Friends and safeguarding: Young people ’ s views about safety and to whom they would share safety concerns

Child sexual abuse prevention strategies typically focus on teaching children ‘ protective behaviours ’ , including telling a ‘ trusted adult ’ . However, disclosure rates are low, and we know little about who they tell. We analysed data from over 3400 young people aged 10 – 18. After viewing hypothetical unsafe scenarios involving either an adult or peer, participants were asked whether – and whom – they would tell someone if such a situation occurred. Most (83.9 per cent) said they would tell someone about concerns involving an adult; fewer (79.3 per cent) would tell if they encountered an unsafe situation with a peer. Across adult and peer scenarios, participants were most likely to say they would approach their mother (about concerning behaviour of an adult, 68.7 per cent; or a peer, 63.1 per cent), a friend (64.4; 57.9 per cent) or their father (52.2; 48.9 per cent). Those most likely to tell a friend were girls and older children. Children in out-of-home care and community welfare organisations were less likely to tell someone about concerning behaviour from a peer/friend than in other organisational contexts. Although organisations must train staff in supporting young people who raise concerns or make disclosures, it is vital to consider the role of parents and other young people in hearing about concerns and building their capacity to respond appropriately.

• If they were to feel unsafe, young people are more likely to a friend, than they are to tell adults in organisations that serve them.
• Train and upskill young people and their parents/carers to know how to respond when children/young people say they feel unsafe or actually disclose abuse.
• Ensure children and young people know that concerns about safety from other children and young people are just as important as talking about when they feel unsafe with adults.

INTRODUCTION Context
An important aspect of child sexual abuse (CSA) prevention and response is ensuring children and young people are comfortable discussing when they feel unsafe and disclosing abuse.CSA can be broadly described as occurring within four contexts: intrafamilial, extrafamilial, organisational and onlineand includes not only adult-perpetrated sexual abuse, but harmful sexual behaviour from other children and young people (Quadara et al., 2015).Across all contexts, the rate and timeliness of disclosures of CSA is poor (Collin-Vézina et al., 2015;McElvaney, 2015).Yet raising safety concerns prior to abuse occurring or making such disclosuresand doing so as soon as possible after the events or concerns occurcan ameliorate negative outcomes and lead to improved psychosocial outcomes in individuals who have been abused (Alaggia et al., 2019;Hébert et al., 2009).Organisational (or institutional) child sexual abuseoccurring within or connected to a youth-serving organisation, such as schools, churches, sports, arts, welfare and youth development (Commonwealth of Australia, 2017) has recently become a major area of research, policy and practice focus across the globe.Spurred on by government inquiries seeking to understand historical abuse and prevent abuse occurring in future, many youth-serving organisations are now working to improve policies and procedures with the aim of keeping children safe.
A key aspect of these changes focuses on ensuring staff (and volunteers) who work in these organisations are knowledgeable, skilled, and confident in preventing abuse and responding appropriately when it occurs.Many organisations are devoting considerable time and money to training staff to ensure this happens.But less attention is paid to whether the right people are being supported in hearing young people's safety concerns and receiving disclosures of abuse.It is vital that we consider the broader social ecology in which children and young people exist, and to whom they are likely to raise concerns or make disclosures, when researching and implementing strategies related to supporting disclosure of CSA as part of a prevention strategy.

Raising Concerns and Delays in Disclosing Child Sexual Abuse
Rates of disclosure of CSA, whether it occurs in organisational or familial settings, are low (Collin-Vézina et al., 2015;McElvaney, 2015).Children often delay abuse disclosure, but in a review of retrospective and forensic interview-based disclosure literature, recantation and denial were found to be less common than often discussed (London et al., 2008).Research findings have helped us understand that CSA disclosure is an iterative process (not a discrete event), the social ecology surrounding a victim-survivorincluding cultural, contextual and familial factorsplays an important role in disclosure, and age, gender and life events affect disclosure timing and rates (Alaggia et al., 2019).Disclosure rates and timing can also be affected by previous negative experiences children, young people and adults have previously had when disclosing abuse or safety concerns (Alaggia et al., 2019).
In a representative sample of Canadian adults, Hébert et al. (2009) found that one in five adult survivors of CSA had never disclosed their abuse, 21.2 per cent of adults abused as children disclosed within a month and 57.5 per cent did not disclose abuse for at least five years from the onset of abuse.In Australia, the Royal Commission into Institutional Responses to Child Sexual Abuse found, from data collected during individual private sessions with survivors, that there was an average delay of 24 years between abuse beginning and disclosure to anyone else occurring (Commonwealth of Australia, 2017).
The likelihood of CSA disclosure by children and young people in formal investigations, however, is reported to be significantly higher than disclosure among the general population.A recent meta-analysis of CSA disclosure in forensic settingsconsidered an important opportunity in the absence of corroborating evidence in relation to legal proceedings against abusersshowed that 64.1 per cent of children (under the age of 18) who have experienced CSA disclose their abuse (Azzopardi et al., 2019).These figures across community and forensic settings point to the importance of understanding the purpose of disclosure (for victim-survivors but also practitioners, organisational leaders, and society as a whole), how and what aspects of CSA are disclosed, the barriers and facilitators of disclosure, and to whom such disclosure are made.
Understanding the barriers and facilitators of CSA disclosure has been an important part of improving disclosure rates and timing.Disclosure however is often obstructed by interpersonal, developmental, contextual and socioemotional barriers (Azzopardi et al., 2019).These barriers include age, gender, disability status, lack of understanding and support from adults, shame and anticipated negative social reactions, and the context and type of abuse.However, researchers have been able to identify facilitators of disclosure.These facilitators include the provision of developmentally appropriate information regarding CSA and being asked or prompted to disclose abuse (Lemaigre et al., 2017;Winters et al., 2020).Similarly, in forensic settings, the most frequently identified facilitators and barriers are the gender and age of victim-survivors, ethnicity, victim-perpetrator relationship, abuse severity, prior disclosure, and interview method (Azzopardi et al., 2019).
Given the well-documented delays and difficulties for children in speaking up, being believed, and having adults respond in appropriate ways to ensure their safety and respond to the risks, we are interested in exploring whether children and young people are likely to speak up about their safety concerns before harm has necessarily occurred.Raising safety concerns, at the earliest possible opportunity, has the greatest potential for protective action to be taken to ensure a young person's safety before child sexual abuse or harmful sexual behaviour from peers occurs.

The Importance of Speaking up
The importance of speaking up about concerns, and making disclosures if abuse has occurred, is central to overcoming the secrecy and silence that surrounds CSA.Adult perpetrators, as part of the grooming process, will often tell victims that they should never tell, and that if they do, they will not be believed, or this will cause reputational harm to the perpetrator, their family, or the organisation (Plummer, 2018).Early disclosure can provide protective actions, to stop the harm from continuing (by interrupting grooming behaviour before it escalates to acts of touching or penetration), to provide therapeutic support for the victim, and to interrupt or reduce the likelihood of future perpetration against other victims (Alaggia et al., 2019;Quadara et al., 2015).
Understanding who children are likely to speak to about safety concerns or experiences of sexual abuse is an important piece of the puzzle.Understanding to whom such concerns or disclosures are made can help develop interventions that target the right people, support organisational staff and leaders to adapt policies and processes accordingly, and ensure children and young people are offered ample opportunity to raise concerns or to disclose harm.A recent systematic review of studies identified that there are developmental patterns in CSA disclosure whereby older adolescents are more likely to disclose to a friend rather than an adult.Findings also suggested that girls were more likely to disclose abuse to friends, with boys less likely to disclose CSA altogether (Manay & Collin-Vézina, 2021).These findings are important when considering safeguarding strategies but whether such patterns are the same for discussing safety concerns before abuse has occurred has not yet been studied.
Historically, much of the work on CSA prevention focused on 'protective behaviour' programmesteaching children about their bodies, trusted adults, saying 'no', and speaking up when they feel uncomfortable.These programmes are, in most cases, implemented in school settings, due to the universal reach of schools and the push towards the use of a public health approach to prevent CSA (Fix et al., 2021;McKibbin & Humphreys, 2020).There is emerging evidence about how CSA prevention education works, which aspects increase children and young people's knowledge and skills related to keeping themselves safe (Trew et al., 2021;Walsh et al., 2015;Wurtele, 2009), and the scope of prevention education uptake across different countries (Russell et al., 2020).However, criticism regarding perceived over-reliance on prevention education programmes being directed solely at children and young people has supported a push towards research investigating the role of parents as protectors (Raising Children Network et al., 2020;Rudolph & Zimmer-Gembeck, 2018).The development of safeguarding strategies that do not rely on children to protect themselves are typically based on Situational Crime Prevention (Kaufman et al., 2019) and Contextual Safeguarding (Firmin, 2020).They emphasise the role of environmental factorseither in terms of physical environmental risks, or organisational or community-level environmental conditions.However, strategies based on these frameworks have not as yet focused on or examined the role of friends and peers as recipients of information about safety concerns, and their consequent role in disclosure or ensuring protective action.

The Role of Friends in Safeguarding
Researchers have paid scant attention to the role of friends in safeguarding.Using a survey design, researchers in Zimbabwe asked secondary school students how they thought they could be best protected from sexual abuse (Gwirayi, 2013).They provided a range of suggestions directly from the participants, but the researchers were also able to identify how girls and older children considered having reliable friends as an important factor in fighting CSA.Qualitative research with young people and adult survivors of CSA showed some individuals have a desire to protect friends (Pearce, 2006) or first disclosing to a friend (McElvaney et al., 2012).After abuse has occurred, friends can be seen as a source of emotional, moral and practical support; as well as intervening to stop abuse, friends act as a 'go-between' helping disclose abuse to adults (Allnock, 2015).However, the extent to which these attitudes and behaviours are more broadly held and used by children and young people is not well known.In focus groups conducted for the Australian Royal Commission into Institutional Responses to Child Sexual Abuse, Moore et al. (2015) reported that young people were clear that peers should not be shouldered with the responsibility of assessing or responding to safety concerns, but rather helping their friend to seek support from a trusted adult.
When asked whether they disclosed sexual experiences that matched the legal definition of CSA, findings from a representative sample from the Finnish population found that disclosures are often made to friends, with adults (including organisational staff and parents), and authorities less likely to be the recipients of disclosure (Lahtinen et al., 2018).A recent systematic review of barriers and facilitators to disclosing abuse suggests that younger children are more likely to disclose to adults while older children are more likely to disclose to a peer (Lemaigre et al., 2017).Despite this knowledge, in the context of organisational CSA, many interventions aimed at increasing disclosures often only focus on training adult workers (and volunteers) within organisations such as teachers (Wager, 2015).
There is little research to understand whether the likelihood of a young person telling a friend or peer about their safety concerns is likely to differ, depending on whether the concerning behaviour is from an adult, or from another child or young person.Although it may be the case that peer abuse is viewed as less serious or problematic by young people, there is evidence that even if so, this may not be only the only reason underpinning low disclosure rates about potentially harmful behaviour from peers.For example, Allnock and Atkinson (2019) demonstrated that there are peer norms creating 'powerful roles' that may prevent disclosure and reporting sexual harm from peers.They argued that

Aim of the Current Study
The aim of the current study was to explore children and young people's intention regarding telling someone about concerning behaviour in organisational (institutional) settings related to CSA.Specifically, we were interested in the role that peers or friends play in young people's intention to raise safety concerns.We sought to answer the following questions: 1. How likely are young people to raise safety concerns in organisational contexts with which they interact?2. Who are the most likely recipients of such safety concerns?3. Does the likelihood of discussing safety concerns to a friend differ when the concerning behaviour is from an adult versus from a peer? 4. Are there age, gender, or organisational differences in the likelihood to raise safety concerns with a friend?

METHOD Design
Data for this study come from the Children and Young People's Safety project, a multi-cohort cross-sectional study of children and young people's perceptions of safety using the Children's Safety Survey (Moore et al., 2016), and of the safeguarding capabilities of the staff and volunteers who work in youth-serving organisations using the Safeguarding Capabilities in Preventing Child Sexual Abuse Survey (Russell & Higgins, 2020).Ethics approval for use of the Children's Safety Survey in the project was given by the Australian Catholic University Human Research Ethics Committee (2018-5H).Participants provided consent to participate in the survey after reading a child-friendly participant information letter.Parents of children aged under 16 also provided parental consent.Participants were unable to disclose abuse within the survey itself as all items in the online survey were fixed response options with no free-text options.However, to support safeguarding, if participants wished to discuss any of the issues that the survey raised, all organisations provided a key contact person within their organisation who is known to participants and whose name is shared at the end of the survey along with a local youth-focused helpline.
Individual youth-serving organisations (or their state/national affiliate or peak body) self-selected to take part in the project.We used an online real-time dashboard to report back anonymous results from their own organisation to leaders to support impactful change to safeguarding policies and procedures.

Participants
We present data from 3417 children and young people aged 10-18 who completed the Children's Safety Survey between 1 January 2019 and 15 September 2021.The study design was cross-sectional, with young people invited to respond only once, but with fieldwork spread out over the 21 months across the different organisations distributing the survey invitation.

Materials
The Children's Safety Survey is a comprehensive survey that draws on previously developed scales, including the Australian Safe Kids and Young People (ASK-YP).The ASK-YP items were developed for the Australian Royal Commission into Institutional Responses to Child Sexual Abuse.The items were created in consultation with children and young people and piloted with over 1400 children and young people (Moore et al., 2016).The Children's Safety Survey is completed online, and items are 'personalised' to the name of the organisation or service to which the child or young person is connected.For details regarding the Children's Safety Survey see https://safeguardingchildren.acu.edu.au/measuring-safety/childrens-safetysurvey.For the purposes of this study, only items relating to participants' sense of safety (see below) were analysed.

Scenarios
To understand their sense of safety, participants were shown two animated 'scenarios' (with voice-over, and subtitles to read the narration) about a concerning interaction, and then asked questions that measure whether the children or young people believe the scenario is likely to occur in the organisation (and, if so, whether it would worry or concern them) and would tell someone if the scenario occurred (and, if so, who they would tell).
In one of the hypothetical scenarios, an adult male behaves in a problematic way, and in the other, a young person (male) behaves problematically.These two scenarios depict experiences that might be concerning for children and young people and 'raise flags'.The scenarios randomly depict a boy or girl experiencing the problematic situation, unless a single-sex organisation requests all 'targets' of the concerning behaviour are the same gender (e.g. an all-girls school).The scenarios depict situations that are 'grey areas' amenable to prevention (i.e.identifying issues early and seeking help), rather than responding after harm has already occurred (to download the scenarios go to the Children's Safety Survey page of the Safeguarding Portal).

Sense of Safety Items
The items about sense of safety are designed to measure whether the children or young people believe the scenarios presented to them are likely to occur in the organisation (and, if so, whether it would worry or concern them).Participants were asked if such a hypothetical scenario were to occur in their organisation, whether they would tell someone if they were in a situation like this (Yes/No).Participants were also asked if they were in a situation like this who would they tell (options were: Mum, Dad, a friend, a sibling, workers in the organisation, a teacher at school, a telephone helpline or someone else).Participants could choose multiple options.

Procedure
Participants were invited to take part in the survey through communications from the organisation that chose to collect data regarding children and young people's perceptions of safety.Organisations used their existing communications processes to send an online survey link containing information sheets, e-consent forms and survey items to children and young people and their parents/guardians.Organisations were responsible for deciding and informing participants whether they would fill the survey out while in attendance at the organisation (i.e. in class at school) or at home.The survey takes between 10 and 20 minutes to complete.

Statistical Analysis
We analysed means and proportions across the demographic data and outcome variables.We used McNemar's test to investigate whether any differences between outcomes between the adult and peer scenarios were statistically significant to answer the first three research questions.To answer the fourth research question, we investigated the percentage of children and young people who said they would tell a friend about the situations presented to them in the scenarios, looking at the different organisational contexts, age brackets, Indigenous status, and gender.We calculated odds ratios using 10-12-year-olds, males, non-Indigenous background, and school as the reference categories in a multiple logistic regression model.We used IBM SPSS Statistics version 27 to analyse the data.

RESULTS
The demographic information about the 3417 children and young people is summarised in Table 1.The mean age of the sample was 13.8 years, with boys accounting for just over half of respondents (51.4 per cent).Participants responded in relation to one of a variety of organisational contexts when completing the survey, however most participants were responding about schools (89.6 per cent).

Would Children and Young People Raise Safety Concerns and to Whom?
More participants said they would tell someone in a concerning adult situation (83.9 per cent) than in a scenario involving a peer (79.3 per cent).Using McNemar's test, the difference was statistically significant ( p < 0.001).When asked to identify as many or as few of the people they would tell if they were in a situation like this, participants chose a range of potential persons they would tell.We also used McNemar's test to identify significance differences (p < 0.05) between the proportion of participants' saying they would tell each target person (e.g., Mum, Friend, etc.) in an adult versus a peer situation (see Table 2).
It was clear for scenarios depicting concerns about either adults or peers, that the top three people they were most likely to tellin orderwere their mother, a friend, and then their father.To each possible recipient, a young person is more likely to tell them about a safety concern from an adult than from a peer, and the differences for most possible recipients are statistically significant.The percentage of respondents was higher for discussing concerns about adult behaviour compared with concerning behaviour from another child/young person.This difference was particularly large in relation to whether they would tell a friend (64.4 per cent would tell about concerns about adults, versus 57.9 per cent for concerns about peers).

Differences between Organisational Contexts, Gender, and Age
We found that older children and young people were more likely to disclose safety concerns to a friend.Girls were more likely than boys to disclose concerns to a friend.This was consistent across both the adult and peer scenarios.It appears that gender diverse participants fell in between.However, there was insufficient sample size to be confident that the differences for gender diverse participants are significant.More research is needed to see whether gender diverse young people may be more likely than boys to disclose concerns to a friend.
T A B L E 3 Likelihood of telling a friend about a safety concern, by scenario type (adult and peer situation), demographic characteristics, and organisational context.

FRIENDS AND SAFEGUARDING: TO WHOM THEY SHARE SAFETY CONCERNS
There was a trend towards First Nations (Aboriginal and Torres Strait Islander) children and young people being less likely, in both situations, than non-Indigenous young people to say they would disclose feeling unsafe to a friend.However, both differences were not statistically significant and so further research with a larger sample size is needed to determine whether this trend is real.
When compared with the likelihood of telling a friend about a concern in a school setting, participants responding in relation to all other sectors were less likely to tell a friend.There was overlap in the confidence intervals from most sectors, for both scenario types, however, when considering concerning behaviour of an adult, there was greater confidence in the data for participants from out-of-home care and the community welfare organisations saying they would be less likely to tell a friend (Table 3).

DISCUSSION
We explored self-reported likelihood of children and young people to raise safety concerns, and the role that peers or friends play.We found that, typically, young people said if they were to experience CSA-related safety concerns in their organisational context, they were likely to raise them.More participants said they would tell someone about a concerning adult situation (83.9 per cent) than in a scenario involving a peer displaying potential harmful sexual behaviour (79.3 per cent).Therefore, an important implication is to raise awareness, not just in parents (both mothers and fathers) and adults (such as teachers and other professionals in youth-serving organisations), but also in children and young people themselves, that concerning or abusive behaviour from peers is not OK, and that speaking up about such matters is just as important as speaking up about safety concerns relating to behaviour of adults.
As noted by Allnock and Atkinson (2019), the normalised nature of sexually harmful behaviour between peers can create norms that may prevent disclosure and reporting sexual harm from peers.Therefore, they argued that simply raising awareness may not be sufficient.A wider barrier to the take-up of safeguarding strategies in organisations that support young people to tell othersadult workers, parents or peers may not be effective unless these powerful peer norms are also understood, and the interventions or prevention strategies contextualised to address these.This places the need for attention from organisations not just on 'people', but on contextual safeguardinglooking at the structures, activities (including access to digital/online communication devices), and supervisory practices that support or disrupt factors that may either enhance or detract from a culture of safeguarding.
We found that the most likely recipients of safety concerns are mothers, followed by friends.Our findings align with those of others which found friends and mothers as the most likely recipients of safety concerns or disclosures of abuse (Manay & Collin-Vézina, 2021;McElvaney et al., 2012).A higher proportion of young people we surveyed indicated that they would tell a friend about a safety concern if their concern was about the behaviour of an adult rather than of a peer, which supports Allnock and Atkinson's (2019) findings that powerful peer norms about sexualised behaviours may get in the way of young people's ability to disclose concerns when they do not relate to the behaviour of an adult.Although this finding has implications, it is important to note that the proportion of young people who said they would share their concerns with any individual (i.e.mother, father, worker, helpline, etc.) was consistently higher when asked in relation to the concerning behaviour of an adult compared to concerning behaviour of a young person.
Our exploration of demographic factors that might influence the intention to tell friends revealed firstly that males were less likely than females (or gender diverse participants) to disclose concerns or feelings of lack of safetyregardless of whether the concerning behaviour was from an adult or a peer.Consistent with findings from Manay and Collin-Vézina (2021), Gwirayi (2013), as well as Hong and Eamon (2012), older youths are more likely to indicate they would disclose concerns to a friendboth for behaviour of adults and peers.This difference was even more pronounced for concerns about peers, where they were much less likely to tell friendsor indeed anyonethan if their safety concerns were about the behaviour of an adult.
It is unsurprising that females are more likely than males to disclose concerns to a friend given gender differences in communication styles and identification of feelings (see Clark, 2011).Sexual abuse of boys is perceived to be a violation of gender norms for masculinity, creating an additional obstacle to disclosure for boys and men (Easton et al., 2014).Children and young people's perceptions of safety often relate to feeling safeas opposed to being safe (Moore et al., 2016).An important implication of our study is to better support boys to discuss their safety-related feelings with friends and other people in their social circle who can support them.Considering the greater likelihood for females (but also the trend towards gender-diverse young people being potentially more likely) to speak to friends when feeling unsafe, it is important that families and organisations help all children and young people to foster positive friendships that can underpin such conversations about concerns.Organisations also have the responsibility to discuss safety concerns, and support avenues for help-seeking for all children, tailoring supports as needed for children regardless of their gender identity.Such an approach fits in with a situational crime prevention approach to CSA prevention as adults are empowered to use the voice of children and young people to understand the context or situation in which such concerns are likely to arise and follow through with risk mitigation strategies at the organisational level to respond to these risks and reduce children and young people's vulnerabilities.
Older youths in our sample (16-18 years old) were more likely than the younger participants to say they would tell a friend about a safety concern.This may be the result of developing friendships that help build trust across time (as such trust would be developed over time between most children and their parents).Organisations could also consider how the same trusting relationships can be built by young people and staff/volunteers in sectors where there is greater risk of high staff turnover or placement instability for young people, such as in out-of-home care.It is ironic that in organisations providing support to some of the most vulnerable children in societysuch as those in out-ofhome care and community welfare organisationsyoung people were less likely to tell someone about concerning behaviour from a peer/friend than in other organisational contexts.Although organisations must train staff in supporting young people who raise concerns or make disclosures, it is vital to consider the role of parents and other young people in hearing about concerns and building their capacity to respond appropriately.
Telling friends about safety concerns was more likely to happen in an education (school) context than in any of the other sectors.This is perhaps not surprising given the stability of friendship groups across a year level and over time (as children often remain connected to their peer groups from one year to the next).Schools account for a significant amount of time in the day compared to all other contextsso it provides more opportunity for children and young people to interact with their peers, build close and trusting friendships and relationships that would facilitate disclosure of safety concerns.Compared with the likelihood of telling concerns that occurred at school, young people in out-of-home care or who were receiving services from a community welfare organisation were less likely to tell a friend, regardless of whether it was a concern about the behaviour of an adult or a peer.This lower reliance on friends for young people in these two contexts may be a result of a lack of trust in others.Young people in out-of-home care or other child-welfare contexts are more likely to have experienced traumatic events, removal from family, and possible disruption to peer relationships (e.g.though placement instability).It is possible that trust in others, in general, is low and that the time needed to develop trusting friendships has not been available to these young people.
In their research on the safety of young people in out-of-home care, Moore et al. ( 2020) identified strategies young people used, including 'protection' (actively stepping in when a peer is at risk), 'guidance' (providing advice to a peer who has safety concerns or has been harmed), and 'supporting help-seeking'.They found that 'most held ambivalent views about workers' capacity to demonstrate empathy or the skills to respond to issues such as violence or abuse.Many reported that they would turn to each other when they had been harmed' (p.130).
An important detail when considering the role of friends and peers in safeguarding, as with the notion of 'trusted adults', is accepting and understanding that friends and peers can be the cause of harm or abuse, as well as a source of support and target for sharing safety concerns or feeling unsafe.Peers as a source of harm is discussed in terms of young people displaying problematic or harmful sexual behaviours (NSPCC, 2019).Older friends can also be a risk factor for sexual victimisation, as they may be old enough to purchase alcohol or other risk-taking behaviours that increase a younger friend's risk of being a victim of abuse or grooming (Davies & Jones, 2013).However, this is potentially observable by friends, and friends may hear about the concerning behaviour from their peers.Developing interventions that promote friends and peers as individuals who can support each other to protect against peer-to-peer abuse has recently been considered (Brodie et al., 2020).We argue that such interventions could also go further to promote friends as support in instances of both adult-and youth-perpetrated CSA.We need further knowledge on how friends and peers can play a positive role in keeping children and young people safe from all forms of sexual abuse.Although many of our findings are supported with strong evidence, the study is limited by the smaller number of participants responding in relation to contexts other than schools, and the underrepresentation of gender-diverse and Aboriginal and Torres Strait Islander children and young people.Further research is needed to investigate the intentions of these groups to raise concerns and to understand the typical people with whom they raise safety concerns.

CONCLUSIONS
Our study is one of the first to use such a large sample to investigate likelihood of sharing safety concerns relating to adult and peer-based CSA in an organisational context.In addition to better understanding the intentions that children and young people have to speak to someone should they experience safety concerns, we have also been able to explore the role of friends and friendship in being told about safety concerns.Our analysis here focused on our research questions about the role of friends in hearing about safety concerns.Further research is needed to look at whether similar patterns emerge when looking at other demographic characteristics, and across different organisational contexts.
Policy makers, organisational leaders, and practitionersas well as parents, families and communitiescan consider the findings in our study when developing strategies and making changes to policy and practice to improve contextual safeguarding strategies and ensure child-safe cultures are strengthened.Our findings should not be considered an effort to shift both the responsibility as well as the focus of interventions solely to self-protection for young people (as has been a criticism of solely child-focused 'protective behaviour' prevention education programmes).Rather, as with a situational crime prevention approach to safeguarding, we suggest that it is important for everyone involved in child safety and safeguarding to consider how we can include young people and their families, in addition to staff and volunteers in organisations, in keeping children and young people safe from CSA.
Likely person they would tell, by scenario type (concerning behaviour from an adult, and concerning behaviour from a peer).
a This category is not included in organisational context comparisons due to the small sample size.Religious schools are included under Education.b The remaining participants (n = 3335) are considered not of Aboriginal or Torres Strait Islander background.T A B L E 2 Note: Participants could choose multiple options, so percentages add to more than 100%.p < 0.05 in bold.a Analysed using McNemar's test.