When research reveals more than intended: Incidental findings in social research with children explained

Author info

26/01/2026

Introduction

At EU level, social research is a core tool for evidence-based policymaking. Its purpose is not merely academic. It underpins how the EU designs, funds, evaluates, and justifies public action. To put it in other words, social research provides the empirical basis that allows the EU to demonstrate that its interventions are justified and appropriately scoped.

Unlike purely technical research, social research often relies on people’s lived experiences, i.e., relationships, wellbeing, family circumstances, and social environments. When such research involves children, or institutional settings (e.g., schools, care facilities), then it may unintentionally reveal information about harm, neglect, abuse, or serious distress. These are not the research objectives, but they may surface nonetheless. This is where incidental findings arise.

To provide a formal definition, in social research, incidental findings refer to information that is unexpectedly discovered during a study and that falls outside the original research objectives, yet may indicate risk, harm, or unlawful conduct. In projects involving children – particularly those from vulnerable societal groups – incidental findings often relate to:

maltreatment or neglect within the family;
abuse or discriminatory conduct by school staff;
bullying or peer violence;
serious psychosocial distress.

Through the example of LET’S CARE project, this blogpost explains how incidental findings work in social research with children, how this differs from incidental findings in medicine, and what can be improved.

Notably, the concept of an incidental finding should be understood broadly. It does not always take the form of a clear or verified fact but may instead consist of signals or indications of a potential issue. Therein lies the challenge of how to differentiate between findings that warrant follow-up and those that can or perhaps should be dismissed and ignored. Any intervention, even when well intentioned, may itself carry negative consequences for participants, including distress or unease. Hence, being overly zealous and treating every hint as requiring action does not benefit the research participants either. In this blogpost, the term “incidental findings” is therefore used to encompass both confirmed findings and preliminary indications, as the procedure of dealing with incidental findings should be activated in both cases, although the appropriate response, and indeed whether any action is ultimately taken, may differ.

Brief description of LET’S CARE project and potential incidental findings

LET’S CARE is a research project that aims to comprehensively understand and improve the caring dimension of educational inclusion and school success. It further aims to address the causes of educational underachievement, disengagement, and school dropout in disadvantaged or vulnerable populations, by identifying key relational variables between various actors of the education process. While trying to create a theoretical and practical framework to break the chain of transgenerational transmission of educational and social exclusion, the consortium opted for conducting a quantitative data collection as a tool for extracting comparable, scalable evidence that can support policy change, and legally defensible conclusions about what works across different Member States.

In this context, students aged from 8 to 18 years old from schools in 6 Member States (Poland, Lithuania, Bulgaria, Spain, Portugal, Italy) have been chosen to answer a series of structured, closed questions about relationships at school and at home. The objective of the data collection is to extract valid scientific conclusions about educational inclusion, not to conduct child protection investigations. Nevertheless, incidental findings could theoretically arise because the questionnaires touch on topics, such as:

family attachment and pressure;
material deprivation and socio-economic stress;
wellbeing and emotions;
relationships with teachers and school staff.

Designing research to minimise – and responsibly manage – incidental findings

Social research involving children should be designed to minimise the likelihood of incidental findings, while remaining suitable to address them if they nonetheless arise. This dual objective reflects a core ethical principle: research should not cause harm or distress to participants and should not actively seek to uncover safeguarding concerns that fall outside its legitimate aims. At the same time, credible risks for children cannot be ignored.

1)Research design as a first line of prevention

The primary tool for reducing the likelihood of incidental findings is careful research design. This includes limiting data collection to what is strictly necessary for the stated research objectives, avoiding unnecessarily intrusive or open-ended questions, and favoring structured or closed formats where appropriate. The design, of course, depends on the goals and the type of research. Quantitative approaches are not always appropriate, and qualitative approaches (e.g., interviews or more direct interactions with research participants) may in certain settings be better suited. LET’S CARE carried out both for different research goals.

Ethical research principles require researchers to avoid causing harm or distress to participants. For this reason, research should not be designed to actively search for signs of abuse, maltreatment, or distress, as these issues are better addressed through established safeguarding processes rather than through research itself. While it is very important that such situations surface, a one-off research activity is not the right setting/context to aim for such surfacing. This should be caught by institutional processes in other settings, e.g., schools should generally have procedures in place to flag and follow up on alarming indications presented by children that they may be the victim of abuse, maltreatment, etc. Research should be beneficial to participants or at the very least not detrimental, or distressing. Clear communication of the study’s purpose and conducting research in familiar, low-disruption environments, such as schools, helps to ensure this.

As such, it is legitimate in social research with children to minimise and avoid incidental findings to start with. However, one should be prepared to deal with them if they arise nonetheless.

LET’S CARE applies these elements:

• Both the type of questions and the content of questions of the data collection activity are aimed at specifically gathering only the information that has been justified as necessary for the research purposes and explained in the consent forms, meaning inter alia that the questionnaires rely mainly on closed questions, in order to reduce the possibility for participants to provide unnecessary information;

• The study’s main objective and focus on school inclusion has been repeatedly explained to participants;

• The research takes place during regular class hours and with the intention of minimally disrupting school routine.

Despite strong research design in the LET’S CARE project (as will be true in many projects), theoretically it remains possible that, in exceptional circumstances, a combination of certain contextual indicators deducted from the answers of the child, in combination with additional information such as participant disclosures or observed distress during the research activity (in LET’S CARE: the answering of the questionnaire), could lead to a safeguarding concern and a potential incidental finding.

This is generally true for many research settings, despite the best design choices. When there are open questions in a questionnaire, when there are questions that read together may imply a finding not actively searched for, or when there is any direct interaction with research participants, unexpected findings may theoretically arise. While research design can reduce the likelihood of such findings, it may also lead to any findings being less clear and certain than they would have been in a setting with more open questions or more direct interactions with research participants. This should however not lead to dismissal of the potential uncertain finding. Indications of issues should always be duly considered.

2) Training and support to the adults present during data collection

If the research activity and questions have been designed appropriately, a second source from which incidental findings may originate is direct interaction with participants. For that reason, it is essential that adults present during data collection are appropriately supported and informed about spotting incidental findings and how to handle them. In school-based research, this typically includes teachers, counsellors, or other educational staff who are familiar with safeguarding procedures and are best placed to respond in real time to situations of distress or indications of abuse or maltreatment.

High-level good practice includes ensuring that these actors understand the purpose and limits of the research, are able to recognise signs of distress, and can rely on existing safeguarding structures (e.g. procedures established in schools where the research is carried out) where necessary. This approach reinforces the principle that researchers should not replace experienced professionals but rather work alongside established systems.

In LET’S CARE:

Data collection is supported by trained facilitators and school counsellors;
School staff has received training to recognise distress;
Reference teachers, being the ones who oversee the data collection process, help manage practicalities.

3) A clear internal incidental findings communication plan for incidental findings that reach the researchers

Even when unlikely, researchers should plan for incidental findings and how to handle them, should they arise. This may happen, among others:

- Because open questions are answered in an unintended way;

- Because incidental findings are observed during direct interaction with research participants, e.g., because researchers are physically present or because these findings are communicated to them by a facilitator acting on their behalf;

- Because incidental findings arise from data analysis, e.g. by putting together data points (e.g., answers to questions) a pattern or issue is revealed that was not intended to be analysed;

The objective is not to transform researchers into investigators, but to ensure that credible risks to children are identified and assessed in a consistent manner, with proper escalation where necessary and justified.

At a high level, an effective incidental-findings framework should define:

roles and responsibilities, identifying who within the project is authorised to assess potential concerns;
An internal escalation mechanism, allowing information to be reviewed promptly and consistently by people with the relevant knowledge and expertise;
decision-making criteria, including the seriousness and credibility of the information that potentially reveal incidental findings, the child’s best interests, and the potential risks associated with disclosure;
clear timelines for internal review and decision-making, to avoid undue delay;
a structured assessment of whether there is a legal obligation to act or report, to whom this obligation applies and what situations trigger this obligation, taking into account applicable national laws and the distinct roles of researchers and school staff;
alternative communication pathways where standard reporting channels may be inappropriate;
documentation and record-keeping for a limited period, ensuring accountability and compliance with data protection principles.

By way of illustration, the LET’S CARE project operationalised these elements through an internal policy providing for the centralised review of potential concerns by a small ethics group, combining project coordination, legal and ethical expertise, and school-level safeguarding knowledge. This structure is designed to ensure consistency across participating schools and Member States, while allowing for contextual judgment and prioritising the best interests of the child. The plan also incorporates defined review timelines, enabling concerns to be assessed and decisions taken without undue delay.

The distinct legal position of researchers vis-à-vis school staff involved in data collection should be considered here. In line with national safeguarding laws, researchers typically do not carry primary reporting responsibilities (unlike teachers, who as professionals working with children tend to have specific reporting obligations to the competent authorities, e.g., child protection services). This must be duly considered, especially where findings are vague traces or indications of a problem compared to explicit statements or even empirical evidence of the same.

Depending on the situation, it may be more appropriate to limit the involvement of research to internal escalation, i.e., at project level, where it is appropriate to inform school leadership, with alternative routes considered in exceptional cases (e.g., when school staff is suspected to be the perpetrator).

A careful analysis of national law is always required, as mandatory reporting obligations differ significantly across countries. In some jurisdictions, these duties apply to any person, not only to professionals working with children. Moreover, both the types of situations that trigger reporting (e.g., only specific crimes vs. broader indications of child maltreatment) and the applicable state of mind/knowledge threshold (e.g., “believe,” “reasonable grounds to believe,” or actual “knowledge of”) may vary from one legal system to another. Moreover, some mandatory reporting obligations require a sense of urgency to justify the reporting (e.g., “imminent harm”). In any case, every disclosure should take into account that this is an act that breaks confidentiality promises made, and must therefore be legally (e.g., data protection, professional secrecy rules) and ethically justified.

As mentioned before, information emerging from questionnaires or interviews in social research is often partial, subjective, and difficult to corroborate. Even direct statements may turn out to be false (e.g., to get attention). Hence, a balance must be sought between not ignoring traces of an issue, yet not being overly quick to act on incomplete or unclear information. In LET’S CARE, this understanding has been translated into a cautious approach to any further action that includes careful consideration by various relevant actors before a decision is made.

Finally, transparency towards participants remains a core ethical requirement. Participants must be informed about the incidental findings policy and particularly about the circumstances under which the promised confidentiality of the data they provided could need to be breached. As with any communication, a balance must be found and the information must be calibrated to the intended audience informing them but not unduly scaring or worrying them.

In this context, the level of detail provided about incidental findings in consent forms should be adjusted to the intended audience and to the concrete risk profile of the study. Where the research design make the occurrence of incidental findings very low, and researcher presence is limited, a proportionate and audience-appropriate approach to transparency can help preserve trust and participation without undermining ethical safeguards. The following graphic summarises the essential elements of an incidental findings plan as described above:

Why incidental findings are particularly important when children are involved

Research with children is not “business as usual.” From the outset, children occupy a distinct legal and ethical position that requires special attention and guarantees. They are not merely research participants or data subjects, but rights-holders whose safety and wellbeing may be directly affected by what emerges in the course of a study. This ethical reality requires researchers to anticipate incidental findings, even where the research is not designed to identify harm, and embed appropriate safeguards throughout the research lifecycle.

Children may normalise problematic situations they are used to experience, without being able to label it as abuse, neglect or maltreatment. Moreover, given that children are not adults, disclosure of incidental findings needs to address an adult (e.g., parent, caregiver, school teacher). However, telling the wrong person – especially if they are the potential perpetrator – may trigger distress, or even retaliation. This creates an ethical responsibility for researchers to define, in advance, how such incidental findings will be assessed and addressed, irrespective of whether they arise during data collection or at the data analysis stage.

In school-based research, educational institutions are often best placed to respond through established safeguarding structures, with alternative pathways available where standard channels are compromised. In LET’S CARE, such alternative pathways should be considered when trusted persons (e.g., school staff) are implicated in the maltreatment concern. In this scenario, bypassing the school and addressing other institutions, such as competent authorities directly, may be an effective strategy, again depending on national applicable law.

Mandatory reporting obligations add a further layer of complexity as they vary significantly across Member States. According to the European Commission’s guidelines on ethics in social science and humanities, criminal activity identified or uncovered in the course of research generally must be reported to the competent authorities, even where this requires overriding the confidentiality or anonymity of participants. In certain circumstances, such reporting is not merely discretionary but constitutes a legal obligation. However, researchers are generally not mandated professionals under national child-protection laws, unlike professions such as teachers, social workers, or psychologists. Legal systems often reserve reporting duties to those professions, considering that they are best positioned to address such concerns. Therefore, especially in cross-border research, it is essential to map reporting obligations in advance and to define the role of researchers in this regard.

Moreover, from a data protection angle, the GDPR does not impose an absolute duty of confidentiality but requires a balancing exercise between the child’s rights to privacy and data protection (Articles 7 and 8 EU Charter) and their right to protection from harm (Article 24 EU Charter). Where national law imposes reporting obligations, or where disclosure is necessary to protect a child’s vital interests, the GDPR provides lawful bases for limited disclosure of relevant personal data [Articles 6(1)(c) and 6(1)(d)]. Any such disclosure must remain necessary and proportionate.

Overall, incidental findings plans that take into account these considerations reinforce ethical responsibility, protect children, and preserve research integrity, strengthening the overall legitimacy of research activities.

As touched upon above, the policy on incidental findings must be communicated to research participants (the children), and when applying parental consent (which is typically necessary), also to the holder of parental responsibility. In that regard it is important to remember that also children themselves have a right to information they understand. Recital 38 of the GDPR recognises that children merit specific protection with regard to their personal data, as they may be less aware of the risks, consequences and safeguards concerned and their rights in relation to the processing of personal data. Furthermore, Recital 58 and Articles 5(1)(a) and 12 GDPR set high transparency requirements. This means that all the information addressed to children about the processing of their data shall be in clear and plain language, adjusted to the age of the recipient, and accompanied by visualisations, where appropriate. This includes the incidental findings policy. Moreover, the special position of children in data protection law must be applied throughout the research design entirely, including which questions to ask, the impact of those questions etc., which relates to the parameters of data minimisation, purpose limitation, lawfulness, fairness and transparency.

How do incidental findings in social research differ from the most traditional use case, i.e., incidental findings in medicine?

People often understand incidental findings through healthcare examples - like when a scan aimed at assessing the situation for one medical issue reveals an unrelated tumor. This comparison is logical as incidental findings originated in the medical context. However, the situation in social science is quite different.

Medical incidental findings usually follow a more structured approach. They involve clinically interpretable information, verification through further tests is routine, disclosure follows established pathways related to professional duties to diagnose/treat, and adopt a clear “patient benefit” logic, in the sense that patients reasonably expect relevant findings to be communicated to them.

On the contrary, social research often involves incidental findings in the form of suspicions or contextual signals that cannot be definitely confirmed in all cases. Verification is possible, but it is not technical or possible to rely on cross-diagnosis as it happens with medical follow-up. Instead, it involves human actors and institutions: parents, schools, social services, or, in some cases, child-protection authorities or the police. These actions necessarily move the situation to a different frame, implying a different context with different actors and stakeholders, e.g., from research to child protection, whereas within medicine example, the frame, i.e., healthcare, remains more constant.

While in medicine escalation would lead to seeing another healthcare provider who brings a slightly different expertise, escalation in social research may trigger safeguarding procedures by schools or reporting to and investigations by competent authorities, such as child welfare or child protection agencies, social support and social work agencies, or law enforcement authorities. Unlike medical follow-up, where the benefit from communicating the finding to the patient is undeniable, these processes can affect relationships of trust, expose children to distress or retaliation, and have long-term social or legal implications. This is an indication of the complexity of stakeholders involved in social research and the often unforeseeable effects that disclosure can have. While disclosure to parents may be appropriate in many cases, it may be problematic where parents are themselves implicated. Communication with schools is usually safer given existing safeguarding structures, but even this may raise concerns in exceptional cases. Reporting to authorities can lead to further verification and protective action, but it also represents a significant escalation with perhaps irreversible consequences. This means that researchers must carefully assess whether disclosure is sufficiently justified.

Avoiding harm therefore includes preventing premature or poorly grounded disclosures that could unnecessarily distress participants, damage relationships, or trigger disproportionate interventions without adequate factual basis. This safeguarding element is part of the researchers’ ethical duties.

To sum up, there are some fundamental differences between incidental findings in social research and in medicine. While medical incidental findings fit within a well-defined clinical logic of diagnosis, verification, and treatment, incidental findings in social research unfold within a far more complex social and legal landscape. They involve multiple stakeholders, rely on contextual interpretation rather than technical verification, and may trigger procedures with profound and lasting consequences. For this reason, the value of incidental findings in social research lies not so much in their evidentiary certainty – this can occur at a later stage if authorities proceed to formal investigation – but in the careful, proportionate, and ethically grounded responses they prompt. Recognising this distinction is essential to ensuring that incidental findings enhance child protection and research integrity, rather than causing avoidable harm through premature or ill-considered disclosure.

Points of attention on how to best address incidental findings in social research projects

• Incorporate a short, child-friendly statement about incidental findings in consent forms and add other measures where necessary

While informing participants that indications of maltreatment may be reported to competent authorities can deter participation – particularly among individuals from disadvantaged or challenging social backgrounds – greater transparency ultimately serves the public interest by strengthening the ethical integrity of research projects. Providing information on incidental findings even in such a context may ultimately reinforce ethical accountability and underline the societal value of impactful research.

In any case and to minimise the risk of disengagement, the information should be framed in clear, neutral, and accessible language rather than in complex legal terms, for example by explaining that if a child appears at risk, confidentiality may be overridden to protect them.

The statement should be adjusted to the target group of the consent form, i.e., parents/caregivers for those children below the age of legal consent, or directly to students who can consent for themselves (as is the case e.g., in Spain).

Alternative measures can be considered to complement or even in exceptional circumstances replace giving the child information in writing (e.g., for very young children or visually challenged children written information is useless). Usually, a combination is the favored approach.

• Map country-specific reporting obligations from day one

A single project-wide policy should be paired with country-specific reporting information so that researchers and school staff know in advance what is expected from them and do not improvise under pressure. Besides providing in person dedicated trainings, it is advisable to have such mapping done on paper and readily available to the relevant parties.

• Plan alternative reporting routes in exceptional cases involving maltreatment by trusted persons

Flexibility is a key factor in comprehensively addressing incidental findings in social research. Institutional actors, such as schools, generally constitute the default reporting channel because of expertise and uninterrupted engagement with children. However, when there are credible indications that a trusted person within the institution, such as a Headmaster, or a teacher, may be involved in the alleged harm, the incidental findings plan should provide for alternative communication routes that avoid alerting the potentially implicated individual.

It should also be recognised that, unlike incidental findings in medicine, signals arising in social research are rarely capable of direct/immediate technical verification and often rely on contextual and interpretative assessment. Verification tends to be a very intense escalation. For this reason, decisions to depart from standard reporting channels should be taken cautiously and only where there is a well-grounded understanding of the individuals involved and the potential risks of disclosure, always prioritising the child’s best interests.

• Make pseudonymisation or anonymisation project rules and practices compatible with safeguarding

Research protocols should ensure that rules on pseudonymisation or anonymisation do not unintentionally prevent appropriate action where credible indications about maltreatment arise. While pseudonymisation and anonymisation protect privacy, other fundamental rights may be harmed if there is no nuance in the approach, taking into account that the ability to respond to incidental findings is closely linked to whether and for how long participants remain identifiable. Research projects should therefore explicitly address how identifiability is managed across research phases.

A phased approach with limited identifiability through pseudonymisation that evolves into anonymisation after the phase of data collection has been completed, may deserve researchers’ attention. Rather than relying on immediate anonymisation of research data, which would make it impossible to assess or act on serious concerns, it may be appropriate to provide for the use of pseudonymisation during the data collection, with a narrowly framed “brief hold” mechanism of the keys of re-identification under strict controls, to be used only where lawful and necessary. This means that for a period of time, e.g., until 3 months after the completion of the data collection, a trusted party holds the keys to re-identify research participants. Researchers cannot identify any research participants but are able to break the confidentiality for serious findings. This procedure can also serve to provide a balance between offering data subject rights (during data collection and for the period of the brief hold of the keys) and offering complete anonymity (after the brief hold has passed), which however implies data subject rights are no longer able to be granted.

Hence, a balance has to be found between all interests, rights, and expectations of the research participants.

Any such mechanism should be justified by the research, subject to clear governance and documentation, and assessed against data protection requirements, the urgency of the situation, and the child’s best interests.

• Provide support and training to researchers and school staff

Handling possible maltreatment is emotionally difficult. Formal supervision and debriefing should be part of research governance, not an afterthought. Moreover, it is important to conduct regular trainings on the handling of incidental findings to discuss emerging difficulties, uncertainties, clarify ambiguous concepts, and update on applicable procedures if ethical and legal rules require so.

Conclusion

Even where social research with children is carefully designed to minimise the likelihood of incidental findings, that risk can never be entirely eliminated. Research that explores participants’ lived experiences, especially those involving vulnerability or adversity, may reasonably give rise to information that falls outside the study’s primary objectives, but nonetheless relates to the safety and wellbeing of participants or others. Such information may carry legal or ethical significance, for example where a participant discloses involvement in criminal activity, and therefore requires careful consideration.

The appropriate response is neither overreaction nor disregard, but the adoption of proportionate, legally grounded, ethically sound, and child-centered procedures that balance safeguarding obligations with respect for privacy, data protection, and the integrity of the research process.

Planning for incidental findings, in the sense that a project anticipates, governs, and handles the respective information through clear legal and ethical frameworks and procedures, ideally embedded in written protocols, strengthens the ethical and legal posture of the research activities in particular, and of the project as whole. In social research with children, where findings may touch upon sensitive issues like maltreatment or abuse, incidental findings policies bridge the gap between the promise of confidentiality and the duty to protect.

By embedding these elements in the project and research design, researchers move beyond "ad-hoc" crisis management, ensuring that when the unexpected occurs, the response is governed by legal and ethical precision rather than individual assessments.

Lastly, having such procedures in place increases the credibility of the research design in the eyes of various relevant stakeholders such as ethics committees, funders, schools, families, and the research participants themselves.

LET’S CARE is funded by the European Union under grant agreement of the project 101059425. Views and opinions expressed are, however, those of the author(s) only and do not necessarily reflect those of the European Union. Neither the European Union nor the granting authority can be held responsible for them.