The Ethics of Gatekeeping: How Guarding Access Influences Digital Child and Youth Research

Digital child and youth research is often conducted in schools involving minors. Corresponding research designs raise two related sets of problems: Ethical issues with regard to working with vulnerable groups like children and adolescents and access to these groups. The latter pertains to the concept of gatekeeping which is an ethical issue in and of itself if certain groups or areas of research are systematically excluded from empirical research and, consequently, from the resulting ben‐ efits. Thus, our study examines how perceived ethical challenges influence gatekeepers’ decisions to grant or deny access to investigate a potentially problematic topic: pupils’ group communication. We addressed this research question empir‐ ically via semi‐structured in‐depth interviews with eight educational gatekeepers in Germany inquiring their attitudes on research in schools in general and on the specific topic of pupils’ group communication via instant messaging as an exem‐ plar of digital child and youth research. Approaching the question from two perspectives (procedural ethics and ethics in practice), we identified hierarchical power structures within multiple levels of gatekeeping and revealed rationales to deny access based on ethical considerations with regard to the given scenario of pupils’ group communication.


Introduction
Nowadays, children and adolescents live in a mediatized world with permanent access to online (social) media (Vorderer et al., 2018). Digital media are not limited to the private sphere, however. They have increasingly found their way into educational settings, for instance in the form of instant messaging (IM) groups as part of the personal learning environment (PLE; Attwell, 2007;Costa-Sánchez & Guerrero-Pico, 2020). IM groups add an online layer to the class' communication space introducing new rules, roles, power dynamics, and a specific netiquette (Knop-Hülß et al., 2018). In order to address questions relating to risks, opportunities, and challenges the digitalization of PLEs poses and to find evidence-based recommendations, empirical research is required. Often, such research is conducted in schools involving minors. This raises two related sets of problems: Ethical issues with regard to working with vulnerable groups like children and adolescents (Davies & Peters, 2014;Nairn & Clarke, 2012), and access to these groups (see Lareau & Shultz, 1996, for an overview on negotiating entry to the field). Concerning vulnerability, aspects of research ethics have to be addressed like informed consent and assent, issues of disclosure, power imbalances, etc. (Phelan & Kinsella, 2013). Furthermore, research in schools relies on access and is therefore dependent on (educational) gatekeepers (Burgess, 1991;Morrill et al., 1999;Morrow & Richards, 1996;Wanat, 2008), i.e., "someone who has the power and control over access to communities and key respondents in a particular location selected for research" (Lund et al., 2016, p. 281).
In Germany, customarily, research projects conducted in schools have to be approved by school officials such as the school board, the principal, and the teachers involved. Identifying and managing these organizational gatekeepers have been described as crucial factors to be considered by researchers wanting to do (qualitative) fieldwork in an organizational setting or, more specifically, in schools (Morrill et al., 1999). Another layer of gatekeeping is added when research is being conducted in public schools where local or state authorities need to clear research proposals before access to schools is given. According to the federal regulations of the German education system, this is the first gatekeeper that every research project has to overcome. Thus, in institutional and/or organizational settings (adult) gatekeepers grant access to the field or deny it, thereby enabling (digital) youth research or rendering it impossible. This relates to research ethics in two ways. First, among others, more practical aspects like the expected disruption and additional workload, the approval of a research project is dependent on its perceived scientific integrity-with research ethics being one crucial aspect. Moreover, the aspect of gatekeeping is an ethical issue in and of itself because the principle of justice is compromised if certain groups (for instance, minorities or underprivileged families; cf. Koschmieder et al., 2021;McAreavey & Das, 2013) are systematically excluded from empirical research and, consequently, from the resulting benefits. If, for instance, scientific research on the causes and effects of digitalization was affected by this bias the digital divide might grow (Rogers, 2001).
Thus, our study deals with the ethics of gatekeeping in digital youth research. We examined how perceived ethical challenges influence (educational) gatekeepers' decisions to grant or deny access to investigate pupils' digital media use. We addressed this subject empirically via in-depth interviews with principals and teachers in their role as gatekeepers (Burgess, 1991;Morrow & Richards, 1996;Wanat, 2008). We inquired their attitudes regarding research in schools in general and on the specific topic of pupils' group communication via IM as an exemplar of digital child and youth research. We chose this example for two reasons: Firstly, IM group communication (e.g., via WhatsApp or Signal) is commonly used in schools in various parts of the world (as well as in Germany) to organize PLEs and to socialize (Costa-Sánchez & Guerrero-Pico, 2020;Dahdal, 2020;Ivanova & Chatti, 2011;Rosenberg & Asterhan, 2018); secondly, both group interaction itself and research of it is ethically challenging-the former poses risks of antisocial behaviors like cyberbullying (Bork-Hüffer et al., 2020), the latter is problematic in terms of privacy and data security.

The Ethics of Gatekeeping
For this article, we draw upon the distinction between procedural ethics and ethics in practice to differentiate between the aspects of access and research ethics (Phelan & Kinsella, 2013;Sherwood & Parsons, 2021). Procedural ethics involve seeking approval from ethics committees, review boards, and institutional gatekeepers like school boards. Dependent on federal-state regulations (in Germany for example in Bavaria and Bremen), pupil representatives, as well as parent councils, also need to approve submitted proposals for school-based research. Ethics in practice, however, refer to the everyday ethical issues that arise while conducting research. Regarding school-based research, this compares to the distinction between seeking official approval (i.e., formal legal power to require compliance) and ensuring informal cooperation (i.e., social power to influence behavior in schools; Wanat, 2008). Thus, there is a two-step process to overcome (or even three steps if the child's assent is considered) when ensuring access to the fieldconvincing (formal and informal) gatekeepers to support the research is a necessary but not sufficient condition as "access does not guarantee cooperation" (Wanat, 2008, p. 207). Morrill et al. (1999) describe two crucial factors for ensuring entry to the field: identifying relevant gatekeepers and determining applicable means to overcome barriers deployed by the organization (such as schools). To facilitate the process, they stress the importance of convincing at least one organizational gatekeeper who in turn might be able to persuade another one on a different hierarchical level to gain access to the field. Additionally, Burgess (1991) points out the importance of building relationships with gatekeepers on lower levels as they in turn closely guard the entities they are in charge of, such as groups of pupils or individuals, and can therefore provide or deny access: "There [is] no individual gatekeeper who could grant or withhold information for the whole school but rather a series of gatekeepers with whom access had to be negotiated and renegotiated" (p. 48).
Educational gatekeepers employ certain "resistance tactics" (Wanat, 2008, p. 203) to prevent successful recruitment and/or cooperation (such as passing responsibility, controlling communication, delaying the process by requesting more information, and forgetting to perform tasks as promised). Stonebanks et al. (2019) found similar stalling methods on the level of official approval. Moreover, the authors identified reasons why teachers themselves would not participate in research projects. Their main concern was not to get involved in a study with an ethically challenging topic. McAreavey and Das (2013) also reported perceived ethical issues (such as privacy concerns and possible harm towards participants) as well as reservations with regard to the validity of the study that led gatekeepers to deny access.
Regarding ethics in practice, all empirical research should adhere to the general principles of respect for autonomy, beneficence, and justice (Beauchamp & Childress, 2019). From them, practical ethical standards are derived like voluntariness and the need for informed consent, confidentiality, and privacy as well as the well-being of the participants (with the latter taking precedence over everything else). These standards need to be addressed throughout the research process while at the same time catering to methodological needs to safeguard reliability and validity (Schlütz & Möhring, 2018). This dilemma between ethical and methodological demands for empirical studies is particularly evident in internet research: With the rise of social network sites and the increase in accessing the internet via mobile devices, researchers see themselves faced with new questions of protecting participants' privacy, ensuring informed consent (both from the participating individuals and from the online communities and system administrators), and managing, storing, and representing the data (franzke et al., 2020). For example, youth internet users are often more inclined to share more (personal) information on quasi-public fora such as social network sites or IM group chats, but nevertheless expect their communication to be private (boyd & Marwick, 2011;franzke et al., 2020). Research ethics in practical research with minors are even more multifaceted because "children are universally treated as a special ethical case" (Nairn & Clarke, 2012, p. 195). Hence, the obligation to protect the rights of the research participants increases if the participants are children and/or minors (Ess & Association of Internet Researchers ethics working committee, 2002). This means, for instance, that not only is there proxy consent from parents to obtain but also the children's own assent (Sherwood & Parsons, 2021). Additionally, notions of power imbalances and representations of the child have to be considered (Oates, 2019;Phelan & Kinsella, 2013). For example, care should be taken to create a research environment where children have agency and are treated as equal counterparts (Phelan & Kinsella, 2013). This may be realized by age-appropriate and child-oriented language in surveys and interviews or by adequate body language when conducting face-to-face interviews or observations. When reporting results, researchers should be sensitive with regard to a fair and dignified representation of children.
In our study, we examined whether aspects of research ethics in general and particularly with regard to minors were anticipated by educational gatekeepers when deciding to grant access and/or participate in a research project. We regard gaining access to conduct empirical research in schools as a complex process influenced by power structures. Gatekeepers on multiple hierarchical levels (i.e., school board, principals, teachers, as well as parents) grant or deny access to pupils as prospective participants. Dependent on the tactics employed, this might introduce a sampling bias, thereby hurting the principle of justice (Groves & Lyberg, 2010). If access is systematically denied, (certain) groups cannot benefit from participating in research. Against this background, we phrase two research questions: RQ1: Who functions as a gatekeeper within schoolbased research and which rationales for granting or denying access do these gatekeepers apply (procedural ethics)?
RQ2: Which ethical challenges do educational gatekeepers perceive with regard to digital child and youth research, i.e., pupils' IM group communication (ethics in practice)?

Method
The subject under study-digital child and youth research and, more specifically, explorations of IM group communication in schools-faces (at least) three intertwined ethical dilemmas: (a) research with minors as a so-called "vulnerable" group necessitating proxy consent and specific rules for interviewing; (b) internet research, calling for privacy considerations; and (c) studying group dynamics and the inherent ethical issues (such as respecting the autonomy and self-determination of all group members). We chose a qualitative approach to adequately address this complexity and to examine the participants' subjective perspective as well as their ethical evaluations of this topic (Colby et al., 1983).
As part of a broader research project, the interview guideline addressed several issues; for reasons of space not all of them can be discussed here. For this article, we will report on the questions regarding research in schools in general, investigating minors, and the challenges of internet research in particular. The question on research in schools in general, for example, was operationalized as follows: Next, I would like to talk to you about the topic of research in schools. Asked in general terms: If research projects have been carried out at your school in the past-e.g., by people from academia or research institutes-how did you experience this?
Furthermore, we will present results pertaining to research ethics with regard to pupils' IM group communication. In this part of the interview, we employed a method called "Mary's Mistake" suggested by Östman and Turtiainen (2016). We confronted the interviewees with the following scenario: Imagine, for example-even though this is not our intention-that we asked pupils to send us screenshots from their group chats in order to analyze them. How would you feel about this and what problems do you think would arise?
The interviewees were then asked to discuss the inherent ethical dilemmas and possible remedies.
We conducted eight semi-structured interviews with German educators in gatekeeping roles. Participants were recruited from the personal environment of the first author (more or less close acquaintances). In the sense of theoretical sampling (Corbin & Strauss, 1990), we strived for maximal variance by interviewing people of different ages and gender involved in the research process in schools, among them three studentteachers who had prior experience with school-based research, and three teachers and two principals who were also members of the school board (as shown in Table 1). All participants gave their informed consent prior to the interview. The interviews had an average length of 27 minutes (range 20-31 minutes). They were conducted in April 2021 via Zoom. The virtual setting was chosen due to the Covid-19 lockdown in Germany. As suggested by Archibald et al. (2019) the interviews were audio-recorded with the integrated recording feature. All interviews were transcribed using assisted F4x automated audio transcription. They were analyzed via qualitative content analysis following the approach of textual structuring using MaxQDA (Kuckartz & Rädiker, 2019). Adopting this method, we systematically described the interviewees' (ethics-related) perspectives on research in schools as well as on the scenario of IM group communication applying a combined theory-driven and data-driven analysis strategy (deductive/inductive approach; cf. Mayring, 2000). All names were pseudonymized to protect participants' privacy.

Results
In the following we will discuss the two research questions formulated above addressing aspects of procedural ethics (RQ1) and ethics in practice (RQ2), respectively. Additionally, we will report on an overarching topic that came up in the interviews unasked, that is the association of research ethics and methodological quality (Schlütz & Möhring, 2018).

Procedural Ethics: Hierarchical Gatekeepers and Their Rationales for Granting Access for School-Based Research
Regarding RQ1, we found that there are gatekeepers to be faced on multiple levels when conducting research in schools (see also Stonebanks et al., 2019). In addition to institutions like school boards, principals take up an important position in the process of granting or denying access (to pupils and teachers, respectively). They can be seen as first-level educational gatekeepers. The (interviewed) principals differed remarkably in their willingness to allow research projects at their schools. In the recruitment process, one principal even refused participation altogether. Another potential participant asked her principal for permission but was denied. Additionally, one of our interviewees reported on the rejectionist attitude of school boards: "And…the announcement from our school was definitely that…they…only allow more elaborate things if the school itself profits from it" (B2). He further reported that the school board rejected all projects requiring informed consent from either pupils or parents because of the high costs in terms of time and human resources. In contrast, a school board member (B6) mentioned his welcoming attitude towards research conducted at his school, provided that the results were made accessible in order to be incorporated into everyday school life. These examples accentuate two things: the general influence principals exert as first-level gatekeepers and the personal attitudes guiding their decision-making.
Additionally, we found that first-level gatekeepers apply various rationales when deciding whether to allow research to be conducted at their schools: One teacher reported that the school board used their fear of "lawyer-parents'' as an argument to deny all research including pupils. Participants also expressed their concern that requests from research projects would be (deliberately) "forgotten" due to the expected organizational expense. Some interviewees further recognized a general skepticism of new things among their school board and cited this as a reason for particularly innovative projects being rejected. According to the interview partners, not only the school board and/or principals but also the teachers themselves act as educational gatekeepers in the research process (second-level gatekeeping). Their willingness to allow research projects in their classes also varied greatly. For instance, interviewees reported that the length of a survey played a crucial role: While short surveys (max. 10 minutes) were described as unproblematic, longer-lasting projects were perceived as interfering with teaching. Additionally, one teacher (B7) suspected younger colleagues to be more open to research as they were more likely to remember the challenges of conducting it themselves. Another aspect that was mentioned was their workload. Interviewees stated that in stressful periods they were more likely to reject requests to avoid potential extra work-regardless of the actual length of the survey or extra work for the teacher: Then, I think it also depends a lot on the current work-load….When I get a request like that and I'm up to my neck in revisions, I shy away from it, even though I don't know how much time it would have taken. (B7) Besides workload, there were other practical issues teachers perceived as potential challenges if they were to grant access. They pointed out how time-consuming working with researchers was stressing the high organizational effort of multiple follow-up loops to ensure sufficient return rates of informed consent forms from parents and pupils, respectively. With regard to concrete implementation, the teachers stated that the time frame (max. 10 minutes) and the age-appropriate wording of questions should be adhered to, especially in the case of interviews. Furthermore, they mentioned the great variance in pupils' backgrounds depending on the type of school. The interviewees expected this to influence the return rate of informed consent forms as well as the pupils' willingness to cooperate.
Concerning the pupils' necessary assent (and possible self-selection bias; Queirós et al., 2017), they themselves can take on a gatekeeping role at the third level. In general, pupils were described as quite open to participating in research projects-even though the interviewees suspected the pupils' willingness to be dependent on their interest in the respective research topic. Moreover, the implementation of surveys during class was stated to offer several advantages compared to asking pupils to fill out (online) questionnaires at home in their own time. Teachers commented that outside of the classroom setting, researchers were reliant on available technology as well as on the participants' willingness to sacrifice their free time.
Taken together, with regard to procedural ethics we identified (at least) three different groups of gatekeepers on varying hierarchical levels: principals and members of the school board, teachers as well as the pupils themselves. Additionally, parents can be seen as fourth-level gatekeepers since they have to give proxy consent for minors and their attitude toward research projects is an important factor in teachers' and principals' decision-making process. For each group, the interviewees gave various rationales for granting or denying access that were largely practical in nature. Besides those, our study aimed specifically at identifying ethical challenges educational gatekeepers perceived when investigating pupils' IM group communication. They will be covered in the following.

Ethics in Practice: Gatekeepers' Ethical Rationales for Granting Access to Pupils' IM Group Communication
In relation to RQ2, we analyzed gatekeepers' rationales regarding their ethical evaluation of research projects. We did so by introducing an ethically challenging scenario concerning a hypothetical research project on pupils' IM group communication (Mary's Mistakes method ;Östman & Turtiainen, 2016). Overall, teachers were highly skeptical about whether or not pupils would be willing to allow access to their chat groups and/or share their group chat histories. Several interviewees suggested talking to pupils about the chats instead of trying to access them directly and hoping for pupils' honesty and willingness to cooperate. As one participant (B6) explained, he expected different levels of willingness to let researchers see pupils' group chats: While access to chats of a deeply personal nature would most likely be denied, chat groups that were used to discuss school-related content only would probably be shared. The interviewees, therefore, identified the perceived level of privacy and/or intimacy of the (content of the) chats as well as the sensitivity of the data as relevant factors.
Additionally, the interviewees saw pupils' willingness to grant researchers access to their IM group communication as highly age-dependent. Older pupils such as teenagers were described as having more experience in regard to group chats compared to younger ones and would therefore be more willing to cooperate with researchers wanting to investigate their group chat communication. It was suspected that younger pupils could still be encouraged to participate-as long as communication before and during the process of data-collection was age-appropriate, the objective of the study was made transparent, and anonymity was guaranteed. These findings align with the literature review where minors are discussed as a vulnerable research group requiring special attention (Phelan & Kinsella, 2013;Sherwood & Parsons, 2021). Congruently, our participants also described minors as a particularly vulnerable group whose privacy required special protection. Besides stressing again the importance of using age-appropriate language when interacting with young children, the interviewees pointed out that researchers should meet adolescents with empathy and respect in order to establish a friendly relationship.
Another aspect that the interviewed educational gatekeepers found to be ethically challenging was the combination of the online setting of the subject under investigation and the fact that it was multiple individuals communicating with each other. Overall, research accessing and investigating pupils' online group chats was described as a massive intrusion of privacy requiring special awareness with regard to a set of (ethical) challenges. Firstly, the interviewees stated that researchers would have to deal with pupils' fear of data misuse such as data collection without prior consent or sharing the content of the chats with third parties. Secondly, they pointed out that researchers would have to obtain consent from all group members (prior to the process of data collection) to make sure that no third-party data were included (Schlütz & Möhring, 2018). Otherwise, this would harm the principles of (informational) self-determination and autonomy violating ethically sound research practices. The participants emphasized the fact that consent was to be obtained before gathering data in order to adhere to ethical principles: Especially because you would be asking retrospectively, I assume that you would be looking at things that the pupils wrote weeks or even months ago. And the chats were written at that time with…the understanding that they would remain in this group. (B4) Correspondingly, they were highly critical of asking pupils to share screenshots of their school-related group chats (as per the scenario we discussed during the interviews) because they originated without the awareness of outsiders (here: researchers) accessing them at some point in the future. Therefore, the educational gatekeepers questioned whether it would be possible for the pupils to give their informed consent regarding the retrospective nature of this type of data collection and advised against it.
Overwhelmingly, teachers and principals in gatekeeping roles regarded the preservation of their pupils' anonymity as the most important prerequisite for all privacy-related concerns. The fact that the scenario we presented investigated pupils' IM group communication was reason for even more concern: The interviewees remarked that not just the (user)names of the people chatting but also the names of people referenced in the messages would have to be anonymized as a prerequisite to obtaining consent to both collect and analyze the chats. Overall, it was made clear that the participants saw pupils as a vulnerable group and regarded their (private) group communication as worthy of a high level of protection from outside access by researchers. The interviewed gatekeepers' main focus was to shield pupils (as their protegees) from harm and suggested creating a trust-based research context in which researchers would ensure the prevention of possible damage by taking appropriate measures (e.g., obtaining consent from all chat group members, adhering to principles of data protection during and after data collection, anonymizing names in the chat protocols).

Ethical Standards and Their Impact on Research Quality
Without being prompted, the interviewees weighed methodological and ethical considerations regarding school-based research. They discussed how ethical standards have to be addressed throughout the research process, while at the same time meeting methodological requirements to ensure reliability and validity (cf. Schlütz & Möhring, 2018). For instance, they named various suitable measures to uphold ethical principles during the research process and raised concerns about the impact of these measures on the validity of the results. One important measure teachers named for establishing research ethics is transparency towards the pupils both in terms of the research purpose and the data collected. However, teachers pointed out pupils may communicate differently when they are aware of being observed (reactivity effect; Schlütz, 2017). The presence of an observer may also lead to changes in the pupils' behavior: It is quite conceivable or perhaps obvious that a "WhatsApp group 8B" will then be maintained for the "research ethics aunt." And then there is the "Class Chat 8B Real Talk" or something. I think it's really hard to get material that is as authentic as possible. I think if you do it ethically, things are always falsified, even if the pupils give information to the best of their knowledge and belief. It's still just filtered and somehow not as transparent as if you could just look in. (B3) As a further measure to uphold principles of research ethics, the interviewed teachers suggested obtaining pupils' informed consent. At the same time, however, they voiced their concern of a resulting bias, for example through the self-selection of particularly relevant and/or extreme cases (e.g., bullying pupils; outsiders). Similar problems might occur when parental proxy consent has to be obtained. According to the teachers, particularly concerned parties so-called "helicopter parents"-"a parenting behavior that is considered to be overly involved, overly controlling, and developmentally inappropriate among parents of emerging adult children" (Love et al., 2020, p. 327)-were most likely to speak out against their child's participation. In order to minimize sampling bias due to a challenging research design (as presented in the given scenario) teachers suggested interviewing individual pupils instead of observing whole chat groups. However, they addressed possible validity problems here as well: "You probably won't have direct access to the pupils' chat histories. That means you then have to rely on the honesty and…the memories of the pupils and…hope that they reveal the information as it really happened" (B1).
Exchanging methods and measures within the given research design (retrospective self-reports rather than real-time observations; cf. Naab et al., 2019) leads to a dependency on the pupils' memory and willingness to report honestly. In addition, the analysis of interactions is rendered impossible. This leads to a loss of authenticity limiting the validity of the study.

Discussion
In this article, we discussed the ethics of gatekeeping by exploring how questions of access influence digital child and youth research. Our research was guided by the question of how perceived ethical challenges influence gatekeepers' decisions to grant or deny access to research with minors in schools. As an example, we used a fictitious study on pupils' IM group communication. We approached the question from two perspectives (Phelan & Kinsella, 2013;Sherwood & Parsons, 2021): (a) concerning the aspect of procedural ethics, we identified hierarchical power structures with multiple levels of gatekeeping; (b) at the level of ethics in practice, we revealed rationales to deny access based on ethical considerations with regard to the scenario of pupils' group communication.
Regarding procedural ethics we found-congruent with extant literature-that access to empirical research in schools is a complex process influenced by a hierarchically structured multi-level-gatekeeping system including principals, teachers, parents, and pupils themselves. While Wanat (2008), who found a similar hierarchical gatekeeping structure at schools, focused on the gatekeepers' tactics of resistance, we examined their prevalent rationales for granting or denying access. In line with Wanat, we found that resistance tactics were mostly based on practical considerations such as the suspected amount of work required and the disruption of school processes. The rationales we found, however, were more to do with the research project itself (object of investigation, research design, and children as research subjects) and the connected ethical questions. Thus, we expanded the perspective on the gatekeeping process as a whole. In doing so, we opened up opportunities to facilitate access to the field by optimizing and, maybe even more important, convincingly communicating information on both aspects (practical considerations and the research design) to the gatekeepers on all levels. As we also found indications that denial of access is systematically linked to the gatekeepers' particular perceptions, experiences, background, and contexts, however, sometimes even the best communication strategy is prone to fail. The ensuing (self-)selection bias goes along with ethical consequences since the gatekeepers are the ones to decide which groups are investigated and consequently, who benefits from the results.
By implementing the Mary's Mistakes method (Östman & Turtiainen, 2016) adapted to the special case of IM group communication we contributed to the research on challenges posed by technological developments of the last few years-for the study of minors, in an online context, and in relation to groups. Our findings provide a starting point for how future research in this complex environment can be approached in an ethically sound way. Furthermore, we contributed to the research on online PLE-a topic strongly influenced by digitalization and prone to constant change. A further, unexpected finding of our study was the reported need of educational gatekeepers to weigh methodological and ethical considerations when evaluating research on pupils" online group communication. This is congruent with the findings of McAreavey and Das (2013) who reported that perceived ethical issues as well as reservations with regard to the validity of the study lead some gatekeepers to deny access. This shows their awareness of the need to balance ethical principles and methodological procedures in order to ensure the quality of data obtained. With this, our study contributes to the much-needed discourse within the field of digital studies "on how to conduct both ethically and technically sound standardized research" (Schlütz & Möhring, 2018, p. 34)-especially when pupils and their online group communication are explored.
Despite our cohesive findings, we have to address some limitations. In addition to the established restrictions of qualitative research (e.g., interview as a social situation, self-selection bias; Queirós et al., 2017), there are further limitations to be considered here. A weakness of our research is that we did not talk to the pupils themselves. This was a conscious decision, however, as we wanted to focus on the preceding gatekeepers in the hierarchical process: the pupils can only become "gatekeepers" once administrative and school officials, teachers as well as parents have granted access to them. Nevertheless, we acknowledge that pupils" gatekeeping rationales are an integral component of the research process and should therefore be considered in following studies. A further limitation of the sample is its homogeneity in terms of geographical and social context, since all our interviewees were teachers at German schools. Further studies should consider a more heterogenous sample including a wider range of geographical and cultural backgrounds (such as researching educational gatekeeping in different federal states in Germany). Furthermore, by using a very specific scenario in the Mary's Mistakes method (namely pupils' IM group communication) we linked our findings very closely to this specific aspect. Since ethically challenging subjects within the field of digital child and youth research are much more widespread, future research should investigate other school-related aspects, for instance other parts of pupils' PLE such as school cloud-based services, World Wide Web's offerings for studying at home, or the digital devices, platforms, and applications used for studying. Since we put our focus on (group) communication among minors, another central and ethically challenging aspect remained unexplored: the IM-based communication between pupils and teachers, acknowledging that this would introduce an even higher level of complexity. Our narrow focus on online group communication also does not shed light on the question of whether or not the respective research topic has an influence on the gatekeepers' decision. This aspect should also be explored in further studies.
Building on our results as well as the reported limitations, future studies should investigate online-related resistance tactics of gatekeepers to adapt the findings of Wanat (2008) to the current state of digitalization. Furthermore, our findings could be a starting point to systematically explore possible ways for researchers to counteract the described resistance tactics and develop a guide with practical suggestions. As shown above, especially in relation to IM group communication, the perspective of pupils in their role as gatekeepers is also worth exploring. Their perspective on the relevance of IM group communication in general and the IM-related research should thus be taken into account. Additionally, we revealed in our study that not only researchers but also gatekeepers themselves see the need to balance methodological and ethical considerations. They even integrate this balancing act in their own decision-making process of granting or denying access. Therefore, our findings might enable future research designs to better meet both the needs of adolescents as participants and challenging research objects. This will ensure that future research including minors is not only valid and reliable, but feasible in terms of research ethics, and thus more prone to be supported by educational gatekeepers.