Presenting Critical Realist Discourse Analysis as a Tool for Making Sense of Service Users’ Accounts of Their Mental Health Problems

Abstract

Making sense of service users’ accounts of their mental health problems requires a method able to deal with complexity. Yet the different underlying epistemological and ontological positions of the methods researchers use, based, for example, on biomedicine or social constructionism, produce highly partial analyses. Addressing this problem, this article offers a method of Critical Realist Discourse Analysis (CRDA) that employs a synthesized discourse analysis, informed by critical realism, to examine the discursive, material, embodied, and institutional factors that might inform how mental health service users make sense of their mental health problems and associated service use. The article describes the epistemological/ontological underpinnings of CRDA and its three-phase methodology, before showcasing the method using, as examples, two data sets from care leavers and mothers. With our CRDA, we demonstrate a method for analyzing the complexity of interacting factors informing service users’ understanding of their mental health problems.

Keywords

Critical Realist Discourse Analysis England (South) mental health childhood service users critical realism qualitative

Introduction

There are important methodological challenges for researchers seeking to understand how people who use psychological support services make sense of their mental health problems. In part, these challenges occur because mental health and illness are contested terms, and used to represent a range of concepts from psychological states to dimensions of health (Lofgren, Hewitt, & das Nair, 2015). In turn, these different concepts are underpinned by different epistemological and ontological frameworks, such as medical naturalism and social constructionism (Cromby, 2016). Medical naturalism (or psychiatric positivism/realism) holds the premise that mental illness has an external reality that can be subsequently identified by experts, while social constructionists argue that we can only know the world via the ways we represent it (Pilgrim, 2014; Roberts, 2014). These ontological and epistemological positions direct analytical attention toward some factors and not others, physiological or discursive for example, limiting the researcher’s perspective (Cromby & Harper, 2009; Pilgrim, 2014). In this article, we offer a route for expanding the researcher’s perspective with a critical realist informed methodology, applied to service users’ accounts of their mental health problems.

Critical realism (CR) is a meta-theory that uses elements from both social constructionist epistemology and realist ontology. Applying a critical realist approach means taking account of the fact that mental health problems are both socially constructed and influenced by external factors and forces that can be real and independent of any one person or social group (Bhaskar, 2014; Sayer, 2000). For example, parent well-being issues can both be a product of fatigue (e.g., Cooklin, Giallo, & Rose, 2012; Dunning & Giallo, 2012) and a consequence of societal expectations (e.g., Brady, Lowe, & Lauritzen, 2015; Sims-Schouten, 2016). Similarly, determining a person’s vulnerability and exposure to mental health risks in society and the differing psychopathological profiles of men and women across the life span can be attributed to interacting factors including hormones, discourses of gender (that in turn impact on experiences of control, power, and dominance), socioeconomic position, roles, social status, and access to resources and treatment (; Cromby, 2016). When considering how people make sense of their mental health problems, CR can, therefore, incorporate consideration of material (e.g., income), institutional (e.g., the availability of mental health services), embodied (e.g., hormonal imbalance), and discursive factors that may structure such sense-making.

This article offers a method, “Critical Realist Discourse Analysis” (CRDA), for doing research in the field of mental health. It draws on and pushes forward a range of developments in discourse analysis (DA) by providing a tool for examining the discursive, material, embodied, and institutional factors that might inform how mental health service users make sense of their mental health problems and associated use of services. Below, we describe the epistemological and ontological underpinnings of CRDA and the methodology for doing CRDA, before showcasing how to apply the method using, as examples, two data sets with mental health service users, one with young care leavers and one with mothers.¹ In doing so, we address the question “how can we better analyze, and thus understand service users’ narratives of their mental health problems?”

The Case for a CRDA

CRDA is driven by Bhaskar’s (1989, 2014) critical realist informed stratified model of reality that conceptualizes the real (exploring causal mechanisms of events, such as hormonal imbalance, trauma, and cuts to services to name a few), the empirical (experienced events, namely, how mental health problems are experienced by people), and the actual (events and processes that occur, in our exemplar case, in relation to mental health support). CRDA does not claim to identify direct causal relationships between these factors, rather it seeks to identify and explore how the real, empirical, and actual may interact in complex, iterative ways that create the conditions of possibility for sense-making. To do so, it combines CR’s retroductive reasoning, which involves making (non-linear and stratified) inferences about underlying structures and mechanisms that may account for an individual’s mental health problems with a synthesized DA, that draws on insights from two different discursive approaches (DA and discursive psychology [DP]).

The starting point of DA is that the discourses people use are culturally available repertoires that structure what they can say, think, feel, and do. Discourses, therefore, have important subjective and affective effects that, in turn, open up or close down possibilities for action (Silverman, 2010; Wetherell, 2013). When researching mental health service users’ narratives, discourse analysts identify the (common sense) discourses participants use and the wider institutional discourses that inform them (e.g., ideas from attachment theory may become part of everyday sense-making). In contrast, with its conversation analytic genealogy, DP focuses on the interactional effects of talk (Wiggins, 2017). For example, how accountability and psychological ideas such as guilt and shame are managed in talk, and analysis of which allows the researcher to see how mental health problems are negotiated in context.

Important synergies are thus produced from synthesizing different forms of DA with CR (S. C. E. Riley, 2003; Sims-Schouten & Riley, 2014). For example, while DP and applied conversation analytic methods more generally may explore how somebody negotiates identities through a close examination of language in interaction (Antaki, 2011), DA allows that talk to be further explored in the context of wider discourses that may not be explicitly oriented to in the talk, while the critical realist element of CRDA allows examination of material realities (e.g., cuts to services) that may also be informing the talk while not being explicitly articulated. In contrast to conversation analysis and DP, CRDA is a more interpretative method, yet, as we show below, in CRDA, this interpretation is done in a systematic way. By bringing analytics from DP to critical realist work, CRDA is also different from other critical realist approaches that try to contextualize respondents’ talk but which focus on structures and mechanisms rather than the interactional elements of talk (e.g., Patel & Pilgrim, 2018). Overall then, CRDA allows the researcher to draw on a wider range of data to inform in their analysis.

Considering the non-discursive and discursive together draws on the recent turn to affect in social-psychological and cultural research, by putting “both affect and discourse back where they should be within emergent patterns of situated activity, making these patterns, as they need to be, the main research focus” (Wetherell, 2013, p. 366). Such work highlights the need for methods that incorporate discursive, material, embodied, and institutional factors so that researchers can better understand how mental health service users make sense of their mental health problems, from which practitioners can better develop their support structures and services. Considering embodied and material factors is essential, given the body of work highlighting the importance of material aspects (e.g., poverty) and embodied factors (e.g., stress and trauma) in mental health problems (Brady et al., 2015; Rivett & Kelly, 2016), we also argue that this allows us to do justice to our participants, viewing them as individuals actively making sense of their experiences within the parameters set by the complex context of their own materialities, institutional and embodied experiences, and discursive resources.

The CRDA method showcased in this article significantly develops our previous work on how a critical realist approach might be successfully combined with discursive approaches to better contextualize participants’ talk (Sims-Schouten & Riley, 2014; Sims-Schouten, Riley, & Willig, 2007). In doing so, it offers a novel method to explore the relationships between subjectivity, discourse, embodiment, materiality, and institutions in relation to how people understand their mental health problems.

The Method of CRDA

To showcase CRDA and its utility across projects, in this article, we apply the method of CRDA to interview data from two separately funded studies that took place between 2014 and 2016: one with young care leavers taken from an evaluation of a life-skills program with a focus on mental health and well-being, and the other with mothers from an evaluative study of the role of children’s centers in supporting the mental health and well-being of vulnerable mothers.² Both data sets comprised of populations considered to be hard-to-reach groups for researchers (Sims-Schouten & Hayden, 2017; Savage et al., 2015; Winter, 2014). Both studies were ethically approved by a university ethics committee (REC reference number: 13/14:01 for Data Set 1; REC reference number: 13/14:15 for Data Set 2), and informed consent was obtained from all the participants and organizations involved. While the primary aim of both these studies was evaluation, a shared secondary objective was developing methodology to better interpret service users’ sense-making of their mental health problems, and it is this objective that the present article addresses. Below we briefly describe the participants in these projects, before outlining and applying the CRDA method.

Participants

A total of 32 participants were interviewed across the two data sets. Given that qualitative research projects tend to have between 10 and 30 participants (Silverman, 2010), this allowed our method to be tested on an appropriately “normal” sized sample, allowing it to be relevant for a range of research projects. The interviews were conducted with mothers in the children’s center, and with the young carers at the center providing the life-skills program. The interviews were semi-structured and were based on talk around mental health (problems and well-being) in relation to early support through children’s centers (for Data Set 1) and in relation to a life-skills project (for Data Set 2). The decision where to carry out the interviews was reached through mutual agreement with the participants and the centers. The interviews were undertaken by Sims-Schouten and two other researchers attached to the projects, and lasted between 30 and 60 minutes, with an average of around 45 minutes.

Across the data set, the participants had a range of mental health problems for which they had received counseling (see Table 1 below). Data Set 1 participants received counseling through the children’s center, either through individual sessions and/or group sessions, and Data Set 2 participants through CAMHS (Child and Adolescent Mental Health Services) when under 18 years, with half of the participants subsequently transferring to adult mental health services and the others not meeting the threshold and/or being supported through other means.

Table 1.

Participant Information and Characteristics.

	Data Set 1: Mothers (N = 10)	Data Set 2: Young Care Leavers (N = 22)
M age	32 (range = 27–36 years old)	18.5 (range = 18–25 years old)
Ethnicity	8 White British; 2 BAME	16 White British (9 male, 7 female); 6 BAME (4 female, 2 male)
Number of children	2 (mean); range of 1–3 children	Five of the participants were parents, all of whom had one child
Marital status	3 married; 5 single; 2 cohabiting	0 married; 10 relationship/cohabiting; 12 single
Clinical diagnosis	6: postnatal depression (3); comorbid postnatal depression with anorexia (1); comorbid chronic depression with generalized anxiety disorder (1); social anxiety disorder (1)	10: depression (4); anxiety disorder (3); phobia (2); ADHD (1)
Participant/center reported mental health problems, undiagnosed	4: anxiety (1), stress (1), depression(2)	12: anxiety (5) stress(5), depression(2)

Note. BAME = British, Black, Asian and minority ethnic; ADHD = attention-deficit/hyperactivity disorder.

The Three Phases of CRDA

CRDA is operationalized in three phases, allowing examination of the wider material, embodied, institutional, and discursive contexts in which the participants’ talk might be located, and explores how they may interact as causal factors that affect patterns of events. We describe this as a linear process below, although we realize that like in most qualitative research, the process is iterative, as information at one stage/phase of the project may require the researcher to return to earlier stages and re-evaluate these in light of the new information.

Phase 1: Identifying Factors That May Scaffold Participants’ Sense-Making

The aim of Phase 1 is to examine, as inclusively as possible, the wider material, embodied, institutional, and discursive context in which the participants’ sense-making might be located.

To start this phase, a thorough and focused review of relevant literature, including research and policy documents, was conducted from as wide a perspective as possible, allowing us to identify the most common reoccurring elements of embodiment, institutions (and institutional and public discourse), and materiality that might impact on mental health and mental health problems of the participants.

Phase 1 is more than just the thorough literature review that is a feature of good scholarship. Rather, we turn to published research, policy documents, and sometimes gray literature with the intension of identifying some of the entities that may combine to form causal factors. As such, this phase adopts a form of analysis with a focus on discovery or abduction, as in this phase, the researcher seeks to develop, as broadly as possible, an understanding of what factors might be relevant to the experiences and sense-making of the participant group, and then to test this iteratively against new information. For example, what may be relevant to both data sets described in this article is the evidence that care leavers are more likely to become teenage parents, and the close association between lack of support networks, low socioeconomic status, teenage mothering, and mental health problems (Boath, Henshaw, & Bradley, 2013; Robb, McInery, & Hollins Martin, 2013). This shows how elements of embodiment, materiality, and institutions overlap and interact together. For example, early experiences (e.g., insecure attachment, child abuse, and neglect), trauma (e.g., domestic violence), individual characteristics, and hormonal imbalance are all associated with mental health problems (Holt, Buckley, & Whelan, 2008; Levendosky, Huth-Bocks, Shapiro, & Semel, 2003). We also identified reoccurring discourses in academic, government, and other relevant institutional bodies that feed into constructions of mental health and well-being. These studies showed, for example, how mental health services were affected by successive U.K. neoliberal government policies constructing ideal citizens as self-reliant and with low dependency on the state (Harvey, 2007; Taylor-Gooby & Stoker, 2011).

Phase 2: Developing Measures and Methods of Data Collection and Extraction

Once key embodied, material, and institutional factors associated with mental health problems with the participant groups were identified, we developed measures and a research design to explore these factors in the participants’ lives. In this case, we used observational methods and existing relevant data (e.g., on local mental health provision) to create a factsheet for all 32 participants that recorded any material, institutional, and embodied conditions made salient in Phase 1. In line with the principles of designing a method flexible enough for research in different contexts, we do not advocate a set template for this phase. Rather, the approach is to use the research question and data from Phase 1 to guide the formation of the factsheet. For our current research, Phase 1 pointed to the need to systematically collect data in relation to the participants’ personal history regarding their mental health problems and related service use in the context of what was available in their area, their relationship networks (e.g., partner, marital status, children, support networks), and information on their work, education, and housing histories (e.g., what kind of accommodation/homeless status they might have or what jobs they did). These data were gathered from the participants, care-providers (with permission from the participant), and through review of service provision (charities, adult mental health services) in each participant’s area. Interview schedules were also created in this phase covering topics identified in the literature, as well as creating opportunities for the participants to introduce topics we had not considered.

Phase 3: Synthesized CRDA

The first step of Phase 3 was a DA of the interview data, transcribed verbatim. The DA focused on identifying discursive repertoires—reoccurring tropes or common sense claims in the participants’ talk about their experience of mental health problems. In this process, extracts were coded for key discursive repertoires in the talk, and those coded for the same pattern of sense-making collated. For example, participants regularly argued that mental health services were important (see Table 2 for details). A second iteration of DA followed, reviewing the collated extracts that represented each discursive pattern, paying particular attention to any affect or emotional content, which often helped to highlight key tensions or aspects of the participants’ accounts, producing what is known as an affective DA (Billig, 2001; Wetherell, 2013).

Table 2.

Discursive Repertoires Reoccurring Across Both Data Sets.

Discursive Repertoires	Example
Importance of resources	Highlighting the importance of “knowing what is out there.”
Resources need to be justified	Giving external validation to their use by “being referred.”
Stigma of mental health problems	Distancing self from mental health problems or service use; downplaying stigma “millions of people”
The need for, but lack of support	Describing inappropriate support “looking for it yourself”; “they don’t help you”
Experience of isolation and loneliness.	Associating mental health problems with isolation “having nobody to talk to”; “completely on my own.”

In line with critical realist ontology, the next stage was to identify links between how the participants made sense of themselves and the wider institutional discourses producing some of the conditions of possibility for this sense-making. Here, we mapped the patterns in the participants’ talk to the wider and/or institutional discourses identified in Phase 1. For example, neoliberal and austerity discourses combine to construct mental health services as limited resources that few should need to use, creating the context of needing to legitimize one’s use of such services. The wider discourses that linked to participants’ talk were coded with each extract to keep an audit trail. We then took each extract coded under a particular common sense discourse and examined the embodied, material, and institutional contexts of the speaker using Phase 2 data, to consider how these might also provide the conditions for sense-making found in specific extracts. To manage the risk of over interpretation at this stage, we looked for orientation to such conditions in the participants’ talk, and also considered whether their account would be viable without that particular context. Through iterative cycles between Phases 2 and 3, we identified patterns in sense-making and in the wider contexts that might support that sense-making (both shared contexts, but also when participants articulating similar discourses had dissimilar contexts), allowing us to develop a detailed understanding of how the material, embodied, and institutional could be implicated as causal factors in individual participants’ sense-making. This process allowed us to identify a smaller subset of extracts representing reoccurring or unusual patterns in the data that offered insight into the research topic. It was to these extracts that the final stage of analysis was applied, from which four extracts presented in this article are used to illustrate the patterns and argument.

The final stage employs analytics from DP that focus on the context in which data were collected (here, research interviews in service providing centers) and the interactional effects of the talk. To do so, we returned to the audio recordings, transcribing in more detail the talk in the chosen, smaller subset of extracts. This transcription³ included aspects of talk such as intonation, pauses, changes in speed, and loudness. How participants described facts; managed blame, stake, interest, or accountability; or used psychological terms in relation to their mental health problems were then analyzed, with attention also paid to “discursive devises”—speech patterns with predictable interactive effects (Antaki & Wetherell, 1999; Potter, 1997; Wiggins, 2017). Any of the interviewer’s talk, such as asking a question or responding to the participant, was also analyzed for its interactional effects and the context of the interview as eliciting particular accounts was considered (e.g., justification for using resources as a service user).

The above procedure was followed by the Sims-Schouten, with Riley acting as critical peer reviewer and co-analyst. Other conventional qualitative quality criteria were upheld throughout, including reflexivity and iterative cycles of data analysis by Sims-Schouten, Riley, and external peer review. Exemplar extracts that offered insight into the research question were then written up using the wealth of information we had generated on the speaker. As with all qualitative research, the number of extracts presented depends on the scope of the publication, but a rule of thumb might be two extracts per discourse. However, given the methodological focus of this article, the role of the analysis presented below is not to give a comprehensive overview of the empirical findings but to showcase the application of CRDA on different exemplar extracts. Within the structure of a journal article, this allowed us to offer four examples below, with each stage of Phase 3 (DA, DP, and CR) done separately to illustrate how CRDA can be applied.

Data Analysis

Example 1: Data Set 1, A Mother Justifies Her Service Use

The participant in the extract below is a White female, who is diagnosed with depression and anxiety; she has two sons. She recently married the father of her younger child. The older son has autism and lives with foster parents; the mother has supervised contact with the fostered son once a week.

1. Participant: I, I h↓ad to, had to put >my h↑ands up and s↓ay, I can’t c↑ope anymore<

2. ° and .hh that’s°, >that’s when things, things fell into place< and that’s .hh where I started to

3. open >↑up my eyes and everything< started to: open up and .hh (3.0) .hh yea::h. it’s been a

4. hh (3.0) long (2.0)

5. Interviewer: Long j↓ourney.

6. Participant: Yeah, it h↑as. I↑ was ref↑erred b↓ut there’s millions of people out there that

7. don’t °even know what’s going o↓n°. So, if you’ve got a disa↑bled ch↑ild, you don’t even

8. know, don’t know what’s out th↓ere until it’s .hh too late .hh or °you go looking for it

9. yourself°, or err you know I was, I was °in a domestic violence relationship as well where he

10. had to be put awa↓::y°, that’s another thing, where I was completely, completely on my

11. own with my ch↓ild.

The above extract was selected as an example of one that used many of the discourses identified in the analysis (see Table 2). In describing a breaking point (“can’t c↑ope anymore” Line 1), a professional referral (Line 6), and a long struggle before getting support (see Lines 4 and 5 where the interviewer and participant co-construct the participant’s experience as a long journey), the participant stresses the severity of her situation, thus justifying her entitlement to resources. This justification is also made in terms of having a partner so violent he is jailed (Lines 9 and 10); that her child has additional needs (Line 7), with the implication that he would require more resources than a typically developing child might ask of their mother; and that she experienced severe isolation (Lines 10 and 11). In addition to describing an extreme situation, she also normalizes the need to use external resources by suggesting that millions of people need them (Line 6), perhaps orienting to the possibility of stigma in her being so needy. Her account is thus a justification of her use of resources. That she needs to justify her use of resources so much can be linked to wider neoliberal and psychological discourses around parenting that construct good citizens as those who do not need state support, and the importance of being a “good mother” who meets her child’s needs (Sims-Schouten, 2016).

Building on this DA, we turn to DP to explore how the participant uses a number of rhetoric strategies to do this justification in interaction with the interviewer. Much of her talk employs extreme case formulations (ECF) (that invoke maximal or minimal properties, Wiggins, 2017); these work to construct her situation as extreme and thus deserving of support service. In Line 1, an ECF is used to describe her state of mind, which she speaks quickly and with stressed rising intonation “put >my h↑ands up and s↓ay, I can’t c↑ope anymore<”. By having to “put her hands up” she engages in what Potter (1997) refers to as stake confession, strengthening her argument by making this personal, as well as describing an extreme and unambiguous situation “I can’t c↑ope.” The severity of the situation is also emphasized with her description of getting help “too late” (Line 8) and the repetition in the ECF of “I was completely, completely on my own” (Lines 10, 11) that emphasizes isolation and lack of support. The participant also orients to the benefit gained from the support she received, evident in the three-part list (using three points to strengthen an argument, Antaki & Wetherell, 1999), “things fell into place . . . started to open >↑up . . . started to: open up” (Lines 2 and 3).

The tone of the participant’s talk—the high levels of justification, many ECFs, stressed intonation, and long pauses—all indicate that she has trouble with the subject (Speer & Potter, 2000). This points to the participant’s emotional response to the subject matter and may also be an orientation to the potential stigma of mental illness, so that her talk works to deflect the potential negative judgment of the interviewer. Further warding off potential stigma or negative attitude from others is seen in her claim that there are “millions of people out there” (Line 6), where she normalizes her need for support—as in this ECF she is just one of millions.

Combining both DA and DP demonstrates how important the justification for resources is for this participant, as it is made through many of the discursive repertoires identified in Table 2 and through specific rhetorical strategies such as ECFs. The above analysis also shows how analytics from different traditions might allow the researcher to identify a pattern they might otherwise have missed—here, for example, the need to negotiate stigma was only identified through the DP analysis, but in so doing, it offered support for the overall DA patterns found in the whole data set analysis that identified stigma as a reoccurring trope (see Table 2).

Adding the critical realist aspect to our synthesized DA above involved looking for any orientation to the non-discursive factors that may be salient in this example. Key issues in the factsheet of information about this participant and her situation included domestic violence, the mental health problems she was being treated for, the additional needs of her son, legislation around child protection, and cuts to services. Exploring these briefly below, we note that her previous relationship came to an end 3 years ago; her ex-husband was subsequently sent to prison for 1 year for domestic violence. Domestic violence is linked to depression and posttraumatic stress disorder (Holt et al., 2008; Levendosky et al., 2003). This mother suffers from depression and anxiety, and is seeing a counselor for this. The last decade saw an increase in child protection cases in the United Kingdom and the implementation of safeguarding measures (e.g., by taking children into care) as a result of domestic violence (Hilder & Bettinson, 2016). Domestic violence is now recognized as a public health issue and a significant child protection concern (Rivett & Kelly, 2016) with associated policies and legislations such as the Domestic Violence, Crime and Victims Act (2012). The extract above is an example of the implications of these policies, as this mother’s son was taken into care due to the perceived risk of the domestic abuse environment and the additional complications of the child being diagnosed with autistic spectrum disorder. This also offers new light on the emphasized “too late” in Line 8, as in this mother’s case, help arrived after her child was taken into care. That help was “too late” and her focus on the fact that people “don’t know what is out th↓ere” (Line 8) also hints at a critique of neoliberal health policies which emphasize being self-sufficient or responsible for identifying resources for working on the self (Riley, Evans, & Robson, 2018), and what Bhaskar (2014) refers to as absence (what is missing in a context or institution/organization) in relation to service provision.

Overall, this mother’s extract can be understood as shaped by a number of non-discursive scaffolding factors impacting on how she constructs, negotiates, and justifies her experience. At the embodied level, for example, there is her emotional response to the subject matter and her history of depression and anxiety. And at the institutional level, there are child protection, neoliberal, and austerity-led policies, impacting on how she understands her need for services and the kind of ethical person and mother she can claim to be.

Example 2: Data Set 1, A Mother Orients to the Stigma and Effect of Mental Illness

This participant is a White woman, who lives with the father of her small son. She has a history of anorexia and a diagnosis of postnatal depression. The extract below is in response to the question of why she is using the services at the children’s center.

1. Participant: >bec↓ause I had a b↓ab↑y<, .hh and e::rm (4.0) I, I got post-natal

2. depr↓ess↓ion so I was kind of pointed in this direction from, er from my h↑ealth visitor

3. (1.0) It helped me with playing with XXX .hh (2.0) ermm (2.0) its h↓elped me:: (2.0)

4. kinda get out the fl↑at, >you know<, rather than be c↑ooped up and st↓ay in the flat, and

5. get depr↓essed and

6. Interviewer: Yeah

7. Participant: Ermm, .hh (1.0), and, s↑o for that purpose it’s been brilliant as

8. well, and (3.0), yeah s↓o

9. Interviewer: Has it helped you manage things at home?

10. Participant: Ah yeah, err I went to a young parents, not a young parents, a new parents

11. gr↑oup that was o:n, went to a toddler’s w↓orkshop to help me w↓ith (2.0) {laughing} XXX

12. sort of becoming a bit terrible two’sies early, .hh so I went on th↓at. Ermm (3.0), there

13. w↑as a err moving on to solids gr↓oup, but >I never actually managed to do th↓at<

14. bec↓au::se °I’ve had °difficulties with° eating mys↑elf, °been anorexic for years° .hh °not

15. not at the m↑oment° but I’ve got, got a history of it so i::t’s (1.0) been quite d↑ifficult

16. °where foods concerned, and°. So I haven’t actually been able to g↑o on that c↓ourse

17. because I just found it too overwhelming, BUT it’s there for people you kn↓ow, and it is

18. meant to be very good.

Applying DA to this extract draws attention to a number of discursive repertoires relating to a justification of the need for recourses, entitlement, stigma, and isolation. First, her relationship with her child, where she stresses how attending the children’s center helped her play with her son in the context of postnatal depression (Lines 1 and 2) and attachment problems (oriented to in Line 3). Second, she focuses on the need to address loneliness and isolation and “get out the fl↑at” (Line 4). Finally, she refers to her own mental health problems, for example, getting “depr↓essed” (Line 5). These discursive repertoires point to positive aspects of using the service.

Explaining her service use on the basis of her son “becoming a bit terrible two’sies early” (Line 12), her talk both normalizes her son’s behavior (“terrible twos” being a common sense developmental discourse) and positions this behavior as the justification of her service use, rather than her inability to cope. As such, we suggest that this mother’s talk can be explored in light of both wider discourses that stigmatize mental illness (Robb et al., 2013) and of the “good mother” where attachment and nurturing are expected to come “naturally” (Sims-Schouten & Riley, 2014). We further read the participant’s orientation to the stigma of mental illness in how she distances herself from her anorexia, describing it as historical (in both Lines 14 and 15) while it continues to affect her (Lines 13 and 17). The role of her mental illness in her service use is thus downplayed, even while constructing her mental health problems as serious.

DP highlights that both her problematic attachment with her son and her own mental health problems are affective topics. Her difficulty in discussing her history with anorexia is evident from the long pause (3 seconds in Line 12 ), as well as the talk around how she has °been anorexic for years°, which is said softly (indicated by the ° in Line 14). Indeed, all her talk around this issue, of “being anorexic,” “having difficulties with eating,” and “where foods concerned,” is said softly (Lines 14–16), suggesting that this is a problematic subject creating an emotional response and/or a need to negotiate the interviewer’s potentially negative judgment to her disordered eating (see Speer & Potter, 2000).

From the long pauses, hesitations, soft tone, and contradictory positioning of her anorexia as both historical and present, the DP and DA analyses suggest that this topic is difficult for her, and her move to discussing the quality of the services on offer might be understood as a move to safer territory. For example, her discussion of the benefits of the “solids group” for those who can go (Lines 17, 18) is emphasized and said loudly, in contrast to her talk around anorexia. Her focus on the quality of the services in the center thus moves the conversation away from her mental health problems.

Turning toward CR, in this participant’s factsheet, we can see a complex intertwining of the material (living in a small apartment in relative isolation), the embodied (long-term disordered eating, depression), and institutional psychiatric diagnoses (anorexia, depression) with a link between postnatal depression and attachment being well-documented (e.g., Brady et al., 2015). Her eating disorder was also evident in how she negotiated the wider interview experience, as the interviewer was asked by one of the center care workers (at the request of the participant) not to offer this mother any coffee or biscuits.

Extracts 1 and 2 offer examples of how service users might take up a position of an ethical subject, able to justify their service use in the context of cuts to services, and wider discourses that stigmatize mental health problems and service users. But Extract 2 also develops the analysis through an exploration of when participants might downplay their mental health problems even while needing to justify their use of resources to support themselves. Here, for example, the CRDA points to both the stigma and effects of mental illness on the person’s life.

Example 3: Data Set 2, A Young Father Who Focuses on Inadequate Help and Support

A young White man who was in care from a very young age (children’s homes then foster care). He became homeless as a teenager and has a suspended sentence after abusive public behavior. He has a diagnosis of depression and now lives in a hostel with his partner (also a care leaver) and their small child.

1. Participant: I can o↑nly use a term that I’ve been a hard-nosed bastard all my life

2. {laughs}, erm: if you (0.5) grow up in the c↓are s↑ystem then there is not m↓uch

3. that can scare you, there is not much that (0.5) can °you know what I mean° put fear

4. into you (1.0). The most sc↑ariest thing I’ve done in my l↓ife, is (1.0) °you know

5. what I mean° is watch my little girl be b↑orn (1.0) err that is the most scariest thing,

6. hh >I mean< losing my m↑um was ups↓etting °but° it h↑appens >you know what

7. I mean<. When you’re in that system, where you’ve got no parental figure, and

8. you’ve got nothing like that and, >you know what I mean< .hh people say they’re

9. y↑our parental figure and don’t actually help y↓ou.

10. Interviewer: Do you feel the c↓are system is disapp↑ointing?

11. Participant: .hh Extr↑emely. I’ve b↓een with f↑oster parents, >I’ve b↓een in

12. children’s homes<, >I’ve been in supported lodgings<, I’ve been in the foyer °and

13. stuff like that° and n↑othing helped m↓e. Erm (2.0) about three years ago I wa::s

14. smoking legal high °and fixated on doing l↑egal°, .hh three years later after losing

15. my Mum, went on a massive bender. .hh I’m no longer smoking, I don’t do dr↓ugs.

16. I’ve got a little g↓irl, and I am looking at moving into my own pl↑ace.

17. Interviewer: So y↑ou have come a long w↑ay th↓en.

18. Participant: Yeah.

19. Interviewer: Wow, yeah. So how would you say your mental health and wellbeing is

20. at the moment?

21. Participant: Erm, I’m ↓on anti-depr↑essants and (2.0) er °yeah I’ve g↓ot°, I’m

22. supp↑osed to °have gone to see erm XXX° but after my last meeting with th↑ose

23. guys, °I don’t want that service°. They’re useless and anything that I could talk with

24. th↑em about, I could talk to XX about. I haven’t had many people in my life who

25. will j↓ust be there.

Our DA suggests that the participant constructs himself as being significantly let down by institutions charged with providing surrogate parental figures. In so doing, he articulates a number of the discursive repertoires identified in this study, such as the importance of resources, entitlement, and his need for support. Billig (2001) describes common sense as a kaleidoscope, and in this extract, we see two kaleidoscopic shifts in the participant’s account, from being let down by the care system to a focus on survival and responsibility when he talks about his daughter, and a shift around the significance of his mother’s death.

Woven into his story is the death of the participant’s mother, which is down-played in its own terms (Line 6, losing my m↑um was ups↓etting °but° it h↑appens) and also in relation to the birth of his daughter which is positioned as more significant (Line 5, “the most scariest thing”). In this talk, the participant highlights how he has been let down by all the parental figures in his life (Lines 7–9), while showing responsibility toward his own newfound role as a parent, something that not only scares him but also commits him to do his very best (unlike the parental figures around him).

Billig (2001) argues that people are contradictory in their sense-making as they shift from one discourse to another. This is evident from how this participant moves from downplaying the importance of his mother’s death to describing it as having significant impact on him (Line 15, “went on a massive bender”). But, within this contradiction is a consistent narrative in which his self-destructive behavior can be explained as not his fault but the outcome of being repeatedly being let down by the parental figures of his childhood—producing isolation and a lack of needed mental health support. The participant thus locates blame in external systems that failed to provide appropriate care for his past self, while positioning his current self as having agency and an ability to turn his life around to be a good father (Lines 15, 16).

DP shows how the two stories, of being let down and of choosing to be a good father identified in the DA, are reinforced through a range of discursive devices. The failure of parental figures in his life is read in stress (see underlined words) and repetition in “no parental figure” (Line 7), “don’t actually help y↓ou” (Line 9), and not having people who “will j↓ust be there” (Line 25). The generalized five-part list that he gives to describe the different care systems he experienced (Lines 11,12) provides a strong warrant for factual accuracy, plus an overwhelming sense of difficulty (as it is far longer than a standard three-part list), which is further emphasized with his concluding ECF that “n↑othing helped m↓e” (Line 13). His vision of being a good parent is evident from the three-part list of “no longer smoking,” “don’t do dr↓ugs,” and “own pl↑ace” (Lines 15, 16), allowing him to show the ways that he has changed.

Applying CR to this participant’s account makes central the trauma and suffering occasioned by his care experiences. Michael Stein (2006) argues that there are three broad groups of care leavers: those who are moving on, survivors, and victims. The moving on group are likely to have had stability and continuity in their lives, made sense of their family relationships, had some educational success before leaving care, and been able to make use of the help offered. Survivors experience more instability, movement, and disruption while in care; leave care earlier; achieve few educational qualifications; and tend to believe in their own self-reliance but often need ongoing support. Victims are the most disadvantaged and have had the most damaging pre-care family experiences; their time in care is characterized by difficulties, instability, and disruption, and after leaving care, they are likely to be unemployed, become homeless, be lonely and isolated, and have mental health problems. The participant clearly fits the “victim” category, but his talk also points to a desire to move to “survivor” in seeking to move “into my own pl↑ace” (Line 16) and meet his responsibilities as a new parent.

His focus on no longer smoking and doing drugs (Line 15) is also linked to his material and institutional context. Currently, he lives with his partner in supervised accommodation, as she became pregnant when they were both homeless. His focus on moving into “my own pl↑ace” can be contextualized within child protection policies and the risk that his child may be taken away from him, as he was from his own parents. This risk is real, with the factsheet identifying that he and his partner are the focus of child protection, with regular visits from social workers who expect him to attend support and counseling sessions.

Example 4: Data Set 2, A Young Woman Negotiates Isolation, Loneliness, and Entitlement

A young woman, originally from West Africa, was taken into care when she arrived in the United Kingdom as a young teenager. She has a history of mental health problems (anxiety) and experienced bullying and racist incidents at school. The extract below is in response to the question about the support needed by the participant.

1. Participant: I get stressed, struggle with m↓oney, ◦and◦ erm (1.0) >sometimes I dont<,

2. I don’t have n↑obody to talk to as well .hh s↓o I was getting (1.0) so much

3. stress and, my hair is f↓alling out, err it’s so hard to handle, you know, l↓ike,

4. for me to have like somebody like, and XXX (1.0) that I can talk to, ◦even if

5. she comes once◦ ↑every two weeks to see me. So, I can talk to her, it makes

6. me feel better.

Above, the participant justifies her use of mental health services in terms of the high levels of stress and isolation she experiences, drawing on the discursive repertoires of loneliness, need for support, and importance of resources (Table 2). The support from her care worker is constructed in terms of alleviating her stress and isolation (Lines 5 and 6). Thus, her entitlement for support is justified by the severity and the outcome, as it makes her “feel better” (Line 6).

DP analysis shows how the participant uses two three-way-list completers to present the situation she finds herself in as multi-facetted (financial, physical, and social) and severe: “struggle with m↓oney,” “stressed,” and “n↑obody to talk to” (Lines 1 and 2); this is followed in Line 3 by the second list, which also contains emphasis on the work stress (“so much stress . . . my hair is f↓alling out . . . it’s so hard to handle”). This severity is used to justify the support she gets and needs, support for loneliness that is presented in an ECF of having “n↑obody to talk to” (Line 2), and eliminated even through the most minimal intervention “she comes once◦ ↑every two weeks to see me” (Lines 4–6).

Considering this account through the lens of CR allows further interpretation, making salient her material context of isolation and poverty produced by child protection policies, bullying by her peers, and the government benefit system, which in turn may be linked to embodied experiences of anxiety and potentially other stress responses. These material, institutional, and embodied factors, which combined have led to her experience of extreme isolation, can be understood as providing the scaffolding for her positive construction of the input of the care worker and having someone to talk to.

Both Extracts 3 and 4 allow us to explore some of the impact of being in care and inconsistent mental health support. Research points to the higher rate of mental health problems among children in care (about 4 times that of the general population of young people) (NSPCC, 2016) and also to the often poor management of transitioning from CAMHS to adult care (Schraeder, Reid, & Brown, 2017). Both may play a causal role in the participant’s, in Extract 4, failure to form any other relationship outside of her care worker.

Discussion

This article proposed a CRDA method for how to better analyze, and thus understand, service users’ narratives of their mental health problems, and demonstrated how it might be applied to data from two groups of hard-to-reach service users with a range of mental health problems, both diagnosed and undiagnosed. The purpose of this article was thus to provide a rationale and step by step guide to performing CRDA and to demonstrate the beneficial outcomes of applying CRDA to service users’ narratives of their mental health problems. This was done through the presentation of a systematic step-by-step guide, which was then applied to four extracts from the exemplar data sets.

The four different examples were chosen to show the use of CRDA across variation in the data sets, coming from both men and women, with different mental health problems, who used different services, and who had different material and embodied contexts. In all four examples, we showed not only the value of a discursive approach in exploring the interactional practices and wider discourses in the participants’ sense-making but also how policy, embodied, and material factors scaffolded this sense-making. For example, the CRDA highlighted an important reoccurring pattern, whereby these service users presented themselves as ethical subjects, responsible and entitled to use the system because their problems were significant but could be attributed elsewhere (their child, their isolation, their care). This points to the power of neoliberal rationality, where service users have come to understand that they are self-responsible and blame is put on the individual (Harvey, 2007). The CRDA allowed this finding to be identified by bringing together the focus on accountability and blame in DP with DA that considers the wider discourses that provide the conditions of possibility for such talk. A further level was added when considering the impact of material conditions (such as being isolated in a small flat, access to services only through referral), embodied experiences (such as stress and trauma), and awareness of a range of policies that might also contribute to a participant’s sense-making, such as child protection and austerity-driven service management. Through the CRDA, we were able to explore some of the complexity that might inform these accounts by contextualizing their sense-making within (“real”) statistical patterns that show causal effects between, for example, domestic violence, trauma, and mental health problems, as well as the material contexts and social position of the participants.

The analysis of the four extracts showed the similarities of the participants’ sense-making, evidenced in their shared discursive repertoires that focused on a need to justify resources; the stigma of mental health problems; the need for, but lack of support; and the role of isolation and loneliness in their mental health problems. The analysis of each extract showed how many of these discursive repertoires could be drawn together even in a short piece of talk, and how they intersected to strengthen the function of the talk, such as legitimizing participants’ claims to use services. The analysis also showed how combining the different aspects of the DP, DA, and CR could synergistically develop analysis, for example, how the orientation to stigma emerged more strongly when combining DA and DP.

The four extracts presented demonstrated how CRDA is also able to explore potentially different realities underpinning the participants’ use of the same discourses. For example, despite using similar discourses to the other participants presented in this article, the participant in Extract 3 differed from the others in that he was institutionally required to participate in psychological support services (in this case, a requirement from social services to undergo counseling)—the others chose to use these services. Such differences show the multiplicity of what service users might negotiate when talking to a researcher. Researchers need to be able to explore such complexity, and we argue that in CRDA, we offer a significant contribution to the study of mental health research with a method for doing this.

The outcome of our CRDA is, we argue, an accessible analysis that engages with complexity by drawing on multiple perspectives offered by methods with different epistemological and ontological standpoints (rather than being limited to only one perspective). As such, our CRDA is different from other linguistic methods or critical realist methods, allowing researchers to better analyze, and thus understand, service users’ narratives of their mental health problems.

Wider Implications and Applications

There is a sense within the wider social sciences community that critical realist approaches are useful because they allow an analysis that recognizes not only the constructive power of language but also the material, embodied, and institutional factors that contain what is possible to say (see, for example, Cromby, 2016; Cromby & Harper, 2009; Houston, 2010; Pocock, 2015). For our research, it meant that we could make a connection between our participants’ talk and their lives, exploring how the mothers and young people in these studies make sense of their mental health and well-being within the context of the complex material and discursive contexts in which they find themselves. As such, we were able to incorporate discursive (stigma, accountability, support) and non-discursive influences (emotional response to trauma, domestic violence, cuts to funding, material context) in our analysis, as factors that scaffold people’s talk regarding their mental health.

Bringing together CR and DA has its benefits, in particular, the ability to contextualize talk in both a deep and wide manner. Yet, it also requires some concessions to be made. For discourse analysts, this means going beyond the interactional effects of the talk or systems of representation and recognizing the existence of the non-discursive not necessarily evident in the data. Framing this within a critical realist ontology and epistemology means taking account of the stratified non-linear dynamic nature of processes at different levels—the real level (namely, causal mechanisms that generate events) and the “empirical” level (experienced events).

We argue such concessions are worth it; through our CRDA, we were able to conceptualize our participants’ talk in relation to their mental health problems as constituting of multiple lines of influence, highlighting the tensions between issues around stigmas and labels (e.g., having to defend that help is needed), loneliness, and personal experiences of mental health, neglect, trauma, and bullying. This also has implications for practice in that we are offering a tool that generates insight into both non-discursive and discursive factors that impact on conceptualizations of mental health and well-being, including how service users make sense of resources and support entitlement, and what Bhaskar (2014) refers to as absence (what is missing in a social context or institution/organization) in relation to service provision. With this article, we therefore offer CRDA as an important contribution to health research seeking to better understand how service users make sense of their mental health.

Footnotes

Acknowledgements

The authors thank Dr. Simon Edwards and Ms. Emma Maynard for support with data collection.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received financial support from the Blagrave Trust and Portsmouth City Council for the research.

Notes

ORCID iD

Wendy Sims-Schouten

Author Biographies

Wendy Sims-Schouten is an Associate Professor in Childhood Studies and Associate Head Research in the School of Education & Sociology at the University of Portsmouth. Wendy is the coordinator of the Mental Health in Childhood & Education (MICE) Hub () at the University of Portsmouth and has researched (and published) in the areas of mental wellbeing of vulnerable children, such as children in care and care leavers; she has also researched issues around bullying and childhood obesity.

Sarah Riley is Associate Professor in Psychology at Aberystwyth University; she is an interdisciplinary researcher who draws on psychology, sociology cultural and media studies, to understand the person in context. Her work covers a range of identity issues including gender, embodiment, health, youth culture and citizenship. She uses a range of qualitative methods including discourse analysis, visual and participatory methods.

References

Antaki

(2011). Applied conversation analysis: Intervention and change in institutional talk. London: Palgrave McMillan.

Antaki

Wetherell

(1999). Show concessions. Discourse Studies, 1(1), 7–27.

Bhaskar

(1989). Reclaiming reality. London: Verso.

Bhaskar

(2014). Foreword. In Edwards

P. O.

Mahoney

Vincent

(Eds.), Studying organisations using critical realism. A practical guide (pp. V–XV). Oxford: Oxford University Press.

Billig

(2001). Discursive, rhetorical and ideological messages. In Wetherell

Taylor

Yates

S. J.

(Eds.), Discourse theory and practice: A Reader. (pp. 210-221) London: Sage.

Boath

E. H.

Henshaw

Bradley

(2013). Meeting the challenges of teenage mothers with postpartum depression: Overcoming stigma through support. Journal of Reproductive and Infant Psychology, 31, 352–369. doi:10.1080/02646838.2013.800635

Brady

Lowe

Lauritzen

S. O.

(2015). Children, health and well-being: Policy debates and lived experience. New Jersey: Wiley.

Cooklin

A. R.

Giallo

Rose

(2012). Parental fatigue and parenting practices during early childhood: An Australian community survey. Child: Care, Health and Development, 38, 654–664. doi:10.1111/j.1365-2214.2011.01333.x

Cromby

(2016). Developing Schizophrenia. Theory & Psychology, 26, 607–619. doi:10.1177/0959354316661430

10.

Cromby

Harper

D. J.

(2009). Paranoia: A social account. Theory & Psychology, 19, 335–361. doi:10.1177/0959354309104158.

11.

Domestic Violence, Crime and Victims (Amendment) Act (2012). https://www.gov.uk/government/publications/domestic-violence-crime-and-victims-amendment-act-2012.

12.

Dunning

M. J.

Giallo

(2012). Fatigue, parenting stress, self-efficacy and satisfaction in mothers of infants and young children. Journal of Reproductive and Infant Psychology, 30, 145–159. doi:10.1080/02646838.2012.693910

13.

Harvey

(2007). A brief history of neoliberalism. Oxford: Oxford University Press.

14.

Hilder

Bettinson

(2016). Domestic violence: Interdisciplinary perspectives on protection, prevention and intervention. London: Palgrave Macmillan.

15.

Holt

Buckley

Whelan

(2008). The impact of exposure to domestic violence ion children and young people: A review of the literature. Child Abuse & Neglect, 32, 797–810. doi:10.1016/j.chiabu.2008.02.004

16.

Houston

(2010). Prising open the black box. Qualitative Social Work, 9, 73–91. doi:10.1177/1473325009355622

17.

Levendosky

A. A.

Huth-Bocks

A. C.

Shapiro

D. L.

Semel

M. A.

(2003). The impact of domestic violence on the maternal–child relationship and preschool-age children’s functioning. Journal of Family Psychology, 17, 275–287. doi:10.1037/0893-3200.17.3.275

18.

Lofgren

Hewitt

das Nair

(2015). Doing fence sitting: A discursive analysis of clinical psychologists constructions of mental health. Qualitative Health Research, 25, 470–485. doi:10.1177/1049732314549479

19.

NSPCC (2016). Children in care Emotional wellbeing and mental health [Online]. Available at: https://www.nspcc.org.uk/preventing-abuse/child-protection-system/children-incare/emotional-wellbeing-of-children-in-care/.

20.

Patel

Pilgrim

(2018). Psychologists and torture: Critical realism as a resource for analysis and training. Journal of Critical Realism, 72, 176–191. doi:10.1080/14767430.2018.1430975

21.

Pilgrim

(2014). Some implications of critical realism for mental health research. Social Theory & Health, 12, 1–12. doi:10.1057/sth.2013.17

22.

Pocock

(2015). A philosophy of practice for systemic psychotherapy: The case for critical realism. Journal of Family Therapy, 37, 167–183. doi:10.1111/1467-6427.12027

23.

Potter

(1997). Discourse analysis as a way of analysing naturally occurring talk. In Silverman

(Ed.), Qualitative research: Theory, method and practice (pp. 144–160). London: Sage.

24.

Riley

S. C. E.

(2003). The management of the traditional male role: A discourse analysis of the constructions and functions of provision. Journal of Gender Studies, 12, 99–113. doi:10.1080/0958923032000088300

25.

Riley

Evans

Robson

(2018). Postfeminism and health: Critical psychology and media perspectives. London: Routledge.

26.

Rivett

Kelly

(2016). “From awareness to practice”: Children, domestic violence and child welfare. Chile Abuse Review, 15, 224–242. doi:10.1002/car.945

27.

Robb

McInery

Hollins Martin

C. J.

(2013). Exploration of the experiences of young mothers seeking and accessing health services. Journal of Reproductive and Infant Psychology, 31, 399–412. doi:10.1080/02646838.2013.832181

28.

Roberts

J. M.

(2014). Critical realism, dialectics, and qualitative research methods. Journal of the Theory of Social Behaviour, 44, 1–23. doi:10.1111/jtsb.12056

29.

Savage

Murray

Hatch

S. L.

Hotopf

Evans-Lacko

Brown

J. S. L.

(2015). Exploring professional help-seeking for mental disorders. Qualitative Health Research, 26, 1662–1673. doi:10.1177/1049732306289224

30.

Sayer

(2000). Realism and social science. London: Sage.

31.

Schraeder

K. E.

Reid

G. J.

Brown

J. B.

(2017). “I think he will have it throughout his whole life”: Parent and youth perspectives about childhood mental health problems. Qualitative Health Research, 28, 548–560. doi:10.1177/1049732317739840

32.

Silverman

(2010). Doing qualitative research (3rd ed.). London: Sage.

33.

Sims-Schouten

(2016). Positioning in relationships between parents and early years practitioners. Early Child Development and Care, 186, 1392–1405. doi:10.1080/03004430.2015.1095187

34.

Sims-Schouten

Hayden

(2017). Mental health and wellbeing of care leavers: Making sense of their perspectives. Child & Family Social Work, 24, 1480–1487. doi:10.1111/cfs.12370

35.

Sims-Schouten

Riley

(2014). Employing a form of critical realist discourse analysis for identity research: An example from women’s talk of motherhood, childcare and employment. In Edwards

O’Mahoney

Vincent

(Eds.), Studying organizations using critical realism (pp. 46–66). Oxford: Oxford University Press.

36.

Sims-Schouten

Riley

S. C. E.

Willig

(2007). Critical realism: A presentation of a systematic method of analysis using women’s talk of motherhood, childcare and female employment as an example. Theory & Psychology, 17, 127–150. doi:10.1177/0959354307073153

37.

Speer

S. A.

Potter

(2000). The management of heterosexist talk: Conversational resources and prejudiced claims. Discourse & Society, 11, 543–572. doi:10.1177/0957926500011004005

38.

Stein

(2006). Research review: Young people leaving care. Child and Family Social Work, 11, 273–279. doi:10.1111/j.1365-2206.2006.00439.x

39.

Taylor-Gooby

Stoker

(2011). The coalition programme: A new vision for Britain or politics as usual? The Political Quarterly, 82, 4–15. doi:10.1111/j.1467-923X.2011.02169.x/full

40.

Wetherell

(2013). Affect and discourse—What’s the problem? From affect as excess to affective/discursive practice. Subjectivity, 6, 349–368. doi:10.1057/sub.2013.13

41.

Wiggins

(2017). Discursive psychology: Theory, method and applications. London: Sage.

42.

Winter

(2014). Understanding and supporting young children’s transitions into state care: Schlossberg’s transition framework and child-centred practice. British Journal of Social Work, 44, 401–417. doi:10.1093/bjsw/bcs128