Exploring internal conversations to understand the experience of young adults transitioning out of care

Margaret Archer’s model of reflexivity suggests that it is our internal conversations, those conversations we have within our own minds, that enable us to decide how we act in relation to the constraints and enablements of social structure to realise a meaningful life. We use the concept of internal conversations to explore the experience of six care leavers. In-depth interviews were carried out, using the interview structure proposed by Archer, to elicit discussion of internal conversations. The research had three related aims: explore the relevance of Archer’s model of reflexivity to care leavers; explore how they make sense of planning and the future; and consider how care leavers make sense of their family and social relationships. There was considerable diversity in internal conversations and in how useful they were perceived to be; We suggest that previous trauma may impact the reflexivity of care leavers and how they engage in planning. Eight themes are discussed including: planning; early trauma and future orientation; making sense of themselves and their past; and the complexity of social relationships. The plurality of attitudes to future planning is of relevance to how care leavers are supported, particularly to pathway or transition planning. ARTICLE HISTORY Received 6 April 2018 Accepted 15 July 2019


Introduction
For the year ending 31 March 2018, 9200 children ceased to be in care in England as they turned 18 (Department of Education 2018). The role of local authority care and support beyond the age of 18 has been the subject of recent political debate as awareness grows of the vulnerability of care leavers and poor outcomes in adult life. In 2015 the National Audit Office noted that: in higher education and poor mental health amongst care leavers (Pryce, Napolitano, and Samuels 2017;Geenan and Powers 2007;Jones, Everson Hock, and Papaioannou 2011).
In the U.K. some positive steps have been made to improve support for care leavers. The Care Leaver Strategy (HM Government 2013) and the resultant Staying Put Duty (HM Government 2014) which enabled care leavers to remain with their foster carers until they were 21 (or 25 if they were in fulltime education) are examples.

Research on care leavers
Growing awareness about the vulnerability of care leavers has resulted in increased research; with a diversification in the research methods used to study the experience of care leavers (see Harder et al. 2011). A quantitative international literature review by Gypen et al. (2017) concluded that whilst there was variation in the experience of care leavers, many struggled with regards to employment, income, housing, health, substance abuse and criminal involvement compared to young people who had not been in care. Two literature reviews of qualitative research, Parry and Weatherhead (2014) and Häggman-Laitilaa, Salokekkilab and Karti (2019), demonstrated considerable variation in the ability of care leavers to negotiate the transition to adulthood. Häggman-Laitilaa, Salokekkilab and Karti (2019) identified two main groups of care leaversthose that saw the transition to adulthood as a new beginning and those who perceived it as a negative change in their life. The first of these experiences was associated with greater self-confidence, hope for the future and carving out a new identity whilst the later was associated with a felt sense of insecurity, social isolation and confusion. Parry and Weatherhead (2014) did not divide the experience of care leavers in the same way but reflected on similar issues through their themes of navigation and resilience; the psychological impact of survival; and complex relationships. Both reviews noted that the complexity of relationships, particularly with birth families, caused stress and anxiety and made negotiating a new adult identity more challenging.
How to support care leavers more effectively is a key question that arises from the research reported above and was taken up by Pryce, Napolitano, and Samuels (2017) who explored the help seeking process in care leavers. They noted that just at the time in their lives when social connection is so crucial, the development of self-reliant identities linked to past trauma might cause them to disconnect from the relationships they seek (Samuels and Pryce 2008). Support provision must therefore take into account the impacts of historic trauma if they are to be effective. The importance of this was also noted by Gypen et al. (2017) who noted that a mentor or supportive figure to support the young person as they aged out of care was an important factor in determining more positive outcomes. Lee and Berrick (2014) explored the potential of identity capital theory to help us understand the differential outcomes for youth aging out of care, highlighting how this model can capture the challenges of adult identity development for this group of young people. Utilising the same model Webb et al. (2017) explored how care leavers were affected by engaging in volunteering. Their conclusions suggest that such opportunities helped care leavers address the key challenges summarised by the reviews cited above, particularly building social capital, increasing personal resilience and developing identity capital.
A further theoretical approach which has been utilised in care leaver's research is Margaret Archer's model of reflexivity (2000,2003,2007,2012). Hung and Appleton (2016) utilised this approach to explore the experience of young people transitioning out of care in London, UK. They found that Archer's modes of reflexivity did not adequately describe the reflexivity demonstrated by the participants but that exploring the internal conversation of participants was helpful in understanding how young people made meaning in relation to the challenges they faced and how they acted towards those challenges. In particular, they noted the participants' reluctance to plan and how this impinged upon their transition into independence.
Previous literature on care leavers has therefore painted a rich picture of the experience of transitioning into adulthood. Now, in the pursuit of supporting care leavers more effectively, researchers are utilising theoretical frameworks to understand how care leavers shape and are shaped by their social economic and environmental context. In this paper we continue this work presenting the results from a small-scale study with care leavers, utilising a novel methodology informed by Archer's model of reflexivity and building in particular on the work of Hung and Appleton (2016). We start by briefly outlining why our theoretical framework is particularly pertinent for care leavers and how the methodology developed.

Internal conversations and care leavers
Archer's theory of reflexivity considers internal conversations as a mechanism that mediates between structure and agency; a way in which individuals are able to make sense of (or not) the social structures in which they live and their ability (or lack thereof) to engage as active agents in the creation of their adult lives: the regular exercise of the mental ability, shared by all normal people, to consider themselves in relation to their (social) contexts and vice versa. Such deliberations are important since they form the basis upon which people determine their future courses of actionalways fallibly and always under their own descriptions. (Archer 2007, 4) For Archer (2007) internal conversations allow individuals to identify their primary 'concerns' those areas of one's life that are personally most important, and then decide how to act upon these concerns in pursuit of 'projects' (important, ongoing activity, e.g. studying for exams, getting married, finding a job). Archer describes how a 'modus vivendi' is established when satisfying and sustainable 'projects' that align with ones 'concerns', can be effectively dovetailed with one another alongside 'an intelligent, though fallible, interaction with those constraints and enablements which are activated during the pursuit of our concerns' (Archer 2003, 150).
As a group known to experience multiple developmental risks and trauma, ongoing social and economic disadvantage (Pryce, Napolitano, and Samuels 2017); and to now be subject to growing policy interest regarding appropriate support provision, care leavers are a key group whose experience may be better understood using Archer's model of reflexivity. The complexity of their situation belies easy understanding but by providing a window into the internal processing of individuals, we are able to capture, at least to some degree, how care leavers try and make sense of their circumstances and express agency given the web of constraints and enablements in which they are embedded. When considering reflexivity in relation to care leavers it is important not to lose sight of their particular vulnerabilities and context. Archer notes that in 'the human experience we are both free and constrained' (Archer 1995, 29). We suggest that care leavers are not only more bound by structural impediments than other less vulnerable social groups, but will also, inevitably have much greater demands placed on their meaning-making capacities (Hung and Appleton 2016).
Archer draws out the difference between 'active' and 'passive' agency. She places considerable importance on the ability of participants to articulate their 'concerns' because, put simply, active agency is not possible if they do not know what they wish to achieve (Archer 2003, 304). The relevance of this point to care leavers is drawn out by Hung and Appleton (2016) who note that despite the focus on life-planning activities by agencies and professionals working with care leavers, some care leavers themselves consider planning an 'anathema' and that the focus on 'pathway planning' in the U.K. is 'lost in translation' for some young people.
Lastly, Archer (2003Archer ( , 2012) goes on to identify four 'modes' of reflexivity: communicative reflexivity; autonomous reflexivity; meta-reflexivity; and fractured reflexivity (see Archer 2012 for a detailed description of these modes). She associated the first three modes with active agency and fractured reflexivity with passive agency noting that fractured reflexives are unable to hold an internal dialogue 'about themselves in relation to their circumstances, which has any efficacy' (Archer 2003, 298). Whilst we are interested in the reflexivity of care leavers, we are reluctant to adopt Archer's typology, or working out where care leavers 'fit' within Archer's modes of reflexivity. This has been explored Hung and Appleton (2016) who found that Archer's 'modes' of reflexivity did not reflect the complexity of how internal conversations impacted upon the agency of care leavers. However, they still found value in the exploration of internal conversations, particularly the explicit exploration of 'what matters' in the lives of participants, as this deepened understanding of how care leavers expressed (or did not express) agency.
As well as the difficulty with the typology of reflexivity Archer's theory of reflexivity has been criticised for minimising the role that social structures play in determining action (Caetano 2015, 63). This is particularly pertinent when researching a population who have already been determined as vulnerable and marginalised and who are significantly disadvantaged compared to other social groups. The risk being that the analysis of their situation becomes overly individualised and the complexity of how structure impinges upon their agency is lost. Caetano (2015, 66) notes that for Archer 'social structures are only present in internal conversations when individuals evaluate them in articulation with their main concerns. The causal efficacy of structures is not directly felt upon subjectivity itself, but on the result of that subjectivity … ' Additionally, Chernilo (2017) building on Archer (2007, 93-99), argues that the way reflexivity is exercised is important because it affects social mobility. There is however, an absence of discussion regarding how social structure affects capacity for, and the nature of, reflexivity. Her theoretical emphasis is on reflexivity as a property of individuals which is autonomous from structural or cultural properties (Archer 2007).
Relatedly, King (2010, 257) recognises that the structure-agency dichotomy on which Archer's theory depends is not compatible with modern network based theories, noting how sociologists 'could start to understand social reality not in terms of an individual confronting a pre-formed structure but in terms of multiple participants negotiating as they interact with and co-operate or struggle with each other'. Vandenberghe (2005, 233) makes a similar point when he notes that 'her theory of the internal conversation is too much of an internal conversation' and that she underplays the importance of inter-subjectivity in the production of internal conversation.
Despite these critiques and the importance of considering contextual and relational factors in the lives of care leavers, we still consider internal conversations helpful in understanding care leaver experience. We have chosen not to use the types of reflexivity Archer identifies as the framework for our analysis because we feel that this would not help us deepen our understanding of care leaver experience. This decision also reflects the findings of Hung and Appleton (2016) who questioned the relevance of that typology to care leavers. Instead, we have chosen to use the internal conversations of care leavers as a way of understanding how they experience and make sense of the transition to adulthood. As such, we have utilised the interview suggested by Archer, and in line with her theory, conceptualised internal conversations as being a mechanism through which the individual interprets and makes sense of / responds to social structure. This 'lens' of analysis has allowed us to explore how participants make sense of and express agency in response to their life circumstances, through the experiences and internal dialogues reported by them. We see reflexivity and the resultant internal conversations, as emergent from the individual in social context. As such, intersubjective interaction within their lives and in the interview itself, will frame the types, themes and content of internal conversations.

Research focus
Our aim in this small-scale research project was to use the internal conversations model to explore how care leavers come to understand, interpret and then act (or not act) in relation to their current context and past experiences. We have three key objectives in doing this: (1) Explore the reflexivity of care leavers as they leave care and how their internal conversations impinge upon their transition into independence.
(2) Examine how care leavers make sense of planning and future orientation, both day-today and long term, and how their future orientation is revealed in their internal conversations and discussion of life projects and concerns. (3) Consider how the participants make sense of their family and social relationships, and what this might suggest about the role of social networks in supporting their transition to independence.
In the discussion section of the paper, we consider what the implications of our findings are for the services that are provided for care leavers.

Sampling and recruitment
We carried out interviews with 6 care leavers all of whom were receiving support from a leaving care team. We recognise this small sample is a limitation of this research but consider it to be appropriate for a study aimed at testing methods of data collection and analysis, which had not been brought together before. Furthermore, the analysis of this study is informed by Interpretive Phenomenological Analysis, which emphasises the need for depth of understanding of individual cases. As such small samples are often a feature of this type of research (Smith, Flowers, and Larkin 2009;Smith 2011).
Initially we aimed to sample randomly from the list of young people supported by the service, however, this was not a viable approach due to difficulty with contacting potential participants and low take up. After a period of slow recruitment, we asked support workers to approach young people they felt would be willing to take part. We realise that this approach meant that the most marginalised young people were unlikely to be recruited into the study. All participants considered themselves to be White British, reflecting the low level of ethnic diversity within the county where 95.2% of the population identified as White British in 2015 (Suffolk County Council 2015) ( Table 1).

Data collection
The data collection process had three elements, completed over two interviews (except for participant 5 who completed all three components in one interview): Interview one: (1) In-depth, semi-structured, interview which focused on their internal conversations (2) Completion of a qualitative social network map Interview two: (3) Completion of The Adult Self-Report (ASR: Achenbach, Dumenci and Rescorla 2003) mental health measure with additional qualitative exploration. The data from this section of the research are not included in the analysis presented here.
The semi-structured interview guide was informed by Archer (2003, 161-162). It introduces the concept of internal conversations, asks specifically about internal conversations that they experience before asking broader questions on life concerns and projects. Participants responded well to the interview, were able to relate to the concept and understood what was being referred to. This section of the interview typically took between 25 and 40 min.
Then the participant was asked to complete a social network map using participatory mapping methodology (for more details see Emmel 2008). There was no requirement to feature certain relations in the mapindividuals added were at the discretion of the participant. All participants chose to have the researcher write the names on the paper but actively picked colours and provided direction about where they wanted the name placed on the paper. The first three participants were asked to think about the people who supported them in their life and consider where they wanted to place them on the map. The word 'support' was thought to be limiting the scope of people being added to the maps. Therefore, the last three participants were asked to think about people who were in their lives and then place them on the map where they thought was appropriate. Whilst we cannot be sure this had an effect, the researcher felt that it facilitated a more open conversation and that the maps better reflected the lived complexity of their networks. We did not seek to gain information about the care history of the participants at any point in the data collection. We wanted to focus on their internal conversations and social networks as they wished to discuss them at the time of the interview. We discuss the implications of this in the discussion.
One participant completed the interviews at a cafe, a second in their flat, while interviews with the remaining four participants took place at the base of the leaving care services.

Data analysis
Our analysis process as informed by that utilised by Hung and Appleton in their study of the internal conversations of care leavers (2015). Our primary qualitative method of analysis was Interpretative Phenomenological Analysis (IPR) (Smith, Flowers, and Larkin 2009;Smith 2011). We utilised this approach due to the attention it gives to understanding experience and meaning-making (Smith 2018) of individuals, before moving on to look at similarities and differences between the participants. However, we did not stick rigidly to the process outlined by Smith, Flowers, and Larkin (2009) but embraced the flexibility that they strongly encourage, informed by Archer's theory of internal conversation, but not determined by her typology. The analysis process involved three researchers (the three authors) one of whom had carried out the interviews. All the researchers reviewed the transcripts and each created a case summary for every participant. These were then discussed as a group. The discussion was recorded, allowing it to be revisited. The final themes and grouping of codes for the group of participants was then decided through discussion and consensus by the researchers including printing out codes and physically grouping them under themes. A detailed account of our analysis is being developed for publication.

Results
We report our findings as a set of eight linked and overlapping themes.
The first three themes: reflexivity and looking to the future; planning; early trauma and future orientation, are all related to thinking about the future. We look at variations in reflexivity, planning capacity and attitude to planning, and how participants interpret the impact of earlier life experiences on future orientation. In effect all of these are aspects of planning but we separate them for analytical clarity and to reflect how they were spoken about by the care leavers.
The fourth theme, internal conversations as unhelpful or ineffectual, highlights that for some care leavers internal conversations could be experienced as unhelpful or even harmful.
The fifth theme, making sense of themselves and their past, considers how internal conversations became a site of meaning making, as they try to develop an understanding of themselves and their past as they move forward into adulthood.
The last three themes all relate to the overarching complexity of social relationships for this group and how this was expressed and worked through in internal conversation. In particular, we explore the lure and risk of the birth family; loneliness, support and socialising; and romantic relationships.

Reflexivity and looking to the future
We observed interesting differences in the reflexivity of the respondents. Matt found it difficult to draw out what was important to him or what he wanted in the future. When asked about his main concern in life he replied 'Her pretty much' nodding towards his girlfriend who was present in the interview. Yet when asked to complete the social networking task he did not want to add anyone onto it and was defensively independent stating 'I talk to myself and myself only really'.
Lacey's internal conversations were rich but were often about big scientific or philosophical questions. There was no evidence of internal conversations informing her actions and she reports being reactive. She also reported feeling misunderstood by those around her so her internal conversations were not validated when she tried to discuss them with others -' … it makes sense to me. But it doesn't to anyone else'. Similarly to Matt, her consideration of what she wanted from life was very limited and her primary concern was her romantic relationship but she notes that this sometimes came at the expense of caring for herself: 'cos some things matter, like my other half, I make sure he's safe and that, make sure he is fed, he's clothed. Things like that so … I tend to put people before myself, I am too kind'.
The other four participants in the study demonstrated greater consideration of the future and greater reflexivity than Matt and Lacey, either through their description of life projects or expression of values that were guiding their current actions. Layla, and particularly Rosie, both seeking sought clarity in discussion with others before taking action. Both of them reported rich internal conversations about their future, for example Layla said: ' … I think about families and stuff … What job I am going to have … Um, finishing my, going back to college … Getting my Cs in English and Maths … Um, going to university … Ah, becoming a play therapist … '.
Layla was pregnant at the time of the interview, but she did not reveal this until the second interview. She reported rich internal conversations about her future and intended career (play therapy). Rosie talked about her reliance on others for decision making. Her dependence on others to validate her internal conversation possibly reflected an unwillingness of her part to take responsibility describing how 'It is very well known that I am rubbish at making decisions. So I seem to just ask everyone else's opinion'.
Both Layla and Rosie demonstrated some degree of insight into their own situations and ways of behaving and thinking. Rosie was aware that she constantly sought out the advice of others, could be lazy at times and needed encouragement from others to get motivated. Layla reflected on how counselling over the previous 12 months had been very useful to her. She simply stated 'I have come a long way' and was able to recognise her own progress.
By contrast, Finley and Luke were more autonomous, with a stronger sense of independence but did consult others when problems arose. This is captured well in this quote from Luke: 'I will make a decision off my own back, I won't really discuss it … Unless I, I feel like I need someone else's opinion … or guidance'.
Both showed concern for the future and a strong degree of intentionality behind their actions. For Finley, his life project was quite vague but it centred on getting somewhere in life and emphasised the importance of work and friendship: Growing up and being an adult … It has really made me change and really made me think well actually, hang on a second, I need to sort myself out and otherwise I am going to be the same as most people … And not get anywhere in life Luke's life project was to join the Army, but when discussing this he became hesitant. It seemed as though he was not quite able to make sense of why he would not take steps towards achieving this, yet in the confusion there is evidence of insight into the underlying reason: What I would like to do is um join the army … um I always put it off … OK and in a way like I can't be bothered to do it this day, I sound kind of lazy, I don't mean to, it's like, it's a dream I've always wanted to do … But I just don't have the confidence Luke also described his wish for a family and a mortgagesomething he describes as 'the traditional sort of settle down'. He was the only respondent to be so explicit about wanting a family of his own although Layla did allude to this.

Planning
None of the participants said that they were good planners. Some participants actively rejected the notion that they planned for things. For example, Lacey stated ' … I don't really plan for the future. Cos, take it day by day basically' and there was little evidence that she did plan.
For some participants there was a disjuncture between what they reported in terms of whether or not they considered themselves to be a planner, and what they evidenced when they talked about how they lived. For example, Rosie demonstrated a strong ability to plan by having held down a place at college, a weekend job and arranged work experience. However, her poor ability to budget and her constant battle to motivate herself to attend college illustrated an interesting tension between present day desires and future aspirations that was prominent throughout her interview: … the whole, do I go to college, do I not and also do I really want to go to uni, what if I don't cope and what if I end up can't hacking it and I am left with loads of debt … it is mostly to do with future I guess.
Finley was able to describe very practical ways in which he planned, such as going to bed early when he had an early start and he described a particular way he remembered things: I had to put my case forward to keep this flat. Because I am in rent arrears. Um, so to help me, I had to write down a, a whole list of things, the reason why I was going to keep this flat. Um, and I done it. And they said I could keep my flat Yet he also reported being bad at planning. Luke had a sense of what he wanted from his future but still reported being focused on the present and resistant to actively planning for the future: I take day by day see what happens. Yeah, my life has been so up and down. One minute I could be really happy, in a job, and the next thing I could get asked to leave and then really down when that happens.
Despite different approaches to planning being demonstrated in their actions, there was an almost universal reluctance to plan for the future or to at least admit engaging in it in the context of a research interview.

Earlier trauma and future orientation
One of the key themes to emerge from 5 out of the 6 participant's (with Luke being the exception) was that harmful earlier life experiences were interpreted as having had an effect upon their willingness to think about the future. The future was seen as risky and unknowable and therefore planning was considered futile: 'Cos the thing is, I used to make a lot of plans but they always blew up in my face so I don't make plans, the further I make plans is maybe a day, two days ahead if that' (Matt).
Matt, on the same topic, also raises the issue of trust in others and the importance of having reliable people: Cos things, well especially (inaudible) other people's lives but things, when we make plans they pop up and pisses it up so I thought there is no point … Unless I know that person is really, really reliable The sense of powerlessness concerning future events and the sense of vulnerability to misfortune were most pronounced for Lacey who actively imagines a variety of potentially bad scenarios: … you don't know what is going to happen in the future, you can plan something but it may not ever happen. So, like, let's say for instance I was, basically say 10 years' time, I could be dead! Or I could be, or I may not be able to see or I may not be able to hear.
She also remembers having planned in the past but ' … have just given up on it basically'. At a later point in the interview she also remarks that she could fall pregnant at any time. There is little sense of the potential of agency on her part despite that fact that when she was younger she had had dreams of working with and riding horses.

Internal conversations as unhelpful or ineffectual
Matt, Lacey, Layla, Rosie and Luke all noted that to some degree their internal conversations could be unhelpful, even harmful at times. Lacey suggested that her internal conversations were occasionally harmful because they could be about death. For the others, the unhelpfulness came from the possibility of overthinking things and the mental chatter that might arise from this (Matt, Rosie and Luke). For example, Matt stated 'There's always things' going round in my head' and Rosie described how ' … the more I think about it [her future], the more it gets confused'.
Furthermore, it was significant that the two least actively agential participants -Matt and Laceynoted how although they may sometimes engage in internal conversations, they do not actually impact upon how they then act. Of communicating with others Matt said 'I talk quicker than what my brain can keep up sometimes'. However, he did not seem overly concerned about this. By contrast, Lacey expressed a strong sense of guilt and shame when she felt that she had upset someone: I tend to say whatever is in my head basically … I try to … hold it in and not say it but … it sort of just comes out … sometimes I feel really guilty … Or sometimes upset, depressed, suicidal sometimes … Cos I, like things I say I don't mean to say it. Furthermore, although she engages in internal conversation, she expressed difficulty with getting other people to understand it, causing further frustration and highlighting a gap between her internal dialogue and her ability to use that as the basis for the expression of agency.

Making sense of themselves and their past
A clear distinction emerged from the data between those participants who actively distanced themselves from their birth families and created a sense of self by positioning themselves as 'different' from their birth families, and those who still sought a sense of self as belong to their birth family.
Matt stood out from the other participants in his strong identification of himself with his birth family despite not having much contact with them. When discussing planning he notes how he has learned that behaviour from his Mum: 'No, just don't, never have done, like my mum, don't plan, she just do what needed to do when she needed to do it'.
In contrast, Rosie and Luke actively sought to distance themselves from who their parents were. Rosie said ' … Me and my parents are completely different, so it sort of makes me want to sort of step away. It makes me want to be completely different' and Luke similarly emphasises how different he is: 'Cos she [his mother] kept telling me oh you're just like your mum and I'm like I am nothing like you at the end of the day, I am nothing, nothing like her'.
Another theme that arose within the sphere of 'making sense' was coming to terms with and finding out the truth about why they went into care. Matt stood out as being actively dismissive of the past: 'Don't dwell, what's in the past is in the past. I don't dwell on it'. Finley in contrast talks about wanting to know the truth: Cos I really want to find out what happened, what went on … I will never know the truth. Everyone has got their stories. Um, so I doubt I will ever know the truth. But I would love to know the truth Finley's interview started with a powerful recollection of an internal conversation that he had had when he was much younger, about being moved from a home into a new foster placement and being unhappy with the decision to move him but being unable to express that. It powerfully evoked the sense of powerlessness and frustration that he felt at the time: I could say it now because it happened. But at the time … I couldn't say it because … It was, I didn't want to hurt no-one's feelings. I didn't want to cause trouble between me and the new carers. I just wanted to be settled. And it didn't happen.
The placement ended very quickly after he threatened violence, believing that to be the only way he would be able to effect change. It is notable that, whereas Luke and Rosie seemed to be more at peace about having been in care and reported largely positive experiences, this was different for Finley meaning his attempts to make sense of the past were more live and emotionally difficult.

Complexity of social relationships
The lure and risk of the birth family There was considerable diversity as to the current state of the participants' relationships with their birth parents, with occasional dissonance between how the participants ranked these relationships during completion of the ASR and how they described the relationship in the interview. This was particularly evident for Lacey who described her relationship with her mum and her dad as both 'better than average' in the ASR but then describes how her mum 'doesn't want to meet up' and that she does not really have contact with her Dad.
Despite having no contact with her parents, Layla remained very fearful of them. Sometimes she thought she would 'see' or 'hear' them, particularly when out in town. She was very anxious about them getting involved in her life and ruining it. She expresses her desire to 'move on from some things' and says ' … it would be good if they don't get in the way'. Whilst she had developed a strong sense of what she wanted in her life, she still experienced her parents as having considerable power to take it away from her. In particular, she fears that her parents will try and ensure that her child is taken away from her: 'they will do anything to take my child away from me … Because they think I am the bad person basically … Telling me I have ruined everything so they will probably do it in return'.
Although not as acute, Finley also mentions avoiding contact with his family so that they cannot have a negative effect on him: I think family is a massive thing as well … They can really pull you down … Um, down to their level with all the crap … cos I just go there and shit kicks off … And then I get blamed for it … Um, um And I can't be bothered with it.
Rosie had taken on something of a parenting role to her own parents stating, 'she [her mother] has a lot of her own issues and so she just gets a bit too much and ends up me being the parent rather than her'. She also noted how she would prioritise her father's needs over her own and how her mother does not want her to go to university because this would involve her moving away. In this way, Rosie's realisation of active agency is constrained or at least impinged upon by her parents through either explicit statements or her response to their need for care as she perceives it.
Whilst it is beyond the scope of this article to consider the relationships of all 6 participants with their birth families, it was evident that the birth families were considered problematic for each participant, albeit in different ways. Of interest in this study is how the parents of several participants actively impinged upon their actions in the social world either through fear (Layla), an enduring sense of duty of care (Rosie) or a desire to be accepted and cared for by them (Lacey). Archer's oversight of the structural influences on reflexivity and agency, are apparent here: we cannot understand the agency of a young person to re-establish contact or avoid contact on the power of their reflexivity alone. Nor can we try and understand the meaning of parental contact without considering how their social experience has impinged upon their reflexivity.
Loneliness, support and socialising Fear of losing connection or having insufficient social connection is raised in several of the interviews and needs to be explored if we are to understand how care leavers experience their social networks. It is most evident for Luke who describes getting a cat and looking for a shared house as way of reducing his isolation 'At the end of the day, I have now got a cat, it's lovely, I actually come home to something, you know what I mean'. Although he has friends, he has a sense of lack regarding close relationships.
Finley, whose interview often mentions the importance of a social life and the different ways that he has gone about building one, indicated that he still does not feel secure in this regard. This was particularly evident when asked about his most pertinent concerns to which he replied 'Where I am going to be … living in 5 years' time. Who I am going to have round me. Cos I don't really speak to my family. So I don't really have hardly anyone'.
Matt stood out from other participants as he seems to be trying to create an identity which is based on being independent from others, and does not report being lonely or fearing the loss of social connection in the same way as other participants. There is a high level of defensiveness in his view of the world and he places great emphasis on how others should respect him: Right, as long as people show me respect, I show them respect, the moment they start disrespecting me then I start getting pissed and then I start losing my temper, that is how I am, I show you respect but if you don't show me respect, if you treat me like a sack of shit, as that is just the way I am.
When discussing the support they get from different people in their lives Rosie, Finley and Luke made clear distinctions between the people to whom they could go to for help or support and those that they considered being primarily for socialising with. Although the social networking task was interesting due to seeing whom participants identified, it was the conversations that the mapping evoked that were most valuable. The presence of names on the maps could mask considerable isolation and loneliness if taken at face value, as well as overlooking how difficult it might be for some care leavers to create meaningful social connections.

Romantic relationships
Four of the participants (Matt, Lacey, Layla and Finley) reported being in a romantic relationship at the time of interview. Matt had a girlfriend who had moved in with him after she experienced trouble at home, Lacey reported being engaged to her boyfriend, Layla was pregnant and in a relationship with the father of her baby and Finley was in a two week relationship with another care leaver who he had met on a tenancy course. It was evident that these relationships quickly became important for them and became the focus of life projects.
Matt brought his girlfriend, who had also been in care, along to the interview and there were a few jokes between them during the interview, although overall, she just sat and listened. He revealed that being in relationship affected his decision making, more so that being on a suspended sentence: 'If it wasn't with her, no I wouldn't, I wouldn't give a shit if I weren't with her but now I have obviously got to think about her'.
The importance of the relationship also became clear when he was asked about what was most important in his life at the moment and he responded: 'her pretty much … when she doesn't drive me up the wall, but yeah, you do that quite a bit don't you?' (to which she laughed but did not comment). The references to this romantic relationship stood in stark contrast to the sense of isolation and defiant independence that was apparent in much of his other dialogue either about his family, friendships or other sources of support.
Lacey also mentioned wanting to take care of the day-to-day needs of her fiancée. Given her reference to abusive friendships and family relationships earlier in the interview, it is possible that the relationship with her fiancé also has similar characteristics although this was not clear. The importance of the romantic relationships was also indicated in Lacey's interview where she directly expressed her fears of the relationship ending: ' … you don't know how it is going to go, so, you've got to take it day by day, you don't know if in the future you are going to split up'.
Layla reported that much of her internal conversation was about her boyfriendimagining how things were going to be between them both in the short term and whether they will be together in the long term. Uniquely, Layla also reports having considerable help and support from her boyfriend's family. A breakdown of that relationship would therefore not only represent a loss of a relationship with him but potentially a wider network, thus magnifying its importance.
Rosie mentioned splitting up with her boyfriend but did not seem overly concerned and was the only participant for whom this did not seem a key interest. Luke, although not currently in a relationship, expressed a desire to be in one and openly explored how his previous girlfriend had accused him of trying to find a 'mother figure' when she ended the relationship, which he admitted was probably true. Wanting to belong and start a family were central life projects.
The importance of romantic relationships amongst the group may simply reflect the age group being interviewed. However, we felt that the intensity of these relationships and the aspect of caregiving were different for this age group more typicallybecoming family rather than romantic partners more rapidly than amongst a less vulnerable population. However, romantic relationships can be a source of vulnerability, and as such, the role of romantic relationships within the social networks of care leavers warrants further exploration.

Discussion and conclusion
From making sense of day-to-day issues such as rent and bills, to how to manage relationships with their families and questioning who they are and who they wish to be, the internal conversations of the care leavers we interviewed were diverse, complex and rich. Yet, for some of the young people internal conversations could be unhelpful and distressing.
Similarly, to Hung and Appleton (2016), we found that the internal conversations and meaning-making of the participants are powerfully shaped by their experiences of trauma and disappointment, which may inhibit their sense of themselves as active agents in the world. Internal conversations were not always supportive of active agency amongst the participants (Munford and Sanders 2015), meaning that a day-to-day perspective tends to dominate and there is a reluctance to make plans. This was the case even where participants did have a sense of important life projects. However, in some cases, participants have shown an ability to reflect on their own situation and respond with creativity and determination despite having conflicting aspects of themselves that make them reluctant to plan and think about the future.
This study has suggested that Archer's model of reflexivity can be useful in its conceptualisation of internal conversations as mediating between structure and agency, helping us to understand how care leavers make sense of and react to their circumstances. The interview in particular, with its explicit focus on internal dialogue and 'what matters', provides an interesting lens through which to look at how care leavers negotiate this transition. Rather than focusing on types of internal conversations (see page 6), this approach develops our understanding of how care leavers try to make sense of their situation within their own minds and how internal conversations might assist or impinge upon their expression of active agency as they transition to an independent adult life. This perspective is thus a valuable addition to the theoretical approaches used in care leaver research because it enables structure to be explored through the lens of internal experience and dialogue, helping to overcome the criticism of Archer's model as 'too internal' and generating understanding about why care leavers are differently affected by their circumstances and show different propensities for planning for the future.
This study suggests that the provision of support for care leavers can only be done effectively if their capacity as an active agent is understood (Munford and Sanders 2015). Given our findings in this study, and those of Hung and Appleton (2016), we suggest that planning activities carried out with care leavers, such as pathway planning (HM Government 2016, 16), may not be useful for care leavers unless delivered with additional support to help them develop planning capacity. This could include helping them to develop reflexivity and educating them about internal conversations, although it is recognised that the unstable context of their lives which may have undermined their capacity to see themselves or planning as effectual will not be easily overcome.
The main limitations of this study are the small sample size and the sampling process that favoured inclusion of less marginalised, more engaged care leavers. We would recommend that future, larger studies using the internal conversations model with care leavers would be useful to increase our understanding of reflexivity amongst care leavers and the potential implications for future support provision. Additionally, in this study no attempt was made to create a full biographical account of the participant's real lives. Collecting the care histories of participants in future studies, potentially through support workers or case notes to avoid re-traumatisation, could help to provide context for the interpretation of internal conversations. This would inform our understanding of how internal conversations of individuals are impacted by their circumstances and therefore how they can best be supported to develop active agency that can support their transition into independence. We also recognise that whilst we have focused on care leavers, the transition to adulthood is recognised as being problematic for many young people so these issues may well have wider applicability.
This study has refined the use of Archer's model of reflexivity in care leaver research to better understand how the agency of care leavers is affected by structure through the lens of internal experience and dialogue. In so doing it has highlighted potential issues with the support provided for carer leavers, particularly where the focus is on future planning. Whilst this study is too small to reach firm conclusions in this regard, we believe the findings are sufficient to recommend this as an insightful theoretical approach worthy of further attention.