Nurturing learning or encouraging dependency? Teacher constructions of students in lower attainment groups in English secondary schools

Abstract ‘Ability’ or attainment grouping can introduce an additional label that influences teachers’ expectations of students in specific attainment groups. This paper is based on a survey of 597 teachers across 82 schools and 34 teacher interviews in 10 schools undertaken as part of a large-scale mixed-methods study in England. The paper focuses on English and mathematics teachers’ expectations of secondary school students in lower attainment groups, and explores how low-attaining students are constructed as learners who benefit from specific approaches to learning justified through discourses of nurturing and protection. The authors argue that the adoption of different pedagogical approaches for groups of low-attaining learners to nurture them may in some cases be fostering dependency on teachers and cap opportunities for more independent learning. Furthermore, more inclusive whole-school learning-culture approaches may better allow for students across the attainment range to become independent learners.


Introduction
Teachers and teaching practices play a pivotal role in students' experiences and outcomes at school. Whilst most schools would affirm their commitment to treating all students equally, variation in teachers' expectations of specific groups of students has been identified across different education systems (Campbell, 2015;McKown & Weinstein, 2008;Rubie-Davies, Hattie, & Hamilton, 2006). Differing teacher expectations influenced for example by students' ethnicity can significantly impact on students' educational outcomes and experiences of schooling (Archer, 2003;Gillborn & Youdell, 2000;van den Bergh, Denessen, Hornstra, Voeten, & Holland, 2010). Furthermore, the act of labelling whereby a student is, for example, identified as being 'at risk' of low educational outcomes may have counterproductive consequences for his/her attainment as the teacher's expectations for KEYWORDS ability grouping; setting; independent learning; teacher expectations; labelling ARTICLE HISTORY received 16 august 2017 accepted 1 february 2018 that student may be influenced by the label of an 'at risk' student (Corrie, 2002;Pianta & Walsh, 1996). In education systems where students are grouped by 'ability' or attainment, attainment groups can introduce an additional dimension through which teachers view students, and may lead to differential expectations depending on the student's placement in the hierarchically ordered attainment groups (Boaler, Wiliam, & Brown, 2000;Ireson & Hallam, 2003;Macqueen, 2013). The notion of 'ability' speaks to an outdated yet pervasive idea of each student possessing an innate academic ability that schools and teachers simply need to tap into (see e.g. van IJzendoorn, Juffer, & Poelhuis, 2005), and hence we prefer to refer to attainment rather than 'ability' grouping.
Grouping students by their prior attainment or 'ability' in specific subjects is a common practice in English secondary schools with OECD (2013) figures suggesting that 95% of students are taught mathematics in attainment groups. Such grouping practices have been challenged for their detrimental impact on student outcomes, potentially lowering educational attainment and self-confidence amongst students in lower attainment groups (Boaler et al., 2000;Francis et al., 2017a;Oakes, 2005;Slavin, 1990). Despite the wealth of research evidence casting doubts on its benefits, attainment grouping continues to prevail in English secondary schools, and has often been constructed as the preferential approach in government policy (see e.g. DCSF, 2005) over other approaches, such as mixed-attainment grouping, which tends to be perceived as risky and untested (Taylor et al., 2016).
Teachers tend to have lower expectations of students in lower attainment groups and students in these groups have access to a reduced curriculum and a pedagogy providing fewer opportunities for autonomous learning than is afforded to students in higher attainment groups (Hallam & Ireson, 2005;Kelly, 2004;Oakes, 2005). Elsewhere (Francis et al., 2017b), we have discussed how together the practices associated with attainment grouping can contribute to a deficit view of students in the lowest attainment groups. This paper examines English and mathematics teachers' expectations of secondary school students in lower attainment groups. It explores how low-attaining students are constructed as learners who benefit from specific approaches to learning justified through discourses of nurturing and protection. It draws on debates around dependency culture emanating from social policy and politics (see e.g. Hartley & Taylor-Gooby, 2013) to put in perspective teachers' construction of students in the lower attainment groups as more dependent on their teachers' support in comparison with their peers in higher attainment groups. We will argue that the adoption of different pedagogical approaches for groups of low-attaining learners in order to nurture them may in some cases be fostering dependency on teachers and cap opportunities for more independent learning. The extent to which individual learners in these groups can become less dependent on their teachers is likely to vary, but analysis of the range of learner needs within the lower attainment groups is beyond the scope of the project. Instead the focus here is to explore some of the potential implications of set placement on the scope for students in the lower attainment groups to be working independently within their classrooms and the complexities of finding a balance between nurturing learning and creating or fostering teacher dependency.
There are four parts to the paper. In the first part, we briefly outline features of the social policy debate concerning the welfare dependency culture thesis and how this has been extended to discussions of underachievement in education in the UK. We sketch out key arguments that have been made in relation to students' agency in formal schooling processes and their dependence on teachers before summarising findings from recent literature on teaching and learning in lower attainment groups. In the second part, we explain our ongoing mixed-methods research project into grouping practices in English secondary schools. The third part presents findings from our survey and interview data in relation to teaching and learning in lower attainment groups. The concluding part reflects on the teacher constructions of lower attaining students as being 'more dependent on people' and the impact these may have on the students' learning trajectories and the self-fulfilling prophecy of students in lower attainment groups being more likely to have lower educational outcomes (Francis et al., 2017b). We also draw attention to the potential for more inclusive wholeschool learning cultures to foster development of positive independent learner identities.

Student agency and teacher dependency in lower attainment groups
The ongoing social policy debate regarding a potential welfare dependency culture is insightful here as an analogy to illustrate aspects of grouping practices and related pedagogical approaches that may unintentionally be instilling a culture of dependency in lower attainment groups. The welfare dependency culture thesis, formulated by Charles Murray (1984Murray ( , 1990 as the propensity of guaranteed welfare state provisions to reduce welfare claimants' incentives and motivation to seek paid employment, has been very influential in policy-making and in shaping popular understandings of the nature of poverty in the western world. Empirical research in the UK, however, has found no evidence to support the thesis (Hartley & Taylor-Gooby, 2013;MacDonald & Marsh, 2005;Shildrick, MacDonald and Furlong, 2016) and there is very limited support for it amongst social policy researchers (Dunn, 2013). The dependency culture thesis is essentially a formulation of the 'blame the individual' type explanation of poverty, which ignores the powerful influence of the socio-economic structures that constrain individual agency in overcoming barriers to, for example, accessing paid employment (Hutton, 1995;Standing, 2011). 'Blame the individual' discourses have, however, gained a strong foothold in UK political and media representations of poverty and life on welfare (Hancock & Mooney, 2013;Pantazis, 2016). Furthermore, individuals experiencing poverty themselves often tend to subscribe to these discourses (Shildrick & MacDonald, 2013).
In the UK, the dependency culture thesis has been more recently linked to underachievement in education through the discourse of 'Troubled Families' that identifies disadvantaged family units as the source of a range of intergenerational social problems (Casey, 2012). Shildrick, MacDonald, and Furlong's (2016) research in two neighbourhoods with a history of industrial decline in the local area focused on families that had experienced significant periods of worklessness. Whilst they found some individuals and families in these neighbourhoods engulfed by a multiplicity of problems including 'failed' schooling, Shildrick, MacDonald and Furlong (2016) robustly challenge the 'Troubled Families' discourse of an assumed negative culture of values and behaviours that is transmitted from one generation to the next (Welshman, 2013). Shildrick, MacDonald and Furlong (2016, 831) crucially interviewed different generations of the same families and found considerable evidence of the younger generations 'actively resisting' the patterns of disadvantage of the previous generation being repeated. In their active resistance, these individuals can, however, be argued to possess only 'bounded agency' (Evans, 2007). This is because their agency is bound not only by broader socio-economic structures, but potentially also by cultural meanings and understandings of appropriate behaviour anchored in local histories (Walkerdine, 2010(Walkerdine, , 2016. These shared meanings may, for example, construct particular types of jobs as 'embarrassing' , and to be avoided even if that means being dependent on welfare (Jimenez & Walkerdine, 2011).
Focusing specifically on the secondary school milieu, students are likewise only able to exercise 'bounded agency' as the 'authoritarian structures of schooling' constrain their autonomy within school institutions (Francis & Mills, 2012, p. 260). Formal schooling systems have long been critiqued for propagating dependency on authority figures for all children at the expense of encouraging emotional and intellectual independence (Harber, 2004;Illich, 1971). In the compulsory phase of education all students can indeed be viewed as being intellectually dependent on their teachers, but the degree of this dependency on teachers is generally assumed to decrease with age so that 'good education at every level ought to inculcate, develop, and support students' ability to think for themselves' (Goldberg, 2013, p. 168). Approaches to teaching and learning that encourage students to become more independent learners have for example been theorised as constructivist (Levin, 2000) or as self-directed learning (Bolhuis & Voeten, 2001). Gorard and Rees (2002) further stress the importance of developing a positive independent learner identity at school as a process that can shape an individual's outlook on learning for life. They argue that the development of a positive independent learner identity is crucial, not only for the student's experiences and outcomes of compulsory education, but also for his/her lifelong learning trajectories.
A recent Education Endowment Fund report, however, identifies that in many schools teaching assistants are being asked to support students in ways that increase their dependency on the support rather than enhance independence in their learning (Sharples, Webster, & Blatchford, 2015). This is of particular importance for students in the lower attainment groups as they are more likely to have statutory education and health-care plans or special educational or behavioural needs and be supported by teaching assistants (Webster et al., 2011). From this programme of research a recommendation arises that teaching assistants should be encouraging more self-scaffolding from the students so that the teaching assistant's 'default position is to observe pupil performance, allowing time and space for them to process, think and try the task independently' (Webster, 2017). Important findings also arise from Mairead Dunne and colleagues' (Dunne et al., 2011) study where they examined 13 case study schools selected for making good progress with low-attaining students, and identified three main ways in which these schools were approaching the teaching and learning of lower attaining students: focusing resources, tailoring curriculum and pedagogies and fostering positive learning environments. In their study, examples of focusing resources included teaching lower attaining students in smaller classes and a greater deployment of support staff, such as teaching assistants and specialist staff. Tailoring curriculum and pedagogies sought to meet the perceived needs of the lower attaining students, and examples of fostering positive learning environments ranged from emphasising strong teacher-pupil and peer relations to making pastoral support available in the school.
The importance of fostering positive learning environments is also highlighted by the inclusive education movement's call for the discrediting of fixed ability mind-sets that can be damaging for all, but particularly stigmatising for lower attaining students (Florian & Linklater, 2010;Hart, Dixon, Drummond, & McIntyre, 2004). The idea of transformability developed by Hart et al. (2004) for example draws on the three pedagogical principles of co-agency, trust and community (learning for all, not just for some students) to illustrate how to enhance students' capacity to learn instead of relying on predetermined notions of how much each student is capable of learning. We will next outline our current research project and the data that informs our discussion on teaching, learning and independent learning in lower attainment groups.

Research methodology
This paper draws on data generated as part of an ongoing mixed-methods study examining grouping practices in English state-funded secondary schools. This large-scale mixedmethods study is funded by the Education Endowment Foundation and explores setting and mixed-attainment practices in English and mathematics, and their potential impact on student outcomes and experiences of schooling (see Francis et al., 2017b for more background to the study). As part of this study a combination of research methods is being used to provide multi-faceted student and teacher perspectives on different grouping practices in action. English and mathematics were chosen as subjects that have historically been prioritised in the national curriculum, but also given the diversity in content and pedagogy. This paper focuses on teacher perspectives on teaching and learning in lower attainment groups in the 126 English study schools that group students by attainment for English and/or mathematics. Schools were recruited to the study through a number of different routes including publicity in teaching and school leadership professional networks and direct contact by the research team (see Taylor et al., 2016 for more details) and allocated to an intervention or a control group for the duration of this two-year study. English and mathematics teachers of Year 7 students at the participating schools were invited to respond to teacher questionnaires over the period from November 2015 to January 2016. The teacher questionnaire items were partly drawn from Ireson and Hallam (2003) and Hallam and Ireson (2005) with new items developed for example to ascertain student grouping and group placement practices and the questionnaire was extensively piloted with Year 7 teachers in spring 2015. The final questionnaire included a total of 30 statements on teacher grouping practices asking teachers to respond using a five-point Likert scale (see Appendix A for these statements). This paper discusses responses to nine of these 30 statements as the items most pertinent to a discussion of teaching, learning and independent work in the lower attainment groups.
The questionnaires were completed online and responses received from 597 teachers from 82 schools. Of the respondents, 63% were female, 36% male and 1% unknown; 84% identified their ethnicity as White, 8% Asian, 2% Black, 2% Mixed, 2% Other, 2% unknown; and 61% indicated that the main subject they taught was mathematics, 29% indicated English, 4% other and 6% unknown.
Ten schools were also selected to take part in qualitative data collection between October 2016 and April 2017 to provide more depth and breadth to the questionnaire data. The sample of schools for the teacher interviews and focus groups with students is geographically spread across England and demonstrates a range of school practices based on the quality of education provision as judged by Ofsted, the English school inspectorate, and the number of attainment-set levels operated within the school. Table 1 provides a breakdown of the sample schools by Ofsted rating, the number of attainment group levels they operate and the number of teachers interviewed. Five of the schools were from the intervention group, and at the start of their participation in the study were asked to follow the project best-practice principles on grouping, including operating at most four levels of attainment groups, not allocating the least experienced teachers to the lowest attainment groups and allowing for movement between groups at fixed points in the year. The other five schools in the qualitative fieldwork sample were from the control group, and at the start of their participation in the study were asked to simply maintain their existing grouping practices.
A total of 34 face-to-face teacher interviews were conducted in the 10 selected schools. Each of the schools was asked to nominate two English and two mathematics teachers currently teaching Year 8 groups to be interviewed. Table 1 provides details of the sample of teachers interviewed, and shows that the sample was evenly split by gender. The majority of the teachers self-categorised their ethnicity as White (n = 23), nine as Asian and two as Black. There was a range in the length of teachers' teaching experience with six interviewees with less than 3 years of teaching experience, 10 with 3-6 years of teaching experience, nine with 7-10 years of teaching experience, and a further nine interviewees with 10+ years of teaching experience.
The individual teacher interviews were semi-structured and on average lasted 40 min. Teachers were asked to talk about how students are grouped in their school and how they find teaching their current classes. Teachers were informed that their participation was voluntary and that the data would be kept confidential. All interviews were recorded, transcribed and pseudonymised. The data were coded using NVivo software (QSR International, Melbourne, Australia), first by coding responses to each interview question, and then analysed for emergent themes arising from, e.g., teachers' descriptions of their current teaching groups. Consistency of the coding was validated by three members of the research team through a process of coding comparison (Bazeley & Jackson, 2013). Although the sample included a similar number of English and mathematics teachers, no discernible subject-specific differences emerged in the analysis of the teacher interview accounts for this paper. There were also no consistently divergent patterns that could be identified in the interview accounts between teachers from the intervention and the control schools, which could partly be explained by significant teacher turnout in two of the intervention schools in particular, resulting in interviewees with limited exposure to the project's best-practice principles as teachers who had attended project training sessions were no longer teaching at the school. Amongst the sample schools there is a range of student intakes in terms of the proportion of students eligible for free school meals and students who speak English as an additional language. Almost all study schools reported placing students in attainment groups strictly based on their prior attainment. Schools in the intervention group perhaps expectedly were more vocal about 'setting on ability rather than behaviour' (Richard, English teacher, School Q) although it should be noted that prior research suggests that the intention of group placement being based on prior attainment is not always realised, with social class for example being a significant predictor of group placement (Archer et al., 2018;Dunne et al., 2007).
In all sample schools, the lowest attainment groups were smaller than the other attainment groups, confirming findings from earlier research (Dunne et al., 2007). There were typically eight to 13 pupils in the lowest attainment group, but one of the intervention schools (School T) reported having up to 28 pupils in the lowest attainment group depending on the size of the cohort.

Teaching, learning and (in)dependent work in the lower attainment groups
Our questionnaire and interview data suggest that in their practice teachers adopt different kinds of pedagogical approaches depending on the attainment level of the class they are teaching. Findings from the teacher questionnaires and the teacher interviews were complementary, and are here discussed in parallel. Table 2 shows how teachers responded to nine of the statements on grouping practices on a Likert scale from strongly agree to strongly disagree (see Appendix A for the full list of 30 statements in grouping practices). Table 2 shows that a clear majority (70%) agreed strongly or slightly that they modify their teaching methods to suit students' prior attainment. Similarly, all teachers interviewed described differences in the kinds of lessons they would typically deliver depending on the attainment-group level with one teacher describing a lesson delivered to a low-attainment group being 'unrecognisable' (Mark, English teacher, School Q) compared with a similar lesson delivered to a high-attainment group.

Structured learning, consolidation and more teacher dependency
In the interviews teachers talked about differences in the pace, content and delivery of lessons delivered to a lower attainment group in contrast with a lesson delivered to a higher attainment group. Whilst the interviewees agreed that teaching and learning in lower attainment groups differed from teaching and learning in higher attainment groups, no clear consensus emerged from the interview narratives about why and how it was different. However, the larger sample size of the teacher questionnaires enables some patterns to be identified in the data. As Table 2 shows, three-quarters (74%) of the survey respondents agreed strongly or slightly that they expect high-attaining students to cover topics in more depth and a majority of teachers (61%) agreed strongly or slightly that they do more repetition and rehearsal with low-attaining students. These findings were reflected in the interview narratives, with Ted (mathematics teacher, School V) for example feeling that he needed to 're-visit' topics more frequently with lower attainment groups and Danielle (mathematics teacher, School U) talking about students in the lower attainment group making slower progress with topics and needing to 'spend more time consolidating' . This perceived need for consolidation was accompanied by more structured work in teaching and learning in the lower attainment groups. As Table 2 demonstrates, 65% of teachers agreed strongly or slightly with the statement that they give more structured work for low-attaining students. In the interviews, this practice was explained as scaffolding learning in the lower attainment groups, and by students in the higher attainment groups being more able to start and maintain independent work over a longer period of time. Just over the majority (52%) of the questionnaire respondents also agreed strongly or slightly that they expect to spend more time on keeping low-attaining students on task. Two mathematics teachers talked about teaching their respective current low-attainment groups as follows: … if I was to say leave them to their own devices too much they would go off task and they wouldn't be able to focus. There is a lot of a lack of confidence in their own ability so you do do a certain degree of teaching but then you have to go around and move around the groups and ensure that they are on task and that they are supported where they need support. (Dominic, mathematics teacher, School T) Also with bottom sets you tend to get more of a behavioural issue so you have got pupils who are disengaged so you have got the behaviour sides of things as well so there is probably a lot more focus on the behaviour and managing the classroom and entertaining them more, you know? Like I always gave out loads more praise points for the bottom set because I was like, 'You sat down. Brilliant!' (Sophie, mathematics teacher, School Q) In their characterisation of students in the lower attainment sets they were currently teaching, seven teachers from five different study schools described a relative lack of independence, constructing students in the lower attainment sets as 'dependent learners' in contrast with students in the higher attainment sets who were viewed as 'independent learners' . For example, an English teacher reflected on students in lower attainment groups as follows: They are more dependent on people, I think, to help them, because they just don't understand things like a higher set would do. I don't know. But, like, I've got a set now that are a bottom set, and there are only a few of them, but they take up more of my time than a higher set. They just need more attention and more support, which is … they still get there in the end, they just need more help along the way with it, really. (Dawn, English teacher, School R) Students in the lower attainment groups tended to be viewed as requiring one-to-one support by the teacher, rather than students in higher attainment sets. Teachers explained that the smaller size of the lowest attainment groups (typically eight to 13 pupils across the school sample) enabled teachers to have more one-to-one time with students so that 'every pupil is spoken to by the end of the lesson at least once about their work' (Sanjiv, English teacher, School S).
Teachers also tended to talk about students in the lowest attainment sets as usually sharing certain characteristics that meant they were likely to require more one-to-one time with teachers. In all study schools bar one, teachers (unprompted) referred to students with special educational or behavioural needs in their descriptions of the lowest attainment sets. This suggested that in the majority of study schools there was a strong overlap with special educational needs and lowest attainment sets. Dunne and Gazeley (2008) attest to the difficulty teachers may face in distinguishing underachievement from special educational needs. In five of our study schools teachers explicitly talked about making the most of scarce resources by, for example, placing students with special education or behavioural needs in the same groups, which would enable 'intervention strategies' to be implemented (Mark, English teacher, School Q), echoing the findings of Dunne et al. (2011) about resource concentration as a common strategy to support lower attaining students.
Students in higher attainment groups were consistently reported by teachers as being expected to work more independently and 'to think for themselves' (Hayley, mathematics teacher, School R). This expectation of more independence from students in higher attainment groups was reflected in the responses to the teacher questionnaire with 64% (see Table 2) of teachers agreeing strongly or slightly that they expected more independent work from high-attaining students. The majority of respondents (65%) further agreed strongly or slightly that they expect high-attaining students to work at a faster rate. This difference in the expected pace of work was also evident in the interviews and echoes findings from the study by Dunne et al. (2011) where a slower pace of delivery was identified as one strategy to enhance low-attaining learners' engagement in learning.
Two teachers were keen to stress that there was variance in teaching and learning in different lower attainment groups depending on the make-up of the group and the group dynamics. One teacher for instance commented that teaching and learning in the lowattainment groups depends 'on the nature of the make-up of the kids that you've got in the group' (Chloe, English teacher, School R). Students in the lower attainment groups were nevertheless described as lacking in confidence and/or resilience as Dominic's earlier quote illustrates. Whilst teachers tended to describe setting high expectations for all students, concerns about over-stretching students in the lowest attainment groups were often expressed in protective or nurturing terms, and on emotional or moral grounds. Students in these groups were described as needing a 'safe environment' (Samira, English teacher, School T) and work that was not pitched too high as otherwise 'it would frighten them to death' (Yolanda, Head of English, School U). Such nurturing and protective stances may partly explain the questionnaire findings of 53% of teachers agreeing strongly or slightly that there are some activities they would not expect low-attaining students to do, and 40% of teachers agreeing strongly or slightly that they provide more practical work for lowattaining students.

Approaches to building confidence and developing independence
Because of the perceived lack of confidence or resilience in the low-attainment groups a third of the teacher interviewees made explicit or implicit reference to strong student-teacher relationships being of critical importance in teaching and learning in these groups. Building strong relationships was viewed as the first stepping stone in the process of instilling confidence and self-belief amongst low-attaining students: I think you've got to work harder on relationships with the less able ones because they've got to trust you because they've been through schooling being told that they're not good enough. And it's that. Whereas with the top sets I'm very clinical. It's, like, just do what you need to do and get out. (Samira, English teacher, School T) You do have to have that caring role where you're gently pushing them forward and not always kind of expecting too much of them. So it's very different to the environment you find in my Set 1 … it tends to be me gently pushing them, checking for understanding a lot more and lots of questioning in there to see what they actually do understand. (Charleen, English teacher, School W) Other interviewees talked more generally about the difference that good teaching and teachers make to low attainers' progress and achievement. Ensuring that the lower attainment groups get 'good quality teaching' (Anthony, Head of mathematics, School U) was identified as important for the progress of students in those groups. Another interviewee suggested that allocating teachers who are recognised within their school as 'the best teacher in the department' (Dev, mathematics, School T) was important in signalling to all students and teachers that low-attaining students were valued.
Some of the interviewees gave examples of how students' independence could be enhanced as part of their learning activities in the lower attainment groups. One example involved a checklist of tasks that students could individually tick off once they had finished each task so that this checklist provided 'almost like a bit of independence scaffolding as well' (Yolanda, Head of English, School U). Webster (2017), however, argues that teachers and teaching assistants should be supporting students through a focus on what they are learning rather than focusing on completing specific learning tasks, as this can enhance the development of independent learning skills.
Furthermore, the apparent emphasis on the teacher-student relationships suggests that the potential for peer-to-peer support as an approach to promoting independence from the teacher may be underutilised. Opportunities for peer support may of course be more challenging to realise in the context of reduced class sizes for lowest attainment groups and the above discussed teachers' expectations and beliefs about students in the lower attainment groups.

Context for teaching and learning in the lower attainment groups
Seven (four from schools in the intervention group and three from the control group) of the teachers interviewed described themselves as constantly challenging themselves in their practice so as not to label learners depending on their prior attainment. One mathematics teacher for example explained that he always looked at the targets for any new students or groups of students 'with a grain of salt' (Marcus, mathematics teacher, School R) to ensure that expectations of individual students would not be capped. Such practices were not, however, evident across the board. Due to external pressures and conflicting demands teachers were not always able to differentiate sufficiently between the needs of all the individual students in the lower attainment groups. A teacher frustrated with a school-wide system of attainment grouping that also took account of behaviour and attitude in student allocation to groups, for example, commented as follows: Some of our Set 6 pupils could easily be in a hierarchical Set 4 class, but if you get put into a Set 6 [bottom] class, you get treated like a Set 6 pupil, and a lot of the work you do is repetitive and dull, and doesn't take you forward. (Max, mathematics teacher, School S) Max suggests that the current grouping practices in his school are not making the most of it as group placement decisions are infused with judgements about student behaviour. His frustration illustrates how a student's placement in the lowest attainment group results in misrecognition whereby his/her placement in the attainment grouping hierarchy can be interpreted by teachers as reflecting the student's innate 'ability' (Archer et al., 2018). There are also echoes of what Hamilton and O'Hara (2011) have insightfully referred to as the 'tyranny' of attainment grouping to highlight the power of attainment grouping as a school organising principle in an education system characterised by high-stakes testing.

Discussion
This paper has explored teaching and learning in lower attainment groups based on teacher questionnaire and interview data collected as part of a large-scale study of student grouping in English secondary schools. Our data suggest that there are specific pedagogic practices being adopted for lower attainment groups. Although there are exceptions, teachers typically do not expect students in the lower attaining groups to be able to follow a similar pace and format to the lesson as delivered to students in higher attainment groups. Teachers tended to think that students in the lower attainment groups were not able to access learning independently without monitoring and support from their teachers and that it was important to continually refresh what had been learnt.
It is possible that some of the apparent variability in attainment group teaching practices arises from differences in how teachers reflect upon, describe and give account of their practice. Nevertheless, the complementary findings from our questionnaire and interview data suggest that, in many schools, teachers are adopting different pedagogic practices to teaching and learning in the lower attainment groups in comparison with higher attainment groups. Furthermore, our findings are in line with prior research that attests to teachers' differential expectations of students and differences in the content and pedagogy of lessons across hierarchical attainment groups (Gillborn & Youdell, 2000;Hallam & Ireson, 2005;Oakes, 2005).
As the work of Susan Hart and colleagues (2004) shows us, teachers' adaptation of pedagogic practices along the attainment grouping hierarchy may restrict the kinds of teaching and learning students in the lower attainment groups are exposed to. In our study, we have found that where attainment grouping and associated different pedagogic practices are used rigidly as a key organising principle for teaching and learning in the school, together they can function as a structural barrier constraining lower attaining students' experiences of teaching and learning. These findings are also evocative of Hamilton and O'Hara's (2011) idea of the 'tyranny' of attainment grouping as a school organising principle. Furthermore, we have identified distinct teacher discourses depicting students in the lower attainment groups as less independent in their learning than students in higher attainment groups. These discourses may reflect school cultures that -intentionally or not -assume teacher dependency in lower attainment groups. Returning to the welfare dependency culture analogy we will suggest that these structural and cultural factors may be limiting the scope for students in the lower attainment groups to develop independent learner identities.
The controversial debates on welfare dependency culture suggest that there are structural, cultural and individual factors at play in generating different welfare and work trajectories in the short and the long term. As accounts of Shildrick et al.'s (2016) interviewees illustrate, personal circumstances clearly impact on individual agency, but structural factors, such as industrial decline in the local area, can increase the likelihood of welfare dependency (Hutton, 1995;Standing, 2011) and communal or cultural factors can shape individuals' responses to welfare dependency across generations (Jimenez & Walkerdine, 2011;Walkerdine, 2010Walkerdine, , 2016. Similarly, there is also a range of past and present factors that may be contributing to a student's perceived low educational attainment (Pianta & Walsh, 1996) and his/her current placement in a lower attainment group. Once a student has been placed in a lower attainment group, the structures of grouping practices and the accompanying school cultures may, however, constrain the extent to which attention is given to each student's past and present factors of low attainment, and crucially how these contributing factors may be countered to enable future learning progress (Hart et al., 2004). As Max's earlier quote suggests, 'if you get put into a Set 6 class, you get treated like a Set 6 pupil' , and this then creates a cycle of restricted opportunity (Wiliam & Bartholomew, 2004). Furthermore, the notion of a 'Set 6 pupil' echoes the simplified media headlines tapping into discourses such as the 'Troubled Families' discourse that construct a unified image of welfare claimants as problem individuals who are ultimately to blame for their need for welfare.
Discourses of dependency were intertwined in the teacher narratives of nurture and support identified as being of particular importance to lower attainment groups. Teachers tended to characterise students in the higher attainment groups as 'independent learners' and students in the lower attainment groups as 'dependent learners' who were not ready for and/or demanding greater independence in their learning. These narratives raise concerns about the extent to which teachers' expectations may be capping opportunities for students in the lower attainment groups to develop their independent learning skills and to enable them to become less dependent on their teachers.
Teachers may feel that they must sacrifice some independent learning skill development to allow for the achievement of other, perhaps more immediately tangible outcomes, such as student progress in standardised student assessments and/or to manage classroom behaviour that may be viewed as incompatible with independent learning (Whitebread et al., 2005). With a longer term view a concern should, however, be expressed around a culture of dependency on teachers being propagated in some lower attainment sets. Attainment group structures and group-specific teaching and learning cultures may be constraining lower attaining students' agency to inhabit and develop positive independent learner identities. As independent learning is considered a key skill in students' later learning trajectories (Gorard & Rees, 2002), its omission may further contribute to the range of factors perpetuating the self-fulfilling prophesy of students in lower attainment sets being more likely to achieve lower educational outcomes (Francis et al., 2017b). Students in some lower attainment groups may be unintentionally directed towards a 'learned helplessness' (Rosenholtz & Simpson, 1984) that can hold back their active participation and engagement in learning over their school career and beyond. Hart et al. (2004) suggest that a student needs confidence to become an independent learner. As we have found in our study, many teachers hope to boost students' confidence and to turn learner identities around through carefully nurturing and protecting students in the lower attainment groups. Interestingly, many of the teachers quoted recognise the damage to pupils' confidence resulting from their designation and labelling as low attainers/'low ability' , and then feel they have to compensate for this through pedagogic practices. As individual classrooms are not immune from the broader school and student hierarchies that continue to shape students' dependent/independent learner identities, efforts may well be better directed towards generating more inclusive whole-school learning cultures that can better accommodate greater fluidity in student identities across the attainment range. We would like to suggest that a key feature of a more inclusive whole-school learning culture has to relate to student agency. All students should have a genuine opportunity to exert positive agency rather than agency being the reserve of students in the higher attainment groups, with students in the lower attainment groups expected to behave more passively as learners, or at worst to express their agency through rebellious or anti-school behaviour. Inclusive whole-school learning cultures are hard to generate (Evans & Lunt, 2002;Graham & Jahnukainen, 2011). Where they prevail they can, however, make an allimportant difference for teaching and learning in lower attainment sets (Hart et al., 2004), and may provide for more students to become independent learners regardless of their past or present attainment level.