Toward ethical praxis in longitudinal research with children : Reflecting on ethical tensions in participatory research

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Toward ethical praxis in longitudinal research with children : Reflecting on ethical tensions in participatory research © 2024 The Author. Published by Elsevier Ltd. Published version Inha, Karoliina Inha, K. (2025). Toward ethical praxis in longitudinal research with children : Reflecting on ethical tensions in participatory research. Research Methods in Applied Linguistics, 4(1), Article 100173. https://doi.org/10.1016/j.rmal.2024.100173 2025 Toward ethical praxis in longitudinal research with children: Reflecting on ethical tensions in participatory research Karoliina Inha Centre for Applied Language Studies, University of Jyv¨ askyl¨ a, Mattilanniemi 2, Jyv¨ askyl¨ a 40100, Finland ARTICLE INFO Keywords: Ethical tensions Young foreign language learners Participatory methods Power differences ABSTRACT The field of early foreign language learning research has witnessed an increasing trend to employ participatory methods, stemming from a paradigm shift in how children are viewed in research. In contrast to more traditional approaches that place children as the objects of research, children are considered active agents and participants in research with expert knowledge of their own lives. In this paper I discuss the ethical tensions that surfaced prior, during, and after a research project that employed participatory methods with young children. I reflect on how the methodology broadened understanding of emerging power differences in the researcher–participant relationship, such as how asymmetry in knowledge and participation alternate and manifest during a longitudinal project, engaging the researcher and the children in recurring negotiation of roles and power. While ethical considerations are often presented as a checklist with do’s and don’t’s for how to proceed, with this article I draw on experiences from my own project and aim to dismantle the complex and messy picture of the ethical process in research with young children. I argue that while the application of participatory methods in early foreign language learning research alleviates ethical tensions to an extent, such as by helping reduce epistemic and institutional power between the researcher and the participants, it is essential to report and reflect on the ethical weighing and reflection undertaken throughout the study as unequal power structures will remain regardless of method. 1. Introduction Research on young foreign language (FL) learners in primary school has grown in consequence of European education policy emphasizing an earlier start in FL education as well as the growing diversity of children’s home languages. The accumulation of studies with young children has led to the reconsideration of ethical issues in research praxis (Pinter &Kuchah, 2021), while ethical matters have also gained more general interest as a topic of inquiry in linguistics (De Costa, 2024). Furthermore, research in applied linguistics has witnessed a paradigm shift regarding how children are no longer perceived as the objects of research, but as active participants who exercise social agency in interpretating the world and producing knowledge (Pinter, 2022). In this article I claim that in addition to methodological considerations, further attention should be paid to the ethics of selected methods and to the relationship built between the participants and researchers. In early childhood education and care, the relationship between methods and ethical practice has moved beyond the consideration of methods to placing a focus on the union of methodology and ethical praxis (Palaiologou, 2014). Simply applying participatory methods does not suffice, but instead ethics should be the focus point. In applied linguistics, discussions of ethics are not a novelty, but E-mail address: [email protected]. Contents lists available at ScienceDirect Research Methods in Applied Linguistics journal homepage: www.elsevier.com/locate/rmal https://doi.org/10.1016/j.rmal.2024.100173 Received 13 May 2024; Received in revised form 22 November 2024; Accepted 26 November 2024 Research Methods in Applied Linguistics 4 (2025) 100173 Available online 30 November 2024 2772-7661/© 2024 The Author. Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). the field has witnessed a recent ‘ethical turn’which has raised new perspectives (De Costa et al., 2021; see also Yaw et al. 2023). Pinter (2022) suggests that ethics in research on FL education studies may not have a consensus on the meaning of ethics in research with children, as ethics is usually only briefly reviewed in publications (Pinter, 2022). According to Kubanyiova (2008), the lack of rigorous reporting may also be due to the fact that research in applied linguistics generally may lead to only minor ethical issues, as opposed to, for example, research in the medical sciences. In consequence to scarce reporting of ethical conduct, Pinter (2022) calls for further “sharing of experiences in the research community to uncover and explicitly examine approaches and principles in research ethics with children, which are currently kept under wrap”(p. 23). In the same vein, De Costa et al. (2021) call for raising understanding of how researchers negotiate ethical tensions prior, during, and after data collection. More and more studies in early FL education make use of participatory methods. The aim of participatory research is to give voice to underrepresented groups by making use of methods that better capture the perspectives of those that yet remain marginalized in research, such as children. Such an approach resembles that of child-perspective methodology, which strives to highlight children’s roles as knowers and actors in their own lives (Honkanen et al., 2018). With children, participatory methods may refer to visual methods or, for example, storycrafting (Karlsson, 2013). The notion of ’participatory’should be understood to exist on a continuum; studies may include features of both participatory and more traditional approaches as regards, for example, who design the tools of data generation (Pinter, 2022). Furthermore, while a power gap between children and adults may be inevitable in research with children (Pinter, 2022), participatory methods can help dismantle possible hierarchical setups. In this article I argue that while the application of participatory methods in early FL learning research alleviates ethical tensions somewhat, such tensions and power differences will remain regardless of method, whereby it is essential to report and reflect on the ethical weighing undertaken throughout the study. In this article, rather than give an elaborate overview of the macroethics of doing research with children, I will focus on the micro (Guillemin &Gillam, 2004;Kubanyiova, 2008), that is, the everyday ethical tensions that emerge during a project. The paper discusses ethical tensions considered prior, during, and after a research project I conducted with 7–9-year-old Finnish learners of English. The aim of the longitudinal case study was to explore the learner beliefs 7–9-year-old Finnish learners of English hold regarding English and learning English. I exemplify each ethical tension with excerpts from transcripts of group discussions held with the participants or with my own reflections during the project. I respond to the following research questions: 1. What ethical tensions emerged prior, during, and after a research project? 2. How did the use of participatory methodology broaden understanding of how power differences emerge in the researcher–participant relationship? Exploring these questions in the context of a longitudinal project enables considering the development and transitions both the children and the researcher experience during a longitudinal project, for example, the maturation of the children, their socialization into the education system, their expectations for research, and the coinciding development of the long-term relationship between the researcher and the children and (see Rutanen et al., 2023 for a similar study). Long-term research relationships with children are complex with inherent power differences and may entail changes as regards, for example, social and epistemic power (e.g., how the children respond to the researcher) (Einarsd´ ottir, 2007). Considering power differences is important because they may affect how and which children participate in generating research data. In this article, I focus on the power differences between children and the adult researcher, although those between children are also acknowledged. Based on what I present in this paper, I suggest that we should focus more on the methods used in child-focused research and how ethical tensions are reported in publications. This paper contributes to the development of ethical, child-oriented methodology in the field of early FL research by addressing some of the successes and missteps in the nexus of ethics and methodology of a longitudinal project. This paper provides an opening and encouragement to share experiences of ethical tensions, as has been done in childhood studies (e.g., Hendry et al. 2024,Rutanen et al. 2023) but are yet scarcely reported in early FL research. Reflection of ethics is an integral part of research. Shedding light on ethical tensions during research is an attempt to strengthen and further develop methodological and ethical integrity in the field. 2. Ethics in child-focused research 2.1. Background The United Nations (1989) article 12 acknowledges that children have the right to express their views on matters that concern them. One of the main principles of the convention is the child’s right to participate in decision-making that involves them. While children maintain that right, it is often the adults who decide who participates in research, as it is the parents’dissent that overrules the child’s consent. Linguistics associations such as the American institution for Applied Linguistics (AAAL) and the British Association for Applied Linguistics (BAAL) have published their own ethical guidelines specifically for research in applied linguistics. Furthermore, handbooks of childhood studies by Christensen and James (2008),Fraser et al. (2004),Sargeant and Harcourt (2012) provide more K. Inha Research Methods in Applied Linguistics 4 (2025) 100173 2 detailed descriptions of what should be considered when conducting research with children, though in the field of second and FL research, such handbooks are rare (Pinter, 2022). A more recently published book by Pinter and Kuchah (2021) focuses specifically on the methodological and ethical issues in research with early second or FL learners. Lastly, in Finland, the Finnish National Board on Research Integrity (TENK) published the first national guideline regulating ethical research with human participants in non-medical research in 2009, which was later updated in 2019 (Finnish National Board on Research Integrity TENK, 2019). In Finland, an ethical review from the ethics committee is mandatory only in specific cases, for example, when separate consent is not gathered from young children’s guardians. However, the ethical and legislative guidelines do not consider ‘situated ethics’(Graham et al., 2016), that is, the researcher’s everyday decisions made during data generation, nor do they give tangible ethical guidelines for what happens prior and after research is conducted. They can be particularly insufficient for more holistic, classroom-based research (Kubanyiova, 2008). Such instances have been referred to, for example, as ethical tensions (De Costa et al., 2021;Kubanyiova, 2008), ethically significant (Kubanyiova, 2008) or important moments (Guillemin &Gillam, 2004), and ethical dilemmas (e.g., Christensen and Prout 2002,Kubanyiova 2008, Pinter 2022), of which the last appears most common. However, Ennser-Kananen et al. (2018),Guillemin and Gillam (2004) question whether the term dilemma truly captures the nature of the everyday issues and ad hoc decisions researchers must make. Dilemma suggests a turning point where the researcher should make a difficult choice between options and weigh advantages and disadvantages (Guillemin &Gillam, 2004), with possibly only ’wrong answers’.Dilemma also suggests that ethics only come into play when something unexpected happens, perhaps because the incidents fall outside the commonplace ethical checklists, reinforcing the idea of research as a process of ticking boxes to meet ethical standards. Instead, such ’dilemmas’should be understood as normal proceedings in a project (Ennser-Kananen et al., 2018). In this article, these incidents are understood as emergent ethical tensions and negotiation of participation, where interaction plays a key role. This does not undermine the significance or criticality of the issues, but instead foregrounds interaction and negotiation. The following subsections outline the background behind the choices I made regarding methodology and ethical conduct. I draw on literature from the fields of childhood studies, anthropology, and research in second and FL education to outline perspectives with significant meaning for research with children. 2.2. Children’s consent to participate Ethical issues in research with children are often considered from perspectives such as informed consent, confidentiality, anonymity, beneficiality, and nonmaleficiance (Kuchah &Pinter, 2021). Further ethical concerns relate to questions of power relations (adult–child, child–child) and the cultural embeddedness of the methodology chosen for the study. In journal articles, researchers traditionally report gaining informed consent from participants and their parents or guardians (Pinter, 2022). While the ethical guidelines in Finland require informed consent from the guardian and the child, guidelines lack further guidance on how children could be informed and their understanding of consent affirmed. Indeed, when conducting research with children, it can be difficult to ascertain children’s willingness to participate in practice, indicating a need for modified practices with children. Especially with children, informed consent should not be understood as a one-time decision marked with written affirmation, but as something that is reflexive (Palaiologou, 2014), ongoing dialogue (Kuchah &Pinter, 2021), and renegotiatable (Butler, 2021;Graham et al., 2016). However, no matter the effort a researcher puts into reassuring children of their rights to refuse, children may feel they are not in a position where they may safely refuse to participate (Pinter, 2022). With young children, the researcher must speak to several stakeholders, such as education and school administration, teachers, and guardians before introducing a study to potential participants. Einarsd´ ottir (2007) notes that such stakeholders, including the researchers, act as gatekeepers to children’s participation (see also Flewitt 2005). Different stakeholders make decisions on behalf of the child, for example, because they may feel a given student or class is not suitable for a study. Indeed, participation is not only a question of who participates, but a question of who does not. Therefore, researchers should consider the impact their actions have on those not given permission to participate, for example, how the criterion of nonmaleficiance is met if non-participating children feel left out (Richardson, 2019). Ethical praxis also relates to the ad hoc decisions researchers make regarding unpredictable ethical tensions that commonly emerge in research with children. Indeed, children’s responses and actions can often surprise the researcher, regardless of how well they have prepared (Pinter, 2022). Such surprises may later be interpreted as ethical tensions that have not been recognized due to the often disorganized nature of data generation with children. Difficulties arise, for example, in how children may digress from the topic at hand or how they express their willingness to continue participating. Children express their assent and dissent in much more varied and ambiguous ways than a simple ‘yes’or ‘no’(Huser et al., 2022). Instead, they may use verbal or embodied ways of signaling dissent, or else by withdrawing to a silent and evasive mode. Researchers should therefore be sensitive to how the children act and react during data generation, including their non-verbal communication and behavior used to express their comfort in participating (Hendry et al., 2024;Pinter, 2022;Sargeant &Harcourt, 2012). Therefore, researchers could provide participants with methods of expressing assent or dissent that do not require verbal expression, such as signs. K. Inha Research Methods in Applied Linguistics 4 (2025) 100173 3 2.3. Methods to empower children’s participation Participatory methods resulted from developments in how children have been conceptualized in research and politics. The 1990s witnessed a paradigm shift in seeing children as social actors capable of expressing their views, as active agents and contributors in their lives, and importantly, worthy of being heard to better understand their lives and perspectives (James, 2007;James &Prout, 1990). Hohti and Karlsson (2014) note that researchers should give space and pay close attention to whose voices they are listening to and legitimizing: those that speak the loudest or most fit in with an adult researcher’s understanding, or the quiet ones or those whose perspectives are not as easy to understand as an adult. For example, Dufva, Alanen, &M¨ antyl¨ a, 2001 question whether when researchers have critiqued children for not finding the ‘correct response’, they have in effect mirrored the idea that children should learn a specific way of thinking and acting before adults consider their views worthy of exploring. The turn of the century led to conceptualizing children as participants and co-producers of knowledge in research and hence to the introduction of participatory methods (Christensen &James, 2008;Christensen &Prout, 2002;Robinson &Kellett, 2004). In the field of second and FL research, data generation methods with young children have ranged from more traditional approaches (e.g., questionnaires and interviews) used with older participants and which often require some level of literacy or skills verbal expression, to more participatory and child-centered approaches. The latter may make use of methods such as group discussions, photo-elicitation, drawings, or storycrafting to enforce ecological validity (Pinter, 2022; see also Bronfenbrenner 1974). Butler (2021) asserts that simply choosing participatory methods does not render a study participatory, but instead attention should be paid to the “social relations between the researcher and research participants that enable the co-production of knowledge”(p. 29), underscoring the importance of considering the power differences at play in research with children. With young children, power differences may stem from physical, emotional, cognitive, psychological, and social factors (Robinson &Kellett, 2004;Sargeant & Harcourt, 2012). Moreover, power relationships can be considered from different aspects, such as social power (e.g., power between children and adults or groups of children) epistemic power (e.g., who is considered as knowledgeable or holding superior knowledge), and institutional power (e.g., the power teachers hold in school) (Christensen &Prout, 2002;Fraser, 2004;Hohti &Karlsson, 2014; Richardson, 2019;Robinson &Kellett, 2004). Researchers may hold (also subconscious) beliefs of how children should act and participate in research, which may be constructed based on previous research frames. The researcher’s expectations pose a challenge for ‘truly participatory’research, as researchers may limit children’s power to meet the aims of or to ensure consistency in a project. Indeed, children hold power over adults and may even exercise it in ways not ‘preferable’to the researcher. While seemingly happy to participate, children may not ‘act accordingly’, by acting out, joking around, or discussing completely unrelated topics. Butler (2021) refers to the negotiation of power, indicating how power is not static but dynamic and socially negotiated. In participatory research, the researcher and children together as participants negotiate how the project proceeds and what their shared objectives are. The flexibility required of research conducted with children is itself a characteristic of doing ethical research, not a threat to the viability or validity of research (Palaiologou, 2014). To bridge the gap in power differences, researchers may try, for example, participating in the children’s everyday activities or by building a relationship with them before conducting a study. In an early study, Mandell (1991) introduced the concept of the least adult role. It refers to the idea that instead of the researcher taking the role of a detached, adult observer, they blend into the children’s world and become ‘one of the kids’, an active participant in the children’s activities with an aim to minimize the differences between children and the adult. Mayall (2008) asserts that it is not possible to minimize the power difference to such a level that children might regard the researcher as ‘one of them’. Instead, power differences should be acknowledged as a central characteristic and be further worked with and deconstructed. Indeed, Mayall (2008) suggests that the use of more child-focused methods that also change the role of the researcher might increase the power the children hold and decrease that of the researcher. Mayall suggests that using pair and group conversations help children feel more at ease with both data generation and the adult researcher as well as could ease their joint reflection of a given topic. While participatory methods have been seen to promote ethics in research with children, researchers have questioned whether the ethical side of participatory methods are taken for granted (Graham et al., 2016). Building on a study on adult-child relationships in research, Palaiologou (2014) extends the discussion of ethics beyond mere methods and asserts that researchers should pertain to ethical engagement throughout the whole process of a research project—ethical praxis. Thereby focus should be placed on the ethicality of the entire methodology applied in a project instead of the ethics of distinct methods (Palaiologou, 2014; see also Flewitt 2005). Considering that ethical research assumes that participants benefit from participating, some researchers have raised concerns whether research with children is ever really ethical, if children do not benefit from it directly (Pinter, 2023). For example, including the participants also in the dissemination of findings and keeping them informed balances out the unequal power between the participants and the researcher (Flewitt, 2005). Butler (2021) asserts that applying participatory methods not only leads to more ethical research, but also has pedagogical merits by ensuring that children also benefit from participating: negotiating power and allowing relations to flux among participants (including the researcher) allows for mutual learning to happen. K. Inha Research Methods in Applied Linguistics 4 (2025) 100173 4 3. The context: a longitudinal research project 3.1. Aim and participants The project is a longitudinal multiple case study exploring the beliefs 7–9-year-old learners of English hold regarding English and learning English and how their beliefs develop as the learners progress from grade 1 to grade 3 of Finnish primary school. The aim is to build knowledge on young learners’conceptualizations of language and language learning at the very beginning of formal instruction of English with regard to how the learners encounter English both in and outside of school. The study builds on earlier research on young children’s beliefs about languages and language learning in Finland (e.g., Aro, 2009,Dufva and Alanen, 2004,Mård-Miettinen et al., 2014). The study was conducted in two rural schools in Western Finland. At the start of the study, the participants (n=39) were 7 years old and in the first grade of primary school. All participants had started learning their first foreign language, English, in the fall term of first grade. The children were given one lesson of English teaching per week. Of the original 39 children, 21 participated in the data generation conducted in 2nd grade, and 18 in the data generation in 3rd grade. According to the university guidelines of ethical research (Finnish National Board on Research Integrity TENK, 2019), with 7-year-- old children, informed consent must be attained from both the guardians and the child. Furthermore, in Finland, consent to conduct research in a school must be attained from the local education administration. I acquired written informed consent from the guardians and the children each year of data generation. 3.2. Data generation My data generation methods build on my theoretical framework, ethical considerations, as well as my previous experiences in research with young children. An equally important, intersecting aspect was to conduct the study in as child-oriented a manner as possible, and thus I decided to use participatory methods: (1) group discussions; (2) photo-elicitation; and (3) drawing. Firstly, I designed my methods to follow my conceptualization of the construct of beliefs. Following the contextual approaches, beliefs are contextual, dynamic, and co-constructed in social interaction (Barcelos, 2003a). Therefore, exploring beliefs would be advantageous to explore using interactive methods that consider the context (Barcelos, 2003b). I chose group discussions instead of individual interviews as my primary method to engage the participants in a wide array of discussion topics. I held the group discussions during class time and afternoon-school activities with each discussion including 2–4 children. This allowed for all those willing to participate to do so during the school day. However, it might have caused some children to feel left out if their parents had not given permission to participate (Richardson, 2019). I formed the groups with the help of the teachers who knew which children work well with each other and attempted to have the same groups each year. Group discussions work well with children, as they may support children in expressing their views as well as increase feelings of comfort during the task (Einarsd´ ottir, 2007;Mayall, 2008). The participants also completed a group task, in which they were given cards representing different classroom activities and were asked to place them in order by preference, as adapted from Kolb (2007). In the first year, the group discussions (N=13) lasted approximately 26 min. In the second and third data generation, I decided to apply visual methods to both provide the children with more varied methods of expression and to prompt discussion on themes that were key to my research. Visual methods have established a prominent role in researching young children’s beliefs and may be especially beneficial for participants who may have difficulties expressing themselves verbally (Kalaja &Melo-Pfeifer, 2019;Mhic Mhathúna &Hayes, 2022). When the learners were in second grade, I applied a photo-elicitation task and asked the children to take snapshots of situations where they heard, saw, or used English outside school. The participants took 4 snapshots on average. After taking photos for a week, we had group discussions (N=7, duration approximately 14 min). I chose one snapshot per participant and asked the children to elaborate on and discuss their snapshots, after which we had a more general discussion on English and learning English outside school based on my loose discussion template. Although having the children choose the snapshots themselves would have been more child-oriented, I decided to do this beforehand to avoid overly long discussions which would not have worked as well with young children (Allen, 2012) as well as to bring variety to the discussions. For the third data generation, I chose to incorporate another visual method and asked the participants to draw themselves learning English to capture an idea of their beliefs about language learning. The drawings were made during class time and the participants were given 15 min to finish their drawings. The children were allowed to discuss freely while drawing. I would hold group discussions (N=6, duration approximately 20 min) a week after the drawings were made and asked the participants to reflect on each other’s drawings and other adjacent themes chosen by me. Another guiding principle in my development of data generation methods was ethical praxis. That is, considering whether my choice of methods and essentially the situations I placed children in would affect how comfortable they felt participating. For example, in a project I had collected data in before the current study, I felt that the one-on-one interviews I had held with the 7-year-olds oftentimes left me questioning whether the participants would express themselves more expansively were they provided with more varied tools of expression. Ethically speaking, I felt it would even be more ethical to engage the participants—young children—in K. Inha Research Methods in Applied Linguistics 4 (2025) 100173 5 activities that were more child-oriented and in activities where they could reflect on matters with their peers and not be put ‘on the spot’by the (strange) adult researchers (see Mayall 2008). In addition to choosing child-focused methods, I piloted each data generation method with two children of the same age group as the participants to revise the instructions and content according to the children’s feedback. I asked the children whether they understood the tasks and whether they enjoyed them, because I deemed it an important characteristic of my methods. 4. Emerging ethical tensions In this section, I will describe ethical tensions that emerged in my project and the resulting self-reflection of power, relationships, and roles. I make use of excerpts from group discussion as well as notes and observations I have made during the project. This section does not aim to be an exhaustive list or thorough analysis of all ethical issues that emerged, but instead to provide an account of select incidents (1) I identified when going through the transcripts from all data generations, (2) that I had noted down during the project and felt uneasy about, (3) as well as changes that I identified to be successes. 4.1. Prior: (the limits of) rapport building I started designing my research plan from the very first stage of introducing myself and my project to the children. In a similar previous project, our project team had contacted children and parents by sending out a research notification slip with the help of the teachers or else by taking the slips to the children alongside a brief spoken description of what the project would comprise. As I visited the classrooms in the previous study, I noticed that around only a third of the children in the class would participate in the study. As my aim was to recruit as many participants as possible, I felt uncertain as to whether all the children, including those that chose to opt out of participating, understood what they were given the opportunity to participate in. Based on these experiences, I felt I needed to approach the children differently and work on rapport building already before introducing my research. That is, my main concern in this project was ensuring that the children understood what they could or would be participating in, because in my view, participation was an opportunity. Furthermore, I wanted to ensure that my way of interacting with the children felt safe to them and helped build a relationship of trust with them. I felt it important to do my bit to make everyone feel seen, heard, and treated as individuals with expertise in their own lives. I began by visiting the school and introducing myself to the children. The idea was to build initial rapport with them. During my visit, I organized a language exploration session (duration around 15 min) with the whole class, during which we would discuss what languages the children could speak or had knowledge of, where they had learned those languages, and where they encountered English. The idea was to approach the children in such a way that they could freely share their knowledge and experiences, with no ‘adult’ judgment or ‘wrong answers’. I hoped they would regard me as a safe adult that encourages them to talk about language, with no predetermined conceptualization of what language is. The children would recount various languages, such as foreign languages (some associated with names of countries), as well as languages such as ‘dog’, ‘cat’, and made-up languages. The children mentioned having encountered English in, for example, games and songs. I then visited the school a second time to discuss with the children what research is, what a researcher does, and the project I was asking them to embark on. I put effort into verbalizing all the steps in the project, for example, what a ‘group discussion’is, how the findings would be used, and why the children were the experts to be heard and experts of knowledge I did not have. The children could ask any questions they might have had. I gave them research notification slips written in plain Finnish I had prepared in advance. I encouraged them to go through the slip together with their guardians and to carefully consider participating. I also tried to make sure they understood they could at any time decide to withdraw from the project and that no one would have to participate. I stressed that everything they told me would be kept confidential. As regards informing the guardians, I put effort into compiling a clear and concise research notification slip with contact details and encouragement to ask for more information if needed. After these visits, I felt I had to some extent connected with the children. I had also put some effort into the way I dressed and portrayed myself as an attempt to reduce the distance and power difference I (perhaps wrongly) assumed to be between myself and the children, reflecting to an extent Mandell’s (1991) idea of the ‘least adult role’. For example, I wore clothing and jewelry with animals and vivid colors. The children clearly responded to this by commenting on my appearance, and some gave me drawings as I was leaving. For example, one child gave me a drawing covered with hearts and the text “Oot kiva”, meaning “I like you”. I also thought I had quite extensively addressed the issue of informed consent with the children and that the only steps left were to make sure the parents received the research notification slips. As I was walking out of the classroom, one of the children stopped me at their side to indicate that they still had a question they wanted to ask in private. I crouched down so they could whisper it in my ear and was rather shocked to hear the question “Does it cost to participate in the project?”. The question threw me, as Finnish school-based research projects do not require a participation fee. While I had thought we had covered all possible questions during the discussions, I realized that what I had felt was a safe, open discussion, was not the case for all involved, and that I had approached the situation with disregard of how children regard any school-based project—aspects that might stop them from even handing out the notification slip to their parents, if they did not want to burden them with yet another thing to pay for. I felt discomfort due to my apparent inability to prepare for the situation and my lacking understanding of the social factors at hand. The question also left me wondering about other K. Inha Research Methods in Applied Linguistics 4 (2025) 100173 6 questions that had been left unasked and whether my approach should have been completely different to allow for more varied questions. I felt that I, as the researcher, had been in a position to design this introductory phase to promote a safe and participatory atmosphere. Instead, I was both very much in control of the situation (and not the children) and had failed to consider possible dimensions of discomfort the children might feel. In terms of informed consent, the most prominent questions in my study related to whether children understood what they were giving their consent to and whether I managed to appeal to and motivate all children. I would attempt to assess the learners’willingness to participate also just before and during data generation, applying the idea of consent as ongoing dialogue and renegotiable (Butler, 2021;Kuchah &Pinter, 2021). I would ask the children whether they would like to participate in the data generation, and if they seemed at all uncomfortable, I would recount that it was not obligatory and that they could opt out if they wanted. Even so, I felt there were children who remained somewhat evasive, but as I did not want to discourage them from participating but to focus on emphasizing that everyone was welcome to join, I felt it was difficult to balance between emphasizing opting in and out. As the project was already ongoing, in retrospect I wonder whether the pressure to generate data might have affected my determination to confirm everyone’s assent, as opposed to looking for clear signs of dissent, especially with those children that may not have felt safe in refusing to participate (see Pinter 2022). To conclude, while I attempted to follow the theoretical underpinnings I had set for my study, notably ethical praxis, I noticed that my preparations, built on the basis of university guidelines and prior literature, had left me with substantial gaps in knowledge as regards the microethics of what happens prior to data generation—rapport building, gaining consent, and increasing children’s participation from the onset of the project—especially as they potentially conflicted with matters such as time constraints and research aims. 4.2. During: choice of methods and my own role For my first data generation with first graders, I decided to use group discussions. I also incorporated an activity card task, where the children would place cards into order by preference. The cards represented different class activities, such as ‘singing’or ‘games’, as well as a ‘mystery card’, for which the children could decide an activity of their choice (such as Minecraft, as one group suggested) (Fig. 1). Fig. 1. The activity cards used in the group discussions held with the first-graders. K. Inha Research Methods in Applied Linguistics 4 (2025) 100173 7 However, I became achingly aware of how while my choice of method led to some children being very vocal, others remained rather quiet. In the following years, I decided to incorporate visual methods in my data generation to allow for more varied tools of expression. To exemplify this, I refer to two participants, L56 and L36. Save for the group discussion held after photo-elicitation in grade two, L56 participated in all three sessions of data generation. In the discussions held in grades 1 and 3, L56 was quiet during the group discussions. The same goes for another child, L36, who was also quiet in the discussions, even though they seemed happy to participate. Both participants would reply to questions posed directly to them mostly with one-word replies. While I attempted to give the children space to opt out, they would say they were willing and took part in the discussions. While these two children were quiet in the group discussions, they were rather active in taking snapshots outside school as part of data generation in second grade. While the children sent on average 4 snapshots, L36 was especially active in this activity: they sent 9 snapshots and one video. L56 sent four snapshots and would further mark the snapshots by using the photo editor and include emojis. That is, the non-committal behavior they expressed during the group discussions did not surface in this activity, but instead the activity empowered them to express themselves by using more varied methods of expression. Other participants also showed excitement when they heard of the task, and one parent gave feedback that their child was particularly excited to document English outside school. Choosing methods not only relates to whether participants have the tools to express themselves, but also to whether they feel comfortable participating. As shown by the literature review, the simple choice to use participatory methods does not suffice to meet ethical standards. Situated ethics is needed to attend to the emotional state of the participants. The following excerpt from a group discussion held with first graders exemplifies what I interpret to be anxious behavior (from L55) that I did not notice during data generation. The group in question was quite energetic and would often digress completely from the discussion at hand. That is, the children’s behavior was rather unpredictable and disorderly. While I wanted to allow the children to discuss topics they were interested in, I still felt I needed to keep to given topics at least to some extent due to my research objectives. In the case of one participant, L55, I found it difficult to keep their attention to the subject at hand. Excerpt 1 L55: I can count to a million…in Finnish R: Okay good. So how about, so the English teacher speaks to you in both English and Finnish? L60: Yeah L59: Yeah R: Okay. What do they say in English? L55: One, two, three, four, five, six, seven, eight, nine, ten…[counts in English] R: What kind of, I think at least when the lesson begins they greet you… L60: …four, five, six, seven, eight, nine, ten R: Okay, what do they in say in English in class? L59: Greeting and the like [L55 keeps counting out loud from 1 to 10 in English for some time] In the excerpt, L55 refrains from the discussion and instead continues to list numbers from 1 to 10. At such intervals, I thought the best solution would be to redirect the conversation. Finally, as the group became increasingly rowdy—most likely disinterested—I decided to end the discussion. I felt discomfort because I felt the data generation had digressed greatly and because I felt uncertain of my conduct in the situation, which in retrospect might have again been affected by a feeling of pressure to ‘succeed’and to generate data. Furthermore, it was clear the children were holding more power over data generation then, which I had aimed for despite my feelings of anxiety. Moreover, while at the time I put my energy into guiding the discussion so that all children were encouraged to participate equally, later as I transcribed the discussion the event felt very different. Instead of the ’vague hassle’it had felt at the time, it felt more as if L55 was showing a reaction resulting from discomfort, something I had not picked up on at the time of the group discussion. L55’s reaction reminded me of a coping method: concentrating on and repeating something specific to ignore the surroundings and growing anxiety. Conversely, next year L55 was once again willing to participate in the project (photo-elicitation and group discussions). After a week of taking snapshots, we would again have group discussions. In the second year, however, L55 engaged in the discussion in a very different manner—they were more responsive and enthusiastic about describing their own snapshot. I interpreted that L55 was much more engaged because of the activity. Taking the topics of discussion on a more concrete and personal level seemed to render it more meaningful, articulating the fine shades of what it means for something to be tangible enough to make the discussion engaging. These examples led me to reflect on how young children express assent or dissent. When children behave evasively, the researcher should strive to ascertain whether the child is still willing to participate. The challenge for the researchers is to balance between encouraging participation and pressuring into involuntary participation, as encouragement might be felt as pressure due to power relations. Moreover, when reflecting on-going, fast-paced interaction, how feasible is it for a researcher to survey and interpret the emotional engagement of the participants, especially if they are near strangers to the researcher? Lastly, especially the group discussions led me to reflect on my own role in the discussions (James, 2007), and especially, whether my role as the researcher is something that I claim, that the participants assign to me, or whether it is jointly negotiated by all of us involved. Vice-versa, what was the participants’role? When I originally planned my first data generation, I had thought of myself as the ‘objective bystander’, who would prompt discussion by guiding children to reflect on given points of discussion. My role would then resemble that of the distant researcher, an outsider asking questions who avoids having any impact on what the learners might K. Inha Research Methods in Applied Linguistics 4 (2025) 100173 8