‘I make a lot of the choices myself— I think I've taught myself that through the imbalance of support’: The internal conversations, reflexivity and post-school educational achievement of care-experienced young people

The percentage of care-experienced young people in England progressing to university by the age of 19 currently stands at around 12– 13% with a further 10% of care-experienced adults attending university during their 20s and 30s. This figure remains lower than both the general population and other groups of disadvantaged learners. It is well established that the educational attainment of care-experienced young people can be impacted by a range of barriers to learning. Existing research often focuses on the importance of key adults and their role in supporting young people in care. A small number of studies examine the reflexive capacities of the young people themselves through the lens of sociologist Margaret Archer's model of modes of reflexivity and internal conversations. Archer's theory can be utilised to understand how care-experienced young people navigate their circumstances. The notion of the internal conversation offers a way to understand how some young people growing up in care develop more stable modes of reflexivity, namely autonomous, communicative or meta-reflexive. Here we contribute to new knowledge by considering care-experienced


INTRODUCTION
Approximately 82,170 young people were in care at the end of 2022 in England (Department for Education, 2022aEducation, , 2022b. It is established that academic and personal outcomes for those young people leaving care remain of concern (Department for Education, 2013; Dickie, 2021). Recent data suggest that 13% of care leavers attend university by the age of 19 which despite remaining static since 2017, represents an improvement from the previously reported 6% (Department for Education, 2022aEducation, , 2022bHarrison, 2020). In addition to this, it is worth noting two factors which suggest that the numbers may be higher: firstly that many care-experienced adults attend university in their 20s and 30s and therefore fall outside of the DfE's figures; and secondly that the DfE identifies care leavers as those young people who have been in care for 12 months at the end of year 11 (the final year of secondary school), which excludes many young people who entered care at 14 or 15 (Harrison, 2020). Despite these improved figures and additional considerations, it remains likely that the percentage of care-experienced undergraduates is still significantly below the young people who develop communicative and autonomous aspects to their day-to-day life functioning. This paper draws on findings and analysis from interviews conducted as part of the first author's PhD (2020) which considered the reflexive capacities of care-experienced young people who self-identified as higher achievers. We utilise Archer's modes of reflexivity to explore participants' internal conversations and to develop our understanding of the relationships, experiences and personal skills that underpin successful educational journeys. number of young people not in care who go on to higher education, which stands at around 45% (Department for Education, 2022aEducation, , 2022b. Of particular interest to this paper are the care-experienced students who do progress roughly in line with young people who are not care-experienced. The importance and impact of care placements during the ages 16-21 will be considered. However this paper will largely focus on how care-experienced young people engage with internal conversations, the interplay between care placement and the development of coherent modes of reflexivity. Utilising Margaret Archer's notion of the internal conversation and associated modes of reflexivity we seek to understand how some care leavers, interviewed between 2015 and 2018 (Matchett, 2020) developed a sense of agency which enabled higher levels of academic and personal, process-oriented, resilience (Hines et al., 2005).

THE IMPACT OF PL ACEMENTS
It is important to acknowledge the range of placements young people experience and the impact these placements may have on the development of a coherent internal conversation and related educational plans. Sebba et al. (2015) conducted a large-scale mixed methods study examining aspects of care which contribute to educational attainment. In an innovative research design, for the quantitative arm of the study, children in care were compared with those 'in need' (a much larger population than those in care) and those not in need and not in care. Among many interesting findings, data showed that children entering care before the age of 11 who experienced stable and longer placements did better-in terms of educational outcome-than children who entered care later. Foster care and kinship care are associated with lower levels of placement change than residential care: the lower the level of change, the better the educational outcome.
Continuity of high-quality relationships plays a key role in supporting the academic attainment of children in care. Sebba et al. found that children who had attained at a higher level had often received ongoing educational support from birth families. In addition to support from birth families, relationships with teachers are also understood as crucial to academic success (Comfort, 2007;Gilling, 2014;Harker et al., 2004;Jackson & McParlin, 2006;Rees & Munro, 2019;Sebba et al., 2015;Selwyn & Briheim-Crookall, 2017;Sugden, 2013). Teachers are regarded as prime motivators in terms of educational success who often act as mentors for children who have experienced adversity (Comfort, 2007;Sebba et al., 2015;Sugden, 2013). Remaining in a stable care placement reduces the likelihood of changing school, thereby increasing opportunities for stable relationships with teachers. Sebba et al.'s work challenges the notion that foster care poses an inherent risk to academic success by demonstrating that longer term foster care can be beneficial. However too many children do not experience the type of care placement necessary to flourish. The older a child becomes, the harder they are to place with a foster family with those entering age after the age of 11 experiencing the least stability (Rees et al., 2011). Of England's 82,170 children in care, around 39% of those young people entering care do so between the ages of 10 and 15 with an additional 25% aged over 16 (Department for Education, 2022a, 2022b). The former Children's Commissioner for England, Anne Longfield (2019), reported that older children are six times more likely than those under 13 to live in residential or secure children's homes with nearly half living in the privately run accommodation which has recently come under public scrutiny. Only a small percentage of young people in care experience the longer-term stability that sees them through GCSEs, A Levels (or equivalent) and university.

PL ACEMENTS FOR OLDER TEENAGERS
The Staying Put policy (Department for Education, 2014) seeks to address some of these issues by allowing young people the opportunity to: Remain living with their foster carer/s (former foster carer/s) after their eighteenth birthday on a familial basis, having already lived with, or as, a member of the foster carer/s family over a period of time (Department for Education, 2014, p. 4) The policy allows young people to remain in their foster placement until the age of 21 or after the age of 21 to complete an educational course. However, approximately just 9% of care leavers remain in the Staying Put agreement with their foster carers after the age of 19 (Department for Education, 2022a, 2022b) suggesting that more development and evaluation of this policy is needed. A far higher percentage of young people over the age of 16 find themselves in unregulated care, which can range from self-contained flats and hostels to caravans and even tents. The recent Children's Commissioner Report (2020) identified that 66% of 16 and 17 year olds had experienced some time in an unregulated placement in 2018-19. The report also stated that 651 children aged 14 or 15 had been in an unregulated placement during 2018/19 which compares with 273 children in 2012/13.

INTERNAL CONVERSATIONS AND MODES OF REFLE XI VIT Y
This article now explores the reflexive capacities and related internal conversations of a small sample of care-experienced young people from a variety of care backgrounds who have achieved in line with peers who are not care-experienced. We draw on Archer's (2000Archer's ( , 2003Archer's ( , 2007Archer's ( , 2010Archer's ( , 2012 theoretical model of reflexivity to examine the position of education within internal conversations and the impact of cohesive internal conversations on young people's ability to sustain their commitment to their academic ambitions. For Archer (2007, p. 63): The key feature of reflexive inner dialogue is silently to pose questions to ourselves and to answer them, to speculate about ourselves, any aspect of our environment and, above all, about the relationship between them. (Author's italics) The internal conversation is the practice of reflexivity. Archer explains that in internal conversations we monitor and evaluate ourselves and through this process re-establish our commitments. The internal conversation also enables us to understand and mediate our role within society. We successfully or fallibly align our commitments and plans (or ultimate concerns) (Archer, 2007) to the societal enablements and constraints we understand our natal context and the wider society to hold. During internal conversations we are both subject and object. We speak to ourselves, but we also listen and question. As we question, we revise knowledge gained and consider emotional responses and so this continues until we reach a resolution or abandon the thought process. Archer suggests that this conversation is shaped by 'me, I and you ' (2010, p. 4) and develops through three stages: discernment, deliberation and dedication (Archer, 2000). When encountering a dilemma or choice we may employ discernment by first considering all available options, their merits and limitations. The deliberation stage involves questioning our motivations and potential choices. To question ourselves we evoke 'me' or our past self-our past actions and routines which form the basis of our decision making. For example, we might ask ourselves 'what did I do last time?' Past actions may be compared to our  Archer (2010) termed 'you'. 'I' is our present, questioning self-'I' is the self in reflexive mode, questioning past actions and ideas in the present context, deliberating the best course of action to move towards future goals or ultimate concerns. Through this process we arrive at dedication where we commit to a course of action and this becomes a concern or possibly an ultimate concern (Archer, 2000(Archer, , 2007. It is reasonable to suggest that young people in care experience several significant natal and societal constraints. It is important to understand how more cohesive internal conversations develop and how these might empower care-experienced young people to draw on their personal, natal and societal enablements to make academic progress which is at least in line with that of peers who are not care-experienced. Archer proposed that internal conversations occur in four main modes which vary in efficacy (Archer, 2007, p. 93): • Communicative reflexivity-individuals whose internal conversations require completion and confirmation by others before resulting in courses of action. • Autonomous reflexivity-those who sustain self-contained internal conversations, leading directly to action. They monitor their own circumstances carefully and are guided by the 'reality principle'. • Fractured reflexivity-those whose internal conversations intensify their distress and disorientation rather than leading to purposeful courses of action. • Meta-reflexivity-those who are critically reflexive about their own internal conversations and critical about effective action in society.
Of the four modes of reflexivity, we now discuss communicative and autonomous reflexivity in more depth.

COMMUNICATI VE REFLE XI VIT Y
In Archer's model, communicative reflexives tend to have experienced a relatively stable childhood, marked by high levels of natal continuity. Those operating in this mode are likely to contribute most to their local community; they value family and social continuity. They may seek a career similar to that of their parents or, Archer suggests, 'aspire lower than originally intended' (Archer, 2007, p. 166). When communicative reflexives encounter difficulties or problems they seek the advice and guidance of their closest family or social network, their priority is to reproduce their natal context.
Two participants in the present study experienced communicative reflexivity.

AUTONOMOUS REFLE XI VIT Y: T WO DIFFERENT SCENARIOS?
Archer describes the internal conversations of 'autonomous reflexives' as 'the lone exercise of a mental activity' (Archer, 2003, p. 10). For Archer, the autonomous reflexive does not need and does not want their internal dialogue to be supplemented by external exchanges with other people. Self-sufficiency of the mind, dovetailing of concerns (a personal configuration of what matters, what is important) and autonomy of decisions, goals and courses of independent action characterise the lives of autonomous reflexives. Planning for the future is something that is done independently and 'naturally' as 'part and parcel' of the autonomously reflexive approach. Archer suggests two main scenarios which during childhood can lead to autonomous reflexivity (2007, Chapter 5). The first scenario is a natal context which allows the child considerable freedom to encounter novel situations. Where these situations are positive, they can distance the child from their natal context without damaging child-parent relationships. Over time, the freedom may lead to heightened independence and a sense that their natal context does not hold all the answers to situations encountered. The second scenario resonates with participants in this paper: Archer explains that self-reliance and the associated autonomous reflexivity can be created by dysfunction whereby independence is forced upon subjects at an early age.
Of relevance to the two scenarios discussed by Archer, Hung and Appleton (2016), in a qualitative study of care-leavers in London, describe a reflexive mode closely related to autonomous reflexivity, but with high reluctance to plan for the future. This avowedly self-reliant strategy had been described in US studies of care-leavers by Kools (1999) and Samuels and Pryce (2008), and in the UK by Stein (2006), but crucially, in the London study care-experienced participants were also highly reflexive, with active and thoughtfully selfreflective use of internal conversations, and clarity of concerns (what is personally important), despite the apparent absence of future goal orientation (Appleton, 2024).
In this paper we describe several participants who experienced a range of forms of autonomous/self-reliant reflexivity and future orientation for educational goals.

FR ACTURED REFLE XI VIT Y, METAREFLE XI VIT Y AND MULTIPLE MODES OF REFLE XI VIT Y
Fractured reflexivity did not occur as a primary mode in any of the young adults in this study (but see Hung and Appleton [2016] for a discussion of this reflexivity mode among some care leavers). Meta-reflexivity, as described by Archer, did not seem to resonate with the accounts given by young people in this study (Archer's characterisation of meta-reflexivity goes beyond 'being reflexive about our own acts of reflexivity'; Archer, 2003, p. 255). Modifications to Archer's modes have been suggested and developed. Baker (2019), drawing on the work of Dyke et al. (2012) considered the use of dual or multiple modes of reflexivity and the possibility that their use could be context specific, a possibility not explored by Archer.

RESE ARCH DESIGN AND METHODS
The data presented in this paper was initially generated as part of the first author's PhD study which focused on the educational experiences and reflexive capacities of care-experienced young adults in one local authority in England (Matchett, 2020). Twenty-one young adults participated in the original study, eight of whom are considered further in this paper. Participants were recruited through a variety of networks including the virtual school, local universities and the Children in Care Council. Extensive and valuable support was also given by a Rights and Participation Officer.
The original sample contained both individual and group interviews. While the group interviews were valuable in the sense that they allowed space for discussion, and often consensus, between participants around issues such as the importance of clothing and being treated in line with peers at school, they did not generate the same in-depth discussions around internal conversations, ambitions and motivations that the individual interviews allowed. These outcomes mirror the common affordances and limitations associated with group interviews (Braun & Clarke, 2013). Therefore for this paper, we focus on individual interviews with eight care-experienced participants who identified as achieving at a high level academically and were able to articulate the internal conversations which enabled them to progress through the education system roughly in line with peers who are not careexperienced (Table 1).
It is important to acknowledge the limitation of this sample. Only one of the eight participants was male which compares with 56% of the care population; four participants described themselves as White British which compares with 75% of the care population. Additionally, three participants were in a Staying Put arrangement, which was higher than the proportion among the care population (Department for Education, 2021).
Interviews promoted flexibility, allowing participants and interviewer to focus on interesting points as and when they arose. Participants offered interpretations of their social situation, discussed key relationships and their experience of utilising internal conversations. Reflexive thematic analysis was used to analyse the original set of interviews. All interviews were transcribed verbatim, and care was taken not to formalise speech patterns or sentence structure. This process was time-consuming but incredibly helpful as listening and re-listening to the interview material allowed the start of immersion in the data (Braun & Clarke, 2006. Initial codes were identified at an early stage but revisited in an iterative cycle throughout the remaining interviews and during the literature review process. This was a largely inductive approach with the themes generated from the initial coding driving decisions around further research and reading. Returning to the interview material for this paper was more deductive in analysis orientation and more aligned to Braun & Clarke (2019) description of a 'Codebook' approach. Archer's qualitative descriptions of reflexive modes were used to code the selected interviews but commentaries were continually reviewed to consider theoretical possibilities beyond those suggested by Archer. As Braun and Clarke stated, these processes necessarily involve researcher bias and we acknowledge a pre-existing belief that more nuanced modes of reflexivity exist than Archer describes, which shaped the nature of our enquiry.
Ethical approval for the interviews was granted by the researcher's University and the local authority in 2015. Given the potential vulnerability of the participants and the sensitivity of the conversations, consideration of ethical conduct extended far beyond formal approval. The local authority's Rights and Participation Officer played a crucial role in ensuring participants understood the voluntary nature and purpose of the research and ensuring interviews were held at an appropriate and comfortable location for participants. While all the young people interviewed disclosed mental health concerns, the Rights and Participation Officer ensured no young people experiencing significant challenges at that time were asked to participate.

FINDINGS AND DISCUSSION
Participants understood education as separate and distinct from their experiences of school. Education itself could be flexible and informal, stemming from many sources: reading, peers, key adults as well as teachers. Therefore education offered numerous benefits which included, but was not limited to, achieving academic qualifications. Participants presented as primarily utilising two main modes of reflexivity: autonomous and communicative, although elements of all four modes were evident. Interestingly those participants who aligned most closely to either autonomous or communicative modes of reflexive were also making academic progress most closely in line with peers who are not care-experienced. We start by focusing on those young people adopting a communicative approach.

COMMUNICATI VE REFLE XI VIT Y
Here we consider how young people growing up in care can still develop a communicative mode of reflexivity and how this can support educational progress. Two participants, Brooke and Sal, presented as communicative reflexives with Brooke providing the clearest account. Brooke gave a clear description of support from her birth parents prior to entering care. She explained that her mother valued education and ensured that she maintained a high level of attendance at school: A lot of kids who go into care don't have good attendance-I always had good attendance. I never really had a day off. I do think that makes a massive difference-the values my mum instilled in me. If you don't have that you might just not care.
Brooke developed positive relationships within a foster placement that lasted for several years. She referred to her foster carers as foster mum and foster dad and described herself as 'one of the family'. Her foster mum was a teacher, again providing a level of continuity between pre-care and care. The importance Brooke placed on family relationships and the relative stability she experienced represent key features of communicative reflexivity. Archer (2007) suggested that continuity is key to developing a communicative mode of reflexivity. Although Brooke's relationships experienced disruption, elements of continuity (through contact and shared values) remained. She maintained regular contact with both birth parents while in foster care and the commitment to education shared by all key adults in her natal context. Brooke's narrative offers a way to understand how communicative modes of reflexivity can develop despite the natal disruptions that entering care necessitates. Brooke's experiences might serve as an example of 'linked lives' where shared values between key actors involved in Brooke's care create continuity and contribute to her sense of herself as a capable learner (Brady & Gilligan, 2020  status and employed humour to deflect any unwelcome attention, questions or concerns. She recognised that this sometimes meant that she occupied the role of 'class clown' and this led to some minor disciplinary issues. However, she recalled finding the academic content of lessons quite easy and felt more concerned about the impact her behaviours might have on her friends' learning and progress than that on her own. From Brooke's perspective, this effective use of humour allowed her to maintain a sense of social acceptance that enabled her to thrive at school. At the time of interview, Brooke had chosen to remain in the same geographical location in order to maintain relationships with family and friends; she was also planning to train as a teacher. Brooke's educational success and chosen career path did not appear to indicate a 'lowering of aspirations' described by Archer (2007) as a feature common to those operating in this mode. Indeed, Brooke provides some challenge to Archer's (2012, p. 165) suggestion that the communicative mode of reflexivity does not adequately equip subjects to negotiate between structure and agency. Brooke presented as an individual drawing on a realistic understanding of her social enablements while also enjoying strong relationships with key adults and peers. When we discuss autonomous reflexives, there is a clear sense that the absence of these strong, supportive relationships is keenly felt.
Sal also demonstrated strong elements of communicative reflexivity. She reflected on her willingness to take advice and guidance from key adults: Sal recognised that all stages of her internal conversation are influenced by her relationships with peers and adults. In the discernment stage she experimented with a variety of activities-often following her friends. Although she deliberated her choices after leaving school she was again heavily influenced. However the combination of recognising her ability in the expressive arts and the support of key adults at the children's home encouraged dedication to her talent.
Like Brooke, Sal placed great value on friendships. She also chose to be open about her care status, explaining that she did this because she valued honesty in friendships but also to ensure that friends were able to visit her home. Sal experienced more changes in placement than Brooke, including some time in a residential children's home. Again, like Brooke, Sal described using humour to manage her care identity. She recalled entertaining her friends with stories from the children's home which she admitted were often exaggerated for comedic effect.

AUTONOMOUS REFLE XI VIT Y
The remaining participants broadly aligned to a more autonomous mode of reflexivityalthough often their relationship with this mode was more complex than those more closely aligned to a communicative mode. As highlighted in an earlier section, this is a nuanced and debated mode of reflexivity and the paths taken towards this mode varied significantly as did the ease with which individual participants utilised this approach.
Caroline's interview provided the clearest evidence of autonomous reflexivity: her self-reliance, ambition and commitment to education was demonstrated throughout. Archer (2007) suggested that an autonomous mode of reflexivity develops through the successful navigation of challenges in childhood and certainly Caroline gave a clear account of educating herself independently (in the absence of support) from an early age. She described reading her brother's books and selecting educational toys as a very young child. At the time of interview Caroline was applying for university courses-she expected to study law and to move a significant geographical distance from her natal context. Perhaps most significantly Caroline was able to identify and articulate examples of her inner dialogue: There's a thought that if I stop now I could be like every other care kid and just not do anything, live the easy life but then I look at the people around me that are doing well and I think 'right, well, I'm here now and I need to keep going. I'm supposed to be here and I need to get on with it' and I get back on track.
Interviewer: And it's a conversation you have with yourself? You get yourself back on track-it isn't a case of talking to someone and them helping you get back on track? Caroline: No, I do it.
As this excerpt from Caroline's interview highlights, those operating in an autonomous mode of reflexivity-in contrast to those using a communicative reflexivity mode-do not rely on external validation. Caroline trusted her internal dialogue and, as illustrated above, this dialogue leads to direct action. During her time at school, Caroline experienced significant frustration that her determination and commitment to education (both clear autonomous traits [Archer, 2007], which did not appear to be understood by her foster carers or social workers). Meetings with social workers often occurred during lesson time and focused on communicative concerns such as friendships and health-neither of which presented as priorities (Appleton, 2024) for her at that time. In these meetings, Caroline used the opportunity to re-state her commitment to education and refused to attend some meetings which she did not perceive as aligned with her educational priorities. During her A Level studies, Caroline also had to adapt to a different role within her household as she transitioned from a foster care placement into a Staying Put arrangement. She was required to prepare her own meals and complete daily household chores. While Caroline accepted the need to develop these skills, she expressed resentment about the disparity between her capacity to focus on her studies and that of her peers.
It is interesting to note two key features: Caroline was progressing in line with peers who are not care-experienced, achieving the highest academic grades of all the participants, which resonates strongly with Archer's evaluation of autonomous reflexives; and Caroline felt her mode of reflexivity and associated ultimate concerns were not understood by the adults involved in her care. This was not the experience of Brooke or Sal, whose communicative mode of reflexivity may conceivably be shared by, or typical of, those who choose careers such as social work, teaching or foster care which are so reliant on relationships. This disconnect between potentially communicative practitioners and autonomous young people is worthy of further investigation.

AUTONOMOUS OR SELF-RELI ANT REFLE XI VIT Y?
While Caroline presented as a 'true' autonomous reflexive, other participants appeared to have a more complex relationship with this mode. The remaining participants demonstrated an ability to utilise aspects of autonomous reflexivity either in specific contexts or out of necessity. We start with Nicole, who presented as an effective autonomous | 11 'I MAKE A LOT OF THE CHOICES MYSELF-I THINK I'VE TAUGHT MYSELF THAT THROUGH THE IMBALANCE OF SUPPORT': THE INTERNAL CONVERSATIONS, REFLEXIVITY AND POST-SCHOOL EDUCATIONAL ACHIEVEMENT OF CARE-EXPERIENCED YOUNG PEOPLE reflexive but who also recognised a desire for external support. Her interview is followed by insights from Raz, Kai and Kath, who potentially regarded themselves as autonomous but whom we suggest have developed autonomous skills prematurely and through necessity. Finally, we consider Kate, whose mistrust of adults has contributed to a selfreliant or survivalist approach.
Nicole's account of her educational experiences resonated strongly with aspects of Archer's account of autonomous reflexivity. For example, education and the pursuit of academic qualifications had certainly become an enduring project; at the time of interview Nicole was a qualified nurse, had completed a Master's degree and was planning to undertake a PhD. She recalled a specific event in her teenage years which acted as a critical moment (Munford & Sanders, 2015). The 'I' and 'you' of Archer's theory (2010) is evident in this excerpt as we see Nicole engaged in an effective autonomous internal conversation to protect her future career aspirations: I always remember once when I was in the children's home and they were all going off to nick a car and they were like 'are you coming or what?' and I thought actually this could change my life so I'm not going and I didn't and they were all caught by the police. So that was a pivotal moment in my life because if I had been caught by the police and got a caution it would have been on my record because I was 16 … Everything else around me had fallen apart and that was the one thing [education] that I had routine with-so I didn't want to give up on it, it was the only thing I knew at that time.
Here Nicole monitored her circumstances and was guided by the 'reality principle'. Without family support a criminal conviction may well have spelt the end of many career aspirations. This internal conversation allowed Nicole to identify her educational prospects as her priority or, in Archer's (2007) terms, a dedication to her ultimate concern.
However, Nicole does not fully corollate even to the most nuanced interpretations of Archer's theory. Aspects of Nicole's internal conversation and educational commitment resonated strongly with Archer's description of autonomous reflexivity but could also be understood more as a learned and necessary self-reliance (Hung & Appleton, 2016). Her desire for, and appreciation of, external support could align with Baker's (2019) finding that young people may utilise different modes in different contexts or support Appleton et al.'s (2021) suggestion that some self-reliant young people engage in formal social support. Unlike the participants in Hung and Appleton's research, Nicole was able to form effective longer-term plans and she did not appear to utilise different modes in different contexts. We suggest that Nicole had reluctantly developed autonomous reflexive capacities without ever losing the desire for a communicative and trusted interlocuter. Nicole highlighted examples of professional love and care she received while at school. Her appreciation for this support is tangible: I had a really close relationship with this teacher-she just really looked out for me. She was the cooking teacher actually and she was so kind and so caring. She said 'look Nicole,' I don't know what she could see but she said 'Nicole, just come and talk to me.' So I did and she would spend twenty minutes of her lunch just talking to me … I've met her a couple of times since and I just say thank you to her every time because she just saved me in a way I suppose.
Nicole's interview illustrated the powerful impact of nurturing key adults and their role in enabling more vulnerable young people to achieve educationally. While this teacher offered valuable support, they were necessarily a temporary figure in Nicole's life and we suggest that in the absence of one constant key adult, Nicole learned to operate autonomously. This effective but possibly reluctant ability to operate in an autonomous mode suggests a difference from Archer's (2007) explanation of autonomous reflectivity, where all her interviewees reported a clear preference to act independent of others' guidance or support.
Raz and Kai also presented complex internal conversations. Their interviews captured the desire and difficulties of accepting support following a sustained period of necessary or enforced autonomy. Both Kai and Raz gave accounts of their early lives and the subsequent development of autonomous traits that resonated with Archer's explanation of natal context 'scenario two'. Like Nicole, Raz and Kai demonstrated a firm commitment to education and the pursuit of academic qualifications.
Raz believed that acquiring academic qualifications would fulfil her ultimate concern of creating a different personal identity which would effectively separate her from her natal context. She made clear statements that reflected the autonomous nature of her internal conversations: Just to reiterate that a lot of it came from me and I think I'm starting to reflect on how much has come from me, how much I've had to push myself.
Everything that happened in primary school-my mum's alcoholism-all of that-I didn't tell anyone. At secondary school nobody knew. The teachers didn't know and at secondary school they don't ask those questions. Which I think they should to be honest-but at the same time I understand there's huge amounts of pupils. But I slipped under the radar so I thought 'okay if nobody knows, I can create my own identity'. I didn't have to be an alcoholic's child, I could just be a normal child who just gets her head down and gets good grades.
Here Raz reflects on her own internal conversation and also the external constraints which contributed to the development of her autonomous priorities (Appleton, 2024). In Raz's account there are clear parallels to Baker's (2019) finding that dual modes of reflexivity are more present than Archer's theory allows. However, we return to the apparent hope of finding a trusted interlocutor within the school environment suggestive of a desire to operate in a more communicative fashion (Matchett, 2020). Where Nicole was fortunate enough to receive the external support she needed albeit temporarily, Raz was not and therefore focused on honing her autonomous or self-reliant outlook.
Kai described focusing on education at the expense of any other aspect of school life. He explained that education was his sole focus because it offered a sense of stability and reliability: Kai: It's easier to manage just one thing. Friendships and relationships they involve emotions and managing the two is hard. Interviewer: So why choose education over friendships? Kai: Because it's worth something-it's something that will last forever. Friendships you know, it could just be an acquaintance or it could be a troubling acquaintance-it could make you go off track a bit.
It is possible to interpret this choice as a strong example of autonomous reflexivity; however, throughout Kai's interview he spoke of a lack of support from social workers and foster carers. At the time of his interview, Kai felt unable to maintain relationships with his peers, which bears some similarity to Caroline's autonomous desire to preference study over friendships. Their accounts provide some challenge to Appleton's (2021) finding that self-reliant young people in care may enjoy the benefits of informal social support. It is interesting to note that while Kai detached himself from peer friendships he sought time and conversations with his teachers. Kai regarded teachers as trustworthy role-models, more reliable than the foster carers and social workers who had been involved in his care.
Kai gave a clear account of his reluctant progress towards autonomous reflexivity: Kai: I make a lot of the choices myself-I think I've taught myself that through the imbalance of support. Interviewer: Do you feel like that's something you've taught yourself since you've been 16 or is that something you've been learning for a long time? Kai: A long time, quite some time. But I think when I went into care-it's an environment where you're a lodger, you do feel more mature. I noticed that and my friend says that too-that being in care makes you more mature. I think you read people better and I think it makes you put up a guard and develop resilience. You wouldn't just open up-you can't just unclench your chest and let things happen. You're always on your guard about certain things. It makes you mature but probably not in the best way.
This sentiment was echoed by another participant, Kath, who offered a very clear account of how and why she utilised autonomous skills: Kath: I'll use my carer as an example-you know, when she asks me to do something and I don't do it straight away, she'll say 'I'll just do it myself'-it's like that. You wait so long for people to do things so in the end you just do it yourself. It kind of does make it better for you in the long run-you get more done relying on yourself.
Both Kai and Kath identified the impact of the enforced premature independence which is a common feature of growing up in care. In addition, Kai articulated the difficulty of openly communicating with others and the emotional cost of reluctantly or artificially developing an autonomous mode of reflexivity (Appleton et al., 2020;Hung & Appleton, 2016;Matchett, 2020;Samuels & Pryce, 2008). The final participant Kate had very different experiences of education and perhaps the most complex mode of reflexivity. In line with autonomous reflexives, she was committed to achieving academic qualifications, although for her, this was a slower and more challenging journey. Kate was the only participant to express mistrust of all adults involved in her care and education and a dislike of school. Kate could not recall any teachers who had offered support or a single advantage associated with attending school. This profound absence of support may have contributed to the fractured elements within Kate's internal conversations (Archer, 2007(Archer, , 2010. Here Kate describes a period in her life when she was unable to act or make progress: I wasn't motivated as much as I could have been because of everything that was going on-it was always at the back of my head … I understand that other people were stressed because of the work but the kind of stress I was having meant I was shutting off-not wanting to do anything-like nothing. However this situation appeared to be temporary as at the point of interview she was able to plan in the longer term and gave clear evidence of persevering with the education course she was studying. In line with Appleton et al.'s (2021) finding that self-reliant young people transitioning out of care may draw on the informal support of friends, Kate described a willingness to share problems with other friends who had also experienced the care system: As stated, Kate's internal conversation presented as complex. It is conceivable that she also shared elements of Samuels and Pryce's (2008) description of a 'survivalist' mode where a young person may be scared of taking opportunities or making mistakes.
It is our contention that while many of the care-experienced participants presented as strongly aligned to an autonomous mode of reflexivity, it is borne from necessity and mistrust of others. This finding, while from a limited sample, suggests that adults involved in the care of young people need to listen carefully to the external articulation of internal conversations so that support can be closely tailored to their individual modes of operation. Caroline, who presented as closely aligned to an autonomous reflexive, expressed significant frustration that key adults failed to recognise her ambitions and suggested that she should be more focused on friendships and developing domestic capabilities. It would be interesting to interview more male care leavers to evaluate any gendered expectations of reflexive modes.

CONCLUSION
Evidence from this sample of participants suggests that care-experienced young people with a more fully formed mode of reflexivity enjoyed a relatively smooth path through education, when compared with evidence from the wider literature on educational journeys of young people in care. The young people who did well academically usually found threads of stable support for their concerns and goals, were more able to develop resilience and then benefit from more fully fledged reflexivity. Although Brooke and Caroline had wholly different approaches to internal conversations, relationships and career aspirations, they both made academic progress in line with the general population, beginning their undergraduate studies at 18. All other participants described a potentially reluctant, but effective, development of autonomous and/or self-reliant skills. It is interesting to note that all these participants experienced some delays to their educational journeys ranging from changing degree subject to re-taking examinations.
The quality and stability of placements clearly impact the formation of reflexive modes. Stable, high-quality foster placements can support appropriate contact with birth families and continuity at school. Continuity has clear educational and relational benefits (Sebba et al., 2015), but also facilitates the development of a coherent mode of reflexivity through the sharing of values and a sense of trust that is crucial to communicative and 'scenario one' autonomous reflexivity.
It is clear that teachers continue to act as role-models and valued interlocuters for young people who may not have reliable or permanent access to a trusted adult within their natal context. Five participants did not feel that they had access to reliable support or that their priorities were understood by the key adults involved in their care. While respectfully acknowledging the significant demands already placed on teachers, social workers and foster carers, we would encourage all involved to listen closely to the This study contributes novel findings around the complexity of personal reflexivity among some care-experienced young people who demonstrate clear educational goals. There are also indications of how teachers and others might support both communicative and autonomous/self-reliant reflexive practice by care-experienced students. We also suggest that Archer's modes of reflexivity need to allow for greater complexity for those presenting as self-reliant/autonomous reflexives and for those developed from a 'scenario two' natal context which Archer (2007) described marked by dysfunction and related premature independence. Care-experienced young people may perceive and present themselves as autonomous but our interviews revealed a strong desire for well-attuned external support and a sense of sadness and frustration when it was not found to be available. Further research with a more diverse group of young people would help to explore this theoretical possibility in more depth.

A C K N O W L E D G E M E N T S
We would like to thank the young people who contributed to the original study and whose stories are told here. We also extend thanks to our reviewers who shared supportive and thoughtful guidance.

F U N D I N G I N F O R M AT I O N
This study is not funded.

C O N F L I C T O F I N T E R E S T S TAT E M E N T
There are no known conflicts of interest.

D ATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are openly available in [Open Access BCU] at https://www.open-access.bcu.ac.uk/id/eprin t/9895, reference number [9895].