distribute the questionnaire to their colleagues who were not present in the facility at the time of the study. In these cases, staff posted their completed questionnaires to the researcher.
Measures
An anonymous questionnaire assessing staff awareness of bullying in their facilities was constructed for this study. The questionnaire was constructed with the aim of being a supplement to residents’ self-reports of bullying and victimization collected in the SSBCRC. 4 As such, the questionnaire covered the main areas of the self-report bullying questionnaire that residents completed in SSBCRC. The questionnaire consisted of 20 items and was divided into three parts. The first part provided a description of the aims of the study, instructions on how to fill in the questionnaire, and a broad definition of bullying. A list of 25 direct and indirect bullying behaviors, equivalent to behaviors indicative of bullying included in the residents’ self-reported questionnaire, was provided next (for details on items included in the residents’ questionnaire, see Sekol & Farrington, 2009).
The questionnaire included the following definition of bullying: Residents are being bullied when they are the victims of direct and/or indirect aggression happening at least 2-3 times a month, by the same or different perpetrator(s). Residents are also being bullied when they believe that they had been aggressed toward even if the actual intention of the bully to cause harm, or the imbalance of power, is not immediately evident. Single incidents of aggression cannot be considered bullying until a reliable method for measuring “the fear of future victimization” is established. (Ireland, 2002, p. 26; adapted here for residential settings)
The list of (in)direct behaviors that constitute bullying included behaviors such as “forcing someone to do chores for others,” “taxing,” “intimidating,” “gossiping,” “socially excluding others,” “spreading false rumors about someone,” “theft,” “damaging someone’s belongings or personal space,” “name calling,” “hitting,” and the like. The first part of the questionnaire also collected some basic information about staff, namely the length of their professional experience in the facility and their gender.
In the second part of the questionnaire, staff were asked to state how often they witnessed bullying in their facility, whether they perceived bullying in their facility as a problem and to assess the prevalence of bullies and victims in their facility, having in mind that bullying and victimization must occur at least 2 to 3 times a month or more often. Other questions that were included in the second part of the questionnaire asked staff to assess certain basic characteristics of bullies and victims, namely those related to their age and physical build, the length of their stay in the facility, and their willingness to report bullying to staff. Staff were also asked whether they believed that bullying was just part of the way things work in residential care and whether they talked to bullies when they found out about bullying.
The third part of the questionnaire was open-ended and asked staff to say what they normally did to combat bullying in their facility.
Data Analysis
Of 151 staff questionnaires that were collected, 11 had more than 10% missing data and were consequently deleted listwise. Only 6% of the remaining 140 questionnaires had some missing data on the open-ended questions and these were deleted pairwise. As none of the results reported in the present study were used for further multivariate analyses, these pairwise deletions are justified (Schlomer, Bauman, & Card, 2010). No pairwise deletions were made for the data collected in the first two parts of the staff questionnaire.
Frequency counts were calculated for each item included in the first and second part of the staff questionnaire for N = 140. Items 14 to 16 of the second part of the questionnaire were dichotomized in the same manner as equivalent items were dichotomized for the self-report data. Staff data were then compared with (a) the residents’ self-reported prevalence of bullying and victimization collected in the same study (SSBCRC) from 601 residents, which was considered to reflect the true prevalence of bullying (48.1%) and victimization (61.4%); (b) the residents’ self-reported types of bullying, times and places of bullying, and basic characteristics of bullies and victims; and (c) evidence-based policy implications arising from the residents’ data.
To compare the staff-reported prevalence of bullying and victimization to their self-reported counterparts, each staff member was scored according to the mean of their reported percentage range. The mean and standard deviation were then calculated for 140 staff-reported bullying and victimization scores. The following T statistic was then used, separately for bullying and victimization scores: T = (self-reported prevalence – staff-reported mean) / pooled SE. Where possible, for dichotomized items regarding characteristics of bullies as well as types, times, and places of bullying, crosstabulations were conducted. Overall, as demonstrated in tables and text below, 33 tests comparing residents and staff reports were conducted. Of these, 17 tests were significant at p = .001, two were significant at p = .01, and two were significant at p = .05, which greatly exceeds the number of significant results that would be expected by chance alone. Due to the open-ended nature of the item measuring staff reactions to bullying, the comparison of the strategies used by staff to combat bullying in their facilities with evidence-based policy implications, arising from research on the self-reports of residents, was conducted descriptively.
Given that no significant variations in bullying and victimization by type of facility were found for resident reports in the SSBCRC, as described in Sekol and Farrington (2010), the data were not disaggregated by type of facility.
Results
Staff and Resident Reports of the Overall Prevalence of Bullying
The results demonstrated that almost all staff (95.1%) believed that bullying in their facility was a serious problem. Of these, 22.9% believed that bullying was a serious problem that required urgent policy solutions, 42.1% believed that bullying was a serious problem that had been becoming increasingly more serious, and 30.1% believed that bullying was a serious problem that had been decreasing in frequency and seriousness. Only 4.9% of staff believed that bullying was not a problem in their facility.
More than three quarters of staff (77.8%) stated that they witnessed bullying among residents in their facility 2 to 3 times a month or more often. Yet, when staff were asked to assess the number of bullies and victims, their estimates were much lower than self-report estimates. More precisely, staff estimates of the prevalence of bullying and victimization were significantly lower than residents’ reports (20.5% vs. 48.1% for bullying, t = −17.25, p < .001; and 24.1% vs. 61.4% for victimization, t = −22.0, p < .001).
Staff and Resident Reports of Types of Bullying
The left hand side of Table 1 compares the prevalence of residents who reported being bullied in different ways and the prevalence of staff who believed that the particular type of bullying was the most prevalent. Therefore, (a) staff data in Table 1 do not imply that staff reported a higher or lower prevalence of certain bullying types than residents and (b) no direct comparisons (i.e., crosstabulations) between staff and resident reports could be conducted for types of bullying. As can be seen from Table 1, residents reported that they were most frequently bullied indirectly, while staff believed that verbal bullying was the most prevalent form of bullying. Of all types of direct bullying, both residents and staff reported that verbal bullying was the most prevalent type of bullying, while physical bullying was the least prevalent bullying type.
Self-Reported and Staff-Reported Types of Bullying.
Note. R = the prevalence of residents who reported being bullied in different ways; S = the prevalence of staff who believed that the particular type of bullying was the most prevalent; the percentages in both columns do not add up to 100 because the majority of residents and staff who reported indirect bullying also reported certain types of direct bullying, namely verbal and coercive/intimidating bullying.
Staff and Resident Reports of Times and Places of Bullying
Tables 2 and 3 compare both residents’ and staff perceptions about times and places of bullying, which allowed for direct comparisons between the two data sources. As can be seen on the left hand side of Table 2, while most residents believed that most of the bullying occurred either any time during the day (34.2%) or during the night (22.6%), most staff members believed that bullying most frequently happened in the evening, between dinner and sleeping time (33.8%) or between lunch and dinner time (26.6%). Staff were significantly less likely than residents to report that bullying usually occurred during the night and in the morning before breakfast. Staff were also significantly less likely than residents to report that bullying was likely to occur at any time of the day. Staff were significantly more likely than residents to report that bullying usually occurred between dinner and sleeping time, between lunch and dinner time, any time between dinner and waking up the next day, and any time between breakfast and dinner (for details, see the left hand side of Table 2).
Self-Reported and Staff-Reported Times of Bullying.
Note. In cases where no participants reported a certain item, the Haldane–Anscombe correction was performed by adding 0.5 to all 4 cells. R = residents; S = staff; OR = odds ratio.
p < .10. **p < .01. ***p < .001.
The left hand side of Table 3 shows places of bullying as reported by staff and residents. As can be seen from the table, while residents reported bedrooms as the most frequent location of bullying (36.5%), almost a quarter of the staff (22.7%) believed that bullying occurred equally frequently in bedrooms, living rooms, dining halls, corridors, and yards. Staff were significantly less likely than residents to believe that bullying most often occurred in the bedroom and the yard. Staff were significantly more likely than residents to report that bullying occurred equally often (a) in bedrooms, living rooms, dining halls, corridors and yards; (b) in living rooms and yards; and (c) in showers and toilets. Staff were also significantly more likely than residents to report that bullying occurred outside the home.
Self-Reported and Staff-Reported Places of Bullying.
Note. In cases where no participants reported a certain item, Haldane–Anscombe correction was performed by adding 0.5 to all 4 cells. R = residents; S = staff; OR = odds ratio.
p < .10. *p < .05. **p < .01. ***p < .001.
Basic Characteristics of Bullies and Victims and “Normalization” of Bullying as Reported by Residents and Staff
As briefly described in the introduction and in more detail elsewhere (i.e., Sekol & Farrington, 2016b), the self-report data demonstrated that bullies were not older than nonbullies (mean age = 15.88 years vs. 15.82 years; t = 0.36; r = .02, ns) or significantly more likely than nonbullies to be institutionalized for problematic behavior, including criminal offending (52.6% vs. 46.5; χ2 = 2.25; odds ratio [OR] = 1.35, ns). Bullies had spent more time in their current facility than nonbullies, but this difference (a) was accompanied by a small effect size (mean institutionalization length = 31.67 months vs. 25.00 months; t = 2.44; p < .05; r = .10); (b) was driven by gender differences (i.e., it was nonsignificant for male bullies); and (c) disappeared after controlling for other personal characteristics of both male and female residents. Almost 60% of staff, however, believed that bullies were older than other residents, while 41.7% of staff believed that bullies had spent more time in the institution than nonbullies, and 27.1% of staff believed that bullies were more likely to have a criminal record.
As briefly described in the introduction and in more detail elsewhere (i.e., Sekol & Farrington, 2016a), self-report data demonstrated that male victims were significantly younger than other residents, although this difference was accompanied by a small effect size (mean age = 15.65 years vs. 16.21 years; t = −2.85; p < .01; r = .14). Neither male victims (mean institutionalization length = 27.41 months vs. 22.21 months; t = 1.79; r= .09) nor female victims (mean institutionalization length = 36.87 months vs. 28.47 months; t =1.37; r = .10) had spent significantly less time in their institutions than other residents. However, almost 50% of staff (47.1%) believed that typical victims were younger than other residents, while 39.3% of staff believed that typical victims had spend less time in their current facility than other residents.
Only 5.7% of staff members believed that bullying was just part of the way things worked in residential care. However, 54.1% of all residents included in the sample of 601 residents (i.e., bullies, victims and not involved; for details see Sekol & Farrington, 2009) believed that bullying was just part of the way things worked in residential care. This difference was significant (χ2 = 107.33; p < .001; OR = 19.43). Furthermore, while 88.6% of staff stated that they always try to put a stop to bullying when they know bullying is happening, significantly less residents (40.4%; see Sekol & Farrington, 2009) believed that staff almost always or often try to put a stop to bullying (χ2 = 105.26; p < .001; OR = 11.42). Only 31.2% of residents (Sekol & Farrington, 2009) believed that staff always or often knew about bullying, while significantly more staff (57.1%) stated that they were well aware of bullying incidents in their facilities (χ2 = 33.37; p < .001; OR = 2.95). Finally, almost 80% of staff believed that most of the bullies were also victims, and vice versa, which is in line with the fact that, according to self-reports, bully/victims were the most prevalent group (for details, see Sekol & Farrington, 2009, 2010).
Evidence-Based Policy Recommendations Versus Staff-Reported Anti-Bullying Strategies as Used at the Time of the Research
Table 4 compares evidence-based policy recommendations described in the introduction and elsewhere (Sekol, 2013, Sekol & Farrington, 2009, 2010, 2016a, 2016b) with strategies that staff reported using to combat bullying in their everyday practices with residents. In line with the “whole school approach” in combating bullying in schools described in the introduction, Table 4 divides both evidence-based and staff-reported strategies into four levels. As can be seen from the table, most staff (56.1%) reported using situational interventions, while only 6.8% of staff reported using interventions on an institutional level. Strategies on a group level were used by 42.4% of staff, while strategies on an individual level were used by only a quarter of staff.
Evidence-Based Policy Recommendations Versus Staff-Reported Current Anti-Bullying Strategies.
Percentages in this column are calculated for the number of staff that reported strategies on the level in question, not for the total number of staff.
This protocol is the only Croatian document that addresses the responsibilities of adults in cases of bullying among youth in all settings, mainly in school setting. As such, the protocol is not made to specifically meet the needs of bullying prevention and reduction in residential care settings.
As can further be seen from Table 4, the majority of staff who reported using situational strategies (70.4%) stated that they use rather reactive strategies such as (a) reporting bullies to the police or Centers for social care (31.1%); (b) applying the Croatian “Protocol in cases of bullying among children and youth” (20.3%), which is also rather reactive and not adjusted to residential care settings; and (c) applying disciplinary proceedings and punishments (mainly in the forms of losing privileges, placing bullies in the “special control and supervision unit” and/or transferring them to another institution; 19.0%). Other staff-reported situational strategies mainly included attempts to create homogeneous compositions of residential groups and to increase supervision of residents.
However, while the evidence-based situational recommendations are rather specific in listing where and when supervision needs to be increased or how to make residential groups more homogeneous, staff-reported situational strategies do not specify times and places in which supervision needs to be increased nor do they make specific suggestions on how to make residential groups more homogeneous (for details, see Table 4). Of all the staff who reported using situational strategies, only 10.8% suggested matching roommates and moving victims or bullies to another residential group, while 1.3% reported physically separating residential groups, but without being specific in terms of criteria for either matching roommates or separating residents. The remaining staff-reported strategies, although somewhat more specific, mainly included certain bans and are, therefore, not in line with the evidence-based suggestions about trying to make situational prevention as proactive and subtle as possible.
The majority of staff who reported using strategies on an individual level (63.6%) stated that they either talked only to bullies or to both bullies and victims, but without specifying the main focus of the conversation. An additional 15.2% of staff, who used interventions on an individual level, reported talking to bullies with the simple aim of conveying the message that bullying was unacceptable. Only 21.0% of all staff who used strategies on an individual level (10.6% of the entire staff sample) reported using strategies aiming at deeper and long-lasting changes in residents’ behavior, either through: (a) empathy enhancement and nonviolent problem-solving techniques for bullies; (b) strategies aiming at improving self-esteem and self-worth of victims; (c) improving social and emotional skills of all residents; and (d) talks with residents with the aim of developing trusting relationships between them (for details, see Table 4).
Compared with staff-reported strategies on the individual level, evidence-based strategies are again more specific, and suggest techniques intended to achieve deeper changes in residents’ behavior such as (a) cognitive-behavioral programs aimed at changing attitudes approving of bullying for all residents; (b) techniques aimed at carelessness, lack of self-control, impulsivity, and cooperation skills of all bullies; (c) empathy enhancement programs for male bullies; and (d) programs aiming at improving the self-esteem of victims (i.e., social skills and assertiveness training; for details, see Table 4).
More than half of the staff who reported using interventions on a residential group level (58.5%) stated that they talked to their residential groups, but without specifying the main focus of those conversations. In line with the proposed evidence-based strategies, one third of all staff who used interventions on a group level stated that they used group work and workshops with residents discussing themes of friendship, solidarity, tolerance, interpersonal relationships, and the like (Table 4). However, staff who reported these strategies accounted for only 14% of the entire staff sample. Only four staff members reported educating residents about bullying and no staff members reported making rules about bullying with residents. Three staff members reported structuring residents’ days with the aim of decreasing boredom by engaging them in various activities. However, it remained unclear whether these were necessarily group activities. Finally, strategies on an institutional level were used by only 6.8% of staff and mainly referred to staff committee discussions (Table 4).
Discussion and Conclusion
Almost all staff believed that bullying was a serious problem in their facility and more than three quarters stated they had witnessed bullying among residents at least 2 to 3 times a month. Yet, when asked to assess the number of bullies and victims, staff estimates were significantly lower than resident self-reports. This is consistent with school-based studies described in the introduction (i.e., Bradshaw et al., 2007; Mishna et al., 2005; Pervin & Turner, 1994), but inconsistent with the study by Sekol and Farrington (2013) which demonstrated a good general agreement between staff reports and resident self-reports of bullying and victimization in residential care, although staff reports showed lower agreement with resident self-reports than did peer reports.
However, as mentioned earlier, Sekol and Farrington’s (2013) aim was not to gain a broader understanding of staff views on bullying in their facilities. Rather, their aim was to validate a bullying questionnaire by asking staff to assess residents who they knew very well on a range of items indicative of being a bully or victim. It is possible, therefore, that the good agreement between resident self-reports and staff reports in their study resulted from the fact that a very detailed list of behaviors indicative of bullying and victimization was used and that staff only assessed residents who they worked with daily and knew very well. However, the present study demonstrates that staff seem to have difficulties in estimating the larger picture of bullying in their facilities. In other words, it seems that staff are rather accurate in rating residents as bullies or victims when they are asked about specific residents who they work closely with, but that they seem to have difficulties in estimating the overall number of bullies and victims in their facilities.
The fact that staff reports of the overall prevalence of bullying and victimization in their facilities were significantly lower than resident self-reports, regardless of the majority of staff regularly witnessing bullying and perceiving it as a problem, might tentatively suggest that staff believe that most of the bullying is performed by small numbers of bullies and includes only a small number of victims. In line with the findings of Ladd and Kochenderfer-Ladd (2002), on peer reports of bullying in a sample of preschool to middle school children, it is indeed possible that, when thinking about the overall prevalence of bullying in their facilities, staff limit their attention to the bullying and victimization of a small number of residents whose experiences they know best or whose behavior is the most noticeable and disruptive.
Although staff were relatively well aware of the most and least prevalent types of bullying, staff perceptions of the times and places of bullying significantly differed from residents’ perceptions. Overall, staff underreported bullying that happened at times when they were absent from the facilities (i.e., during the night when only the “night staff” were present) or when they were just starting their shifts (i.e., in the morning before breakfast). They overreported bullying that happened at times when the majority of staff and residents were present in the facility (i.e., in the afternoons and evenings).
Staff underreported bullying that happened in bedrooms and in communal locations such as living rooms, yards, and recreation areas, but overreported bullying that occurred equally often in different combinations of various locations. Compared with residents, who were rather specific about places where bullying occurred, staff were much less certain about the specific locations of bullying in their facilities. Overall, it appears that staff believed that bullying occurred in various unspecified locations at times when they were present in the facility.
The fact that staff underreported bullying that happened in specific communal areas such as living rooms or yards either means that staff do not supervise these areas appropriately or that a lot of bullying is happening despite staff being present in those communal bullying locations (Sekol & Farrington, 2009). Both possibilities might occur as well. Indeed, in a qualitative study of peer violence in Croatian residential care (Sekol, 2013), residents noted that staff tended to spend a lot of time in “the staff room” and spent time with residents as a group mainly when there is a problem. Similarly, as self-report data demonstrated that indirect, less noticeable types of bullying were more prevalent than direct types of bullying, it is possible that, in cases when staff do spend time with the residential group, they tend to overlook indirect types of bullying.
In line with school-based research described in the introduction (i.e., Mishna et al., 2005; Nicolaides et al., 2002), staff tended to hold stereotypical views about basic characteristics of bullies and victims. Although this was not supported by resident self-reports, a significant number of staff believed that (a) bullies had criminal records, they were older than other residents, and they had spent more time in care; and (b) victims had spent less time in care and were younger than other residents. As mentioned in the introduction, the SSBCRC demonstrated that, rather than age and length of institutionalization, other factors contribute to bullying and victimization in care. These included (a) individual characteristics of residents; (b) characteristics of the residential care environment; and (c) the nature of the residential care social world, marked by constant changes in the residential group dynamics through discharges and new admissions. It is the latter that probably contributes to the large proportion of self-report bully/victims in care, which staff did recognize to be prevalent. It is, therefore, somewhat unusual that staff hold traditional views about the characteristics of bullies and victims, even though they might be aware of some residential processes that shape the large prevalence of bully/victims.
Compared with residents, staff were significantly less likely to believe that bullying was just part of the way that things work in residential care, and significantly more likely to state that they always tried to stop bullying and that they were well aware of bullying incidents in their facilities. In fact, while more than half of residents believed that bullying was just part of the way things worked in care, virtually no staff believed this. Together with the fact that staff were more likely than residents to believe that they always intervened in bullying situations, this might mean that, unlike residents, staff do not perceive bullying in care to be normal. Although this would certainly be admirable, it raises the question of how it is possible that most staff perceive bullying to be a serious problem, which they witness often and are well aware of, and still do not believe that bullying might be part of the way things work in care. Future research should examine whether staff truly believe that bullying in care is not normalized or they only tend to provide socially desirable answers.
Given qualitative evidence based on resident accounts, according to which in Croatian care institutions some staff sometimes use residential group hierarchies and violence among residents as a means of controlling residents (Sekol, 2013), it seems more likely than not that staff do perceive bullying as normal to a certain degree. Unlike residents, however, such a perception might be more subconscious for staff. Rationally, staff know that bullying is unacceptable so, when they are asked whether it is normal, they might be likely to state that it is not, despite witnessing it often and perceiving it to be a serious and growing problem. This is, of course, speculative and future research could address this topic of underreporting in more detail. The results of the present study, however, do indicate that staff and residents’ perceptions of how normal bullying in care is, how aware staff are of bullying, and how often staff intervene in bullying situations are radically different.
Most anti-bullying interventions reported by staff differed from the implications of evidence-based policy recommendations described in the introduction. Compared with evidence-based policy recommendations, the anti-bullying strategies used by staff tended to be intuitive, reactive, and unspecified, rather than proactive, specific, and based on residents’ needs. The majority of staff reported using reactive situational strategies, such as reporting bullies to the police, applying disciplinary proceedings or protocols, and isolating bullies. Other situational strategies reported by staff included increased supervision of residents, certain bans and restrictions, and attempts to create homogeneous residential groups, but without specifying where and when supervision needed to be increased or how homogeneous residential groups could be created.
Strategies on a residential group level were the second most preferred interventions used by staff, and these mainly included unspecified group talks with residents. More specific and proactive strategies on a group level, such as those suggested by evidence-based recommendations (i.e., education of residents about bullying, rule making or workshops about tolerance and friendship), were used by few staff. Strategies on an individual level were used by only one quarter of staff and in most cases referred to rather simple unspecified conversations with individual bullies and victims. Extremely small numbers of staff reported using individual strategies targeting at topics identified as important in evidence-based recommendations, such as empathy and self-esteem enhancement for bullies and victims, respectively (each reported by less than 2% of staff). Strategies on an institutional level were virtually nonexistent, implying a lack of institutional clarity regarding anti-bullying policies in Croatian care institutions.
The current study is limited by the fact that staff reports collected by questionnaires depend on the honesty of participants’ answers. Although assured of anonymity, staff were concerned about the possibility of their identity being revealed, which might have made them underreport bullying problems in their facilities and provide socially desirable answers when referring to anti-bullying strategies they used or wished to use. It would be useful to validate staff reports of intervention strategies against resident reports in future research, as well as to use social desirability scales. A further problem is that staff reports of anti-bullying strategies were collected by answering open-ended questions. This resulted in staff replies being rather short, and quantifying them might have resulted in further loss of information. Future research could qualitatively examine anti-bullying strategies used by staff to address this issue. Another limitation of this study is that it only compared staff reports and residents’ reports on a small number of rather descriptive factors. Future research should aim to examine whether there are any further individual and contextual factors that contribute to the discrepancies between staff and residents’ reports. Finally, the current study was conducted in Croatia and its results may not be applicable to other countries. A comparative study of staff reports of bullying and their comparison with resident reports would be needed to replicate the findings of this study.
Despite its limitations, the current study contributes to residential care bullying research and raises many new questions, which need to be examined in the future. It is the first study to quantitatively examine staff views on bullying in residential care by comparing staff views with residents’ self-reports. In doing so, the current study illustrated that staff seem to be unaware of the true prevalence of bullying and victimization. The study also demonstrated that staff predominantly (a) use anti-bullying policies that seem to be intuitive, rather than evidence-based, and (b) believe that current anti-bullying policies could be improved by additional reactive situational interventions. As such, the present study has implications for residential care policy.
First, given that staff views on life in care seem to significantly differ from residents’ perceptions, it is essential that staff try to put themselves in residents’ shoes and take residents’ perspectives more often. Staff need to take a broader view of life in care and realize that life in care continues at times and places when they are not present. Efforts need to be made to change classic shift work and provide residents with continuity of care. Because staff are not present in the facility for most of the day, shift work does not allow staff to fully understand life in care or realize the true prevalence of bullying. Continuity of care is important not only for bullying prevention specifically but also for developing strong and secure emotional attachments with significant adults, which is a prerequisite for healthy emotional development.
Second, education of staff about bullying is needed. Staff should be taught about a definition of bullying applicable to residential care, types of bullying in care, and consequences of bullying not only for the victims but also for the entire residential group and therapeutic processes that residential facilities aim to achieve. It is particularly important that staff are made aware that bullying in care is much more prevalent than they think and that they should not limit their attention to only a small number of residents whose behavior is most noticeable or disruptive. It is also essential that staff understand that bullying in care is shaped by more complex factors and processes than school bullying and that stereotypical views of bullies and victims do not seem to apply to residential care. Staff need to be made aware of the true personal characteristics of bullies and victims in care, not only to be able to recognize bullies and victims more easily but also to tailor intervention strategies according to the actual needs of bullies and victims. All education efforts need to be shaped according to research evidence about bullying in residential care. To ensure staff are motivated to participate in staff training, staff education could be counted as their working days. Similarly, a reward system on the adequate management of bullying could be another motivational mechanism for staff to face the bullying problems.
Finally, staff need to be taught the principles of an effective anti-bullying strategy and trained in more proactive approaches to dealing not only with bullying but also with other challenges of working in residential care. Staff need to be made aware that “. . . the control exercised should be the minimum control necessary for young people to grow and learn for themselves” (Elliot & Thompson, 1991, p. 126) and that an overreliance on reactive, punitive situational strategies may be counterproductive, conveying the message that residential placements are dangerous places and encouraging violent behavior as a way of “self-protection.” The supervision in residential care should, therefore, be performed in a way which would allow residents to perceive the presence of staff as a natural part of residential living, not as “external control.” Instead of overly relying on situational strategies, more emphasis needs to be placed on strategies aiming at both deeper changes in residents’ behavior and reforms of residential care systems as a whole, namely through interventions on individual, residential group and institutional levels. Only reformed residential care systems for youth, in which young people would be cared for in the same manner as other young people are cared for by their own families, could prevent bullying among young people in care and could also help residents to develop their full potential and become integrated members of society. This is a social justice prerequisite that is in everyone’s interests.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
