Understandings of mentoring in school placement settings within the context of initial teacher education in Scotland: dimensions of collaboration and power

ABSTRACT This study critically addresses mentor and mentee understandings of mentoring primary education student teachers within existing operations of power in the context of Scottish Initial Teacher Education. Semi-structured interviews of mentors and student teachers were used to elicit relational understandings of the mentoring process within an instrumental, collective case study research design. Findings indicate that participants understood mentoring as a multifaceted process aimed at supporting the professional learning of student teachers. The article addresses the main mentoring relationship between the class teacher mentor and student teacher mentee. Analysis of their responses suggests an understanding of mentoring as involving both personal and professional dimensions. From these dimensions emerge the focus of this article: implicit collaboration and conceptions of power as a relational duality situated within a more Foucauldian ‘flux’ form. Critical discussion of findings extends understandings of the complexity of the mentoring process with reference to perceptions of collaboration, power and their co-existence. Conclusions focus on the need for a quality, consistent mentor education programme to promote an informed knowledge and understanding of the complex nature of mentoring in order to improve the quality and consistency of mentee learning experiences.


Introduction
Mentoring is a recurrent theme within the literature on initial teacher education. Current discourses of a 'knowledge society' emphasise the importance of lifelong learning and school improvement where achievement is shown to be linked with quality teaching and learning (Forde et al. 2006). This connection suggests that the processes of learning and teaching are significant in influencing student teacher success and mentor development. This means that more wide-ranging, complex knowledge, skill and competence bases are necessary to foster a future generation of teachers of appropriate quality, and has implications for the mentoring practices employed to foster such quality.
Within the context of Initial Teacher Education (ITE), this article is derived from an empirical study aimed at exploring mentor and mentee understandings of mentoring primary education student teachers within Scottish school placement contexts, and their perceptions of the use of formative assessment principles and practices to support professional learning within the mentoring process. This article builds on previous work (Mackie 2017) about class teacher mentor (CT mentor) and student teacher (mentee) understandings of mentoring primary education student teachers. While it may seem obvious that the CT mentor/mentee relationship is key in influencing student teacher success and mentor professional development, the study highlights the extent to which this relationship is critically important, having potentially long-term consequences for student teacher success and the quality of education being provided in the Scottish educational context. Analysis of CT mentor and mentee responses suggests an understanding of mentoring as involving both personal and professional dimensions. From these dimensions, elements of collaboration and different conceptions of power emerge. These elements are the focus of this article. It contributes a Scottish perspective to the international body of literature on mentoring within ITE contexts. It extends understandings of the complexity of mentoring relationships with reference to perceptions of collaboration, power and their co-existence, and highlights the importance of quality and consistent mentoring education opportunities. Additionally, it emphasises the role of mentees as well as those of the mentor, filling a gap in current literature (Ambrosetti 2010). Scottish education policy is used to frame and exemplify points made alongside those from international literature to inform discussions and suggest recommendations for future mentoring policy and practices.

Mentoring in initial teacher education
Learning to teach occurs within the realms of relationships with others (Harrison, Lawson, and Wortley 2005) therefore having a mentor who is part of the teaching community is essential (Hargreaves and Fullan 2000). Effective mentoring practices require a mentor who understands the complexities of the mentoring process (Harrison, Lawson, and Wortley 2005) such as the mentor/mentee relationship and its varied forms and functions influenced by external and internal contexts (Ambrosetti 2010).
Definitions of mentoring abound (Ambrosetti and Dekkers 2010). It may be viewed as where a more experienced colleague supports, challenges and facilitates the professional learning of another (Pollard 2005;Carnell, MacDonald, and Askew 2006). It involves a complex array of social interactions including an interpersonal element (Yeomans and Sampson 1994) which fosters an ethos of openness to individual and joint constructions of knowledge and understanding (Hargreaves 2010) through its emphasis on a more educative perspective on learning, one focused on independence and empowerment (Jones and Straker 2006).
The complex nature of mentoring means that there is no one 'recipe for success' (Harrison, Lawson, and Wortley 2005, 425) and is influenced by intercultural performances within learning environments (Carnell, MacDonald, and Askew 2006;Kemmis et al. 2014). For example, a 'quality assurance' context restricts mentoring to supervision emphasising skills (Rix and Gold 2000). Relationships are prescriptive, directive, normative and nondialogic (ibid.) and traditional separations (Sachs 2000) of expert/novice prevail. An opposing constructivist view sees collaborative school cultures employ more nondirective, educative approaches to develop autonomous, self-regulated learners (Iancu-Haddad and Oplatka 2014).
Collaboration is recognised as a key concept in contemporary mentoring practices (Hargreaves 2000;EPPI 2008) and is the act of working together to achieve something (Dictionaries 2015). It is central to mentoring as it improves teaching and addresses issues of social justice and equality through practices such as critical reflection, active trust, selfregulation, respect and reciprocity (Hargreaves 2000). However, all learning environments are influenced by power relationships (Seddon, Billett, and Clemans 2004). In the context of mentoring, traditional hierarchies of power may inhibit collaborative, constructive dialogue (Graham 1999) in that mentors are viewed as more experienced and may use this perception to direct dialogue, sanction particular actions and associated reasoning (Ritchie, Rigano, and Lowry 2000). Within this sort of traditional power duality, an oppositional dynamic of powerful/powerless becomes apparent (Seddon, Billett, and Clemans 2004) which may lead to mentees feeling exposed (Sewell et al. 2009), bullied (Maguire 2001) and relationships breaking down (Kim and Danforth 2012).

Study context
In Scotland, prospective teachers are required to undertake either a four-year undergraduate or one-year postgraduate qualification within a university provider followed by one year of induction (or equivalent) to evidence competence against the Standard for Full Registration set out by the General Teaching Council of Scotland (GTCS 2012). Mentoring quality within Scottish ITE has been reported as inconsistent in ITE review reports for some years (for example, Deloitte & Touche 2001;Education Scotland 2015). To promote quality provision, the latest review of teacher education recommends careful mentor selection based on their knowledge, understanding and skills of mentoring and associated assessment processes gained through appropriate mentor education (Scottish Government 2011). However, mentor education is currently only available to a small minority of mentors who oversee post-ITE induction year teachers (Education Scotland 2015). This disparity is problematic given the complexities of mentoring (Hall 2003) and the positive correlations between mentor education and the quality of subsequent mentoring practices suggested by research evidence (for example, Hennison et al. 2008;Hoffman et al. 2015).

Methodology
As noted previously (Mackie 2017), the methodology employed in this study is underpinned by constructivist epistemology. Whilst acknowledging the variety of conceptions of constructivism, two are prevalent in education (Fosnot 1996), namely cognitive constructivism, concerned with individual construction of knowledge and understanding, and social constructivism with its emphasis on constructing knowledge and understanding through social interaction (Bredo 2000). Constructivist research is exploratory and process-oriented concerned with detailed investigation of the perspectives of participants within their social and historical contexts (Littledyke and Huxford 1998;Jonassen 2006). These epistemological conceptions are appropriate for this study as it addresses mentor and mentee understandings of the mentoring process. This process is subject to both individual and social constructs whereby knowledge and understanding are developed both independently and collaboratively.
An instrumental, collective case study design was adopted: instrumental to examine an overarching case, the process of mentoring primary student teachers, in order to comprehend the phenomenon of mentor and mentee understandings of mentoring; collective to provide a more holistic view of that phenomenon in that individual cases is examined but situated within a collective study (Stake 2005). Case study research is reported as suited to the study of a single case rather than multiple cases (Simons 2009), however, it has also been argued that both are appropriate with regard to the use of multiple cases to promote increased understanding of a single case (Gillham 2000;Gomm, Hammersley, and Foster 2000). Cresswell (2007) advises that four or five cases are sufficient so analysis can be carried out within and across cases. Within the overarching collective case of the mentoring process, four individual cases were investigated: class teacher mentors (CT mentors), mentees (student teachers), school management mentors and local authority mentors. University tutors were not involved in the study as it is concerned with mentoring within school placement experiences and the everyday reality of that process. Within Scottish ITE in general, and the specific teacher education institution involved in this study, 'teamed models' of mentoring, where teacher mentors, university tutors and mentees collaborate in the process of mentoring the student teacher (Fenimore-Smith 2004), are not evident. University tutors from the institution in this study make one school placement visit to the student teacher unless there are issues with progression in learning to teach. Any other contact is carried out over email and tends to be limited. The placement visit gives a snapshot of the student teacher's progression so is insufficient in developing a substantive mentoring relationship with the mentee. Its function is a formative one where the tutor does not make any summative judgement, rather offers support to school mentors and student teachers where required.
A purposeful sampling strategy was selected for all participants. Those pertinent to CT mentors and mentees follow as the understandings of these participants are the focus of this article. Six years three student teachers of an undergraduate primary education degree were recruited. These students had experienced mentoring on a previous placement and would be able to employ experiences from this study to expand and extend professional learning, and to employ it in their final year four placement. The six CT mentors of these student teachers were recruited to make up six mentoring pairs (twelve participants in total). The structure of these pairs was to facilitate comparative analysis of responses where appropriate. CT mentors were all qualified primary class teachers situated in six Scottish primary schools. These schools were all of reasonable size, at least one class at each stage, which gave student teachers access to a variety of staff members. Schools were situated in both urban and rural areas within two local authorities. All of these schools mentor student teachers on a regular basis.
The British Educational Research Association's Ethical Guidelines (2011) were used as the main reference document for ethical considerations. Key aspects relevant to this study are informed consent, confidentiality, accuracy of reporting and positionality. For example, student teacher participants were from the researcher's own university programme thus awareness of potential issues was important such as traditional conceptions of university tutor/student hierarchy. This may lead, for example, to the 'interviewer effect' where interviewees say what they think is expected as opposed to giving an honest account (Denscombe 1998). To address such issues, it was stressed to mentees that the researcher's role in this school placement was not as a university tutor but as a doctoral student. Fears about students feeling unable to be open in their responses proved to be unfounded. In fact, the researcher's own attitudes and beliefs about the importance of establishing positive relationships with students were evident in that they appeared to feel at ease and therefore confident to be honest in their interview responses. This was evident through their non-verbal body language in appearing relaxed and able to share laughter and sharing amusing stories. Mentees' confidence in wanting to be honest was evident in the content of their responses, for example, in being critical of mentors and willing to share their thoughts on how the university and school might improve school placement experiences.
In research studies, any possibility of bias should be acknowledged as it may influence the objectivity of conclusions (Hancock and Algozzine 2006). A strategy of 'detached honesty' was adopted by the researcher in continually questioning and challenging all aspects of the research undertaken (Gillham 2000), for example, seeking advice about aspects of the research process from others and engaging in data analysis that looks for and discusses that which does not fit with emergent themes (ibid.). Throughout the process of the study, colleagues were an important resource in discussing the reasonableness of findings and subsequent interpretation, for example, at the coding stage of analysis then in synthesising codes to abstract key themes and in making sense of them. Further, critical analysis of findings was undertaken with due acknowledgement of the possibilities of multiple interpretations through reading of literature and communications with participants.
Methodological congruence is vital in any study so that an appropriate approach to data collection, analysis and interpretation is selected which is consistent with the researcher's epistemological position and the research design (Thomas 2009). In accordance with constructivist epistemology, these processes are viewed as active (Esterberg 2002) and shared where participants co-create data and meaning. Given the case study design of this study, the partially structured format of the semi-structured interview was appropriate as it fosters elucidation of data particular to individual cases as well as comparative analysis of data sets (ibid.). To explore their understandings participants were asked a variety of questions about mentoring based on a review of apposite literature. In addition, probing questions were used as the flexibility of this type of interview presents opportunities to explore further during the interview with regard to participants' understandings of their lived experiences (Kvale 1996) thereby promoting depth of analysis (May 2001).
Approaches to data analysis and theory generation were derived from constructivistgrounded theory as it offers detailed guidelines for analysis where theories are constructed from the data (Charmaz 2006). The classic notion of a researcher as 'silent' objectively 'discovering' data based on pre-existing theory is replaced by one who constructs participants' understandings (Hallberg 2006). A systematic approach to coding was employed (Charmaz 2006) within and across cases to promote the identification of themes and sub-themes. A collective case study requires within case analysis to identify key themes and foster thick description of individual cases prior to cross-case analysis (Simons 2009). Three stages of coding were adopted: initial, focused and theoretical. Initial coding is where data is analysed in detail through close reading of word, line and segments of data (Charmaz 2006). It helps researchers to interpret what is actually in the data and to avoid preconceived ideas being applied (ibid.).
Focused coding is a subsequent mechanism to refine codes, it 'uses the most significant or frequent initial codes to sort, synthesize, integrate, and organise large amounts of data' (Charmaz 2006, 46). This was carried out within each case and involved refining initial coding by examining codes for commonality and difference. Within each case, initial codes from each participant were compared for each interview question. Analysing codes for commonality was determined to be an effective basis for justifying construction of focused coding and subsequent theoretical coding/themes. In this respect, initial codes evident from over 50% (four out of six) of participants within each case were utilised in constructing focused codes whilst also paying attention to salient differences/individual voices. The constant comparative method inherent in analysis situated in an interpretive paradigm entails being flexible to revise codes whilst visiting and revisiting data as the research process progresses (ibid.). Revisiting data from initial coding while focused coding led to the emergence of new codes but also revision of existing ones.
Theoretical coding is the merging of concepts into groups (ibid.). It assists researchers to identify connections between codes so theories begin to emerge (ibid.). Focused codes were examined within each case with themes and sub-themes noted. These were then compared across the four cases in a cross-case analysis strategy to further refine themes and sub-themes ascertaining both commonalities and differences. Throughout the process, research memos about emerging themes, absences and interesting aspects were noted. This fostered construction of meaning from data at a more holistic level alongside finer detail derived from coding. The qualitative nature of this study means that traditional positivist notions of generalisation are rejected in favour of a conception of 'fuzzy' generalisations where no absolute social truths are stated. Instead, possibilities are suggested in terms of the possibility of research findings being more widely applicable with regard to influencing policy and practice through dialogue between relevant parties (Bassey 1999).

Findings and discussion
Participants understand mentoring as a multifaceted process comprising a range of relationships that support the mentoring of student teachers within school placements. Findings suggest that the CT mentor/mentee relationship is key. Analysis of CT mentor and mentee responses suggests an understanding of mentoring as involving both personal and professional dimensions. Referencing Yeomans and Sampson (1994) model, the professional dimension focuses on mentee professional learning in the process of learning to teach and associated mentor roles (ibid.). The personal dimension is concerned with interpersonal elements and involves mentor roles such as host, friend and counsellor (ibid.). Within the dimensions of this close working relationship, elements of collaboration and different conceptions of power emerge. These findings are summarised in Figure 1.
Findings of collaboration and different conceptions of power are the focus of this article. Within and cross-case data is summarised in Tables 1 and 2 followed by presentation and discussion of findings with reference to a selection of participant responses. Cross-case data is used for purposes of illustrative exemplification. It is selected as a representative of participants' experiences and to provide a variety of examples under each finding. For each example an overview of CT mentor and mentee understandings is provided then specific respondent data cited to provide further detail. Participants were accorded pseudonyms: CT for class teacher mentors, ST for student teachers, for example, student teacher A is STA.

Implicit collaboration
With regard to implicit collaboration, CT mentors and mentees noted having discussions as a central element, before and after teaching episodes, that were both formal and informal. In this respect, collaboration may be identified and described as both planned and spontaneous (Williams and Prestage 2002). While spontaneous collaboration is more effective in fostering the professional learning of beginner teachers (Patrick et al. 2010) due to the use of continual professional dialogue as the key support mechanism (Williams and Prestage 2002), structured collaboration is successful in improving mentoring practice as consistent structures are in place (ibid.).
From a social constructivist perspective, dialogue is vital in fostering learning and helps organise thinking, re-frame or build new understanding within social contexts (Vygotsky 1978). It is significant to beginner teacher learning (Schwille 2008) in terms of considering difficult situations, addressing anxieties and specific aspects of teaching (Hargreaves 2010). These conversations are an outside-the-action component of mentoring where mentor and mentee discussion takes place before and after teaching episodes in a cyclical form of dialogue and reflection (Schwille 2008). This educative mentoring is collaborative where mentors assist mentees to interpret and understand teaching and learning in order to  . Although advantageous in mentees' gaining practical experience, outside-the-action mentoring may promote inferior quality teachers in that the mentor role is on the periphery during teaching episodes and can take a supervisory, rather than educative, position depending on the nature of postlesson conversations (ibid.).
CT mentors and mentees indicated that both sides instigated discussion topics. Topics focused on mentee learning in emphasising planning and teaching practice. CT mentors talked about topics based around lessons and associated resources while mentees noted aspects such as preparation for the following day and any questions/issues that had arisen. I think they just came about because (participant's name) asked or maybe it was something that had come up if I'd looked at her folder. Or something that I thought of. CTD . . . a kind of mixture between the two of us but it was mostly situations that would arise during teaching or during the class time, even after school. And she would bring things up as well 'I noticed you did this, you could have done that better by doing . . . ' or 'how do you think you could have done that better'. So it was both of us really. STA Joint instigation of discussion topics reflects mentors taking an interactive role where topics of conversations are initiated by either party and are responsive to both mentor and mentee needs (Young et al. 2005). An interactive role requires equity in the mentoring relationship where mentors and mentees acknowledge each other's unique and valuable input (ibid.). Such relationships are indicative of more non-directive methods of mentoring where the focus is on facilitating mentee understanding through mentor scaffolding using a variety of educative strategies in order to foster responsibility for learning (Carnell, MacDonald, and Askew 2006). Such practices are collaborative, indicative of both individual and co-constructed learning (Hargreaves 2010). Encouraging self-evaluation was viewed as a mentor role within mentoring conversations. CT mentors viewed this as important in fostering the independent and reflective practice required of teachers. Given that mentees tended to be quite harsh in their selfevaluations, it fostered opportunities to encourage mentees to focus on strengths as well as development points. CTE noted self-evaluation in terms of assisting the mentee to engage in balanced reflection as well as specifying next steps in her teaching and for the children's learning. Mentees saw being encouraged to self-evaluate as significant in developing abilities to think for themselves in order to progress in their teaching practice. STB recognised that the CT mentor was encouraging her to think about her teaching practice in relation to her own experiences and knowledge.
. . . she would get me to think about it so I came up with an answer and she would say 'yes' or 'no, have you thought about this'. Getting you to enquire and think about your own experience and your own knowledge and how you can put that into practice. STB In encouraging mentees to engage in self-evaluation, CT mentors were again adopting an educative role in respect of challenge as opposed to instructional support. Challenge involves aspects such as encouraging experimentation, sharing insight, asking questions, encouraging reflection and co-planning (Certo 2005). Such practices are collaborative and educative in that mentors facilitate mentee understanding in terms of encouraging them to consider a variety of perspectives (ibid.) and to develop the professional autonomy (Harrison, Dymoke, and Pell 2006) vital for entry into a profession involving both pupil academic and pastoral care where teachers need to take responsibility for their own professional learning in order to foster quality learning and teaching (Hudson 2013). Patrick et al. (2010) maintain that mentoring should not be entirely instructive but promote mentee capacity to critically reflect on practice and encourage them to evaluate a variety of teaching for effective learning strategies. In opposition to this study's findings, Certo (2005) reports that previous studies indicate a lack of challenge and predominance of instructional support. This may be explained in that challenge is argued to be appropriate when a basis of competence has been achieved (Harrison, Lawson, and Wortley 2005). However, this assertion seems inaccurate as beginner teachers do have capacities in areas associated with challenge even with limited teaching experience (Eraut 1995). If challenge is not evident mentees may fail to develop the broad range of knowledge, understanding and skills required. Instead, compliance to current procedures may dominate and result in stagnation of practice (O'Brien and Christie 2005).
Contrary to this study, a variety of others indicate that encouraging self-evaluation was not a dominant mentor role (see, for example , Certo 2005;Harrison, Dymoke, and Pell 2006). Instead, mentors led and dictated mentoring conversations, which were focused on their own thoughts and experiences (Certo 2005). Indicative of an educative approach, CT mentors in this study viewed self-evaluation as important in getting mentees to think independently as required by qualified teachers. Mentees also recognised this skill in terms of progressing in their teaching practice. This dual focus on current progression as well as future practice is important in assisting beginner teachers' professional learning (Schwille 2008).
This study also centred around wider educational contexts and mentee professional learning, for example in terms of highlighting the positive influence of mentorship (Orland-Barack and Hasin 2010) given the unpredictable nature of teaching and learning and the limited experience of beginner teachers (Patrick et al. 2010). Further, given the attention accorded to collaboration within the current Scottish education context, influenced by the wider educational discourses and inherent agendas of a knowledge society, study participants may have been expected to be more explicitly recognisant of this element. In Scotland, several significant reforms evidence a focus on collaboration through recommendations about collegiality and engaging in broader partnerships. For example, the McCrone Agreement on teacher pay and conditions promotes a cultural shift from compliance to collegiality (MacDonald 2004), and the most recent review of ITE advocates professional learning in mentoring within a model of enhanced partnership between schools, universities and local authorities (Scottish Government 2011).
The lack of explicit recognition of collaboration may be a question of participant understanding about the complexities of mentoring due to a lack of provision of professional education opportunities but also of the aforementioned education reforms. Confusion may be evident given the tension between reform rhetoric and reality where collaboration is promoted but within omnipresent managerialist external agendas not conducive to collegiate practice in their focus on accountability, competence and pupil attainment. However, it may be that teachers do understand these reforms and that they are engrained in school cultures and structures so are established practice. Study findings are indicative of implicit collaboration suggesting that school cultures were positive in this sense. Strong school cultures champion collaboration where teachers are able to discuss the nature of learning and teaching whereas schools with individualist cultures are not conducive to such practices (Williams and Prestage 2002). Beginner teachers who work within collaborative cultures are reported as experiencing greater professional learning and personal fulfilment (ibid.). Collaborative classroom and mentoring practices follow from collaborative school cultures.

Power
In opposition to the notions of explicit and implicit collaboration, conceptions of power emerged from CT mentor and mentee understandings of mentoring. This study showed that power is interpreted as a traditional opposition (duality), where one party is perceived as powerful and one as powerless, and as a more Foucauldian 'flux' form where no one person is viewed as owning power, rather actors can be both powerful and powerless in the same context (Foucault 1979). Within and cross-case data is summarised in Table 2 followed by presentation and discussion of a selection of findings.
Power as duality All learning environments are influenced by power relationships (Seddon, Billett, and Clemans 2004). In mentoring relationships, a power duality manifests as a perception of mentors as the 'expert' and of mentees as the 'novice' (Berliner 2001). This is due to mentors being positioned as authoritative with the school as qualified teachers and because of their greater experience, assumed knowledge, understanding and skills (Garvey, Stokes, and Megginson 2009). Berliner (2001) argues that this is inaccurate as expertise does not necessarily equate with experience: some teachers may never progress beyond a level of basic competency. Mentoring is indicative of a training approach, where more directive strategies and skills are employed (Yeomans and Sampson 1994), symptomatic of a perception of teaching as simplistic (Hargreaves and Fullan 2000) as opposed to the realities of its complexity. Such methods are mentor focused and encourage mentee dependence suggestive of managerialist notions of compliance (Harrison, Lawson, and Wortley 2005).
An understanding of CT mentors in positions of power and mentees as powerless was evident in the data. For example, the mentoring styles of CT mentors encompassed a directive element. CT mentors noted that mentees require a lot of guidance as they are learning to teach. CTB felt that student teachers could not be left to their own devices. CTD noted that the class was her responsibility so she had to retain control over learning and behaviour. In addition, CT mentors' style of mentoring was seen as dependent on the mentee's teaching confidence and competence. CTB noted that her mentoring style differed according to the capacity and character of the mentee but also felt that they should be specifically positioned as learners.
Well I'd probably be more inclined to . . . be more directive . . . so that the student would realise . . . that they don't actually know it all. They have to learn . . . you would be slightly different depending on the ability and character of the person that you're the tutor for . . . some would need a lot more help than others. CTB Responses indicate that as mentee competence improved, mentor confidence increased and so mentoring styles became less directive. The importance of giving mentees the freedom to try things out as part of their learning was acknowledged and that making mistakes was important. STE noted her CT mentor as being more directive at the beginning of the placement as she was in the process of ascertaining her mentee's teaching capacity. This decreased as she developed more confidence in the mentee's abilities. STE linked this to being trusted in a professional sense.
The directive element of CT mentoring styles positions mentees as learners. This was attributed to them being beginner teachers and CT mentors having responsibility for the learning of pupils. This positioning aligns with literature regarding mentor issues with adopting more non-directive methods (Langdon 2014) and the resultant danger of mentees feeling bullied due to mentors' demands for practice to be carried out in specific ways (Maguire 2001). However, mentees may desire direction (Strong and Baron 2004), employing mentors as sources of information and skills to foster performance as opposed to learning through constructing and co-constructing understanding (Hargreaves 2010). Further, direction may be required based on mentee learning needs so the demands of more non-directive methods might not be conducive to their professional learning. A directive approach can be disadvantageous in that it may be restrictive rather than promotional of mentee capacity (Young et al.;Hargreaves 2010). Conversely, it can also assist mentees to develop some degree of understanding of the academic and pedagogical facets of teaching (Schwille 2008) but this may be limited to a conception of teachers as classroom technicians as opposed to an educative view of those able to reason and exercise their professional judgement based on informed practice.

Power as 'flux'
It is difficult to avoid traditional hierarchies of power as they pervade school cultures and structures (Fenimore-Smith 2004). However, considering other conceptions of power provides alternative interpretations and bases for reflection. Power may be viewed as resistance (Foucault 1979). In this study, mentees noted that they felt the need to be compliant in mentoring conversations and that they could not communicate differences in opinion due to the mentor being more experienced. STA remarked that, due to her mentor's experience, she did not feel able to disagree if the opportunity arose but would ask questions in a respectful manner. In such a situation reading on the topic helped her to view it as a difference of opinion rather than right or wrong. This may be viewed as resistance to CT mentors' enactment of power. This resistance distorts the traditional power duality of being silenced so therefore powerless. Therefore, a conception of power as 'flux' is apposite here where participants may be interpreted as being both powerful and powerless in the same context (ibid.). In this sense, power is viewed as enacted within interactions, rather than something that is possessed (Foucault 1979), and is in a constant state of change (Graham 1999).
With regard to learning, mentees are positioned as being both learners and being learned from, powerless and powerful, respectively. To exemplify, regarding characteristics of an effective mentee, mentees are positioned as powerless in the role of learners in the classroom of a more experienced teacher. A willingness to learn by trying out new things was viewed as vital by CT mentors. CTE noted that mentees need to learn how to improve through failure but should not take such experiences personally. CTF maintained that taking ownership of learning was important. Mentees noted being open to new ways of teaching and guidance from mentors as key characteristics of an effective mentee. STB remarked on learning as happening progressively and the importance of having your own 'knowledge' but being open to other ideas. In addition, learning was viewed in the realms of advance preparation for being in class.
You have to learn, and I think it is something you acquire over time. The most important thing is obviously have the knowledge . . . about your role as a teacher but being open to observe, to get instruction, to take things in that are going on around you rather than having the mind-set 'I am going to do this my way' . . . You have got to put the background work in before you go into the class . . . STB In opposition to the above, mentees are positioned as powerful in responses about the benefits of mentoring for mentors. They are viewed as being learned from and as making mentors reflect more on their own practice. CT mentors viewed mentees as sources of learning on different aspects of teaching, such as curriculum, pedagogy and assessment, based on their university learning and other teaching experiences. CTF remarked that she was aiming to improve her own teaching by observing and talking to the mentee. I like the idea of they are coming out with all these brand new ideas . . . Being able to use some of the ideas that they are coming in with . . . really I am looking to improve my practice by taking on anything that I have learned from watching the student teach or professional dialogue that we have together . . . CTF Mentees thought that they could bring new ideas for curriculum and teaching strategies if mentors were receptive to learning from them. STA noted that mentors may be encouraged to look at things differently rather than being stuck in a rut of the same practices.
Mentees are also in a position of power in that their presence in class makes the mentor reflect on their own practice more than usual. CT mentors talked about seeing themselves as role models for mentees so felt it was important to think about their teaching practices carefully. CTD explained that it made her consider not just what she did in class but why. In terms of professional learning, CTF noted that this heightened reflection could highlight areas for development in her own practice. Mentees felt that having responsibility for a student teacher made CT mentors think about their own practice in more detail and to question it. STA remarked that this depth of self-evaluation might not be a usual occurrence given the demands of daily classroom practice. However, by engaging in more in-depth reflection STC noted that CT mentors could be learning from the mentee.
Mentor willingness to position themselves and be positioned as learners is demanding as being seen to alter a perspective may be viewed as a weakness (Ulvik and Sunde 2013). In this respect, personal and professional trust is important and can facilitate openness to cognitive conflict around existing ways of working and thinking. This conflict opens up the opportunity to learn and understand teaching in new and different ways leading to a more symmetrical mentoring relationship appropriate for co-constructed knowledge and understanding. Mentees are empowered with confidence and competence through this position. This context can also be beneficial for mentors if they have identified development needs that can be addressed through engagement in mentoring (Ambrosetti, Knight, and Dekkers 2014). It may also challenge them to question existing school cultures and their inherent understandings, beliefs and values (Ulvik and Sunde 2013). This conception of the mentor/mentee relationship is one of the reciprocity (Ambrosetti and Dekkers 2010;Ambrosetti, Knight, and Dekkers 2014) and, as such, is in opposition to the traditional notion of mentoring based on mentor knowledge (ibid.) so may assist in avoiding stagnatory practice. Collaborative school cultures are more cognisant of learning for all and so more effective contexts for beginner teachers (ibid.) because learners occupy positions of visibility (Long et al. 2012). However, it is important to note that in the process of power differentials becoming more equal, mentoring relationships may suffer as the mentee gains in confidence to question mentor capacity (Garvey, Stokes, and Megginson 2009). Maintenance of an effective relationship is dependent on how this is approached by both parties with collaboration and positive interactions as key.

Conclusion
This article addressed mentor and student teacher understandings of mentoring within Scottish ITE school placement contexts. Findings indicate that participants understood mentoring as a multifaceted, complex process aimed at supporting the professional learning of student teachers. This article extends understandings of the complexity of mentoring relationships evident in current research in its contribution to interpretations of collaboration, power and their co-existence within a Scottish ITE context. Additionally, it emphasises the role of mentees as well as those of the mentor, filling a gap in existent literature noted by Ambrosetti (2010).
It may be concluded that the aforementioned complexity highlights the importance of quality and consistent mentoring education opportunities to promote an informed knowledge and understanding of the complex nature of mentoring in order to improve the quality and consistency of mentee learning experiences. Mentor education is widely reported as significant in fostering such mentoring experiences (for example, Kwan and Lopez-Real 2005;Pogodzinski 2012;Langdon 2014). However, it is variable with regard to its quality and availability (for example, Bubb, Earley, and Totterdell 2005;Harrison, Lawson, and Wortley 2005). Therefore, all teacher education providers and student placement schools should attend to the provision of effective mentor education opportunities in order that student teachers are adequately supported in their professional learning. As noted in the previous work (Mackie 2017), student teachers also require such opportunities so that they are more prepared to effectively engage in the mentoring process.
In considering the quality of mentor education, it is important that content addresses the complexities of mentoring rather than an oversimplification of the process which leads to a proliferation of undifferentiated mentoring techniques underpinned by a lack of understanding of substantive rationale for their use. The variety of complex mentoring roles, relationships and potential tensions need to be attended to alongside appropriate practices to address them (Jones 2009;Fransson 2010;Achinstein and Davis 2014). Elements in tension, such as those of collaboration and power identified in this study, are particularly significant given that movement to more equitable mentoring relationships brings the matter of their co-existence to the fore. In this study collaboration and power coexist within the professional dimension of mentoring, for example, within mentoring conversations where mentors give feedback but also encourage self-evaluation. Quality mentor education programmes are vital in assisting mentors and mentees to moderate such conflicts in order to promote effective professional learning experiences.

Disclosure statement
No potential conflict of interest was reported by the author.