Reflections on Peer Facilitation of Graduate Teaching Assistant Training

This paper outlines a recently designed programme of training for graduate teaching assistants (GTAs), focusing on peer elements embedded in the programme. In particular, we describe our approach to co-facilitation of training, with sessions facilitated by the GTA programme lead and an experienced Peer GTA. Peer GTAs have at least one year of teaching experience and are able to provide practical, contextualised and discipline-specific input, which helps to address the challenge of balancing generic and discipline-specific training in GTA programmes. We describe a small case study of co-facilitation and reflect on the benefits of this approach for new GTAs and for the Peer GTAs themselves. The paper is co-authored by the GTA programme lead and a Peer GTA, who provides first-hand reflections on her experiences.


Introduction
There have been attempts to professionalise the training of postgraduate students who teach, often called graduate teaching assistants (GTAs), for over 30 years now. In North America, in particular, there has been a focus on establishing GTAs as trainee academics and as 'junior colleagues' since the 1980s and 1990s (e.g. Nyquist, Abbott & Wulff, 1989;Wulff & Nyquist 1996), with GTAs occupying recognised positions simultaneously as postgraduate researchers and as teachers, usually as the first apprentice-like step towards an academic career.
This drive to raise the status of teaching has led to increased provision of training and support for GTAs. In the UK, for example, higher education institutions (HEIs) have seen a rapid increase in the prevalence of pedagogic training courses offered to (GTAs). By 2010, over 50% of GTA courses at UK universities were compulsory, 57% included some form of assessment, and there was a general trend towards aligning course content with the UK Professional Standards Framework (UKPSF), thus supporting GTAs to gain fellowship of the Higher Education Academy (HEA)/Advance HE, usually at Associate level (Lee et al. 2010).
In many GTA training contexts, particularly in North America, programme designers have embedded peer learning and teaching as key parts of the process of learning to teach in higher education. At Cornell University, for example, the College of Engineering trains experienced graduate students to facilitate pedagogic training sessions for new GTAs on the College's Teaching Assistant Development Program (Hollar, Carlson & Spencer, 2000). Similarly, at the University of Toronto, the Teaching Assistant Training Program (TATP) has been in place since 2002, with sessions facilitated by experienced graduate students (Rouf, 2012). In this paper, we provide an overview of a recently designed GTA programme at a research intensive university in the UK, focusing on peer elements included in the training. The paper is co-written by the programme designer and an experienced GTA, who offers reflections on her experiences of co-facilitating the core sessions of the programme. We begin with a brief review of the literature on GTA programme design.

Features of GTA Programmes
In the UK, the Dearing Report (Dearing, 1997) highlighted the need for those teaching in higher education to be suitably qualified to teach. Thirteen years later, the Browne Review (Browne, 2010) increased scrutiny of teaching further by recommending that UK HEIs publish information about the teaching qualifications of academic staff. Alongside this political pressure, there have been efforts to learn from the, in some ways further developed, North American context since the early 2000s. In a detailed review of the literature on GTA training in North America, Park (2004) called for more comprehensive training in UK HEIs. As Park explains, "training means the process of bringing the GTA to an agreed standard of proficiency by practice and instruction" (Park, 2004, p. 351), which therefore calls for a coordinated approach by stakeholders in the HE community, whether that be within one HEI or across the sector as a whole.
One such coordinated approach is outlined by Hopwood and Stocks (2008) at Oxford University. Based on evidence from 'apprenticeship-style' programmes in North America, Hopwood and Stocks present their Development of Learning and Teaching programme (DLT), in which GTAs work with an experienced academic mentor and attend a systematic training programme, rather than one-off, isolated workshops. In their evaluation of the DLT programme, analysed in the context of activity theory (cf. Engeström, 1987), Hopwood and Stocks report that all participants in their study placed great importance on formal recognition of teaching, and the opportunity to obtain teaching qualifications. Also of importance was the use of mentors and the establishment of a teaching community, partly through reading groups related to teaching and learning. Another institutional insight into the development of GTA training is presented by Chadha (2013). She describes the process of reviewing a GTA programme at King's College London, which initially consisted of seven core sessions and two optional workshops. Feedback from GTAs revealed that the programme did not succeed in preparing GTAs for the teaching duties at hand, as demonstrated by the following: According to the students, sessions were directed towards gaining understandings of conceptual frameworks related to teaching and learning, whereas, as novice teachers, their interest lay in surviving the face-to-face interactions in the classroom. For example one of the participants had commented that 'the sessions rely on a lot of theory with very little in the way of substance -there is very little of use'. Chadha, 2013, p. 210 Chadha concluded that GTAs need to learn about the practicalities of functioning in teaching and learning settings, before they are able to critique their practice against the backdrop of pedagogical theories. This led to a redesigned programme with two core sessions about general teaching and learning and discipline-specific teaching and learning; the second session is team-taught by a member of the academic practice team and an academic from the discipline in question (Chadha, 2013). In a subsequent impact study of the programme, Chadha (2015) found that GTAs benefited from attending training in several ways, such as becoming more reflective practitioners and employing a more student-centred approach in their teaching. However, Chadha also raised the need for greater input on disciplinary differences. This point was raised by Park almost a decade earlier: "A thoughtfully designed GTA training programme … should include both generic and subject-specific elements" (Park, 2004, p. 356). However, it is highly challenging to achieve this balance of generic and subject-specific input. Educational developers engaged in training of GTAs cannot be experts in all disciplines, so a collaborative, interdisciplinary approach is required. Chadha's (2015) approach, combining input from academic practice colleagues and academics from specific disciplines, offers one way of including greater discipline-specific content. Another option is to draw on peerfacilitated GTA training, which is more prevalent in North America than in the UK, as alluded to in the Introduction (e.g. Rouf, 2012;Hollar et al. 2000).
From the wider literature, it now seems commonplace for GTA training programmes to include opportunities for practical application of teaching skills acquired, often in the form of microteaching sessions, which involve GTAs leading a short (usually 10-minute) lesson segment in front of a small number of peers and a facilitator, with immediate feedback given by the group. Such sessions are usually filmed so that the GTA can review her/his own teaching (e.g. Boman, 2013;Meadows, Olsen, Dimitrov & Dawson, 2015). The practical focus in GTA training was recently explored by Becker et al. (2017), who organised sessions around key teaching techniques. These techniques were explained and modelled by training facilitators, and then GTAs practised the techniques while their peers acted as students. The researchers found that, for techniques which were merely discussed but not practised, there was a lower rate of adoption of these techniques in the GTAs' teaching practice. According to Becker et al. (2017, p. 12), this shows that "targeted practice appears to be a prerequisite for mastery of an instructional skillset". This raises the question of how to create opportunities for targeted practice in the design of GTA programmes. Combining Becker et al.'s (2017) approach, with microteaching sessions focusing on particular teaching techniques in isolation, facilitated by more experienced GTA peers, might be one way of creating such opportunities.
When it comes to the amount of training required by GTAs, findings from research are inconclusive. For example, Postareff, Lindblom-Ylänne and Nevgi (2007) posit that training programmes must be in excess of 30 hours in order to have a substantive effect on the effectiveness of GTAs' teaching practice. However, findings from Meadows et al. (2015) showed that the effectiveness of GTAs' teaching was improved after a 20-hour programme, as evaluated on the basis of several metrics, including not only self-efficacy questionnaires but also teacher observations, focus groups, and interviews with GTAs. It therefore seems that the length of time spent in training is not the central factor here; instead, the opportunity for GTAs to practise new skills over a longer period of time and then to reflect on their practice, is a key element in effective training programmes (Meadows et al. 2015).
As shown by this brief review, there are many approaches to GTA training in a broad diversity of contexts, ranging from accredited programmes leading to Associate Fellowship (or similar certified status) and the award of a Graduate or Postgraduate Certificate, to one-off introductory sessions about teaching and learning in the particular institutional context. Nevertheless, there are several key aspects which seem to feature in many programmes across a variety of institutions, namely: a practical focus and opportunities for GTAs to practise the teaching techniques discussed; opportunity to gain professional recognition; sufficient opportunities to engage with training over a prolonged period of time to foster a reflective approach to teaching; inclusion of generic and discipline-specific input; and peer-led initiatives, particularly in North American programmes. These aspects were incorporated into the design of the GTA programme outlined in the next section, with the addition of peer facilitation to address the challenge of balancing generic and discipline-specific input.

Overview of our GTA Programme
In addition to reviewing the literature, we also surveyed GTAs and academics to gather views on the previous GTA training offered, and to ensure we were designing the programme with the institutional context in mind. The programme, GTAP, was designed around two core courses: Introduction to Learning and Teaching and Introduction to Assessment and Feedback for Learning, with a variety of optional courses to enable GTAs to attend flexibly and to tailor sessions to their own disciplinary needs and context (see Table 1). In line with findings from the literature and from conversations with GTAs and academics, the programme was designed with a practical focus. Input on educational theories was limited in face-to-face sessions, but an online GTA toolkit was compiled with theoretical content as well as practical information about programme and module structure at our institution; information about GTA roles and responsibilities; professionalism and boundaries: pay; information about mental health and wellbeing; an overview of Advance HE/HEA and the benefits of applying for Associate Fellowship. Opportunities for GTAs to practise teaching were also factored in by running several microteaching sessions throughout the year, where GTAs can plan and deliver a 10-minute lesson segment and then receive feedback from their peers and the tutor, in addition to self-reflection using the video file from the session. GTAs may attend microteaching sessions as many times as they wish, and they are encouraged to practise a different technique or element of their teaching practice each time. At the heart of the application process is a reciprocal peer observation, which involves pairs of GTAs observing each other, providing reciprocal feedback, and then writing a peer reference in support of each other's AFHEA application. This is one of the peer elements embedded within the programme design. The other key peer element is peer facilitation of the core courses. In 2018-19, we piloted this approach, which involved experienced GTAs and the GTA Programme Lead cofacilitating training sessions for new GTAs. We describe this study and our reflections in the remainder of the paper.

Reflective Case Study
After the GTA Programme had been designed and piloted, we started to implement co-facilitation of the two core courses, which involved the GTA Programme Lead team teaching the courses with experienced Peer GTAs. We trialled this approach in our biggest faculty, the Faculty of Engineering, which consists of 10 different departments covering subjects such as Design, Computing, Aeronautics, and Chemical Engineering.

Peer GTAs
We placed an advertisement in the 10 departments of the Faculty of Engineering requesting applications for, what we call, Peer GTAs. Applicants were required to have completed GTA training provided by the College and to have gained at least one year of teaching experience in their department. Applicants were considered to be particularly strong candidates if they also held Associate Fellowship (AFHEA) and had been recipients of one of the College's annual GTA awards, administered by the Students' Union. Applicants were shortlisted according to these criteria, and references were sought from doctoral supervisors and Teaching Fellows who had a good knowledge of the GTAs' teaching work. Shortlisted candidates were then invited to an informal interview with the GTA Programme Lead in order to discuss the role further.

Co-facilitation of GTA Training
The GTA Programme Lead met with each Peer GTA to discuss the two core courses in the GTA Programme, and to decide which parts the Peer GTAs would like to lead during the sessions. At these meetings, it was also agreed which aspects of the courses could be tailored to each Peer GTA's department. For example, in the first core course, Introduction to Learning and Teaching, there is a section on learning outcomes and session planning. Some Peer GTAs decided to lead this part and to use examples of real learning outcomes from modules in their department for the basis of discussion and activities. Similarly, in the Introduction to Assessment and Feedback for Learning course, some Peer GTAs brought sample marking schemes and grade descriptors from a variety of assessments in their departments.
In the following section, Hannah reflects on her experiences of co-facilitation in her role as a Peer GTA.

A Peer GTA's Reflections on Co-facilitation
Co-facilitation of GTA training courses presented an opportunity for a very different GTA experience to the usual formats of laboratory and tutorial teaching, which was very much appreciated by all the Peer GTAs who took part. In this section, I reflect on the benefits of such a teaching arrangement to both the new GTAs being trained and the Peer GTAs, and will consider the role of the Peer GTA and the potential improvements to the co-facilitation scheme.
The presence of a Peer GTA in training courses had a clear benefit to the GTAs taking the courses, as was established by informal verbal feedback during and after the sessions. Many GTAs remarked to me on the usefulness of a Peer GTA in establishing context for the training, as a lot of the concepts, such as active learning, can seem somewhat abstract without somebody to provide concrete examples of their application in the subject context. This went some way to answering the call from some researchers (e.g. Chadha 2015; Park 2004) for greater discipline-specific input in GTA training programmes. It was also very important as a Peer GTA to make the training relatable to those taking the course, thereby increasing engagement in the session and subsequent application of the taught techniques.
Also highly beneficial was the opportunity for GTAs, who are often new PhD students, to meet a Peer GTA with some experience of both teaching and research who they can later recognise and approach if they so wish. Indeed, many GTAs from the sessions I have co-facilitated have subsequently talked to me around the department and college. The experience of meeting new GTAs in my role as Peer GTA has also promoted ongoing pedagogical discussions, which I believe will have a positive impact on new GTAs' implementation of the teaching techniques discussed and practised during the training courses (cf. Becker et al. 2017).
As well as enhancing the learning experience of the GTAs, co-facilitation had a number of benefits to me as a Peer GTA. The classroom dynamic when teaching peers is very different to that when teaching undergraduates, with a much larger cohort of students than I am used to. Both of these factors served to create an environment much more similar to that of a conference or workshop, settings in which as a researcher it is vital to be able to present effectively. I therefore had the opportunity to practise and develop my peer communication skills in a relaxed situation, and have found this invaluable when presenting at conferences since. As a Peer GTA my main role was to provide examples from my own teaching experience, and generating relevant anecdotes was an excellent exercise in reflection, which has in turn been very helpful in preparing my AFHEA application. I also very much enjoyed meeting the GTAs, and other Peer GTAs, since it is not always easy to meet other researchers outside of one's own research group or department, and it was very interesting to compare experiences and learn form each other. Similar to the study by Hopwood and Stocks (2008) outlined above, the experience of peer facilitation has helped to establish a teaching community of GTAs within and across departments. This is an element of the Peer GTA role that could be developed further, in order to create further opportunities for learning and teaching discussions across disciplines.
At this stage, the role of the Peer GTA in the preparation and delivery of the training courses is still being shaped, and I appreciated the freedom granted by the module lead to interpret it as I wished. For example, I saw myself as a peer educator sharing my experiences with other PhD students on a similar level to me, and found the duality of identifying both with the GTAs and the module lead quite satisfying. However, from informal discussions with other Peer GTAs, some saw themselves as sitting somewhere between the student and lecturer, and preferred to define themselves as more of a discipline-specific expert, giving anecdotes from this perspective. Others preferred to adopt the role of a mentor to the GTAs, drawing on their experience to support and encourage the learners. I enjoyed leading parts of the sessions, for example a section on dealing with various student behaviours, to which I was able to add personal anecdotes and advice, whilst other GTAs preferred to share their experiences at opportune moments as the session progressed. The flexibility for each Peer GTA to make the role their own allows the Peer GTA to feel comfortable and relaxed and this therefore contributed to the success of the cofacilitation.
To improve the co-facilitation of GTA training further I, as a Peer GTA, would like to be more involved in the planning and preparation of the training sessions. This would help me to have more ownership of the content of the sessions, and would enable me to contribute more in terms of relevant examples and reflections. However, I recognise that given the already large demands on PhD students' time, this may not always be possible, and given the short time anybody is likely to occupy the role of Peer GTA, it would not be viable to enter into a wholesale redesign of sessions with each new co-faciitator.

Conclusions and Future Work
In this paper, we have presented a recently designed GTA programme which incorporated elements of peer learning and teaching. A key consideration was the challenge of balancing generic and discipline-specific content in pedagogic training. To explore this, we piloted co-facilitation of GTA courses, with sessions team taught by a member of staff and a Peer GTA -a GTA with at least one year's teaching experience. From this pilot, the new GTAs benefited from being taught by a peer, as the Peer GTA is in a position to place concepts, such as active learning, in context, providing examples of how particular teaching techniques are employed in the department in question. The Peer GTA's presence in the courses also provided space for new GTAs to become familiar with an experienced colleague and peer, which in turn promoted ongoing teaching-related discussions as part of a teaching community.
There are also benefits for the Peer GTA. One such benefit was the opportunity to teach in a different format compared to their usual teaching, which is mostly in small-group tutorials and labs. The experience provided the Peer GTA with skills which are transferable to other professional settings, such as presenting at conferences and discussing research with colleagues in meetings. The opportunity to co-facilitate sessions about learning and teaching also provided the Peer GTA with the impetus to reflect on teaching practice which, in the UK context at least, was seen as useful preparation for subsequent applications for teaching fellowships (AFHEA or FHEA).
A key learning point from this study was the need to be flexible when conceptualising the role of a Peer GTA. It is important to enable Peer GTAs to lead appropriate sections of the training, where this is requested by the individual Peer GTA. However, we need to keep in mind the transient nature of the GTA role, which means it is not possible, or desirable, to completely redesign courses when each new cohort of Peer GTAs joins the teaching team. It is also important to remember that any work undertaken as a Peer GTA is done in addition to teaching work carried out in the GTA's department. Furthermore, it goes without saying that the priority is for the GTA to complete her or his doctorate in a timely manner.