The university role in new teacher learning – why it matters: Teach First trainee perspectives

An emphasis on school-led initial teacher training (ITT) in England has marginalised the university role and led to instrumental models of new teacher learning. Rather than commit to continued university involvement in new teacher learning like the rest of the UK, England appears to be following in the footsteps of the USA, where new graduate schools of education (nGSEs) train teachers without university input. School leaders and academics have sought to articulate the value of the university role in initial teacher education (ITE), but there is little understanding of trainee teacher perspectives. This article presents findings from in-depth qualitative research with Teach First trainees at IOE, UCL’s Faculty of Education and Society (University College London, UK). Trainee


Introduction
A turn to the practicum in teacher education, with an emphasis on apprenticeship models, has marginalised the university role, and some argue that this has led to restricted and instrumental models of new teacher learning (Vanassche et al., 2019).Academics, including those working within initial teacher education (ITE), have argued to retain the university role and the role of theory and research in ITE (Vanassche et al., 2019;Furlong, 2019;Winch et al., 2015), but there is little understanding of the experience of those learning to teach.Teacher education is said to be 'desperately under-researched' (Menter et al., 2019: 75).There is a dearth of research exploring student teacher perspectives (Murray et al., 2019), and little understanding of their perceptions of their chosen route into teaching (George and Maguire, 2019).
This exploratory study, conducted from an insider perspective, sought to understand trainees' experience of the Teach First programme, one example of an employment-based initial teacher training (EBITT) programme, run in partnership with IOE, UCL's Faculty of Education and Society (University College London, UK).There is limited research exploring the secondary Teach First programme (Rauschenberger et al., 2017), no evaluation of the primary and early years programme beyond reports by the Office for Standards in Education, Children's Services and Skills (Ofsted) and no comprehensive review of the two-year Teach First Postgraduate Diploma in Education (PGDE), established in 2017.
This research sought to understand the perspectives of primary and early years trainees on the roles of the three partnership organisations (Teach First, university and school) in supporting their learning.The article will focus on perceptions of the university role and programme as a whole, before considering the implications for the future of the university contribution to new teacher learning.As a university teacher educator at UCL IOE, I was working on both the university-led Postgraduate Certificate in Education (PGCE) and employment-based Teach First programme at the time of this research (Teach First, 2021).

ITE and ITT in England: a reduction in the university role
The English teacher education landscape has been subject to a lengthy period of reform (Furlong, 2019).A relentless drive to raise education standards in the post-war era included a focus on recruitment and retention of high-quality teachers (Vanassche et al., 2019).The 1984 establishment of the Council for Accreditation of Teacher Education marked the beginning of accreditation and a sustained period of regulation in ITE (McNamara et al., 2008).It was followed by the development of the Teacher Training Agency and ITE National Curriculum in 1993and 1998respectively (McNamara et al., 2008).Until then, ITE was largely led by higher education institutions (HEIs).While courses included school placements, they were often critiqued for being overly academic and distanced from professional practice (Menter et al., 2019).Successive governments arguably (Blake et al., 1997) considered ITE led by HEIs to be the problem, rather than the solution, for improving standards.From 1992, 'on-the-job' training became statutory, with two-thirds of course time to be spent on placement (Peiser et al., 2019), andin 1993, School Centred Initial Teacher Training was introduced, encouraging groups of schools to establish teaching schools and award qualified teacher status (QTS).This move to school-based training was contrary to many European countries, where universities retained a fundamental role (Blake et al., 1997), although it followed similar policy shifts in the USA (Zeichner, 2014).The university role in new teacher learning 3 In 2010, the then education minister Michael Gove (2010: n.p.) claimed that 'teaching is a craft ... best learnt as an apprentice observing a master craftsman or woman' in school.This arguably (Wrigley, 2019) reductionist perspective moved policy and practice towards a technical discourse, with a view of teaching as technically simple -a skill to be learnt through on-the-job apprenticeship.The 2010 White Paper (DfE, 2010) sought a rapid increase in school-led provision (Menter et al., 2019), and programme diversity was encouraged.Provision became increasingly fragmented, with the development of 'branded and localised models of teaching', such as Teach First (Menter et al., 2019: 68).Power and resources for training new teachers moved away from universities and towards schools (Ellis and Spendlove, 2020), with 80 per cent of ITE places situated in universities in 2011/12, decreasing to 47 per cent in 2018/19 (Ellis and Spendlove, 2020).As in the USA, where new graduate schools of education (nGSEs) accredit initial teacher training without university involvement, the conception of new teacher learning has been fundamentally challenged (Cochran-Smith et al., 2020), and the university role in ITE threatened.
With ever increasing choice and diversity in the marketplace of ITE, or perhaps fragmentation and incoherence, as George andMaguire (2019) claim, the Teachers' Standards (DfE, 2011) were introduced to provide a common assessment framework.However, this arguably led to an accountability culture (Lunt, 2008), with a focus on professional compliance, serving to reduce teacher autonomy (Mayer and Mills, 2020).The standardisation agenda continued, and in 2019, the ITT Core Content Framework (CCF) and Early Career Framework were introduced, aiming to provide three years of structured professional development for new teachers (DfE, 2019a).Aligned with the Teachers' Standards, they outline what trainee and early career teachers should learn, framed around 'learn that' and 'learn how to' statements (Mayer and Mills, 2020).Mayer and Mills (2020: 11) argue that, by providing a list of content to teach trainees, teacher education is defined by 'content to be mastered', and compliant professionalism is promoted.The CCF was met with caution by the university sector, which seeks greater autonomy to adapt course curricula.Indeed, the Universities' Council for the Education of Teachers (UCET, 2021: n.p.) argued that 'the DfE [Department for Education] must not confuse consistency with uniformity'.The CCF was followed by the government-initiated ITT market review in 2021, which aimed to ensure that the sector 'maintain(s) capacity to deliver enough trainees', and to ensure 'high-quality training' for all (DfE, 2021b: n.p.).The review has mandated all existing ITT and ITE providers to apply for reaccreditation in line with new requirements (DfE, 2021c).Simultaneously, a new National Institute of Teaching was established, proposing to bring 'diversity and innovation to the existing teacher development market', and to become 'England's flagship teacher training and development provider' (DfE, 2021a: n.p.).UCET argued that a case for the review had not been made, with Ofsted only recently reporting 100 per cent of ITE partnerships as 'good' or 'outstanding' (UCET, 2021).They opposed the plans for the National Institute of Teaching, claiming that it would threaten the viability of existing high-quality providers with a transference rather than an increase in overall new teacher applicants (UCET, 2021).This rapid policy change is arguably based on politics and ideology, rather than on robust evidence of effective practice (Parker and Gale, 2017;UCET, 2021).It is perhaps comparable to the situation in the USA, where nGSEs were established with no evidence of effectiveness (Cochran-Smith et al., 2020).With considerable sector change and little understanding of trainee perspectives on school-led programmes (George and Maguire, 2019;Murray et al., 2019), this research seeks to understand how trainees interpret the university contribution to their learning and, in turn, to contribute to dialogue regarding the future of the university role in the sector.

The Teach First programme
The English incarnation of the US Teach for All movement, which was founded in 1990 (Zeichner, 2014), Teach First is a charity which aims to address educational disadvantage through a programme of teacher education developed in partnership with universities and schools.Following six weeks of intensive training (Summer Institute), immersion in the class teacher role is swift.Placement schools are situated in areas of socio-economic disadvantage, and most primary and early years trainees have their own class from the outset.This differs from the one-year PGCE, where students are unpaid, supernumerary to the class teacher, and spend approximately one-third of their time in university and two-thirds on a wider range of school placements.While there is time away from school for Teach First trainees to critically engage with theory and research evidence, and to reflect on their practice, this is significantly reduced compared to HEI-led provision.The terms ITE and ITT are indicative of the varied routes into teaching in England, which include HEI-led ITE, School Centred ITT and EBITT programmes (Rauschenberger et al., 2017).Learning to teach in England is increasingly conceptualised as ITT (Rauschenberger et al., 2017;Vanassche et al., 2019), with new teachers often referred to as trainees, rather than as student teachers.In 2015, the Carter review referred to all provision as ITT, suggestive of policy emphasising training over education (Mayer and Mills, 2020).An EBITT programme with a postgraduate qualification, the Teach First programme perhaps embodies a hybrid model (Robinson and Mogliacci, 2019).Although a salaried EBITT programme, it was perhaps mistakenly perceived by Mayer and Mills (2020) as having minimal university involvement.At the time of this research, the university was one of three bodies supporting trainee development and providing dedicated teacher educator roles: the employing school and school mentor; Teach First and the development lead; and the university and university teacher educator.School mentors were defined as 'highly effective teachers' who ensured that trainees have successful school placements and make a positive contribution to student learning.Development leads were employed by Teach First, with responsibility for coordinating trainee support and mentor development.They provided advice and guidance, often focusing on the technical aspects of teaching, in line with Teach First policy (Craster, 2019).University teacher educators were defined by their role in awarding QTS and PGDE, as academic specialists supporting master's level teaching and learning, and as experts in teacher development.
While taught in partnership by Teach First and the university, the university was responsible for the majority of teaching on the PGDE.Over the course of four academic modules, trainees developed understanding of professional practice, subject knowledge and subject-specific pedagogy, child development and philosophical perspectives on education.Critical reflection on practice was encouraged, and there were opportunities for trainees to actively engage with research in their classrooms, aiming to address the theory-practice gap (Furlong, 2019;Perry et al., 2019).Winch et al. (2015: 202) explain a need to develop new teachers' situated understanding, technical knowledge and critical reflection.Similarly, BERA-RSA (2014) assert that teachers need to combine practical knowledge with conceptual understanding of teaching and learning and research evidence in order to support critically reflective practice.These aspects of new teacher knowledge are reflected in successive reviews of ITE and ITT (Perry et al., 2019;Carter, 2015).However, tensions remain between those who hold instrumental views of teaching 'as a craft -a set of difficult and complex techniques to be picked up' (Marland, in Richards, 2018: 8) and those who perceive teaching to be a complex intellectual activity (Mayer and Mills, 2020).From an insider perspective, these philosophical differences were reflected in contrasts between Teach First and the university, the two partners responsible for curriculum development.
An instrumental view is evident in the teaching of key technical skills on the Teach First programme (Craster, 2019).Literature emphasising technical strategies and direct instruction informed by cognitive science is promoted (Edgcombe, 2019), and trainees are taught to apply the techniques of Lemov (2011) and Rosenshine (2012).Initially, this is without underpinning theory, and it reflects Winch's (2017: 67) 'executive technician' model of the teacher who is 'handed theory derived procedures that he is trained to put into effect'.Winch (2017) argues that while the executive technician reflects some of the qualities of an effective teacher, the model is underpinned by a mono-dimensional, reductionist view of teacher knowledge.Such a model is said to promote a behaviourist approach to learning based on teaching standards frameworks (Heilbronn, 2011) and to encourage professional compliance (Mayer and Mills, 2020).Conversely, the university tends to align with a constructivist conception of ITE, focusing on developing learner agency, with understanding constructed by the trainee, rather than being passed to them by the teacher educator (Totterdell and Lambert, 1998).While not a clear division, Mayer and Mills (2020) might suggest that Teach First provides training, and the university provides education, highlighting the perceived ITT-ITE dichotomy (BERA-RSA, 2014).
While technical aspects of teaching are deemed a key element of new teacher learning (Wilson and Demetriou, 2007), they then formed just one aspect of new teacher learning on the Teach First programme.The Teach First partnership provided an ITE programme in which the core curriculum involved subject and pedagogical knowledge development, engaged teachers in practical enquiry in the classroom, and provided opportunities for trainees to apply and evaluate theory in practice.The programme in 2020 aimed to go beyond Sachs's (2011: 165) 'traditional training approach', seeking to develop trainees' critical engagement with research and theory, considered key to effective new teacher learning (Lange and Burroughs-Lange, 2017).However, this research came at a pivotal point in the programme's history.At the point of data collection, along with a major northern university, The university role in new teacher learning 5 UCL IOE announced that it had not tendered for the Teach First contract beyond 2022 (UCL IOE, 2020), and that it planned to discontinue its role with school-direct salaried and school-centred programmes, reaffirming a commitment to developing research-literate teachers, and to work with schools to support outstanding practice through critical research engagement (UCL IOE, 2020).The decision was also said to be influenced by changes to the Teach First programme, which was moving away from HEI partnership to a 'sub-contracting model' (UCL IOE, 2020).Since 2021, partner universities no longer award QTS, and they have reduced involvement in curriculum planning and teaching.This exploratory research sought to understand trainees' experience of learning to teach on the programme before this further reduction in the university role.

Research design
Drawing on rich qualitative data, this small-scale doctoral study sought to understand trainee perspectives on the role of the partnership organisations in supporting new teacher learning at a time of significant sector change.The article focuses on trainee perspectives on the university role and implications for the future position of the university in new teacher learning.Analysis of data from eight semi-structured interviews with first-and second-year trainees illuminates the trainee experience.A convenience sampling approach was adopted, with all first-and second-year trainees invited to share their perspectives.The sampling approach aimed to garner sufficient interest to enable a purposive and broadly representative sample.The sample in Table 1 was equally representative of first-and second-year trainees and employing schools in London, including a range of academies, local authority-maintained schools and those of religious character.Although fairly representative of gender and age, trainees from Black, Asian and minority ethnic backgrounds were unfortunately not represented.Trainee motivations for taking part could be questioned as the whole cohort was invited.However, interview data show that participants were agentic, showed a high degree of loyalty towards the Teach First programme and vision and were constructive in their thoughts for improvement involving all partners.While methodological questions might be raised when seeking to draw conclusions from a small number of in-depth interviews, this research acknowledges sample size limitations.It was small in its aims, seeking only to begin to explore Teach First trainee perspectives at a crucial time of sector policy change, and it indicates a need for a larger study of its kind.Eight in-depth semi-structured interviews were conducted, lasting between 45 minutes and an hour.Interview questions sought trainee perspectives on the roles of the three partnership organisations in their learning, the academic elements of the programme and how this informed their practice, as well as consideration of PGDE teaching by different partnership organisations.There was flexibility to be responsive, and to enable trainees to add further comments and raise questions of their own.
The research adhered to British Educational Research Association (BERA, 2018) guidelines and received university ethical approval.The need for reflexivity to prevent my insider position from compromising participants' ability to disclose information (Creswell, 2014), and to limit the capacity for bias (Sikes and Potts, 2008), was acknowledged.My role as a university teacher educator on the The university role in new teacher learning 6 programme, and my relationship with participants, had the potential to affect outcomes, reducing authenticity (Bryman, 2016).To minimise the potential for response bias (Creswell, 2014), I was explicit about my position as a researcher rather than a tutor, and about the purpose of the research -to establish trainee perspectives on the programme in order to further consider how the university can provide the very best teacher education.Confirmation of anonymity and the option to withdraw aimed to give participants confidence to speak freely.To mitigate against challenges of role duality (Sikes and Potts, 2008), I remained neutral in my responses during interviews, so as not to reveal my own values or potential biases and influence respondents (Cohen et al., 2018).While acknowledging potential ethical challenges, my position supported a deeper understanding of perspectives shared, and afforded a 'more insightful exploration of the issue[s]' than an outsider might have had time and capacity for (Charmaz, 2009: 40).
Thematic analysis was used, following Braun and Clarke's ( 2006) six-stage model, which allows for a more organic approach to coding than the positivist coding reliability (Boyatzis, 1998) or structured codebook approach, and which includes the formation of key criteria for the inclusion or exclusion of data under themes (Braun and Clarke, 2019).Following transcription of the interviews, detailed notes were made on emerging themes to familiarise myself with the data, and to make links between data sets.Individual coding supported reliability (Huberman and Miles, 1994), and using this process for several scripts enabled me to identify similarities and differences across responses.I sought to code inductively before using theoretical and insider knowledge to search for underlying themes.Thematic analysis is seen as an iterative process which involves the researcher moving back and forth between stages (Nowell et al., 2017).This was the case here, with themes naturally emerging through the coding process.
Reflexive thematic analysis emphasises the researcher's active role, with subjectivity and creativity seen as key to the production of knowledge (Byrne, 2019), and perhaps presenting a challenge to an entirely inductive approach (Byrne, 2019).Indeed, the value of the insider may be lost if attempting such an approach.While recognising the value of my insider position in helping to identify relationships between data sets (Byrne, 2019), I was acutely aware of the need for reflexivity during data analysis (Sikes and Potts, 2008), attempting to use the data to tell the story, rather than allowing my own values and beliefs to influence findings.Repetition of data helped to identify the significance of emerging themes (Bryman, 2016), and informed the dominant themes reported here.
A discussion of trainee perspectives on the role of the university teacher educator and university in their learning is followed by an exploration of two superordinate themes which relate to the whole programme: 'time to learn effectively' and 'pedagogy and partnership'.This leads to a consideration of the future role of the university in new teacher learning.The majority of trainees perceived university teaching and academic course elements as fundamental to their development.

Quality of teaching
Seven out of eight trainees (T) reflected on the quality of university teaching, and the impact of this and enquiry-based assignments on their practice: 'the quality of the lectures was just brilliant ... they make you think ... you've got the academic stuff in there and [we] reflect and think about evidence-informed practice ... in a very practical way' (T7).
Several compared university teaching to Teach First training days, with one remarking that the Teach First content: 'felt wishy-washy compared to the academic focus of UCL ... Teach First training days almost seemed ... like ... something they had to kind of tick off ... it was less substantial' (T5).

Tutor expertise
Reflections on the quality of university teaching often related to trainee perceptions of tutors as 'experts' (T3): 'the amazing UCL tutors ... when you have the training it's really good' (T8); 'the quality of the education that I received was excellent' (T7); 'the university are the ones that know the most about it' (T3).
One trainee (T2) commented on feedback provided by different support roles, noting a difference in the depth of questioning and discussion with their university teacher educator.

Enquiry-based assignments: linking theory to practice
The majority of trainees valued the opportunity to swiftly apply theory to their practice.This was noted as an overall programme strength, although comments often related to university teaching or tutor support: 'I think my tutor was very good at ... giving practical strategies but also giving readings ... and saying, like, this might be a good thing to implement or this is a good theory to look at ... this is why this might be a good idea' (T4).
Responses support the view that theory is most valuable when integrated with practice, and when swift application to the classroom is enabled (Waege and Haugaløkken, 2013).A second-year trainee (T6) gave a specific example of how theory and learning about subject-specific pedagogies had informed their understanding of the development of quality classroom talk, while another spoke of how learning at university had supported their understanding and dialogue with school colleagues: 'I found that me and [Teach First peer] have been able to talk to colleagues really confidently about theory, and I feel like I've got a lot to share ... that's been really influential' (T3).Such responses support an argument for developing research literacy among new teachers (BERA-RSA, 2014) and indicate the importance of the university role in establishing and supporting this throughout teachers' careers.

Academic qualification and guidance
Half of those interviewed explicitly noted the importance of the university role, explaining that: 'it has to work with a university.I think you have to have ... the rigour of the PGCE or PGDE ... teaching is now very much more evidence informed ... it's an academic job with a massive dose of practicality' (T7).Another felt that the university provides 'more training ... to really distinguish yourself from others' (T8).Several remarked on the university's role in providing clear academic guidance and readings to support reflective practice: 'I see my uni tutor as the person to talk to about essays ... more academic things like readings on topics ... very knowledgeable about things like that' (T4).
However, frustrations were shared about the academic assignments caused by conflicting advice from Teach First and the university.Trainees felt that the university provided clarity, partly because they are responsible for academic assessment.This relates to the overarching theme of pedagogy and partnership, which will be explored later.

The Teach First partnership: programme-level themes
Beyond themes relating to the role of the university were two broader programme-level themes: time to learn effectively and PGDE pedagogy and partnership (see Figure 1).

Swift immersion into class teacher role and responsibilities
While most trainees welcomed the swift immersion into teaching, several shared a need for more support, further opportunities to observe effective practice and a more gradual increase in class teacher responsibilities.This is perhaps to be expected, and it was reflected in the Teach First recruitment literature at the time of the research, which highlighted that, while not necessarily the right approach for everyone, swift immersion into practice is a good way to ensure quick progression.

Observation of effective practice from the outset
Some reflected on the value of observing their mentor teach, and others highlighted the value of their alternative key stage and/or contrasting school placements.Pring ( 2012) raises concerns about the potential of EBITT programmes to limit trainee understanding to the pedagogies and practices of their employing school.Recent research suggests that there can be pressure to 'mould' trainees to a particular school, limiting their access to a wider of professionals and contexts (Czerniawski et al., 2019: 180).Teach First trainees complete a minimum of one week on placement in a contrasting school.If combined with the option of an alternative key stage placement, this rises to four weeks in another setting.This ensures time to observe and reflect on practice, and to apply further learning in a different context.Often not completed until the spring or summer of the QTS year, findings indicate that this would be beneficial earlier in the year, with one trainee stating that they wish they had completed this earlier (T4), and another explaining how they had 'spent the whole first term not observing pretty much anyone' (T3).
A need for increased opportunities to observe from the outset was often linked to a desire for a slower induction into class teacher responsibilities.There is a need to plan for meaningful observation of experienced teachers to support new teacher learning.Burn et al. (2015) argue that new teachers find it hard to notice the modelling in expert teaching, and Korthagen (2017) talks of a need to be guided in reflection in order to understand the choices of the expert teacher, and the relationship between theory and practice.On the Teach First programme, opportunities for guided observation and reflection are often limited to those identified as requiring additional support, rather than being planned for all.The speed of immersion has the capacity to significantly hamper new teacher learning, sometimes leading to an early sense of failure.One trainee explained that they were 'drowning in the sheer enormity of what the job is' (T7), and another (T5) that they 'didn't like the feeling that I was failing the school, failing the kids ... the expectation was too high.I just didn't know what I was doing.'This concern was echoed by T1, who claimed that the Summer Institute could not sufficiently prepare trainees for class teacher responsibilities.

Time to learn away from the school context
There are significant challenges to ensuring a high-quality ITE/ITT curriculum alongside school practice on EBITT programmes.While time for guided observation of good practice is deemed crucial (Korthagen, 2017), trainees were acutely aware of the need for time to learn and reflect outside the classroom.If trainees replicate what they observe, there is, as Vanassche et al. (2019: 485) assert, a concern that 'we accept and recreate rather than transform and renew current schooling', stifling critical and creative thinking.T2 reflected on the importance of being able to 'move between a totally immersed setting and a detached academic setting fairly regularly ... you can constantly be kind of gathering ideas and thinking about how to feed them back'.T3 considered how regular reading and reflection aided their development, and appreciated the additional time for this afforded by the pandemic.While T2 valued the distance afforded between university and school, they sought more time to engage with learning away from the classroom, arguing that there is 'too heavy a balance away towards the immersive setting such that your days to think ... are few and far between'.Some reflected on insufficient learning time within university sessions, with T5 noting that content had been 'crowbarred unrealistically' into too few sessions.However, the content was welcomed, and several considered how the university role might be extended, with more opportunities for the university to work closely with the school and trainee to support practice and classroom enquiry.This is a model not unlike that in place prior to the implementation of the PGDE in 2017, when university teacher educators visited trainees fortnightly in school and had more involvement in their formative development.Others The university role in new teacher learning 9 sought further university input to support understanding of subject-specific pedagogy, with several valuing the subject specialism in the second year, while also lamenting limited coverage of certain subjects and areas of professional practice, such as music and PE, and supporting students with special educational needs and disabilities.

More gradual and scaffolded support for academic writing
While trainees valued PGDE curriculum progression, and all trainees commented on the importance of enquiry-based assignments, several remarked on the challenge of completing an assignment during their first term.They claimed that 'the first essay was hard to do' (T2), and that they felt 'overwhelmed in the autumn term' (T3), as if they were 'drowning' (T5) and just 'getting through things' (T4).This suggests that a more scaffolded and staggered approach to the QTS year would be more supportive of new teacher learning.One trainee, who emphasised the importance of teacher research engagement, suggested that assignments might be more manageable if 'condensed ... into smaller projects' (T8).

Flexible learning opportunities needed
Several trainees (T1, T5, T7) sought more flexible learning opportunities which would take into account trainees' varied experience, school contexts and workloads.One trainee, who highlighted the importance of teacher research literacy, noted that 'there has to be more time out of the classroom to allow you to focus on doing that properly' (T5).This linked to workload concerns, with this trainee stating they were not alone in 'working every single weekend'.Others reported having to complete unnecessary administrative tasks that served little developmental purpose.Recent research has explored the significant impact of workload on teacher retention (Worth and Van den Brande, 2020), highlighting the importance of ensuring sufficient time for new teacher learning.

PGDE: pedagogy and partnership -partnership communication and conflicting pedagogies
A reduction in the university role in ITT has arguably caused tensions in school-university relationships (Hordern, 2014;George and Maguire, 2019).All but one trainee reported partnership communication issues, and while two cited issues between all three partners (T2, T3), all remarked on those between Teach First and the university.Responses indicate that these issues were mostly caused by fundamental differences between university and Teach First perspectives on teaching and learning, with one trainee reporting a 'weird dynamic between school, Teach First and UCL ... it doesn't quite gel' (T1), and another explaining 'there were differences massively with the theory that was put across from the two' (T8).A third trainee remarked that 'most of the Teach First training we had was separate from our more academic training' (T2).There was a clear perception of pedagogical differences between Teach First and the university: I think UCL teaches you more of a social constructivist approach ... more opportunities to try new things and allow the children to explore it for themselves ... they're more child centred, whereas Teach First is maybe more the teacher being in charge and dictating, more than being a facilitator.(T6) This trainee was not alone, with four others noting the Teach First focus on teaching key technical skills in a directed fashion without theoretical underpinning, in line with Winch's (2017: 67) 'executive technician' model, and reflecting an instrumental approach to teaching.Some shared their thoughts on Lemov's (2011) 'Teach Like a Champion' techniques and the Teach First elements of the ITE curriculum, with several noting university tutors' critique of Teach First content: ... some of the UCL tutors were really disparaging about parts of the Teach First course.Teach First ... were using this Doug Lemov book as their bible ... perhaps understandably some of the tutors ... were clearly not of the mind that this should be a brand new teachers' bible and that there was more to it than just being able to clap and point and ... the strategies that Lemov propagates.(T5) Some recognised a value in being taught key technical skills by Teach First from the outset: Just try this and it will work.That was the Teach First approach.And it wasn't all bad ... it became apparent why they do that in those first few weeks in September.I could have read everything about humanism and behaviourism and ... and it just might not have helped because ... you need to know how to put it into action and get control quickly.(T5) However, the same trainee noted that Teach First sessions 'felt wishy-washy compared to the academic focus of the UCL delivered content'.
Perceived philosophical and pedagogical differences have implications for the realisation of Darling-Hammond's (2006) shared beliefs about good teaching, considered as key to effective ITE partnerships.Trainees' perspectives indicate that Teach First focuses primarily on a technical model of training, while the university tends to align to a constructivist conception of ITE, with understanding constructed by the trainee (Totterdell andLambert, 1998). Fenstermacher (in Shulman, 1987: 13) states that 'the goal of teacher education is not to indoctrinate or train teachers to behave in prescribed ways, but to educate teachers to reason soundly'.New teachers articulated the need to develop the ability to critically enquire about their practice, drawing on theory and research, to establish the most effective practice for their context (Tatto and Furlong, 2015).Otherwise, there is a danger of creating replica teachers who follow what is prescribed, rather than renewing or challenging the current system (Vanassche et al., 2019).
Trainee reflections overall indicate that the programme as it stood in 2020, with significant university involvement, did problematise teaching and learning, and encouraged trainees to engage in critically reflective, evidence-informed practice.While findings indicate a need for greater curriculum coherence and more time to learn effectively, there is much in trainee responses to suggest that in its 2020 iteration, the Teach First programme developed the 'pedagogically thinking, reflective and inquiry-oriented teachers' argued as the ideal ITE outcome (Toom et al., 2010: 339).Indeed, without that critical thought, trainees may not have reflected so thoughtfully on the philosophical and pedagogical differences between organisations, nor considered the implications for their practice.This raises the question of the implications of the further erosion of the university in new teacher learning.

Discussion: Moving forwards -towards a model of professional autonomy, growth and retention
While they were positive about the programme overall, to hear of trainees feeling 'overwhelmed' and 'drowning' at the earliest point in their career is deeply concerning.Workload and challenging school contexts have a significant impact on teacher retention (DfE, 2019b;Worth and Van den Brande, 2020); this level of stress and fear of failure (T5) in the first year is disturbing.New teachers need more time: to engage with literature, expert colleagues and practitioners; to engage in classroom enquiry and reflection; and to become research-informed, engaged and autonomous professionals who gain satisfaction from their work.The BERA-RSA (2014) inquiry argued a need to equip teachers with the skills to conduct their own classroom-based research, and Milton et al. (2020: 29) suggest that a 'culture of enquiry, innovation and exploration' is key to professional growth.All trainees in this research valued engaging in practitioner enquiry, and some explicitly shared the benefits of professional collaboration to develop research-informed and innovative practice to meet the needs of students in their contexts.Some were given more freedom than others to engage in practitioner enquiry, and there were mixed views on schools' capacities to provide supportive learning cultures, which Milton et al. (2020) argue can have a significant impact on the professional formation and growth of new teachers.
Trainees in this study are calling for a model of initial teacher training and education that enables swift immersion into practice, but with a more gradual move to whole-class teacher responsibilities and academic course elements.They appear to be seeking a model not dissimilar to the PGCE, but via a salaried route.This aligns in many ways with the centrist position promoted by Totterdell and Lambert (1998).Although established some time ago, this is an apt model, because it aligns with trainees' desire for a balanced approach which goes beyond an instrumental model of learning focusing on professional compliance, to one of professional growth and autonomy, where teachers assume 'responsibility for continuously enriching and extending their craft both conceptually and practically within educational communities that foster the common good' (Totterdell and Lambert, 1998: 356).The emphasis is on increased collaborative enquiry, which requires effective partnership between teacher educators: The university role in new teacher learning 11 universities, schools and wider ITT organisations.These partnerships should be founded on shared beliefs about how new teachers learn (Darling-Hammond, 2006), and shared goals for professional growth that can support the development of professional judgement, rather than a chiefly technicist model.Such a partnership would both train and educate.Mayer and Mills (2020) argue strongly for the promotion of professional trust and autonomy, and for the importance of collaboration to develop new knowledge and understanding.Recent research by the National Foundation for Educational Research highlights the relationship between teacher autonomy, professional development, motivation and, ultimately, retention (Worth and Van den Brande, 2020).Of the 1,144 teachers surveyed, approximately half of those with the lowest perceived autonomy voiced an intention to stay in teaching in the short term, compared with more than 85 per cent of those with the highest reported levels of autonomy (Worth and Van den Brande, 2020).The research report recommended that the DfE 'embed the principles of teacher autonomy into the Early Career Framework', so that it acts as a 'menu for early career teacher development, rather than a prescription' (Worth and Van den Brande, 2020: 4).While recognising a need for further research over time, as teacher intentions are not necessarily synonymous with outcomes, it suggests that instrumental models of teacher training which promote compliance have implications for teacher retention.In 2016, a study found that three years after qualifying, only 43 per cent of Teach First trainees remained in the profession, compared to 61 per cent of those on HEI-led full-time routes (Allen et al., 2016).If we are going to retain teachers and stop servicing the revolving door (Brill and McCartney, 2008), then time and opportunities for professional growth and learner autonomy are vital (Worth and Van den Brande, 2020).

Implications for the university role in new teacher learning
According to Mutton et al. (2017), there is an urgent need to move away from the binary view of ITT and ITE, viewing training and education, practice and theory, as separate, and to move towards an integration of research-based and professional knowledge, and a conception of ITE as collaborative endeavour.Peiser et al. (2019: 15) assert that ITE curricula and mentor training should be developed collaboratively by universities and schools, playing to the strengths of both, with universities focusing on developing student teacher and mentor knowledge of theory and research, 'whilst school colleagues promote coaching skills that draw on professional craft and contextual knowledge'.Perry et al. (2019) argue that universities need to reassert themselves as experts in educational research, developing students' ability to engage critically with research to inform reflective practice.Trainee perceptions in this research support Orchard and Winch's (2015: 27) argument that the university environment is 'more conducive to sustained discussion and sharing ideas away from the pressures of the immediate workplace'.Universities are arguably best placed to develop student teachers' critical research literacy because they are more likely to employ staff with the 'necessary scholarly and pedagogical expertise to undertake this task' (Orchard and Winch, 2015: 22).Equally, the BERA-RSA inquiry (2014) found that some non-university-based ITT programmes have struggled to prioritise the role of research.This study indicates a need not only to reassert the university's position as educational research expert, but also as an inspirational learning environment for teaching professionals at the start and throughout their careers, so that, as T5 articulates, teachers do not 'stagnate because there is no distraction from the day job', but continue to grow and innovate collaboratively and, crucially, maintain their commitment to the profession.
It is of national and international interest that, while the university role in new teacher learning recedes in England, the rest of the UK shows a continued or renewed commitment to the role.ITE is university-led in Scotland and Northern Ireland, and Wales has recently shown a renewed commitment to the university contribution to ITE (Furlong, 2019).Since this research was undertaken, the requirement for universities to award QTS has been removed in England, and a new process for ITT accreditation has been initiated (DfE, 2021b).With growth in EBITT programmes with limited or no university role, including the new National Institute of Teaching, England appears to be an outlier in teacher education policy in the UK (Menter et al., 2019), instead following in the footsteps of America's nGSEs (Cochran-Smith et al., 2020).American nGSEs are characterised by a view of new teacher knowledge as 'embedded in practice itself, specifically in the actions of highly effective teachers' (Cochran-Smith et al., 2020: 28).These new graduate schools have positioned themselves as competitors of university-based provision in the USA (Cochran-Smith et al., 2020).Further research in the USA is ongoing, and should The university role in new teacher learning 12 yield important insights that will be of relevance to developments in England, including the National Institute of Teaching.
Initial teacher education continues to be 'desperately under-researched' (Menter et al., 2019: 75), and the National Audit Office (2016) has raised questions regarding a lack of evidence base for DfE teacher education policies.Menter et al. (2019: 75) argue that there is a need for 'large-scale, longitudinal work that really gets to grips with questions about quality and effectiveness in different routes and contexts', although they recognise the challenge of this in times of rapid policy change.While this small-scale study just scratches the surface of trainee perspectives on one English EBITT programme, it indicates that these new teachers not only value the university role in their learning, but also seek more, not less, university involvement.Before irreparable sector damage is caused (University of Cambridge, 2021), and the university contribution diminishes further, there is a need to further explore the valued practice that already exists through the eyes of those learning to teach.After all, they are the ones whose attrition from the profession is of national and international concern.