The importance of public and social network confidentiality (Hill, 2005 (in Greene & Hogan, 2005)) is highlighted when members of the research team are from the same or linked communities to the participants. There may be inherent risks attributable to established social relationships and existing power dynamics with interviewers from the same community, which are beyond the awareness of an ‘outsider’ researcher. These may not be readily resolved, even if the risks are recognised, as selecting interviewers from other communities may see aspects of the research relationship, including those related to confidentiality, influenced by historic rivalries or different social backgrounds.

Parents and others may be interested in the information that has been collected and consequently they may ask about the data collected or the content of interviews, which can put stress on the child and the researcher. However, respecting the privacy of the child requires the researcher to keep information confidential and not pass it on intentionally or inadvertently (for example, by making comments among research team members in open space) to family members, friends or others known to the child.

Finding a location that allows confidentiality can be further complicated by societal concerns about protecting children from abusive adults, which leads to an unease with research being conducted by lone adults in settings that are not public (Barker & Smith, 2001; Matthews, Limb & Taylor, 1998). It may therefore be advisable to use locations that are visible to others but not within their hearing. However, tensions can arise if there are risks associated with research participation for children which are increased by being visible. The considerations have to be carefully weighed, in the unique context of each research study, to ensure that the risks are minimised and benefits maximised by whatever course of action is taken regarding privacy.

An additional consideration is that some children may wish to have parents, siblings or friends present during the information gathering process. In such instances, it may be difficult to provide assurances of confidentiality and may also impact on how much information individual children wish to share. Respect for the child’s wishes and autonomy would ideally provide guidance, however, this could be challenging and inappropriate in certain cultural contexts in which this is not the accepted custom. The reality is that researchers may not always be able to accommodate children’s wishes if they are at odds with the parent’s wishes, especially if the researcher needs to negotiate their position as a guest in the child and parent’s home (Alderson & Morrow, 2011; Mayall, 2000; MacDonald & Greggans, 2008; Sime, 2008). The reality of the power dynamics in most situations is that when child and adult wishes conflict, children’s wishes are usually subordinate to those of the adults.

When parents remain present for interviews, a range of covert and overt techniques can be used to minimise parental involvement (Bushin, 2007). These may include, limiting eye contact with the parent, specifically using the child’s name when asking a question, raising subjects it is more likely only the child would know about, and reiterating the importance of ascertaining the child’s opinions. These suggestions are subject to an awareness of the local context and, if necessary, discussion with community elders, to ensure that any minimisation techniques are culturally sensitive.

The researcher may need to adapt their expectations of data collection in response to parents who are reluctant for researchers to interview children privately or insist on being present. Respect for children’s privacy regarding how much information they wish to share, which may be affected by who is present, should be privileged over the researcher’s wish to elicit more information. A degree of reflexivity and flexibility is required on the researcher’s part to ensure that they do not encroach on children’s privacy in their desire to attain quality data. Some of the challenges that may be encountered in respecting children’s privacy are highlighted in research with children with disability, in which traditionally parents or other adults have been expected to act as proxies representing children’s voices and interests. In her ERIC case study, Berni Kelly discusses challenges to children’s privacy in this context.

The challenge involves the prioritising of ethical principles. The principle of respect for children’s dignity and preservation of confidentiality underlies arguments against reporting issues such as suspected child abuse. Other arguments centre on the principle of beneficence (King & Churchill, 2000), on the basis that reporting will decrease the risk to the child, which can be perceived as a beneficial outcome (Knight et al., 2000).

There are numerous views about confidentiality and reporting concerns in research involving children. Some researchers recommend ensuring rather than guaranteeing confidentiality, and informing participants explicitly, prior to starting data collection, of the limits of confidentiality, the actions that will follow any safety concerns and which organizations will be involved (Duncan et al., 2009; Meade & Slesnick, 2002). A consequence of being explicit about the limits to confidentiality may be the loss of participants at the stage of recruitment, and attrition during the study, as well as impacting on the completeness and quality of the information obtained.

However, researchers’ ethical commitment to respecting children is not diminished by applying limits to confidentiality and being explicit in relation to these allows for respectful conversations with children about the relevant issues. Researchers, ideally, should be able to talk with children before acting on their concerns and, where possible, plan the safest and most effective ways to take action, thus applying good practice in supporting children’s participation which promotes the safety and protection of children (Feinstein & O’Kane, 2008; see also ERIC Case Study by Tim Moore). Ultimately though, it is the researcher’s responsibility to ensure that the safety of children is prioritised. The ERIC case study by Lorraine Radford discusses ethical dilemmas related to protecting children and addressing confidentiality, in a UK study about children’s experiences of violence, abuse and neglect. It describes a system that was put in place with alert, review and referral processes.

The age of the child and the familial, as well as societal, context may have some bearing on the approach taken by the researcher in regard to child protection. The researcher response needs to take the evolving capacities of the child into account in relation to the perceived concern, potential reporting action, inclusion of the child in deciding this and the possible outcomes. For example, it seems unlikely in some situations that older children or youth would agree to participate in research if they thought that such participation would result in them being referred to the authorities of whom they (and perhaps the researcher) are mistrustful. In addition, researchers need to be aware of both law and customary law on when adulthood is reached, recognising that there are differences between countries in relation to these. This is particularly pertinent with regard to reporting safety concerns regarding children. In peer research it is important that training is given, parameters of confidentiality are clear and that support mechanisms are in place for young researchers. The ERIC case study by Clare Lushey and Emily Munro looks at challenges that arose in peer research, with regard to differing degrees of concern and opinions about confidentiality in relation to a specific incident.

Researchers need to be aware of requirements with regard to mandated reporting. The decision to report concerns or knowledge regarding harm or potential harm to children may be a legal one as well as an ethical one (Fisher, 1994). Legal requirements with regard to reporting suspected abuse or child maltreatment differ across international contexts (Williamson, Goodenough, Kent & Ashcroft, 2005). In some areas legal or professional requirements may mandate reporting of suspected child abuse (for example, by health professionals, teachers and psychologists), even if researchers as a professional group are not specifically mandated to report. However, these requirements are not consistent internationally, or even nationally in some cases.

Some ethics review committees and institutional review boards require mandated reporting of suspected child abuse, and this mandatory approach is recommended by some researchers to provide clear guidelines, prioritise child protection and ensure uniform research practice (Allen, 2009; Steinberg, Pynoos, Goenjian, Sossanabadi & Sherr, 1999). The lack of consistency across international contexts, within countries and across ethical review boards underscores the importance of researchers considering the issue of reporting prior to starting data collection and creating a plan or protocol to follow if required. Making a contingency plan before the research starts, for providing support or referral to appropriate services for children who reveal themselves to be at risk of harm, provides some relief from the dilemma.

The needs of the research must not subsume ethical responsibilities toward children who may be at risk. However, some researchers intentionally use methods to prevent disclosure, avoid reporting child abuse and assure confidentiality (Socolar, Runyan & Amaya-Jackson, 1995), such as limiting response options, warning participants not to tell, and masking the responses so that the data is anonymous, thus avoiding researcher knowledge of individual disclosure of concern. These methods avoid compromising confidentiality, but clearly the principles of beneficence and justice and the researcher’s obligation to act in the best interests of the child are neglected. In addition, children who may have an expectation that disclosure of concern in a research context would lead to help and support, despite assurances of confidentiality, are disregarded.

However, waiving anonymity presents a challenge for researchers who take the approach that participants are best not identified, for example, in research that is sensitive in nature. One consideration is the extent to which it may be difficult for children to assess the potential risk and/or long-term implications in using their own name. They may not have the experience to know where the material will end up and how it may be distorted by the media (Laws & Mann, 2004). However, this has to be balanced against individual children’s ability to make a sound judgement and the reality that children may be better placed to assess the risks to themselves in familiar contexts. Discussing the concerns with children is an important step in this decision-making process.

Anonymity is a means of avoiding harm, however, precautions against recognition of participants (such as removing identifiers) are not always necessary, and may even impede appropriate recognition. Researchers also need to be aware of the importance of ensuring that children’s participation is recognised and valued in forms that are discernible and meaningful to the participants. In participatory research where children are actively involved as researchers they may want their names included to recognise and value their important contributions. In such contexts children can be encouraged to identify any potential risks and advantages (now or in the future) of their names being included, so that informed decisions can be made.

The use of photographs in the dissemination of research raises issues around privacy and anonymity, both at the time of dissemination and also in the future, when the child may have a different opinion about the use of their image. It also raises issues related to power and representation of children (Phelan & Kinsella, 2013). Researchers may use ways to hide the identity of children involved in their research (or photographed by children involved in their research) in order to ensure they are not identified and maintain their anonymity. Strategies may include pixilation, blurring the face or image, or perhaps using only text in representation of the findings (Nutbrown, 2010). However, Nutbrown contends that this raises an ethical debate about children’s voice — arguing that, if consent or assent has been obtained from children to use their images, might researchers be silencing children in attempts to protect them? The issue of anonymity in relation to children’s authentic representation raises the tensions between protection and participation that are not easily resolved, particularly with the added dimension of speculating about future (unknown), as well as present, contexts. As a community of researchers, we continue to discuss and debate such issues in the ERIC expert blog (see 2021 blog contributions from Harry Shier and from Kay Tisdall, Marlies Kustatscher and Helen McAneney).

The issues are further compounded in research that uses information and communication networks such as the internet, instant messaging and social media for data collection. The expectations research participants may hold of privacy online may be exaggerated, if not illusory (Lobe et al., 2007). Diverse modalities of online internet communication have different features and, consequently, different practicalities in regard to privacy and confidentiality. For example, some internet forums are intentionally public and therefore anyone can read messages that are posted without leaving a trace of their presence. However, people’s presence in chat rooms is more apparent and real time communication makes it awkward to observe without interacting (Lobe, Livingstone, Olafsson & Simões, 2008).

The complexities of obtaining online informed consent, the ability of researchers and participants to assume anonymous or pseudonymous identities, and the potential for them to have multiple online identities, contribute to the ethical concerns in online research with children (Lobe et al., 2007). There is the risk that people not belonging to the group of selected child participants, may give false details and participate under false pretences, threatening the privacy and safety of the group (Alderson & Morrow, 2011).

Research using new technologies requires consultation with communities (including online communities) and children, prior to undertaking research, to reduce risks of breaking confidentiality (WHO, 2011). In addition, the challenges around protecting images of children and data collected (for example, children’s artwork or photographs) on the internet are an increasing concern.