‘It’s never okay to say no to teachers’: children’s research consent and dissent in conforming schools contexts

This article examines the limits to children giving research consent in everyday school contexts that emphasises their conformity to comply with adult expectations, and highlights children’s competence and agency in navigating this conformity through different practices of dissent. It draws on research into children’s agency, using a multimodal ethnography of Year 1 classrooms in two English primary schools. The article includes a reflexive methodological focus, exploring the extent to which I counter the schools’ emphasis on conformity. This includes creating visuals for children to practice consent; positioning myself as the researcher in a non-teacher role, as ‘least adult’ and the one who ‘least knows’; and designing interview spaces markedly different from classrooms. The article examines how children navigate conforming discourses by finding different ways to dissent in the research. Firstly, children demonstrate a sophisticated awareness of the cultural norms of indicating refusals beyond saying the word ‘No’. Secondly, children achieve unnoticeablity, by which they absent themselves from the ‘on-task’ classroom culture, and by extension the research process. Thirdly, they engage in playful dissent, demonstrating their political knowingness of the classroom social order. The article discusses the implications for educational research when the values of consent are in conflict with a schooling focused on conformity. This includes emphasising the limits of consent procedures, paying closer attention to how researchers recognise and respond ethically to children’s multiple practices of dissent, and using research to disrupt problematic power structures in education settings that limit possibilities for children’s consent.


Introduction
This article examines the limits to children giving consent in everyday school contexts that emphasise their conformity, drawing on an ethnographic study of children's agency in two English primary classrooms. I begin with an overview of the literature on research consent in school-based research. Next, I examine the ways in which schools emphasise children's conformity to ensure that children learn to fit into society, rather than to question or challenge adults, and how this creates tensions for researchers offering children the option to refuse to participate in their adult-initiated research activities. I outline my methodological approach to the study, and detail how compulsory, which makes them 'something of a captive audience' for researchers (Denscombe and Aubrook, 1992, p. 129). Secondly, dominant educational aims emphasise conformity through academic attainment (learning existing knowledge and skills) and socialisation into behaviours considered necessary for learning (Biesta, 2009). Both are important and support individuals to operate within existing sociopolitical configurations, but neither emphasise pupils questioning, withdrawing from or transforming their schools in some way. Sahlberg (2015) draws attention to a globalised education movement that pursues attainment in ways that emphasise efficiency and accountability, and minimise experimentation, alternative pedagogies and risktaking in schools. In England, this has translated into delivering a knowledge-based national curriculum (DfE, 2014), as well as a strong emphasis on 'good' behaviour (Ofsted, 2015(Ofsted, , 2019) that helps to keep pupils focused on curriculum learning. This includes a demanding bodily conformity, in which children must spend much of the day sitting silently, listening to staff, discouraged from expressing difficult feelings that might disrupt the primary focus of lessons (MacLure et al., 2012;Kulz, 2017;Kirby, 2019). In such classrooms, a concern with judgement can silence children. An awareness of one's place in the hierarchy and anxiety over delivering on what is expected-including correct answers and behaviour-are not the effects of the classroom, but the means through which children are educated. Currently, education rarely involves a transformative purpose, identified by Biesta (2009) as 'subjectification', allowing for the possibility of pupils to shape who they might be and what they can do in school institutions and beyond.

Attending to children's research assent
Within the literature, there has been an overarching focus on how to address the difficulty of ensuring informed consent; a school context, limiting such possibilities, draws attention to the researcher awareness required for the on-going monitoring of participants' assent (Cocks, 2006). The BERA (2018) guidance stresses that where 'capacity, age or other vulnerable circumstance' might limit how much participants 'can be expected to understand or agree voluntarily to participate', they should be 'supported to participate with assent in the research' (p. 15). Cocks (2006) defines assent as the expression of agreement to (continue to) participate in research, which relies on a relationship of trust between the researcher and the researched, in which the child demonstrates an 'acceptance of the researcher's presence' (p. 258). It does not require that the child demonstrates 'adult-centric attributes such as maturity, competence and completeness', instead it 'accepts the child's state of being' (Cocks, 2006, p. 258). This understanding recognises that children 'may give or withhold consent for a number of reasons unrelated to the kind of "rationality" assumed by informed consent' (Gallagher et al., 2010, p. 479).
Dissent and assent are recognised as a non-binary 'continuum from clear refusal, through degrees of ambivalence, to clear and willing consent' (Holland et al., 2014, p. 415). Establishing children's assent requires researchers' 'time and constant effort' to attend to children's many ways of communicating (Cocks, 2006, p. 257). Bourke and Loveridge (2014) highlight the need for further accounts of how researchers make judgements about 'less easily "heard"' dissent (p. 154), illustrating the need to listen Children's research consent and dissent 3 to children's verbal and nonverbal expressions, such as the 'reddening of face, hand twisting, movement in the seat, head dropped and decreasing eye contact, slower responses ' (p. 159). Listening to children, suggest Nolas et al. (2018), currently relies too 'heavily on cognitive, conceptual, and rational models of idealised and largely verbal forms of communication' (p. 1). They stress that researchers 'engaged in the act of hearing' attend to the multisensory, embodied and affective moments of encounters with children (p. 4). There are few such examples of how researchers respond to children's multiple and unpredictable forms of dissent.

Relational understanding of children's agency in research
Children's agency is both 'constrained and enabled' by research contexts (Atkinson, 2019, p. 199), entangled as it is within broader, shifting discourses, practices and complex relations of power at school. Biesta and Tedder's (2006) understanding of agency is the achievement of an outcome 'where individuals are able to exert control over and give direction to the course of their lives' (p. 9) that is temporarily and relationally situated. As such, agency is not restricted to individual or group capacity, and 'consent can never be freely or independently given' (Loveridge and Cornforth, 2014, p. 465). This is markedly different from an understanding in the participatory research literature, identified by Gallagher and Gallagher (2008), of children as 'coherent, knowing, autonomous beings. . . imbued with agency' (p. 502). Instead, children's competency to consent is developed through social interactions; to understand the possibilities and limits of children's consent, researchers must pay attention to how 'power relations shape, mediate and transform research relationships' (Holland et al., 2010, p. 363). For the child, this includes how they are situated within the broader context of the school, in which they are subject to relationships with peers, parents and teachers, as well as institutional hierarchies (Gallagher et al., 2010).
This focus on how children are embedded within their wider social and cultural contexts requires the type of reflexivity that attends to the researcher's own positionality-and how this is reconfigured throughout the research (Atkinson, 2019)-as well as the positions available for younger participants to take up. This includes how the researcher and their techniques might contribute to producing conforming governable subjects (Loveridge and Cornforth, 2014, p. 469). Social relations of the classroom may constrain children's agency in giving consent, and the research process may support or further limit possibilities to consent, while recognising that children's agency is enacted in multiple ways. A commitment to promoting children's consent requires us 'to ask what kind of agency they have, how they obtain and exercise it, how context shapes it, and how their agency relates to others' (Abebe, 2019, p. 6).
This article draws on a study that asked where and how children achieve agency in the primary classroom, plus the conditions that support or limit their agency (Kirby, 2020). As the fieldwork and analysis advanced, I also became increasingly aware of children's agency in navigating research dissent, in ways I had not foreseen. Below I document how I aimed to support children's informed consent, before examining three different ways children indicated their dissent without giving an explicit refusal. I highlight how paying close attention to the children and myself in the research 4 P. Kirby encounter, as well as the wider school context, helps to identify constraints to children's consent and amplifies their dissent.

Methodology: disrupting educational conformity?
In this section, I provide a brief background to the research design, and then examine the extent to which the research approach disrupted a conforming educational discourse to support children's on-going consent.

Background to the study
For the study, I conducted a multimodal ethnography of Year 1 classrooms in the south-east of England, in two schools within predominantly middle-class catchments, where the majority of parents were educated to degree level. I undertook research over the course of a year at Daleview Primary, 1 rated 'good' by the Office for Standards in Education (Ofsted). The research consisted of several phases: 1. I observed the whole class over several weeks, from the end of Reception through to their first weeks of the Year 1 autumn term. 2. I interviewed 29 children individually. This included using drawing and feeling faces to explore classroom experiences, and asking children to help a teddy bear to undertake a puzzle in which bricks must be joined up, to explore their orientation to challenge. 3. I shadowed six carefully selected core children for a minimum of two days each, including a visit to their homes, during which time I conducted informal interviews using a range of multimodal methods (e.g. drawing, photography, walking, talking, crafting, feeling faces). Discussions focused on their classroom experiences and how they understood themselves as 'learners'. 4. All children in the class were included in group activities and discussions at four points across the year, using multimodal methods. 5. Afterwards, I undertook a week-long 'rapid ethnography' in another Year 1 class at Clifftop Primary, rated as 'outstanding'. This used a similar ethnographic approach but focused on detailing a few aspects of the classroom and children's responses that emerged from the initial in-depth study (e.g. pupil talk). I began by observing the whole class, before interviewing children individually (13) and then conducting group interviews (22 children).
I used multimodal methods that emphasise children's agency and competence, supporting them to recall, bring to consciousness, articulate and reflect on their emotions and thoughts, allowing ideas to emerge. The chosen activities and discussions focused on the research topic but also responded to what children expressed they wanted to do and talk about, as well as my growing understanding of what might be of personal interest and relevance to their classroom experiences.
The core participants were carefully selected to ensure a diverse mix of gender, social and ethnic background, current academic achievement, birth order and children's agentic orientations. The aim was to include cases that were information rich, rather than representative of the school catchment; selection was based on school Children's research consent and dissent 5 information, as well as my observations and initial interviews. I also interviewed 10 staff across both schools to understand school priorities and the purposes of learning activities. I interviewed the parents of the core children at Daleview to understand their social and cultural background.
In this article, my analytical process is impressionist, and includes reflexivity as an ethnographic tool. I focus on observable behaviour and participants' voices, interpreting possible symbolic benefits and deeply entrenched structures of belief where participants may never be fully aware of them. This is done through understanding the school cultural context and the forces that produce and condition different practices and feelings (Mitchell and Elwood, 2012). My analysis includes what children say, but also their silences that demand an attentiveness (Mazzei, 2009) and offer opportunities to look beyond surface meanings, possibly telling us more than children's words (Spyrou, 2011).

Disrupting an educational rapport
In this section, I examine the extent to which my research practices disrupt (or reinforce) an 'educational rapport' (David et al., 2001, p. 363), which is shaped by discourses, practices and materiality that frequently emphasise children's conformity and their place in existing hierarchies/structures in school. I focus on how I negotiate consent, my positionality and interview approach.
Negotiating consent. Opt-out parental consent is used for all children in the two classes to allow for greater inclusion of diverse groups (Boddy, 2013), and because whole-class involvement is limited and confined to the types of activities used in the classroom (e.g. drawing, puzzles). The fieldwork was conducted before the introduction of the 2018 General Data Protection Regulation (GDPR) (Information Commissioner's Office, 2018) that requires opt-in consent, allowing for greater parental control; instead, the school, acting in locus parentis, provides this agreement. For the six key informants more fully involved in a range of research activities, opt-in consent is sought using forms for parents and children to sign. All adult and child participants receive information leaflets stressing that children do not have to participate and can withdraw at any time without consequences. The children's leaflet includes a similar instructional question and answer format to that used by David et al. (2001), together with an invitation 'to explore with me what helps us to learn?', reiterating rather than escaping an educational discourse. I repeatedly explained the research, including consent, to children using a fold-out booklet, a hand-drawn version of the type of books used by Pyle and Danniels (2016); the children come to know this well, but discrepancies in whether we achieve a shared understanding are generally 'hard to spot' (Gallagher et al., 2010, p. 479).
I involve all the children in making hand-printed consent cards, informed by methods employed by Thompson (n.d.): a red hand on one side, together with the word STOP, and a green hand on the other, with the word GO. This offers an enjoyable activity that helps to build my relationship with the children, and offers an opportunity to explain and practice consent. The cards appear to work well when I ask children if they would like to be interviewed-I place the card with the red side face up, and the children are invited to leave it or to turn it over to green-and only one child refuses. I use the cards when asking if I can observe children engaged in table activities in Daleview Reception, where children are frequently engaged in child-led activities, and in the Year 1 Clifftop class where I spend only a week and do not worry about being a disruptive presence; several children indicate refusals. In Daleview's Year 1 classroom, however, I am initially concerned with finding a way of being least disruptive to the teaching staff, given that I am to be there over the year. I come to sit at the side of the carpet or on a chair set slightly back when children work at tables, and I do not feel able to interrupt activities to ask for consent.
I also discover that traffic-light colours are part of the educational discourse, so wonder whether the consent cards might (sometimes) generate a different effect and affect than I had intended. For example, there is a traffic-light poster: green reads 'I can do this' (with a smiley face); amber reads 'I'm getting there' (neutral face); and red reads 'I need help' (sad face). James, one of the six key children that I shadow, understands red to signal threat and anger: after being given a warning by a teacher, he tells me he 'hates' his classroom and dislikes teachers because 'they tell me off'. He picks up a small piece of red paper telling me it is 'the warning RED', saying the word loudly.
Researcher positionality. During the research, I attempt to position myself with the children in part as 'least adult' (Mandell, 1988), but particularly 'least teacher' (Dunne et al., 2005, p. 62), aimed at reducing social hierarchies which may offer greater possibilities for children to withdraw consent. For example, I use my first name and sit alongside the children, emphasising that 'I'm not a teacher'. I say to the children that I will not tell others (children, carers or teachers) what they say, to let them know that they can say things that are not allowed or might feel negative about school. I also stress that they do not have to say anything, and I explain the limits to confidentiality in the case of safeguarding concerns. I am a 'sort of adult': one unusual for school, who asks children their views, laughs and plays, and rarely reprimands (James, 1999, p. 103). There is a limit to how much I can escape teacher subject positions in this context (Atkinson, 2019), and I come to understand my position more as the 'adult who least knows'. I introduce myself as a 'learner', showing pictures of 'my teachers' (doctoral supervisors) and the university, and invite the children to investigate 'what helps us to learn'. I work hard to demonstrate my lack of knowledge about the school cultural and social context, following Mayall's (2000) emphasis on the researcher's lack in relation to what children know. When children ask for help with something to do with the expectations of the classroom, generally I explain that I am also unsure, asking 'what do we do when we don't know?' I respond to requests for help in English and maths, but must tread carefully not to provide an answer, knowing the school expectation that children find out independently, while not wanting to sound like a teacher by telling them to work it out for themselves. In the interviews, I also invite the children to help a teddy bear to understand how he might experience the classroom, to support them to distance difficult feelings; he is less knowing because I ensure he does not observe lessons.
Interview approach. Spyrou (2011) identifies that a 'structured and highly controlled space of the school encouraged children to provide the "correct answer"' (p. 155), in Children's research consent and dissent 7 other words their conformity to meet with adult expectations, compared with fieldwork undertaken in neighbourhood playgrounds. I used types of research activities (e.g. drawing, puzzles, teddy bears) that are also used in classrooms, but below I explore how I aimed for my interviews to be different from children's classroom experience. I also examine the limits of what it is possible to disrupt when conducting research in a school setting.
Firstly, I make the space comfortable with a multi-coloured rug and cushions, and include various playful and creative materials. Children volunteer that they enjoy this space, although a more relaxed and intimate spatial context raises 'moral dilemmas of the exploitative potential of such interviews', in which participants may feel lulled into opening up (Sin, 2005, p. 289). In both schools, my interview space is in view of passing staff and children-a recessed room opening onto a corridor and a glassed room off the classroom-which is important for safeguarding, but also acts as a reminder that we are engaged in an 'interview' located in and about school.
Secondly, I move away from using predetermined categories or prompts, which might suggest a 'right' response (Katsiaficas et al., 2011)-the initial puzzle activity was an exception which I realised felt too test-like. I use open multimodal methods that emphasise sensory, embodied and emotional dimensions of experience, and that offer children a degree of choice about what they want to talk about and to do. The interviews become 'conversations with a purpose' (Kvale, 1995), usually undertaken informally while immersed in crafting or playful activities. There is a difference in how I conduct the research at home and at school, raising a tension between my desire to distance but benefit from an educational rapport. When I visit children in their homes, I observe and engage in conversation and play, led by the children's agendas who show me their out-of-school interests and family life. When interviewing in the school environment, I adopt a directive/authoritative researcher role, one that fits more with how adult/child relationships and activities are framed in this context. In these interviews, I probe more explicitly the children's experience of the classroom, watching closely for signs where children need to move on, pause or stop, while also staying with some discomfort. I am firstly a researcher, guided by an ethic of on-going consent as well as minimising harm and making sure activities are not too difficult, but also-and particularly in a school context-I am in part pedagogue: encouraging effort when inviting the children to make meaning of their classroom experiences: 'Meaning is the work of the will. This is the secret of universal teaching' (Ranci ere, 1991, p. 56). This is a difficult role to navigate. When interviewing James, his 'soon' below suggests that my continued probing provokes difficult feelings that he works to manage in the interview, as he does similarly in the classroom.

Children's research dissent
In a context where attendance is compulsory and staff do not seek children's consent about their participation in school activities, navigating research consent in a way that honours children's agency becomes complex. When I tell a teacher, at Daleview, how I plan to explain consent to the children, she instantly responds: 'In this school, we teach the children it's never okay to say "No" to teachers'. This lack of consent includes examples related to their bodies, such as the frequent instruction to wear coats outside without being asked if their active bodies feel cold. At Clifftop, I thank children for coming along to a group discussion. Kate tells me: 'I always do what people ask' and I explain that she does not have to; she replies: 'I like to. . . (pause) sometimes'.
The complexity of research consent under such conditions is worth exploring further, in order to support researchers to reflect on what we might attend to in the field and beyond. Most children agree to participate in the research, but I observe three distinct types of dissent, through which children demonstrate their agency in navigating conformity. In this section, I demonstrate young children's awareness of the cultural norms of indicating refusals, how they avoid participation through their 'unnoticeability' and engage in playful dissent.
Children's normative dissent Kitzinger and Frith (1999) examine the complexity of dissent within conversational interactions: how 'refusals are awkward to perform, and that (polite) rejections are often done inexplicitly' (p. 298). Most of the children demonstrate an implicit understanding of these norms, by avoiding saying 'No' in class and in their interviews with me. Children use the cards to indicate their consent to being interviewed but once in an interview, despite my encouragement, they do not use the cards to pause or stop the interview. Instead, they indicate their dissent in other ways: they become distracted, look away and during the puzzle some deviate from the rules, playfully making pieces 'fly' into place or telling me the puzzle is complete when it is not. It can be hard to read their signals, with some children working at the puzzle beyond 10 minutes, without signs of dissent, but clearly relieved when I ask whether the teddy bear would like to stop. One child, Kitty, terminates the interview, but rather than saying she wants to stop, she adopts a more culturally sensitive and normative approach: 'Well, I really want to go back to the classroom'. Kitzinger and Frith highlight the strategies for dissent used by adults and young people: delays in response to questions using pauses and fillers (e.g. 'uh', 'well'), palliatives (e.g. token agreements) and accounts (i.e. explanations/excuses). Alfie illustrates young children's expertise in each of these.
Pause, palliative and account.

PK:
We haven't got a lot of time, would you, are you going to want to do the castle puzzle or not? Alfie: Mm, I might in a little bit, I'm just going to write my name and then I might.
Palliative and account.
PK: I was going to ask you, would you draw your classroom? Alfie: Yes. PK: That would be great. Alfie: (I want to draw. . .) PK: Can you draw your classroom, there's pencils as well? Alfie: I want to draw a picture of the beach. (Interview transcript) The few children who explicitly decline research activities are currently low achieving in school and demonstrate anxiety about their 'performance' in research activities. Having noticed how Nikita complies and does not question staff even when given contradictory instructions, I am surprised when in the interview she suddenly announces 'That's enough'. She feels unconfident in these activities, telling me 'No, I can't do it' when I suggest we do the puzzle. She vocalises the desire to avoid challenging activities, whereas in class, I observe children's more covert coping strategies: for example, missing out pages when expected to read a whole book. In both schools, only the boys diagnosed with autism repeatedly say 'No' in class. When a teaching assistant tells Preston, 'Now you've got me, isn't that nice?', he says 'No', which prompts the response, 'I'm going to be upset now', signalling a cultural unacceptance of dissent and being honest when it might offend and, more particularly, the normative pupil as one who does not contradict staff. Preston's honesty-something staff rather confusingly also advocate-deviates so far here from the normative pupil-subject that perhaps, as Youdell (2006) suggests, it creates for staff a 'psychic disturbance' that must be 'censured' (p. 128).

Children's unnoticeable dissent
During my fieldwork, occasionally, I unintentionally exclude certain children, alerting me to how being good is more than simply a way of getting noticed; it also achieves 'unnoticeability', a form of anonymity that offers safety from discipline (Waksler, 1991, p. 110). In the classroom, children adopt 'good' subject positions, in which they sit quietly and still; this position is most frequently rewarded by staff as it serves the dominant classroom pedagogy, emphasising that children listen to and acquire teacher knowledge. Roma, currently in lower maths and English sets, is a child who frequently sits quietly with her finger on her lips, and I do not notice her until the second week of fieldwork: I'm sitting in the Reception classroom and overhear the teacher talking to a girl named Roma. Roma? I have spent over a week in this classroom and never have I heard anyone refer to Roma; I look at her face, but have no memory of her at all. (Field notes) I continue to misplace Roma and it feels I am not alone in keeping her at the margins. A woman serving in the dinner hall initially misses Roma as she stands waiting for her food. A teacher who works with her regularly is 'not quite sure how I would describe her as a learner at the moment. . . [I'm] still getting to know her, I think', and another, who has not worked with Roma for a few months, cannot remember her name: 'I can't even 10 P. Kirby remember the quiet ones now'. I begin to understand that children such as Roma perfect the art of avoiding challenge by appearing to be 'on-task', to mask a lack of understanding and engagement in challenging learning activities, which does not serve them educationally. By extension, they can absent themselves from the research process. Unnoticeability helps to make the classroom liveable, but it limits children to signalling a preference to absent themselves. Several months after leaving the field, when I am finishing off transcribing, I realise that I failed to invite Anna to an individual interview. I cannot easily explain how this happened, as I had a list of all the children that I checked against, and yet, knowing how I misplaced Roma, I am unsurprised the child I exclude is Anna. She is another child who sits quietly, although she submits herself to the discipline of the classroom and is currently achieving highly. When I speak to her in a focus group, she is sometimes reticent to express a view in relation to questions where there are not right or wrong answers, similar to some others who are high performing in class, saying only 'I don't know'. Anna is exercising her right not to speak, which must be respected, but it is important to foreground how this might become a preferred option in a context that emphasises success through demonstrating the acquisition of knowledge and meeting expectations for quiet. Not commenting keeps her out of trouble when there are no clues to the expected answers to my questions about her experiences, feelings and opinions. This resonates with other research that concludes children perceived as high ability do not question a teacher's definition of knowledge, or make sense of it in their own terms, contributing 'in large measure to their educational achievement' (Keddie, 1971, p. 156).

Children's playful dissent
During individual interviews, we usually find an easy balance between engaging in the research agenda and having time to craft, play and talk about things not clearly relevant. Although not with James. He is clear he wants to spend time with me (rather than in the classroom), but wants to play rather than answer my questions. He also repeatedly challenges my authority, for example by drawing on a cushion even when I tell him to stop. James's ambivalence towards the research is troubling, as I do not want to force him to do anything that he does not want to do. I repeatedly communicate that children can return to class should they wish to discontinue their participation, but am unsure what to do when James wants neither to return nor to engage in research activities. I consider whether to let the play evolve to see what emerges, similar to the work of Varvantakis and Nolas (2019), who are frequently 'recruited' (p. 6) into children's play when interviewing in their homes. I feel that I owe it to the research, with its specific classroom focus, as well as to other children who would like to be included in the core group, to stress we cannot only play (cf. Holland et al., 2010). I am also working under structural constraints that limit when and for how long I can take a child out of the class, so I want to ensure our time together includes a clear focus on the research agenda.
In a school context, my interview activities are what Bernstein (1996) would call weakly framed, resulting in an 'invisible pedagogic practice' (p. 28). The instructional discourse is implicit, as I ask James to reconstruct rather than reproduce knowledge, and I use an ambiguous regulative discourse, telling him what I would like him to do/ Children's research consent and dissent 11 talk about (adult directed) while suggesting he has choices (child directed). I sense James is trying to work out the rules, so I try to clarify these. In the interviews, I attempt to negotiate a 'deal' with James, where we can both 'talk and play', one that needs repeated reiteration as he continues to resist the talking part; 'why does it has to always talk?' James frequently tells me 'I want to work with you all day'. Once, frustrated that I could not find a way to engage him, I answer 'but I ask you lots of questions'. His response is telling, 'yeah, and I listen'. His words are a creative form of resistance; he is agreeing with me rather than pushing back, signalling I'm playing your game to a degree. The school emphasis is on children listening to adults; James does not fully appreciate perhaps that it is me who wants to listen to him. While I am an adult, I am somehow different from others in school; James is working out how to be with me, as well as me with him. These interviews-which take place on different days over several months-become a process of constant negotiation, rapidly shifting between different activities, and they require a more intense on-going reflection than with other children. They highlight the non-binary of consent/dissent, with James very clearly assenting to spending time with me while shifting rapidly along the assent/dissent spectrum in response to different research questions/activities. Our interviews foreground how power is invoked in this research relationship, with both of us trying to give and get something in return. James is finding ways to 'act beyond the limits prescribed' by my participatory techniques (Gallagher and Gallagher, 2008, p. 507), while succeeding in maintaining the benefits of the interview compared with that afforded by his social position in the classroom (Holland et al., 2010).
James occasionally makes it explicit that 'I don't really like talking', but more often his dissent is signalled playfully: James recruits the toy here to avoid having to talk, upending my attempts to use the bear to encourage children to be more open, a case in which the interview for him is becoming more 'interrogation' (McWilliam et al., 2009, p. 70). His little laugh is infused with emotion, as he avoids speaking of adult anger. James likes to make people laugh and, on different occasions, he tells me he lives in Spain, has a swimming pool, went in a pond with a king and queen, and comes to school in a helicopter. Here his narrated history 'moves away from the "real"or rather it pretends to escape the present circumstances: "once upon a time there was. . ." ' (de Certeau, 1984, p. 79), in a context where James is in the currently lowest achieving groups in his class and is also amongst the minority as a working-class child in this school. During an individual support session, when I am observing James reading a book aloud, he pauses to ask 12 P. Kirby me if I would 'go on' the dinosaur in an illustration. I say 'No', and the teacher adds 'I would only if he was buried in the mud'. James tells her, 'I'd put you in the mud and bury you'. James's humour, similar to that of the children researched by Nolas et al. (2018) who laughed at the researchers' expense, establishes an 'inversion of hierarchies', demonstrating an 'agential power by debasing adult positions' (p. 11). During an interview, James finds some coloured counters and creates his own version of Snap for us to play; he instructs me in 'the rules of the game' that allow him to be 'good at this game' and to win every time.
In the group interviews, children also frequently resist my participatory activities; at times they hide behind chairs, write on the whiteboard, play with toys on shelves and make repeated irritating noises, 'doob, doob, doob'. In interviews with the 37 children, across both schools, their definitions of 'learning' never include mention of movement or talking, only seven refer to art and one mentions play. Unused to playing and making sounds in the classroom, my expectation that the children do so while also keeping focused on the research agenda is confusing; I have become positioned as a type of teacher/play worker with a low personal rule frame, offering an opportunity for children to be 'otherwise'. Through their everyday inventiveness, the children are engaged in what de Certeau (1984) might call 'tactical' manoeuvres, that 'accept the chance offerings of the moment' (p. 30), as a way of 'making do' in the classroom (Gallagher, 2008;Atkinson, 2019). Thinking with Ranci ere (1991), we might add that their dissenting play and humour identifies their political knowingness: they challenge the perceptual and epistemic order of the classroom, acting in these moments with the presumption of equality to move, to play, to laugh and to talk at will.
Children are competent participants in the school culture, whereas I have an ambivalent relationship with it. I am what Lave and Wenger (1991) call a 'legitimate peripheral participant'. Legitimate in the sense of wanting a stake in it, so that the children participate in my group sessions and do not draw staff attention to my inability to keep the children quiet and focused. While also peripheral, not wanting anything to do with it, and in particular not to jeopardise my mutually trusting relationship with the children, which is integral to consensual research. When James continually fails to follow the ground rules we had agreed in a group session, proving disruptive and getting into conflict with Alfie, gently but firmly I tell him he can go back to the classroom if he feels unable to follow the rules today. When he becomes visibly upset at this idea, with his face hidden in his folded arms, I find myself backing down; I am worried his apparent sadness and anger will compromise a relationship I have spent time building, while feeling I am failing the group. Alfie later confirms his anger that James remained, drawing my attention to my prioritisation of one child over another. I feel stuck in this 'liminal space', where 'the implicit order that structures normal conduct does not apply' (Motzkau and Clinch, 2017, p. 271), seeing no way here to juggle a requirement for conforming behaviour, participatory methods and the assent of all children.

Discussion
The values of research consent are in conflict with a schooling emphasising conformity. Introducing the possibility of choice to participate in school-based research and Children's research consent and dissent 13 expecting children to speak an unequivocal refusal to participate is counter-productive both to everyday conversational patterns and to school discourses and practices that demand children comply. This article demonstrates that even young children understand and reflect a normative cultural avoidance of saying the word 'No'. Below, I discuss the implications for researchers working in educational settings.

Recognising limits to consent procedures
A first consideration is to expand existing research guidance to recognise the limits of consent procedures in school settings. BERA (2018) recommends carefully considering situations in which potential participants may not be in a social position to give unrestrained consent, illustrating with examples where the researcher is the participant's teacher or family member. This article suggests there is a need to extend this guidance to include all (young) pupils in English schools; it is not enough to 'reassure' pupils that 'non-participation is acceptable' (BERA, 2018, p. 14) in a context where there exist deeply embodied structures of belief about the requirement to conform.
It remains important that we continue to seek children's informed consent, as well as to observe assent. Further research innovation may support children communicating informed decisions about the extent to which they want to participate in research, while recognising that there is no simple 'right' technique (Gallagher et al., 2010, p. 479). The hand-printed cards went some way to offering a means of expressing a refusal, but their use was limited in part by my own pressure to conform to classroom requirements. Consent procedures are aimed at managing risk. Participants are invited to weigh up the relevant benefits and risks of their involvement, and researchers mitigate risks to themselves, for example by requiring written rather than verbal consent. Managing risk might be appropriate for what can be foreseen, but can 'result in constructing a sense of security' where the process and outcomes are not known; an ethical stance instead requires an 'attention to unknowns' (Scoones, 2019, p. 9). Variability in expressing consent is integral to research relationships, rather than a problem to be overcome; a more fluid, uncertain research approach opens up possibilities for dissent to surface.

Attending to children's multiple dissent
Close attention is needed to how researchers recognise and respond ethically to children's multiple practices of dissent. In this article, I identify three examples of children's dissent in a school-based study: the variability and shifting nature of consent/ dissent demands the researcher's on-going reflexive attention to the relational and situational encounter.
There is not always a clear demarcation between consent and dissent as children, like researchers, are working to give and get something in return, so that the research process involves 'considering, and constantly re-negotiating, various forms of expanded consent' (Loveridge and Cornforth, 2014, p. 469). This demands researcher openness to the confusion of not knowing how to act, together with a willingness to respond to what emerges. Attention must also extend beyond the research encounter itself: my late discovery of Roma, the recognition that I failed to interview 14 P. Kirby Anna, an emerging analysis that children maximise moments to play in interviews, all shine light back onto moments of children's practices of dissent.
A tension exists in deciding when examples of children's disruption might (ethically) be included as data or excluded on the basis that they are engaged in acts of dissent. By defining the rules of research participation as both playing and talking, I potentially constrain possibilities for how children might act, and I concur with Gallagher and Gallagher (2008) that fascinating insights emerge from 'children acting in unexpected ways: appropriating, resisting or manipulating our research techniques for their own purposes' (p. 508). The authors suggest that ethnographically, these become instances of 'potential data', instead of being viewed negatively as 'instances of non-compliance' (p. 508). The challenge of navigating between dissent where children redefine the boundaries of their research participation (i.e. what it is possible for them to do and to be) and dissent that signals a withdrawal of consent to participate. This uncertainty requires an acceptance and openness to our researcher vulnerability, supported by an 'immature attitude of creative experimentation' (Gallagher and Gallagher, 2008, pp. 512-513), bringing us closer to the children who are also finding their way, more than having confidence in our knowledgeable superiority in the complex process of consent.
Children's agency is bound up with school and wider relations, so becoming familiar with different discursive levels and material relations through which researchers and children create meanings around consent is important (Spyrou, 2011, p. 159). Dissent is embodied, affective, and negotiated moment to moment, and only sometimes involves verbalisation, raising the importance of how children utilise available materiality (e.g. consent cards, toys, chairs to hide behind), including their own bodies (remaining so still as to be unnoticeable). Additional attention is needed to the ways in which particular children may be vulnerable to different aspects of the research-including how their responses are circumscribed by factors such as current educational achievement and gender-resulting in diverse practices of dissent.

Promoting participation rights in educational settings
It is important that researchers model consent practices and that we use research as a means to disrupt problematic power structures in educational settings. Working with schools, we must collectively draw attention to the culture in which children cannot consent. This calls for a re-engagement with educational purpose and attention to how participation rights are integral to children's protection: opportunities to make informed choices help to keep children and young people safe in research, schools and beyond (Whittington, 2019). At Clifftop, I go into the assembly hall to invite a girl to be interviewed; she agrees to come with me, until a teacher tells me she is not in the research class. Afterwards, I express my embarrassment to the teacher who laughingly voices my concern: 'I'm glad we teach them so well about talking to strangers, "Yeah, I'll come along".' provoking conversations and valuable comments on an earlier draft. Thank you to the children, parents and staff who shared with me their thoughts, feelings and ideas about the research focus and process.

Funding information
This research was supported by the Economic and Social Research Council (ES/ J500173/1).

Ethical guidelines
The research was approved by the University of Sussex's Social Sciences & Arts Cross-School Research Ethics Committee (C-REC).

Conflict of interest
The author has no conflicts of interest to disclose.

Data availability statement
The data is not open access to ensure anonymity and confidentiality given the complexity of de-identification.