The unintended consequences of targeting: young people’s lived experiences of social and emotional learning interventions

In the past twenty years there has been a proliferation of targeted school-based social and emotional learning (SEL) interventions. However, the lived experience of young peoples’ participation is often elided, while the potential for interventions to confer unintended and even adverse effects remains under-theorised and empirically under-explored. This paper reports findings from a qualitative case study of students’ participation in a targeted SEL intervention, the Student Assistance Programme. Data was generated with four secondary schools in Wales, with 41 students (age 12 – 14) taking part in the study. Findings indicate that students’ identification for participation in the intervention and their reaction to the group composition may lead to harmful effects. Four iatrogenic processes were identified: (1) identification may be experienced as negative labelling resulting in rejection of the school (2) the label of SEL failure may serve as a powerful form of intervention capital, being employed to enhance students’ status amongst peers. Possession of this capital is contingent on continued resistance of the intervention (3) targeting of discrete friendship groups may lead to the construction of intervention ‘outsiders’ as students seek safety through the reification of pre-exiting relationships (4) students may seek to renegotiate positioning within targeted friendships groups by ‘bragging’ about and reinforcing anti-school activities, leading to deviancy amplification.

Thi s v e r sio n is b ei n g m a d e a v ail a bl e in a c c o r d a n c e wit h p u blis h e r p olici e s. S e e h t t p://o r c a . cf. a c. u k/ p olici e s. h t ml fo r u s a g e p olici e s. Co py ri g h t a n d m o r al ri g h t s fo r p u blic a tio n s m a d e a v ail a bl e in ORCA a r e r e t ai n e d by t h e c o py ri g h t h ol d e r s .

Introduction
Modern childhood and adolescence has been associated with a proliferation in poor health and educational outcomes across many societies. These include the escalation of mental health disorders, with a prevalence rate of between 20% and 25% in developed countries (Patel et al., 2007). Current estimates indicate that 10% of children in the UK have a clinically diagnosed disorder (Green et al., 2005). Equally, concerns abound the apparent deterioration in subjective wellbeing (Bradshaw et al., 2007;Currie et al., 2012). International studies also indicate continued problems with academic achievement and disengagement across educational settings (Aud et al., 2010). Inadequate social and emotional competency has been increasingly postulated as an underlying source of these problems, while social and emotional learning (SEL) has been promoted as a potential solution (Weissberg and O'Brien, 2004;Durlak et al., 2011). Social and emotional competency may be defined as an individual's ability to 'recognise and manage their emotions, appreciate the perspectives of others, *Corresponding author: DECIPHER, School of Social Sciences, Cardiff University, CF10 3BD, UK. Email: EvansRE8@cf.ac.uk The copyright line for this article was changed on 4 July 2014 after original online publication. establish pro-social goals and solve problems, and use a variety of interpersonal skills to effectively and ethically handle developmentally relevant tasks' (Payton et al., 2000, p. 179). In practice this competency has been operationalised as interrelated skill sets that include: self-awareness; self-management; motivation; empathy; and social skills (Goleman, 1996). Evidence indicates an association between possession of these skills and a range of mental health outcomes including depression, anxiety and conduct disorders (Waddell et al., 2007). A causal relationship between emotional intelligence and academic achievement has been established (Qualter et al., 2012), which may be partly explained by its psychologically stabilising influence during intensive periods of study and the enhancement of individuals capacity to learn (Durlak et al., 2011;Qualter et al., 2012). Educational institutions that prioritise SEL can also become more inclusive settings, supporting positive developmental and learning opportunities for those with social and emotional behavioural difficulties (Mowat, 2010;Reicher, 2010).
As schools constitute an influential institution within the early phases of the life course, they have been increasingly viewed as central sites for the delivery of SEL, both within pedagogic practice and formal intervention packages (Cigman, 2008;Greenberg, 2010). Programmes include three key intervention approaches. Targeted programmes, such as the Change Project (Kendal et al., 2011), focus on individuals exhibiting risk factors. Universal, classroom-based programmes often involve the systematic teaching of skill sets, with notable examples including the PATHS curriculum (Greenberg et al., 1995), FRIENDS (Stallard et al., 2007), and Incredible Years (Webster-Stratton et al., 2007). Complex, multi-dimensional approaches aim to transform the educational context. Although often a nebulous construct, contextual restructuring predominantly entails: enhancement of learning opportunities; improved school ethos; and weakening of boundaries between schools, parents and the community (Kidger et al., 2010). Whole school approaches have received particular support within the UK, with the English government investing £40 million in its flagship SEL intervention, the Social and Emotional Aspects of Learning (SEAL) programme, from 2007 to 2011. By 2010 the SEAL intervention was delivered in more than 90% of primary schools and 70% of secondary schools in England . Similar programmes have been promoted internationally. The Australian government invested substantial resources in the mental health promotion initiatives Kidsmatter and Mindmatters (Wyn et al., 2001;Graetz et al., 2008). In the USA, effective approaches include Communities that Care (Hawkins et al., 1992) and the Caring School Community Programme (Battistich et al., 2004). A number of systematic reviews comprehensively synthesise the effect of a range of SEL interventions (Weare and Nind, 2011;Durlak et al., 2011).
Despite evident support and investment in SEL, the election of the Conservativeliberal Democrat Coalition Government in 2010 has transformed the political landscape in the UK and instigated the marginalization of the agenda. Indeed, a reorientation of educational priorities towards academia has arguably precipitated a drift away from complex, resource intensive programmes, and a move towards less costly, one dimensional approaches, which often include intervention with targeted individuals . While previous reforms of school inspectorate frameworks had seen the inclusion of 'wellbeing' as an assessment standard (Ofsted, 2009;Estyn, 2010), in 2013 it was eradicated from Ofsted's framework in England, with the Education Minister Michael Gove disregarding expenditure on such 'peripherals' (Northen, 2012). Equally, in 2011 the English Department for Education withdrew endorsement of the SEAL programme amid claims that it was prudent to prevent further resources being expended on an intervention that demonstrated limited effectiveness (Bywater and Sharples, 2012). Evaluation of the programme found that the loose enabling framework ensured numerous translational difficulties, culminating with insufficient implementation and equivocal effects (Humphrey et al., 2010).
Given this political backdrop, the potential of targeted interventions in educational settings has received increased interest. The effectiveness of the approach has been identified in a number of systematic reviews of SEL interventions and positive youth development programmes more broadly (Beelmann and Losel, 2006;Shucksmith et al., 2007;Payton et al., 2008), culminating with recommendations for combining targeted intervention with whole school approaches (Weare and Nind, 2011). However, there is some indication that targeting may be ineffective owing to the identification of 'false positives' (Shucksmith et al., 2007), and may even be harmful to those categorised as 'in need' (Farrington and Ttofi, 2009). Yet despite this equivocality, there remains a dearth of critical examination of targeted SEL interventions, and a lack of qualitative exploration regarding the conditions under which they cause iatrogenic effects.
This oversight may be partly attributed to insufficient evaluation of SEL interventions, with criticism suggesting that the evangelical promotion of the agenda has ensured the presumed irrelevance of evidence of effect and the discounting of unintended or adverse consequences Ecclestone, 2012). Indeed, the adoption and delivery of SEL interventions is often motivated by belief and personal predilections as opposed to evidence (Durlak et al., 2011). Equally, where evaluation has been conducted, focus on students' responsiveness and adherence has ensured the continued elision of the lived experience of participation. In particular, there remains limited theorisation and empirical examination of how the interaction of educational intervention and individual is mediated by the broader social, cultural and material contexts of participants' lives. As Greenberg (2010) suggests, the perspectives of self-aggrandising 'experts' have often been privileged over the voices of children and young people, with their views being reduced to peripheral, reflective activities. Yet their voices are central to intervention outcomes, as they take up an active role in their subjectivity, negotiating both their identity as an intervention participant and the intervention experience (Hall et al., 2004). This paper explores young people's lived experiences of participating in a targeted SEL intervention, and considers the potential for targeting to confer unintended and even harmful effects.
The unintended effects of targeting: negative labelling, deviancy training and the 'failing' subject position Iatrogenic effects of targeted interventions have been empirically explored in numerous health and educational domains, with studies theorising the mechanisms through which harms may occur (Dishion et al., 1999;McCord, 2003;Cho et al., 2005;Bonell and Fletcher, 2008;Wiggins et al., 2009). Two dominant iatrogenic processes The unintented consequences of targeting 3 have emerged. First is the negative label associated with the subject position of intervention participant. Identification for intervention may prove to be stigmatising (Farrington and Ttofi, 2009), while the label of 'at risk' may diminish an individual's expectations for the future, concluding with an increase in the very behaviours the intervention seeks to mitigate (Wiggins et al., 2009). Second is positive deviancy training, which is the consequence of aggregating 'deviant' peers (Dishion et al., 1999). Interventions may provide fertile sites for participants to trade stories of rule breaking, exchange advice and proffer encouragement (McCord, 2003;Cho et al., 2005;Bonell and Fletcher, 2008). Through exposure to new forms of deviancy and on receipt of support for such activity, deviancy may escalate.
There is also limited theorisation of adverse intervention consequences within the domain of SEL. In their seminal critique of social and emotional competencies, notably emotional intelligence, Matthews et al. (2004) suggest that the privileging of such skills encourages the construction of an 'emotional elite', akin to the intellectual elite, that are differentiated by their IQ. As elites can only exist in relation to the failing subject position, the introduction of the 'emotional elite' necessarily brings the 'emotional failure' subject position into being (Matthews et al., 2004;Craig, 2007;McLaughlin, 2008). Intervention targeting may have the capacity to exacerbate the construction of this binary, with the emotional failures being clearly identified for intervention so that their social and emotional idiosyncrasies may be eliminated.
Benjamin (2002) provides a useful departure point for understanding how students' may react to assignment of the 'failing' subject position, primarily through the act of resistance and rejection. The author's ethnography of young women deemed to have special educational needs describes how individuals positioned outside of dominant discourses of achievement work to construct identities that privilege alternative forms of success. This includes the adoption of subject positions that are diametrically opposed to the normative values of the school discourse, and even involves the repulsion of the education system (Benjamin, 2002;Hall et al., 2004). Craig (2007) draws upon the example of the SEAL programme to indicate how similar reactions may prevail within SEL interventions. The author argues that the limited effectiveness of the programme's pilot (Hallam et al., 2006), was because it provoked resistance and rebellion amongst participants who objected to the undue interference, and resented the coercion to emote in alignment with normative and uniformly prescribed criteria. However, at present such assertions remain speculative rather than empirically substantiated.
This paper seeks to explore young people's lived experiences of participating in a targeted SEL intervention, the Student Assistance Programme, in order to generate new theoretical and empirical insights into the manifestation of iatrogenic effects within this educational domain. The concepts of negative labelling, positive deviancy training, and the 'failing subject' position scaffold this exploration. The Student Assistance Programme (SAP) was first developed in the USA and has been incrementally introduced to Wales since 2003. It has been recommended by the Welsh school inspectorate as best practice in managing challenging behaviour (Estyn, 2006). The programme aims to provide a developmentally appropriate and supportive context where children and young people may develop social and emotional competencies (Watkins, 2008). Intervention includes a targeted student support group. The group comprises two trained facilitators and 8-12 students. Eight sessions are delivered per intervention course. Sessions are scheduled on a weekly basis and last one to two lesson periods. The intervention is theoretically underpinned by social cognitive theory (Bandura, 1986) and the social development model (Catalano and Hawkins, 1996).

Methods
This paper presents qualitative data from a case-study of the Student Assistance Programme. Data were generated from four mixed-sex secondary schools (ages 11-18) in Wales, across two Local Education Authorities. Three schools were from the post-industrial South Wales Valleys town of Pen-y-Cymoedd: Ysgol-y-Dyffryn, Ysgol-y-Glyn and Ysgol-y-Cwm. One school was from the rural area of Castell Mawr: Ysgol-y-Foryd. Schools were purposively selected on the grounds that they implemented the intervention and were socio-economically and academically diverse.
Students were recruited to the study through their participation in the intervention. One student support group was identified in each school. Support groups were selected to ensure comparability across students in terms of academic year group. Students were drawn from Year 8 (age 12-13) and Year 9 (age 13-14). The three schools in Pen y Cymoedd recruited mix sex groups, while Ysgol-y-Foryd recruited boys only. Students in these groups were invited to take part in the study. Parents and students gave consent and all group members participated. A total of 41 students took part (21 male and 21 female). All students were from a white background.
A number of qualitative methods were used to explore the intervention experience. Participant observation of all support group sessions (n = 32) was carried out (Junker, 2004) in order to explore students' reaction to the programme in situ. A participatory approach to observation, whereby the researcher partakes in intervention activities alongside students and engages in similar levels of emotional disclosure, may diminish some of the power differential between the researcher and the researched. Structured observation schedules were completed immediately following each session. These schedules recorded: reach; duration; quality of delivery; adherence and adaptations to programme materials; student interaction and responsiveness. Supporting ethnographic field notes documented additional observations.
On conclusion of the observation phase focus groups were conducted in each school (n = 6) to provide students with an opportunity to reflect on their experience and to ensure participant validation of observation data (Kitzinger, 1994). Students expressed a preference for this method as it preserved the permissive and secure context created by the intervention process. However, it should be acknowledged that while focus groups may provide a productive space to meaningfully explore experiences, they also risk censoring deviations in opinion or exacerbating the presentation of endorsed views or behaviours (Kitzinger, 1994). Focus groups lasted 60-75 minutes, and were audiotaped and transcribed in full. Pseudonyms for education authorities, schools and students are used throughout. Cardiff University's School of Social Sciences Ethics Committee provided ethical approval for this study.
Analysis of the data drew upon a thematic approach, encompassing analytical techniques associated with grounded theory (Glaser and Strauss, 1967;Strauss and Corbin, 1990). In the first instance the corpus of data was read and coded to identify the The unintented consequences of targeting 5 presence of a priori themes under consideration, notably students' lived experiences (Strauss and Corbin, 1990). Themes that had not been pre-specified were identified in a second reading of the text. Following this process, axial coding was conducted, whereby codes were causally linked to build a category (e.g. intervention referral) by explaining the conditions that give rise to the category, the context in which it is embedded, the strategies through which it is managed, and the consequences of these strategies (Strauss and Corbin, 1990). Data collection and analysis were conducted iteratively, with emergent themes being explored in future observation sessions and interview schedules. N-Vivo software supported the process of data analysis.

Findings
The data suggest four key processes through which targeting may influence young people's lived experiences of SEL interventions. These processes are related to (1) the experience of being identified by the targeting criteria as in need of SEL intervention (2) the composition of the intervention, which results from the application of targeting criteria.

Identification by targeting criteria
The SAP stipulates four referral routes for the intervention: self-referral; teacher referral; parental referral; community services referral. In practice all schools relied on the school coordinator to make referrals. School coordinators are staff who oversee programme delivery. Despite employing the same referral strategy, schools employed different targeting criteria, with the conferred 'failing' subject position reflecting the dominant educational discourse. Schools espoused two main discourses: (1) Care. This was manifest in all schools in Pen-y-Cymoedd: Ysgol-y-Dyffryn, Ysgol-y-Glyn and Ysgol-y-Cwm. In this instance schools were seen as offering a 'sanctuary' amid the chaotic lives of students located in a socio-economically deprived community. As one head teacher stated: I don't want to make either value judgements or gross generalities, but where this is economic deprivation, attached to that is going to be certain circumstances where children don't have settled lives. And in a number of instances the school as an education institution offers stability, and normalcy, and constancy, and a sort of regulation to their lives that is needed. (Mr Griffiths, Ysgol-y-Dyffryn) Targeting criteria identified students experiencing social and emotional problems, particularly within school and the family. Intervention was intended to offer additional support and attention to overlooked students. Effort was made to construct a positive targeting experience, with students being offered a choice to attend, and the school downplaying the targeting process in order to avoid potential stigma.
(2) Discipline. This was espoused by the school in Castell Mawr, Ysgol-y-Foryd. It also existed, to a lesser extent, in Ysgol-y-Dyffryn. This approach involved schools drawing upon SEL intervention in order to eliminate the problematic behavioural repertoires of challenging students who compromised the academic 6 Rhiannon Evans et al.
learning of the wider student population. As one staff member said of recruited students: And the eight boys are renowned for being uncontrollable in a classroom setting. And that's when they're split up. And to bring them all in one room, all together, is never thought of by some members of staff in the school. (Mr Rees, Ysgol-y-Dyffryn) Thus students assigned the 'failing' subject position were those deemed undesirable within the classroom. There was no evident effort to provide a positive targeting experience. Both observation and focus group data indicated that different targeting criteria, and the educational discourses embodied in these criteria, had a differential impact on students' experiences.

Negative labelling: inspiring resistance and rejection
Identification by the targeting criteria was experienced as negative labelling by some students. This was evident in Ysgol-y-Foryd, where the educational discourse ensured that targeting was reserved for 'naughty' students. Although not explicitly informed of the reason for being targeted, intervention participants were able to decipher the criteria because of the fact that a discrete friendship group, governed by 'naughty' behavioural norms, had been identified. Conscious of an educational system and discipline policy built around the eradication of naughtiness, students were aware of the negative connotations of this label and interpreted their targeting as indication of the school's desire to exclude them: Jayden: They want us out. Neil: They want us out of lessons anyway. (Focus Group, Ysgol-y-Foryd) There were detrimental consequences of this unarticulated but evident process of negative labelling. Primarily it served to engender resistance to the school, as illustrated by the exacerbation of anti-school attitudes and behaviours. Students offered increasingly visceral responses to questions regarding school and teachers, regularly describing the latter as 'cunts', 'niggers', 'pricks' and 'slags'. Such sentiments were accompanied by reticence to modify ostensibly naughty behaviours and a determination to amplify anti-school activity through more disruptions or truancy: Rhiannon: Right, so do you feel any different about school since you've come here? Callum: No it's still shit. Neil: I think it just makes us think more though that the teachers hate us. Jayden: Yeah, for picking us to be in this group.
The unintented consequences of targeting 7 Neil: Everyone shares their opinion of the teachers and then it just gets stuck in your head. Liam: All of us lot should just go out of school. (Focus group, Ysgol-y-Foryd) Understanding of this process of negative labelling, and the role of educational discourses in framing this process, may be further supported by contrasting Ysgol-y-Foryd with the case of Ysgol-y-Glyn. Here the discourse of 'care' ensured that the intervention was framed as the school's effort to be more attentive and supportive of students. As a result the elite/failing subject was not brought into being. Students consistently spoke of how fortunate they were, and how they finally felt visible in a place where they had often been overlooked. This was evident during a follow up session with students where they reflected on their transition into the intervention: Faye said she felt lucky and special to have been chosen and it was a really good mix of people. She also felt that being with people she didn't really know made her realise that you don't know who has a problem, and in fact everyone has some problem or other. She said this made her feel better about herself and made her think that she wasn't the only one who was dealing with things. (Field notes, Ysgol-y-Glyn)

Coveted labelling: claiming intervention capital
While labelling was experienced negatively by some students, it could simultaneously be interpreted as a much coveted process; students perceived the 'failing' subject position to be both undesirable and desirable. Indeed students' initial effort to reject the label of the 'naughtiest lot' in Ysgol-y-Foryd was quickly subsumed by efforts to claim the conferred subject position. They recognised that the label could be used as a resource to renegotiate and even strengthen their positioning within the broader peer group, as identification by the targeting criteria provided a signifier of their uncontrollability and anti-authoritarian attitude, both of which were considered desirable within the school context. Leighton: Seriously it just gives you a little bit of respect. Rhiannon: Being naughty? Leighton: Yeah, you're walking around the school and other people [are David: 'How] you doing man?' Leighton: And you can walk out and they're like 'BOOM' [snaps fingers]. (Focus Group, Ysgol-y-Foryd) The desirability of the label was also manifest in debates over who had the right to claim intervention membership, and which students could not justify referral. One student became a particular point of contention, with other participants concerned he had not cultivated a 'naughty' persona and so was marring the group's reputation: Jayden: That's what I really want to know. Why Gareth? Rhys: Is it really though? Jayden: He's never got in trouble. Oh.] There is no naughty people in our year than us. Jayden: But Gareth. He listens. (Focus Group, Ysgol-y-Foryd) Thus for these students, labelling as intervention participant offered an important form of 'intervention capital', which served as a powerful commodity within the school context, being employed to garner respect, status and even fear amongst peers. However, as with negative labelling, this positive labelling also has the potential to carry iatrogenic intervention effects. This is because the intervention capital conferred by identification could only be retained through unwavering resistance to intervention messages, combined with a refusal to change; the notoriety of the 'naughtiest lot' was contingent on continued anti-school attitudes and behaviours. Where students sought to move away from the failing subject position through engagement with the intervention they risked losing this elevated status.

Peer group composition
Application of different targeting criteria in each school ensured that the group composition within the intervention also differed. There were three different configurations. In Ysgol-y-Glyn and Ysgol-y-Cwm identification and consultation of students with various social and emotional problems meant that a range of students were referred. The group comprised weakly bonded and even unfamiliar peers largely owing to the peripheral social position of many students. In Ysgol-y-Foryd, targeting of 'naughty' students ensured identification of a pre-existing peer group that existed in a long yet tumultuous friendship that was closely bonded through endorsement of such behaviours. The dual discourses in Ysgol-y-Dyffryn meant a mixed composition: closely bonded anti-school students and marginalised, socially withdrawn peers. These complex configurations created highly politicised and conflicted group dynamics, which could potentially lead to adverse outcomes for all intervention participants.

Seeking safety: privileging the familiar over the strange
In Ysgol-y-Foryd and Ysgol-y-Dyffryn the presence of an entrenched friendship group ensured that some students often privileged these relationships over the intervention, and sought to protect them within the alien context of the SAP. Indeed for many of the students SEL offered a previously un-encountered educational experience, with many scaffolding their identity with notions of 'roughness' to describe their lack of ease with exploring emotionally intimate relationships. Equally, students would retreat into their friendship groups in order to defend them from the 'uncool', socially isolated peers who had also been targeted.
The process of seeking safety within the familiarity of friendships was most evident in the case of Ysgol-y-Dyffryn. The pre-existing friendship group, led by Samantha, continually demonstrated their latent resentment of forced association within ostensibly unpopular peers. They were vociferous in complaining about who they sat next to and employed exclusionary tactics when asked to move into dyads The unintented consequences of targeting 9 or triads with non-friends to complete a task. Primarily, this included emphasising social activities they shared outside of the group and expressing contempt for the outsiders' own interests: Andrea asked that we tell our partner about a time when we were happy. I heard Nathan tell his partner about an online gaming community he had recently joined. His partner rolled her eyes and started to chat to a friend who sat next to her. Increasingly frustrated by her lack of interest or knowledge he became aggressive. Turning his chair he physically removed himself from the pairing. Joyce, the other facilitator, said she didn't understand what he was going on about either, and he needed to explain himself better if people were going to listen. (Field notes, Ysgol-y-Dyffryn) Attempts to make Nathan aware of his position outside of the dominant friendship group culminated with Samantha's explicit instructions not to associate with him.
As we came to the end, Andrea asked if anyone wanted squash and biscuits. As they crowded around the table I overheard Nathan ask if the squash was sugar free, before concluding that he would just have some water. At this point a huge argument erupted and Samantha starting shouting 'Nathan's got diabetes, everyone stay away from him, he's got diabetes'. Four of the girls quickly moved to sit with Samantha at the other side of the classroom. Nathan turned to me, and running his hand over his stomach said 'I'm not diabetic. I'm just trying to lose some weight'. (Field notes, Ysgol-y-Dyffryn) While these exclusionary activities were significant in affirming the friends' insider status, and allowed them to demonstrate their commitment to these safe and familiar relationships, they had detrimental impacts for those constructed as outsiders. The intervention offered a negative and alienating experience, even culminating with their withdrawal. Nathan eventually refused to take part, claiming he would rather be in lessons.

Deviancy amplification: re-negotiating friendship hierarchies
The presence of entrenched friendship groups within the intervention had an impact on its members as the internal dynamics of the group were renegotiated, with this process frequently manifesting in deviancy amplification. Indeed for many this political activity was at the fore, with the intervention offering a site where their positioning could be augmented. This jostling for position was clearly evident in Ysgol-y-Foryd.
As the group commenced, the majority of students were engaged in activities and tentatively offered emotional narratives. Mid-way through the course Jayden joined the group. He was described by facilitators as 'the most dominant. He has this reputation of being a bit of a leader'. His presence was a significant influence on his friends, who increasingly sought to gain his favour by complaining about activities when he said 'there is no point in doing it', and mocking the facilitators when he said they were 'boring'.
Attempts to elevate status also saw students increasingly brag about the high risk, and often illegal activities that were esteemed by the group. Jayden instigated this bragging effect, telling the group that he had broken into school following heavy cannabis use because he had 'serious cotton mouth' and wanted to steal from the canteen vending machine. Following a positive response to Jayden's account, the boys were 10 Rhiannon Evans et al. able to estimate the stories that would improve their status, and sought to develop a similar portfolio of tales: As soon as they came through the door Neil and Leighton could not wait for everyone to know what they had done. They tried to appear very relaxed, with one swinging on the chair and the other leaning against the table, but both spoke very loudly about the 'crazy shit' they had been up to. Eventually, on mentioning they had smoked nine joints a day, the others took an interest. Talking over each other, they were quick to tell that they had been so stoned they had walked four miles to Tesco for munchies but it was shut and Leighton fell asleep on a bench outside. Neil called for a taxi, but when it arrived Leighton would not wake up so it drove off and they had to walk all the way home again via McDonalds. (Field notes, Ysgol-y-Foryd) The role of deviancy in renegotiating status within hierarchically structured friendships can be further illustrated through juxtaposition of this example with students in Ysgol-y-Glyn. Because of the discourse of 'care' focusing on individual needs, participants in this group were diverse. None of these individuals were bound by closely bonded friendships or entrenched behavioural norms. Deviancy amplification was not a feature of this group. Rather students behaved differently from how they habitually acted within the broader context of their lives. Gemma was described by a facilitator as 'the biggest bully in the school', and had been subjected to several attempts to permanently exclude her. Yet in the absence of friends Gemma recast herself as a pro-social student who supported her peers, suggesting deviancy reduction rather than amplification.
Faye said that some friends she had argued with had been really horrible that morning and had asked her what she was wearing. She had told them that she was wearing the same as always and they said she looked awful .... Once this had finished and nothing else seemed like it was going to emerge Emma asked us to excuse her while she went to get some paper. When she left Gemma turned to Faye and told her to stand up to the girls in order to make them stop. She said 'the next time they say that you should tell them you're wearing clothes and then walk off'. There were some furious nods and 'yeahs' around the group. (Field notes, Ysgol-y-Glyn)

Discussion
This paper reports the lived experiences of young people participating in a targeted SEL intervention. Until now the potential for iatrogenic effects within such interventions has remained under-theorised, with a dearth in empirical examination. Through the qualitative exploration of students' identification for the Student Assistance Programme, this study has identified four key processes that may facilitate the emergence of unintended or adverse effects. These mechanisms resonate with and refine existing concepts, while offering new insights.
The first two processes were the result of students being identified by schools' targeting criteria. Negative labelling, which is an established phenomena within targeted interventions (Wiggins et al., 2009), was clearly evident. However, while labelling has predominantly been theorised as negative, invoking notions of stigma and exclusion, this study provided a more complex and nuanced account of its manifestation. For some students the failing subject position actually held positive connotations, with The unintented consequences of targeting 11 the increasingly coveted label conferring a degree of intervention capital that could be drawn upon within the peer group to elevate status. Yet despite the apparent positive reaction to targeting in this instance, a similar decrement in outcomes was apparent, with students resisting and rejecting the intervention in order to retain their acquired positioning.
The second two processes relate to the composition of intervention groups, which is a result of the targeting criteria employed. In a number of schools targeting according to a specified behaviour identified entrenched friendship groups that were characterised and bonded through endorsement of such behaviours. These friendships were frequently privileged over the activity of programme participation; within the unfamiliar context students would seek safety in the familiarity of these relationships. Commitment was often signified through the rejection of ostensible 'outsiders', who in turn experienced the intervention negatively. Students also sought to jostle for position within their hierarchically structured relationships through the amplification of deviancy. This process of amplification may be explained as a 'bragging effect' (McCord, 2003). Through the repeated exchange of stories of deviancy participants were able to estimate the reports that would garner the most positive responses. They would then seek to develop a portfolio of tales that they could share during future encounters in order to enhance their status.
Although the latter two processes provide some indication of an iatrogenic group effect, they differ substantially from those explored in other studies (Dishion et al., 1999;McCord, 2003;Cho et al., 2005), specifically because of the nature of the relationship between group members. Weiss et al. (2005) suggest that positive deviancy training should only occur when individuals are exposed to unknown peers and novel forms of deviancy, with Cho et al.'s (2005) concept of 'peer contagion' suggesting new incidents of deviant relationships. However, as this study has shown, it is the very fact that peers exist in closely bonded friendships prior to commencement of the intervention that may be important. Anti-school and anti-intervention behaviours were not the result of novel advice or knowledge being exchanged, but rather were a consequence of students' conducting vital identity work. Deviancy formed part of the political activity of friendship. Thus there was a phenomenon of deviancy amplification as opposed to deviancy training, with this amplified activity serving as a mechanism for achieving both emotional and reputational security.
To further understand the emergence of these iatrogenic effects it is vital to explore pertinent contextual features that serve as supporting or inhibitory factors. In this study adverse outcomes were not an inevitable consequence of targeting. Rather they were contingent on the educational discourses espoused, with the discourses of care and discipline having a differential impact. Where there was a supportive context students often indicated their pleasure at being identified for additional learning as they interpreted it as an act of care that was conducive with the broader organisational ethos. Equally, the schools' focus on individual needs rather than collective group behaviours ensured that the intervention was not dominated by hierarchically structured friendship groups and thus students enjoyed exposure to new peers. Conversely, where there was a discourse of discipline, students interpreted the intervention as a routine act of punishment for their non-normative behaviour. This evident rejection by the school only served to strengthen resistance and further alienate students, all of which culminated in the exacerbation of anti-school and antiintervention attitudes.
The emergence of iatrogenic mechanisms suggests a number of implications for targeted SEL programmes, and targeted interventions more broadly. This is not to suggest the abandonment of targeting per se, as in some instances it may constitute the most appropriate intervention approach (Weare and Nind, 2011). Rather it is to acknowledge and accommodate the fact that young people may have complex and conflicted responses to the subject positions conferred during the targeting process. In the first instance there is an evident need to explore these multifarious reactions more fully through comprehensive qualitative work that avoids adultist interpretations and privileges the meanings and perspectives of those targeted (Greenberg, 2010).
Second, the educational discourses espoused by schools must be attended to. Alongside formal delivery of intervention curriculum, there is a need for supplementary work that introduces contextual restructuring when the more distal domain of organisational influences fails to support programme processes and outcomes (Spratt et al., 2010). In order to achieve this end, development and evaluation of targeted interventions may benefit from the integration of a socio-ecological model that theorises the multiple domains of influences that structure students' lived experience of intervention (McLeroy et al., 1988).
Third, there needs to be careful consideration of intervention composition. Existing literature has indicated that positive deviancy training is theoretically underpinned by diffusion, and the dissemination of deviancy through the peer group (Cho et al., 2005;Weiss et al., 2005). This suggests that targeted interventions should avoid the aggregation of unfamiliar peers because of the risk of knowledge exchange. However, as this study has shown, diffusion may not be the central process at work here, as the aggregation of entrenched peer groups may be equally conducive to the emergence of iatrogenic effects. As a result the composition of targeted interventions needs to be meticulously crafted, striking a balance between exposing participants to new peers that bring novel forms of deviancy and recruiting them as part of established friendships.

Limitations of the study
Although the intention of this study was to theorise the conditions under which iatrogenic effects may manifest within targeted SEL interventions, the data are based on a limited number of cases. Although there is no evidence to suggest that the four schools were atypical of secondary schools within Wales, the local educational authorities were selected for being the most socio-economically diverse regions implementing the intervention. However, despite some limitations in generalisability these cases were instructive in the development of theoretical propositions which could be explored across other educational contexts (Yin, 1994).
The most methodologically and scientifically robust approach for verifying these theoretical propositions and testing identified iatrogenic mechanisms is to conduct subgroup analysis of outcome data from randomised controlled trials. Subgroups may comprise the targeting criteria employed and the subsequent intervention configuration. Data from process evaluations will be required to map the targeting process.
The unintented consequences of targeting 13