To disclose or not? Children’s tendency to disclose peer victimisation in elementary school

Abstract

This study aimed to shed light on the prevalence of chronic peer victimisation among Dutch elementary school children and factors associated with (non-)disclosure of such experiences by victims. 5,961 students from 73 schools participated (51.5% male; M_age = 9.96; 77.7% native Dutch). Results showed that 12.3% of all children were victimised chronically, of which 29.4% did not disclose. Multilevel logistic regressions indicated that girls, older and native Dutch children were more likely to disclose. Moreover, experiencing depressive symptoms prompts disclosing victimisation. Anxiety, high emotion regulation skills, and perceptions of cohesion in the classroom were negatively associated with disclosure. The disclosure was not related to frequency or duration of victimisation, self-perceived social acceptance, self-worth, impulse control, or perceived classroom climate. Our study reveals information on the prevalence of peer victimisation and its disclosure, based on a nationwide study conducted in 2016–2017. It gives important insights into factors associated with disclosing victimisation experiences.

Keywords:

• Peer victimization
• school bullying
• disclosure
• elementary school

Bullying among children in elementary schools is a pervasive problem all over the world. Numerous studies have shown that peer victimisation affects children’s well-being, psychological and physical health, and academic outcomes (e.g. Arseneault, 2018; Juvonen et al., 2011). Given these severe and sometimes long-lasting consequences, it is not surprising that bullying is considered a serious public health issue (WHO, 2010). The first and essential step in providing adequate help is knowing who actually is being victimised by their peers. However, many of the victims never disclose their victimisation (Bjereld, 2018). To date, we have limited knowledge of the prevalence of disclosure, as well as what characterises victims who do (not) disclose their victimisation, especially among younger children. The goal of the current study, therefore, was to shed light on the prevalence of chronic peer victimisation in Dutch elementary schools and the factors associated with (non-) disclosure of such experiences by victims.

Disclosure of victimisation

There have been several reports on the prevalence and severity of bullying and victimisation across the world (Inchley et al., 2018; Modecki et al., 2014). In the Netherlands, ∼10% of children in elementary schools are victimised by their peers (Scholte et al., 2016). These prevalence reports are often based on anonymous and confidential surveys. It might be that children who indicate that they experience peer victimisation in these reports, remain silent to people in their immediate daily environment (e.g. classmates, friends, parents, teachers) and thus remain unidentified.

And indeed, studies have reported low disclosure rates of victimisation (Blomqvist et al., 2020; Petrosino & Guckenburg, 2010; Shaw et al., 2019): between 36 and 67% of youth who are victimised disclose their problems to friends, parents, or an adult in school. These studies have predominantly been conducted with adolescents in middle or high school (see, for an exception, Blomqvist et al., 2020). In one study in the Netherlands, Fekkes et al. (2005) asked children who are victimised whether they told their parents or teachers about their victimisation. They found that 48% of the children who experienced victimisation told their teacher and 62% told their parents. All in all, these studies show that roughly a third to half of youth who experienced victimisation actually do not disclose their victimisation to responsible adults, with higher numbers in adolescence.

A substantial group of victims thus appears hesitant to disclose being victimised to others. Reasons for keeping silent may vary from denial, feelings of shame, a sense of autonomy, and self-reliance, to distrust in adults and their abilities to alter the situation (Bjereld, 2018; deLara, 2012; Newman & Murray, 2005). This hesitance to disclose is highly unfortunate as talking about victimisation is considered essential for both solving the bullying situation and coping with negative feelings following the harassment (Mishna & Alaggia, 2005). Moreover, many of the existing intervention programs focus on talking about bullying as an essential key component, so that teachers and peers can support victims (Salmivalli, 2014; Veenstra et al., 2014). The success of such intervention programs has been mixed (Gaffney et al., 2019) and even with an evidence-based intervention a substantial percentage of children remains victimised (Kaufman et al., 2018). This may partly be due to non-disclosure of victimisation, resulting in a lack of awareness among peers and teachers that certain students are victimised. It is, therefore, crucial to obtain more insight into the characteristics and circumstances of children who do and those who do not disclose their victimisation to others.

In the past decade, more and more schools worldwide have the legal obligation to take care of the social safety and well-being of their students (Council of Europe, 2017; GOV.UK, n.d.; Stuart-Cassel et al., 2011). In the Netherlands, the law on ‘Social safety in school’ was implemented in 2015 (Wet Veiligheid Op School, 2015). Since then, all schools have to monitor the safety and well-being of their students regularly (e.g. by using questionnaires) and must make efforts to tackle bullying (e.g. by implementing an intervention).

We know little about victimisation and disclosure rates following the legal obligation of monitoring bullying and victimisation in the Netherlands. In the current study, we, therefore, aimed to investigate the prevalence of chronic victimisation (i.e. children repeatedly exposed to peer victimisation) and disclosure among elementary school children, using data from a large-scale nationwide study conducted in the academic year 2016–2017 survey.

Who is likely to disclose?

In addition to knowing the percentage of chronic victims who disclose their victimisation to others or not, it is also important to know who is more likely to disclose in which context. The likelihood to disclose peer victimisation may be related to demographic characteristics, frequency and duration of victimisation, children’s social-emotional functioning, and their perception of the classroom context.

With regard to demographic characteristics, we know that there is an age effect in talking about personal and peer problems. In general, younger children are more likely to seek help from others than older children as they perceive less social pressure to solve problems independently and believe more strongly that disclosing is an effective strategy (Bauman et al., 2016; Boulton et al., 2017). In addition to age, we examined the role of sex and ethnicity. It repeatedly was found that ethnic minority group members are less likely to seek help for a wide variety of problems, but evidence for bullying victimisation is mixed (Consedine et al., 2007; Eliot et al., 2010). Regarding sex differences, social and cultural expectations acquired at an early age, for instance, girls being socialised to focus and rely on relationships, may make girls more inclined to share problems with others compared to boys (Connolly, 2018).

In general, individuals are more willing to talk about problems when these appear more uncontrollable or emotionally distressing (Bauman et al., 2016; Newman, 2008). In this regard, the degree of exposure (e.g. frequency and duration) to peer victimisation may play a role in disclosure too: children who suffer from victimisation frequently and chronically may be more inclined to disclose their experiences (Shaw et al., 2019). Yet, on the other hand, these children may be more sceptical about the positive effects of disclosure (Unnever & Cornell, 2004). The evidence is mixed. In some studies, disclosure was associated with the frequency and/or duration of bullying (e.g. Blomqvist et al., 2020; Petrosino & Guckenburg, 2010). In other studies, the chronicity of victimisation was not related to talking about it (e.g. Hunter et al., 2004; Shaw et al., 2019).

There are also reasons to assume that children’s social-emotional functioning is related to the likelihood that they will disclose their victimisation. In this study we focus on aspects that were particularly addressed in the anti-bullying interventions implemented and examined in the larger nationwide study on the effectiveness of the anti-bullying program in The Netherlands, that is social anxiety, depression, self-perceived social acceptance and global self-worth, impulse control, and emotion regulation (see De Castro et al., 2018 for more information on the interventions).

First, it has been established that social anxiety, depression, emotional problems (Reijntjes et al., 2010), low self-esteem (van Geel et al., 2018), and low emotion regulation skills (Riley et al., 2019) are risk factors for being victimised by peers. Second, evidence exists that victimisation has important effects on social-emotional functioning: victims show lower levels of social-emotional functioning compared to non-victims (Arseneault, 2018). Third, many anti-bullying interventions have a strong focus on social-emotional learning to prevent or tackle bullying and victimisation (Espelage, 2015). Insights into the role of social-emotional functioning in the disclosure may further enhance intervention effectiveness and to further reduce chronic peer victimisation. Finally, it has been suggested that children and adolescents are prone to disclose victimisation when they are emotionally upset or confident in their peer relations, but not when they blame themselves, have low self-esteem, or can regulate their emotions (e.g. Boulton et al., 2017; Newman, 2003, 2008). For this reason, we examined whether the likelihood to disclose victimisation differed depending on various aspects of children’s social-emotional functioning.

Finally, the classroom context might be affecting disclosing victimisation. When children feel more comfortable and supported, they are more inclined to disclose problems to others (Eliot et al., 2010; Mishna & Alaggia, 2005). In classes with a safe social climate, children may thus be more likely to disclose their victimisation. In a recent study, it indeed was found that children were more likely to tell an adult about being victimised when they perceived peer support to the victims of bullying (Blomqvist et al., 2020). Hence, we examined whether the perception of children who are victimised on different aspects of the classroom context, indicated with feelings of comfort, cohesion, conflict, cooperation, and isolation, is related to disclosure of victimisation.

In summary, in this study we investigated: (1) the prevalence of chronic victimisation in Dutch elementary school children, (2) the prevalence of disclosure among children who experienced victimisation, (3) demographic and social-emotional differences between non-disclosing and disclosing victims, and (4) characteristics associated with the likelihood to disclose victimisation.

Methods

Participants

Data used in this study came from a large nationwide study of the effectiveness of anti-bullying interventions in schools (De Castro et al., 2018). This study was initiated by the Dutch government to examine (a) the prevalence and severity of bullying and school safety, and (b) the effectiveness of available anti-bullying prevention and intervention programs in the Netherlands.

In total, 5,961 students (51.5% boys; M_age = 9.96; SD_age = 1.20) from 288 classrooms (grades 3–6) and 73 schools participated in the study (response rate 97.5%). The majority were native Dutch (i.e. both parents born in the Netherlands), 22.3% of the students were from various minority groups (Dutch Antillean, Moroccan, Turkish, Surinamese, or other). This is representative of the Dutch population (Statistics Netherlands, 2016). The scores of 11 students had to be excluded as their answers were unreliable (i.e. SD across self-reports = 0, not taking into account reversed items).

Procedure

Schools throughout the Netherlands were recruited via recruitment letters and follow-up phone calls. Schools for children with special educational needs were not invited. Schools were asked whether they wanted to participate in a nationwide study to monitor bullying and victimisation across the school year. Four cluster randomised and quasi-experimental trials were conducted. In each trial, schools either carried out one of the universal (i.e. targeted to all students) anti-bullying programs (intervention condition) or were on the waiting list (control condition). Data were collected twice; once at the beginning of the school year before the start of the interventions (T1; September-October 2016) and once at the end of the school year at intervention termination (T2; June-July 2017). We used data from T1 to avoid results being influenced by intervention effects.

Schools and teachers provided active consent. Parents received a letter in which the purpose and procedures of the assessment were described and parental consent was asked. In total, 71 parents objected to their child’s participation. Children were asked to give informed assent at the start of the questionnaire. None of the participants declined to fill out the questionnaire before or during the assessment. Procedures were approved by the Institutional Review Board of the [Radboud University in Nijmegen] (ECSW2015-1210-345).

Children completed an online questionnaire during regular school hours for ∼45 min. Videos were used to inform children about the goal and setup of the study. Participants were informed that the data would be pseudonymized (i.e. by using an inconvertible ID number) and handled confidentially. They were asked to keep their answers to themselves while filling out the questionnaire and to be truthful in answering all questions. Participants read and answered the questions independently. They were not allowed to talk to classmates during the assessment, but they could ask questions to their teacher. Children could stop participation at any time.

Measures

Frequency and duration of victimisation

A Dutch version of the widely used and validated revised Olweus bully/victim questionnaire (Olweus, 1996; Veenstra et al., 2020) was used to measure the frequency and duration of victimisation. Videos were used to explain different types of bullying, the intentional and repetitive nature of bullying, as well as the meaning of power imbalance. Next, participants were asked ‘Now you know what is meant with bullying, how often since the beginning of this schoolyear have you been bullied?’ (e.g. frequency of victimisation). They responded using a 5-point Likert scale, ranging from 1 ‘I have never been bullied’ to 5 ‘I have been bullied multiple times per week’. If children indicated that they had been bullied, they were asked about the duration (e.g. ‘How long were you bullied’) using a 3-point Likert scale (1 = since this school year, 3 = multiple years). Participants who indicated that they had been bullied at least twice a month (e.g. frequency of victimisation ≥3) and already before this school year (e.g. duration of victimisation ≥2) were labelled as chronically victimised (Bowes et al., 2013).

Disclosure of victimisation

Only children who indicated that they were victimised since the beginning of the school year were asked ‘Whom did you tell about the victimisation?’. They could name as many or as few of the following persons: 1 = your father, mother, or caretaker, 2 = brother or sister, 3 = a friend, 4 = the teacher, 5 = somebody else (Blomqvist et al., 2020; Veenstra et al., 2020). Children who disclosed their victimisation to at least one of these persons were labelled as disclosing victims. Children who indicated that they did not disclose their victimisation to any of them were labelled as non-disclosing victims.

Social anxiety

Social anxiety was measured with a Dutch version of the Social Phobia Screening Questionnaire (Furmark et al., 1999). We used seven items from the original questionnaire that were appropriate for this age group (e.g. ‘I am afraid to talk with somebody I do not know’) (Gren-Landell et al., 2009; Veenstra et al., 2020). Children could answer on a 5-point Likert scale (1 = never, 5 = always). Mean scores were computed with higher scores indicating higher levels of social anxiety. The internal consistency of this scale (α = .72). is comparable to psychometric properties in other studies (e.g. Gren-Landell et al., 2009; Kaufman et al., 2018).

Depressive symptoms

The severity of self-reported depressive symptoms was measured with the Dutch version (Veenstra, et al, 2020) of the Major Depression Disorder subscale. This is part of the Revised Children’s Anxiety and Depression Scale and turned out a suitable instrument for use among school-aged children (Chorpita et al., 2000; Kösters et al., 2015). Participants answered nine questions (e.g. ‘I do not really like anything’) on a 4-point Likert scale (1 = never, 4 = always). Mean scores were computed with higher scores indicating more depressive symptoms (Cronbach’s alpha = .76). Psychometric properties in other studies were comparable (e.g. Bouman et al., 2012; van der Ploeg et al., 2015).

Self-perceived social acceptance and global self-worth

The Dutch version (Veerman et al., 2004) of the self-perception profile for children (Harter, 1988) was used to measure self-perceived social acceptance and global self-worth. Self-perceived social acceptance (e.g. ‘I have a lot of friends’), as well as global self-worth (e.g. ‘I am happy with myself’), were measured with six questions. All questions were answered on a 5-point Likert scale (1 = not true at all, 5 = very true). Mean scores were computed for both subscales with higher scores indicating more feelings of social acceptance and self-worth. Cronbach’s alpha was .66 for self-perceived social acceptance and .77 for global self-worth, which is satisfactory and in line with psychometric properties found in other studies (e.g. Boor-Klip et al., 2016; Reijntjes et al., 2013).

Impulse control and emotion regulation

The Dutch version of the Difficulties in Emotion Regulation Scale (Gratz & Roemer, 2004; Veenstra et al., 2020) was used to measure impulse control and emotion regulation. To measure impulse control, we used the six items of the Difficulties Controlling Impulsive Behaviours when Distressed-subscale (e.g. ‘When I’m upset, I feel out of control’). Emotion regulation was measured with the eight items of the Limited Access to ER Strategies-subscale (e.g. ‘When I’m upset, it takes me a long time to feel better’). All questions were answered on a 5-point Likert scale (1 = almost never, 5 = almost always). Mean scores were computed with higher scores indicating more difficulties in impulse control and emotion regulation. Internal reliabilities were good (α = .80 for impulse control and α = .79 for emotion regulation) and similar to reliabilities found in other studies (e.g. Neumann et al., 2010; Weinberg & Klonsky, 2009).

Perception of classroom context

The Classroom Peer Context Questionnaire (Boor-Klip et al., 2016) was used to measure children’s perceptions of the classroom context. This questionnaire consists of five subscales: comfort (e.g. ‘In this class, I feel comfortable’), cohesion (e.g. ‘In this class, everyone belongs to the group’), conflict (e.g. ‘In this class, children are mean to each other’), cooperation (e.g. ‘In this class, children help each other’), and isolation (e.g. ‘In this class, some children are often alone’). Every subscale has four questions that are rated on a 5-point Likert scale (1 = not true at all, 5 = very true). Mean scores were computed for all subscales with higher scores indicating more feelings of comfort and perceptions of more cohesion, conflict, cooperation or isolation in the classroom. Cronbach’s alpha was .83 for comfort in class, .70 for cohesion, .81 for conflict, .84 for cooperation, and .72 for isolation, which is comparable to other studies (e.g. Braun et al., 2020; Laninga-Wijnen et al., 2021)

Analyses

We first examined (a) how many children reported to be chronically victimised, (b) whether or not they told someone about their victimisation, and (c) to whom they disclosed their victimisation. Next, we examined whether non-disclosing and disclosing victims differed from each other in terms of demographic characteristics, frequency and duration of victimisation, social-emotional functioning, and perceived classroom context, using Independent samples t-tests. Lastly, we examined whether victims’ likelihood to disclose was associated with demographic characteristics, the frequency and duration of victimisation, children’s social-emotional functioning, and the perceived classroom context. As children were nested in classrooms, and the intra-class correlation of .064 indicated that some variation in disclosure across classrooms exists, multilevel binary logistic regression techniques were used.

Independent variables were grand mean centred to avoid multi-collinearity (see Supplementary Appendix 1 for bivariate correlations). Girls and native Dutch participants were used as a reference group. Significance tests were performed two-sided. All analyses were performed in SPSS 25.

Results

Disclosure of victimisation

Descriptive analyses showed that 12.3% (n = 741) of all children indicated being victimised chronically. Of these victimised students, 29.4% (n = 218) did not disclose their victimisation. The remaining victims (70.6%; n = 523) disclosed their victimisation to their parents (75.9% of disclosing victims), followed by friends (56.4%), teachers (49.9%), siblings (27.7%), and another person (20.5%). Further, 30.7% of the disclosing victims only told one person about their problems, whereas 69.3% disclosed to more than one person.

When comparing non-disclosing victims to disclosing victims (Table 1), disclosing victims turned out to be a few months older than non-disclosing victims, but both were on average 9 years old. It also can be seen that there were more children with an ethnic minority background in the group of non-disclosing victims. Regarding frequency and duration of victimisation, disclosing victims were more frequently victimised than non-disclosing victims. Yet, they did not differ in terms of the duration of victimisation. Some differences in social-emotional functioning exist. Disclosing victims showed more depressive symptoms and worse emotion regulation than non-disclosing victims. Lastly, disclosing victims perceived more conflict and less cohesion in their classroom compared to non-disclosing victims.

Table 1. Comparisons of non-disclosing victims and disclosing victims.

	Non-disclosing victims	Disclosing victims
n	218 (29.4%)	523 (70.6%)
Age (in years, M with SD)	9.31 (1.10)	9.60 (1.15)	t(739) = −3.09, p = .002
Sex (% boys)	59.6	52.6	χ^2(1,741) = 3.08, p = .079
Ethnicity (%)^a
Native Dutch	65.1	74.8
Minority background	34.9	25.2	χ^2(1,741) = 7.08, p = .008
Victimisation
Frequency	4.20 (0.89)	4.34 (0.85)	t(739) = −2.04, p = .042
Duration	2.55 (0.50)	2.61 (0.49)	t(408.22)^b = −1.54, p = .125
Social-emotional functioning
Social anxiety	2.13 (0.81)	2.10 (0.81)	t(739) = 0.49, p = .626
Depressive symptoms	1.88 (0.57)	1.98 (0.56)	t(738) = −2.16, p = .031
Self-perceived social acceptance	3.11 (0.91)	3.16 (0.89)	t(739) = −0.62, p = .535
Global self-worth	3.82 (1.03)	3.78 (0.94)	t(739) = 0.56, p = .579
Impulse control	3.40 (1.01)	3.28 (1.06)	t(736) = 1.39, p = .166
Emotion regulation	3.52 (0.92)	3.34 (0.94)	t(736) = 2.38, p = .018
Perceived classroom context
Comfort	3.49 (1.12)	3.42 (1.07)	t(736) = 0.76, p = .446
Cohesion	3.01 (0.96)	2.80 (0.90)	t(736) = 2.85, p = .005
Conflict	2.84 (0.93)	3.01 (0.90)	t(736) = −2.28, p = .023
Cooperation	3.47 (0.92)	3.39 (0.87)	t(736) = 1.14, p = .256
Isolation	2.67 (0.93)	2.78 (0.90)	t(736) = −1.57, p = .117

Note. ^aCategorisation following Statistics Netherlands (2016).

^bVariances were not equal across the two groups. The calculation is based on un-pooled variances and corrected df.

Likelihood to disclose

As can be seen in Table 2, older children (OR = 1.282, p = .003) were more likely to disclose their victimisation experiences than younger children. Boys (OR = .636, p = .014) and children with a minority background (OR = .664, p = .030) were less likely to report victimisation than girls and native Dutch children. The likelihood to disclose victimisation was higher when children experienced more depressive symptoms (OR = 1.617, p = .026). Yet, children were less likely to disclose when they were more anxious (OR = .772 p = .041), had better emotion regulation strategies (OR = .736, p = .045), or when they perceived their classroom as cohesive (OR = .750, p = .034). Disclosure of victimisation was not associated with the frequency (OR = 1.178, p = .108) or duration (OR = 1.116, p = .546) of victimisation. There was also no association with self-perceived social acceptance (OR = 1.262, p = .053), self-worth (OR = 1.118, p = .350), and impulse control (OR = 1.127, p = .350). In addition, no associations with feelings of comfort in class (OR = 1.153, p = .251), and perceived conflict (OR = 1.106, p = .448), cooperation (OR = 1.009, p = .951), or isolation (OR = 1.014, p = .920) in the classroom were found.

Table 2. Disclosure among chronic victims as a function of demographic characteristics, frequency and duration of victimisation, social-emotional functioning, and perceived classroom context.^a

				95% CI
	B	SE	OR	Lower	Upper
Intercept	−1.540	.822	.214	.043	1.076
Age	.249**	.083	1.282	1.089	1.510
Sex (0 = girls)	− .452*	.184	.636	.443	.913
Ethnicity (0 = native Dutch)	− .409*	.189	.664	.459	.962
Victimisation
Frequency	.164	.102	1.178	.964	1.439
Duration	.110	.182	1.116	.781	1.595
Social-emotional functioning
Social anxiety	−.259*	.127	.772	.602	.990
Depressive symptoms	.481*	.215	1.617	1.059	2.468
Self-perceived social acceptance	.232	.120	1.262	.997	1.596
Global self-worth	.111	.119	1.118	.885	1.413
Impulse control	.119	.128	1.127	.877	1.448
Emotion regulation	−.306*	.152	.736	.546	.993
Perceived classroom context
Comfort	.142	.124	1.153	.904	1.469
Cohesion	−.287*	.136	.750	.575	.979
Conflict	.101	.133	1.106	.852	1.437
Cooperation	.009	.145	1.009	.759	1.341
Isolation	.014	.135	1.014	.778	1.320
Random effects (var)
Class	.197 (.171)
−2 loglikelihood	3357.476
N	738

Note. *p < .05, **p < .01.

^aResults of block 1–3 are presented in Supplementary Appendix 2.

Additional analyses

To provide better insight into whether the perception of the classroom climate interacts with victims’ personal characteristics in the decision to disclose victimisation, additional analyses were conducted with interactions between victims’ demographic characteristics and social-emotional functioning on the one hand, and the perception of the classroom context on the other hand (see Supplementary Appendix 3). No statistically significant effects were found (range p = .162–.806). These results thus do not indicate that the effects of victims' personal characteristics on disclosing victimisation vary across their perception of the classroom climate, and vice-versa.

Discussion

The aim of the current study was to shed light on the prevalence of chronic peer victimisation among Dutch elementary school children and the factors associated with (non-) disclosure of such experiences by victims, using a recent large-scale nationwide study on victimisation rates. It was found that around 10% of the children were victimised severely—at least monthly—and over a long time. Strikingly, one-third of the children who are victimised chronically did not talk to anyone about their experiences. The hesitance to talk about victimisation is worrisome, given that knowing of the problem is a prerequisite for intervention and support. For this reason, we investigated which characteristics were associated with the likelihood to disclose victimisation.

Findings reveal that demographic characteristics were associated with the likelihood of disclosure: girls, native Dutch, and older children were more likely to tell their friends, family, or teachers about their experiences. With regard to the age difference, it should be noted that the difference between disclosing and non-disclosing victims in our sample is small: both were on average nine years old with about three months difference. The difference may thus have been statistically significant—probably due to the large sample size—but hardly meaningful. Compliance to social and cultural norms may play a role in the child’s decision to disclose victimisation. It has been argued that for girls seeking help is more socially accepted, whereas boys are more often expected to solve problems on their own (Connolly, 2018). Another explanation why the likelihood to disclose was higher among girls may be that victimisation among boys is often more overt and visible and as such more easily recognised by teachers and peers (Haataja et al., 2016). It thus might be that male victims are identified before they disclose their experiences. Although this explanation is in line with the study by Haataja et al. (2016), it is inconsistent with a study by Shaw et al. (2019) who found that physically victimised students are more likely to talk. Based on that study, one would therefore argue that boys would have been more likely to disclose their victimisation. Given these inconsistent and inconclusive results, future studies on disclosure of victimisation should therefore examine gender differences and different types of victimisation in more detail.

The finding that native Dutch children are more likely to disclose victimisation may be explained by cultural differences. In some cultures, sharing feelings, thoughts, and emotions with adults may interfere with cultural communication styles (Verhulp et al., 2013) and more hierarchical structures in parenting styles (Halgunseth et al., 2006; McAdoo, 2002), which in turn may make children from these ethnic minority group less likely to talk to adults in their school (Marsh & Cornell, 2001). However, it has also been suggested that it is not the ethnic status in itself that determines the likelihood to keep problems to yourself, but it is the interaction with the socioeconomic status that determines disclosure. In the Netherlands, minority groups generally have a lower socioeconomic position than native Dutch (Statistics Netherlands, 2016, 2020). Restrictive socialisation patterns that are associated with lower socioeconomic status environments may promote successful adaptions to these environments (Bradley et al., 2001) but may also result in relational styles that do not reward self-disclosure and related reliance on other people (Consedine et al., 2007). Future studies could take socioeconomic status into account.

Characteristics of the victimisation (i.e. frequency and duration) were not related to the likelihood to disclose. Although it has been suggested that people are more willing to talk about more severe problems (Bauman et al., 2016), previous evidence regarding bullying experiences was mixed (e.g. Hunter et al., 2004; Petrosino & Guckenburg, 2010). Possibly, the frequency and duration of victimisation do not necessarily indicate the perceived severity of the experiences, but instead, the type of victimisation is important (Chen, 2015). This is in line with the finding that students exposed to physical bullying were more likely to talk to an adult in school than students exposed to other types of bullying (Shaw et al., 2019). Another possible explanation is that variations in the frequency and duration of victimisation were relatively low in our study because we only involved children who are victimised chronically.

Social-emotional functioning appears to play a role in the disclosure of victimisation. Lower social anxiety and less adaptive emotion regulation skills were associated with the likelihood that children mentioned their victimisation to at least one person. This is in line with the theoretical assumption that children are more inclined to talk about their problems when they are emotionally distressed, but less likely to do so when they blame themselves, fear negative reactions from others or can cope with negative emotions (Boulton et al., 2017; Newman, 2003, 2008). The finding that higher levels of depressive symptoms also prompts the probability to disclose is somewhat unexpected given that children suffering from depression tend to blame themselves for interpersonal difficulties (Krackow & Rudolph, 2008) and to avoid their emotions and socially withdraw (Kahn & Garrison, 2009), which would make them less likely to share their problems with others. It might be that children who suffer from depression and are victimised chronically are particularly emotionally distressed, and feel like they have nothing to lose, that the victimisation is too severe to blame themselves for it in any way, and that things can only get better when they share their negative feelings and experiences. Another explanation might be that parents, teachers, or peers are more alert to what is happening when they notice that somebody is feeling unhappy. They might start conversations more regularly, or ask about experiences and emotions to monitor these students’ well-being. Possibly, this monitoring makes the threshold to disclose victimisation lower. Suffering from depression and being victimised often are co-occurring problems. Unfortunately, in this study we could not disentangle causes and consequences due to the cross-sectional nature of our data, nor could we test the proposed mechanisms.

Finally, our results emphasise the importance of classroom context. When children who experienced victimisation perceived their classroom as more cohesive, they were less likely to disclose their victimisation to others. This is at first sight a surprising result. We expected victims who perceive their classroom climate as more positive, safe, and supportive to be more likely to confide in their classmates and teachers to stop the bullying (Blomqvist et al., 2020; Unnever & Cornell, 2004).

On further thought, this finding might make sense in several ways. First, in interpreting this result we should keep in mind that the disclosure we measured is not only directed at classmates and teachers but also at important others. It might be that disclosure to teachers and classmates is positively associated with perceived classroom cohesion. Second, it is possible that victimisation is less frequent or intense in cohesive classrooms in ways that our measures did not capture, and hence there might be fewer reasons to report in such classrooms. It is also possible that victims who see their classroom as cohesive feel an even larger distance between ‘them’ and ‘me’ and are more lonely and isolated. That is, seeing the classroom as cohesive (most children like each other and play together during breaks) and still being (the only one) victimised, may make children who are victimised feel more alone in noticing the bullying, and even less hopeful that anything can be done stop it (Garandeau & Salmivalli, 2019). Whether the reason is more objective (a function of the actual nature of the bullying) or more subjective (a function of the victim’s experienced reality) is a question for further research that we were not able to determine at this point. It is clear that studying children’s reporting of their victimisation is a critical issue for further research.

Strengths, limitations, and directions for future research

The current study makes a substantial contribution to previous studies on the disclosure of victimisation. It provides recent information on the frequency of chronic victimisation and disclosure rates among elementary school children and gives important insights into the factors that are associated with talking about victimisation experiences. It is important to understand with whom children share their victimisation experiences and which children are more inclined to do so. With this understanding, school bullying can be better addressed as knowing who is victimised is necessary to provide adequate help.

Aside from these strengths, some limitations should be taken into account. We were unable to draw causal conclusions due to the cross-sectional nature of our data. It may very well be that disclosing victimisation leads to more depressive symptoms, less feelings of anxiety, or better emotion regulation. Yet, using only the pre-test data enabled us to give a large-scale exploration of disclosing victimisation among elementary school children without the results being possibly affected by the interventions.

The findings of this study were based on self-reports. Although focussing on actual disclosing behaviour is preferable above the use of vignettes, the use of self-reports alone has some drawbacks. Victims who do not disclose their victimisation do not necessarily remain unidentified by parents, teachers, or peers. Likewise, we do not know whether children who reported being victimised were also seen as victims by their peers and teachers. When victimisation is not acknowledged by their environment, children might be less likely to disclose it. Studies with a multi-informant approach would validate self-report data and further contribute to our understanding of parents’, peers’, and teachers’ awareness of a victim’s situation without being explicitly told.

We were unable to draw conclusions about the reasons children might not tell adults about their victimisation. Previous studies have suggested that such reasons may include denial and embarrassment, a sense of autonomy and self-reliance (deLara, 2012), fear of worsening the situation (Newman & Murray, 2005), and distrust in adults’ ability to intervene effectively (Bjereld, 2018). However, these studies were mainly based on qualitative data from adolescents or vignettes. Future studies should examine these reasons in elementary school children as well.

In addition, contextual factors at the classroom and school levels were not taken into account. For instance, strong anti-bullying norms within the classroom and at schools are known to prevent and diminish bullying and may thus also promote disclosure of victimisation (Saarento et al., 2015). Relatedly, studies have found that children are more likely to report victimisation to teachers and experience fewer behavioural problems if they believe that the teacher will actively intervene (Cortes & Kochenderfer-Ladd, 2014; Troop-Gordon et al., 2021). However, a recent study looking at teacher’s actual responses instead of children’s expectations about teacher’s handling, found that students’ disclosure could not be predicted by teacher’s active responses towards bullying nor by more passive responses (ten Bokkel et al., 2021). This shows that teacher’s (intended) actions against bullying do not necessarily transfer to similar perceptions among the children. Future studies should therefore take these contextual factors into account to better understand and promote disclosure of victimisation.

Our study also does not provide insight into the potentially beneficial or harmful effects of disclosing victimisation. On the one hand, talking about being victimised might be an effective strategy to receive emotional support and tackle the bullying situation (Mishna & Alaggia, 2005). On the other hand, disclosure can enhance victims’ social-emotional maladjustment, for example, because they are afraid to be seen as different or weak (deLara, 2012), or because of reliving the negative experiences after talking about it. Future studies would move the field ahead by investigating the consequences of disclosure in terms of actions undertaken and victims’ well-being, especially because adult responses were not always perceived as helpful by victims (Bjereld, 2018). Moreover, future studies should investigate how victims who do not disclose victimisation could be supported.

Practical implications for intervention

Despite the fact that many anti-bullying interventions encourage children to tell others about what is happening to them, our findings show that one-third of children who are victimised chronically do not disclose their experiences. It also appears teachers are not the first person children talk to when victimised (Blomqvist et al., 2020; Boulton et al., 2017). Teachers and peer bystanders play a crucial role in addressing bullying in school (Salmivalli, 2014; Veenstra et al., 2014). Because teacher interventions and peer support can make a difference for the victim it is important that interventions include appropriate efforts to encourage disclosing victimisation on the one hand and to increase awareness of the social dynamics in the classroom on the other hand. This is particularly important as studies have shown that children who experienced victimisation are more inclined to seek help when they trust their teachers (Bjereld, 2018) when their teacher takes a clear stance against bullying (Cortes & Kochenderfer-Ladd, 2014), and peers are perceived as supportive (Blomqvist et al., 2020). Creating a safe classroom climate in which bullying is clearly disapproved thus is an important step to promote talking about being victimised.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This project was supported by a grant from The Netherlands Initiative for Education Research (NRO; 405-15-781).