Dealing with policy expectations of mentoring newly qualified teachers – a Norwegian example

The purpose of this study is to contribute to knowledge about school leaders ’ and mentors ’ perceptions, experiences, and legitimizations of national policy expectations about the organizing of mentoring of newly quali ﬁ ed teachers (NQTs). The analytical framework is based on perspectives of governing with attention to policy by expectations combined with theory of making policy solutions thinkable, calculable and practicable. The thematic analysis of interviews of school leaders and mentors suggests that the organizing of mentoring NQTs is understood as a multitude of practices, the policy initiated NQT mentoring becomes a precarious practice within the school organization, and the legitimization of NQT mentoring reveal diverging purposes of such a practice. The ﬁ ndings raise concerns about how such perceptions, experiences, and legitimization can contribute to professionalizing and ensuring teacher retention.


Introduction
The phenomenon in this article is how school leaders and mentors deal with policy expectations regarding mentoring for newly qualified teachers (NQTs).Both internationally and in the Norwegian education policy context, which is the empirical example here, the induction (of NQTs) is a primary measure to prevent teacher attrition (Europeancommission, 2010;Fletcher & Mullen, 2012;Norwegian Directorate of Education, 2018).Moreover, induction focused on mentoring is considered significant in terms of contributing to the policy aim of professionally developing NQTs (Ingersoll & Strong, 2011;Tang & Choi, 2005).The purpose of this study is to contribute to knowledge about school leaders' and mentors' perceptions, experiences, and legitimizations of national policy expectations about organizing for mentoring of newly qualified teachers (NQTs).
In the current study, we utilize both "induction" and "mentoring" as concepts, where we understand induction as the wider process of introducing and developing NQTs into the profession and mentoring as the practice where a more experienced teacher support a less experienced teacher toward professional development.
In a Norwegian education context, mentoring NQTs is part of a policy expectation toward school leaders' organizational responsibility (Jacobsen et al., 2023).Generally, school leadership is essential for how policy initiatives are enacted in schools (Coburn, 2005;Hargreaves & Fullan, 2000;Leithwood et al., 2004).Influenced by an international governance trend of quality demands, school leaders have experienced increased pressure in their leadership practices (Verger et al., 2019).Policy pressure emphasizes challenges concerning student outcome and wellbeing, professional development for teachers, teaching quality as well as issues of teacher retention to name a few.These challenges are also arguments underpinning why mentoring in induction is considered important (Aspfors & Bondas, 2013;Europeancommission, 2010;Fletcher & Mullen, 2012).Policy expectations can lead to performativity pressure in schools, which may produce fabrications similar to what is described by Hobson and McIntyre (2013): establishing "fake-practices" to encounter policy demands.In a Norwegian education context, policy expectations concerning NQT mentoring is characterized by low stakes accountability and much leeway for practitioners to enact policy expectations (Gunnulfsen & Møller, 2021).Such a context is relevant to scrutinize as policy demands regarding mentoring here differ from other contexts where NQT mentoring is compulsory and associated with teacher registration programs like e.g., in Australia, England and the US (Hobson et al., 2009;Kemmis et al., 2014).
In the analysis, we utilize theories of governance and accountability presented by Theisens et al. (2016) and Hopmann (2008), allowing us to investigate how mentoring as a measure to promote professional development and secure a good induction period for NQTs is perceived, experienced, and legitimized by school leaders and mentors.
The present article displays Norway as an illuminating example of how policy expectations for mentoring NQTs are perceived, organized, and legitimized.In 2017, Norwegian educational authorities launched a national framework with principles and guidelines for mentoring NQTs to provide a tool for the institutionalization of mentoring NQT (Ministry of Education and Research, 2017).The Norwegian educational policy context both encompasses local autonomy and a responsibility for reaching national goals (Hopmann, 2008).The approach of the Norwegian authorities has focused on presenting ideas about mentoring NQTs, trying to make these ideas available for local practitioners like school leaders, and trusting them to find ways of practice (Hopmann, 2008;Jacobsen et al., 2023).Education policy in Norway is characterized by "soft accountability" (Maroy & Pons, 2019), which indicates low or moderate accountability stakes.Mentoring of NQT in Norway is an example of how "soft" accountability works, as the main means to enact such a policy initiative provided by the authorities is guidelines, agreements and encouragement, and no legislative measures to support the effort of schools and local educational authorities in organizing for NQT mentoring.
Although much is known about several aspects of mentoring, little is known about school leaders' and mentors' perceptions and professional experiences of organizing for policy initiated NQT mentoring in Norway.The present study focuses upon upper secondary schools.Upper secondary schools are often considered particularly resistant to reform efforts (Leithwood, 2016).
The study research questions: 1. How do school leaders and mentors perceive the policy demand for mentoring NQTs? 2. How do school leaders and mentors experience the school's organization of the practices of mentoring NQTs? 3. How do school leaders and mentors legitimize the school's organization of mentoring NQTs?
We use the terms school leaders, principals, and middle-level leaders when addressing the different actors from a formal leadership level who are engaged in mentoring.Leadership is also utilized to describe the practice of these actors regarding mentoring.School leadership is understood as a collective activity directly linked to educational purposes, in which professional school actors work to organize and develop the school's practices (Gunter, 2016).Leadership is understood as the way in which power and conflict are regarded as parts of the processes of communication between individuals and groups in an organizational context (Ogawa & Bossert, 1995;Sørhaug, 1996).
Below, we first present concepts and relevant research regarding practices of mentoring of NQTs, which is followed by a presentation of the analytical framework.We present the methodological approach and, subsequently, the analysis of the findings.In the discussion and conclusion, we address issues that need further attention in future research on organization of mentoring in upper secondary schools.

Mentoring NQTs in policy and research
The mentoring of NQTs has been a priority issue for policymakers, teacher educators, and researchers for decades (Hobson et al., 2009;Long et al., 2012;Spooner-Lane, 2016).The report "A Flying Start" (OECD, 2019) encouraged mentoring to be regarded as an integral part of a "coherent system of initial teacher preparation that can serve as the foundation for a process of continued development throughout the full duration of a teacher's career" (OECD, 2019, p. 11).This report showed policy expectations at an international governance level.
For our purpose here we display the definition of mentoring provided in the Norwegian policy documents 'as a structured and a planned, systematic process carried out individually and in groups'.We acknowledge that such a policy description differs from other definitions e.g., the definition suggested by Hobson et al. (2009) as support of a novice (mentee) by a more experienced practitioner (mentor), designed primarily to assist the development of the mentee's expertise and to facilitate their induction into the profession and into the specific local context (p.207).The term "mentoring" has caused debate over the years, being contested as to what it really is, what practices it may include, and what entails effective mentoring (Haggard et al., 2011;Hobson et al., 2009;Hobson & Malderez, 2013;Ingersoll & Strong, 2011;Kemmis et al., 2014).Kemmis et al. (2014) presented how mentoring NQTs is perceived differently in three different countries (Australia, Finland, and Sweden) in not only how mentoring is understood, but also about how it is enacted and who is involved (Kemmis et al., 2014).
A recent evaluation report from Rambøll 1 (Halmrast et al., 2021) stated that just over 50% of all upper secondary schools in Norway provide mentoring for their NQTs.Hence, mentoring practice is not a highly prioritized activity in upper secondary schools, though mentoring as a solution to problems of teacher attrition and a lack of connection between theory and practice in teacher education has been strongly encouraged in Norwegian policy documents dating back to the 1990s (Bjerkholt, 2013).The results from TALIS 2018 in Norway, as well as the evaluation results from Rambøll (Halmrast et al., 2021;OECD, 2020), indicated that there are some inconsistencies in how school actors perceive the policy demand for NQT mentoring in Norway.Moreover, Norwegian NQTs do not experience mentoring according to government intentions, while school leaders believe they facilitate adequate arrangements for NQTs (Carlsten et al., 2020;OECD, 2020).School leader's engagement is important to enable and facilitate the structures that support the induction process (Ingersoll & Strong, 2011;Long et al., 2012;Tiplic et al., 2020).Other contributions show that school leaders are central for NQT mentoring, including coordinating and facilitating mentors and the recruitment and retainment of teachers (Hobson et al., 2009;Ingersoll & Strong, 2011;Wood, 2005).Other studies suggested that an increased understanding of school leader's role identification and clarity related to mentoring programs and work with NQTs are crucial for mentoring program success (Kutsyuruba, 2020).This implies that school leaders must acknowledge that mentoring is "the way to go" to ensure the professional development and inclusion of NQT.A study regarding a teacher registration reform in Sweden (2011) showed how principals struggled with balancing between assessment and support of the NQTs (Gerreval, 2017).Regarding the same reform, Frelin and Fransson (2019) showed how principals were creative translators of policy intentions which is supported by a Norwegian study on school leaders enactment of state reforms (Hall et al., 2023).Both Frelin and Fransson (2019) and Hall et al. (2023) however presented challenges the school leaders met in dealing with tensions emerging with such policy enactment.A Norwegian study from 2014 illuminated that few of the school leaders considered the 1 Norwegian consultancy company.
mentor role as a professional activity and neither did they perceive education for mentors as necessary (Sunde & Ulvik, 2014).Willis et al. (2019) found that mentors responsible for mentoring NQTs are in unacknowledged roles as middle leaders in their schools, with central tasks of organizing the professional development and induction of NQTs.Despite the emphasis in educational policy and research on teacher retention and the anticipated function of mentoring, the mentor role as a leader is unacknowledged (Lejonberg et al., 2019;Margolis, 2012;Willis et al., 2019).Willis et al. (2019) underscored the role of the mentor as an interpreter and enactor of policy at the micro and meso levels.These findings provide valuable perspectives regarding the mentor's role in organizing for the mentoring NQTs.However, the findings have not addressed the perceptions of the many actors involved in the organization of mentoring of NQTs and the interaction between them concerning policy initiatives.

Analytical framework
The analytical framework helps us to elucidate the Norwegian education governing context, and how policy demands regarding mentoring NQTs are played out in practice.
The policy must be thinkable, calculable, and practicable Theisens et al. (2016) provided three analytic concepts suggesting that, to realize a policy proposal, the idea must be made thinkable, calculable, and practicable to actors in relevant fields.This means that policy ideas must first be presented in a certain language to be thinkable to appear intelligible and amenable to change.Second, governing authorities or other central actors may provide instruments for practitioners to realize or make calculable policy ideas, and these instruments may be seen as the embodiments of these ideas.Third, the "proof" of localized policy enactment is if the suggested ideas become practicable, that is, to what extent actors in the practice field avail themselves of the provided instruments.Supposedly, the educational authorities provide means or instruments that will enable the enactment of the policy idea or, in other words, the suggested solution.The trilogy of Theisens et al. (2016) provide an analytical lens for investigating empirically how policy intentions are perceived, experienced, and legitimized by school leaders and mentors.

Governing by expectations
Accountability is a crucial concept concerning the perceptions, experiences, and legitimations of policy ideas.We apply Hopmann's (2008) framing of accountability as a tool for social change, in which retaining NQTs are a part.Accountability is not a universal phenomenon in different contexts because it is deeply engrained "constitutional mind-sets" (Hopmann, 2008, p. 417) like cultures that define the relations between institutions and the public.Accountability, may in some contexts be described by management of expectation, which is described as "more or less welldefined expectations of what has to be achieved by whom" (Hopmann, 2008, p. 424).
In a Nordic context, schools' enactment of policy initiatives has been described as "muddling through," meaning that policy initiatives involve no real stakes and enactment takes shape as arbitrary activities and processes with inconclusive outcomes (Hopmann, 2008, p. 442).Regarding the policy initiative of mentoring NQTs in Norway, this description may be well suited because the initiatives can be described as encouragement and persuasion through policy documents and local policy governing with no real stakes.Such, governing by expectations delimits the responsibility of institutions by making ill-defined problems into better defined problems, which are easier to manage (Hopmann, 2008).We find the combination of governing by expectations with Theisens et al.'s (2016) trilogy, of thinkable, calculable and practicable as relevant when investigating school leaders' and mentors' perceptions, experiences, and legitimations of the policy demand of mentoring.Such a combination may illuminate approaches to making policy ideas available for practice and how this plays out in an accountability setting characterizing the policy expectations of mentoring of NQTs in this Norwegian example.

Methodology
The present study is a qualitative multi-case study design (Yin, 2018) based on 13 individual interviews with school leaders and mentors in four upper secondary case schools located in different parts of Norway.

Criteria for selecting the schools
Selecting school leaders and mentors as two groups of informants validates statements coming from school leaders about the organization of mentoring NQT in parallel with experiences from the mentors.The four case schools contribute to a range of perceptions, experiences, and legitimations of arranging for the NQT mentoring policy intention.The upper secondary level was relevant because this school level has provided less mentoring of NQTs than lower-level schools in Norway (Halmrast et al., 2021).The criteria for selection include geographic distribution and schools with practices of organizing the mentoring of NQTs.The school size was a concern for ensuring the informants' anonymity because Norway is a small country and the local school communities are transparent.The individual interviews were carried out digitally via Zoom due to pandemic restrictions for school visits.The research project was approved through Norwegian Agency for Shared Services in Education and Research.Informed consent was sought from all informants.Video data from Zoom was deleted immediately after the interview, and only audio data was utilized in the analysis.All data have been transcribed and anonymized.The informants were asked questions about perceptions of both national and local educational policy concerning mentoring of NQTs, how, where and why mentoring was organized in their school and who was involved.Table 1 presents an overview of the case schools and informants.

Process of coding and analysis
The transcribed interviews were subjected to thematic analysis (Braun and Clarke (2012).Analysis was carried out inductively.The reading of the material followed a four-step process to identify themes relevant to the three research questions regarding perceptions of policy signals, experiences of mentoring organization, and legitimization of these practices.Data coding and analysis followed an iterative process of inductive coding until we identified analytical themes representing some patterns or meaning.These themes were developed across the cases, and we explored whether the themes worked in relation to the data (Braun & Clarke, 2012); see Table 2).The analytical framework was activated in the fourth step of the analysis and in the discussion to elaborate on and further discuss the findings.To illustrate how we as researchers worked in the analytical process, we read through the material individually, and thereafter we met to discuss the themes and categories and planned for the next step of the analysis.We repeated this process through the three readings, making sure to focus the analysis directed at the purpose, the research questions, and the analytical framework such as the notion of the thinkable, the calculable and the practicable.The analysis and interpretations were discussed with other researchers as part of the communicative validity process (Kvale & Brinkman, 2009).Although the findings are not generalizable, the analysis and discussion provide transparency and may offer an analytical understanding of a specific phenomenon (Kvale & Brinkman, 2009).

Presentation and analysis of data
In the following section, we present findings related to the investigation of how school leaders and mentors perceive, experience, and legitimize national policy expectations about organizing mentoring of newly qualified teachers (NQTs).We identified five thematic areas representing the findings across the three research questions: (a) unawareness of policy documents, (b) policy enactment as window dressing, (c) mentoring NQTs understood as a multitude of induction practices, (d) NQT mentoring as a precarious practice and (e) NQT mentoringa practice with diverging purposes.

Perceiving the policy demand for mentoring
When asked about how the informants perceived the policy signals from the national authorities regarding the mentoring of NQTs, the responses covered various aspects of policy demands: First, the principals, middle-level leaders and mentors reported that they perceived there is a general policy demand to take care of newcomers and that professionalizing NQTs and ensuring teacher retention is a recognized matter of development.
Several respondents from the leadership level perceived the need to induct the NQTs as a collaborative responsibility, not a matter of individual set-ups between one chosen mentor and the NQT.The notion of a common responsibility is conceptualized by the principal at School 3: When we think of mentoring the NQTs, we [the leadership group] base it mainly on how we instruct the teams that have a main responsibility in the inclusion of the newly employed.[…] There is a clear expectation of the How are policy ideas about mentoring of NQTs reflected in the practice of organizing for mentoring?
team to include the newly appointed teacher, there should be no doubt in a new employee that it is ok to ask colleagues for help, and such a mandate for the team gives a sense of reassurance.This is critical.(Olav, Principal, School 3) The excerpt illustrates a perception of the policy expectation of arranging for mentoring practices as a collective responsibility for inducting NQTs and making them feel safe and included, not so much a specified practice of mentoring NQTs.The collective responsibility of mentoring-or rather induction of NQTs or newly employed-is especially directed at the teacher team being closest to the NQTs in daily practices.The principal at School 3 expressed professional practice should be the main strategy in welcoming NQTs.Such expression illustrate that there is a general understanding of the need for induction, but not a particular consciousness about mentoring as the most central measure which is the expectation forwarded by the national educational authorities.
The second aspect of perception concerns the lack of knowledge regarding policy demand described in policy documents like "the national framework for mentoring NQTs".The principal at School 4 stated, "I would be exaggerating if I said that I know it well.But I have … I have seen it [the national framework for mentoring NQTs]".This principal only knew the policy demand on the surface.This was true of all the principals we interviewed.Interestingly, this does not imply that the school did not have a structured organization of NQT mentoring, because in the case of school 4 and school 2 it did.
The mentors had slightly better knowledge of the national policy demands regarding mentoring NQTs.The coordinating mentor at School 1, Christine, expressed, "I don't remember exactly what and how much mentoring [the framework directs], but they are entitled to have it [mentoring] in their first period as a teacher."The informant expressed that the NQTs have a right to receive mentoring, which was a general perception among the mentors.This perception is not completely accurate because the national policy demand is a strong recommendation, not a legal right, but it illustrates how the perception of the policy demand was present when the mentors were asked about the policy demand regarding mentoring.
The data show that the school leaders and mentors were unaware of the detailed content of the national political expectations expressed in central policy papers, though the respondents were aware of why mentoring NQTs is important, not how the schools were supposed to enact policy expectations.
The third aspect of perception concerned how the informants perceived policy signals coming from regional authorities. 2According to the principal, the aim was to make a design suitable for the school: … then you draw up your own local plan … as all schools are different.And you can make your own design.We made our own mark by cutting out two meetings from the recommendations [from regional authorities] … we thought, how can we do less?That's what our contribution was about (laughter).(Olav, Principal, School 3) One of the principal's approaches was to cut down meetings compared with what the regional authority had expected in their proposed guidelines, indicating that the policy demand for NQT mentoring is not taken too seriously by the school leadership.They did establish an arrangement for a mentoring practice, which may appear as a symbolic response to the expectations of the regional educational authorities, a way of "window dressing," which is a notion applied to illustrate how the mentoring practice does not involve any structural change (Lin, 2010), but is more a matter of what is displayed.Such a finding was supported by what we found in another statement from same school: This [the mentoring of NQTs] has been established because the principal was directed to do this from the regional educational authority, who said that you ought to provide this.And then the school leaders got 2 Upper secondary schools in Norway are governed by national and regional educational authorities.

SCANDINAVIAN JOURNAL OF EDUCATIONAL RESEARCH
together and created a small arrangement.… No one else in the staff is involved in this (Hedda, department head and mentor under training, School 3).
It seems that Hedda believed the mentoring practice was arranged first and foremost because the schools were expected to comply with policy expectations, and the easiest way to organize the mentoring was chosen: not include too many formal dispositions.Here, Hedda seemed to express it as an issue that mentoring was not included in other structures and practices within the school.

Experiencing the organization of mentoring
In all schools we found organization of NQT mentoring as a response to policy expectations.The mentoring was however carried out differently in each school.Generally, the organization can be categorized into two different set-ups: (1) organized by a coordinating mentor who provides mentoring individually or in groups.The mentor had a mentoring education and released time to perform this role, or (2) the mentoring was organized in group meetings lead by the leadership team outside of working hours.Two of the schools, Schools 1 and 4, performed the first set-up, while schools 2 and 3 performed the second set-up.None of the schools received any external funding to organize mentoring locally.
In addition to such policy related organization, all the schools provided all types of different setups for NQTs which applies more to the categorization of induction, like "buddies", two-teacher systems, teacher teams as responsible for induction of NQTs and databases with practical information.For instance, one school leader in School 3 described that: " … mentoring is first and foremost carried out by the teacher team (Kristine, middle manager)", and in school 2, the assistant principal states that "we mentor those who we are responsible for as middle managers … ", and in a third example from school 3 the principal says: " … we also appoint "a buddy" to make sure that the NQT feels reassured that there is a person available for the NQT".These practices would be mentioned when the informants were asked about how they organize for mentoring of NQTs indicating that mentoring of NQTs is conceived of as a multitude of initiatives and practices, and not particularly mentoring as it is described in policy guidelines.
The division between NQTs and other new employees also seems to be blurred in many of the schools because it could be difficult to define exactly who could be categorized as NQTs.In one school, the organization of mentoring of NQTs or induction was not necessarily directed at NQTs but also at other newly employed: There are three categories; we have the ones who are completely newly qualified.And we have those who are new at the school, but who may have been a teacher elsewhere … Or there are the ones somewhat in between, who are still in their first years but newly appointed here.And everyone should really have some [mentoring].(Victor,Principal,School 4) Often, NQTs are in temporary positions, which may represent organizational challenges when offering mentoring in accordance with policy guidelines.Challenges may, for instance, regard time and resources to facilitate mentoring: Newly qualified teachers are usually not permanent employees … a typical way into our school is that you start in a small temporary part-time position, and then if you need mentoring, it can simply be difficult to find time.(Oliver, principal, School 1) The principal points out that the NQT often is in a part-time position which creates challenges of finding time for the NQT mentoring.At the same time this quote also indicates another issue from a school leaders' perspective which regards how much resources to invest in part-time workers.
The issue of resources is one reason why we address the NQT mentoring as a precarious practice.By precarious practice, we mean that the organizing of mentoring was a fragile practice, not prioritized and not experienced as included in collaboration between several actors.The issue of collaboration and resources is addressed by the mentoring coordinator Christine.She explained how her working hours were not enough to cover all the tasks she needed to take care of: Ideally, I would have liked to be on a mentoring team with others [at the school], I'm pretty alone in this practice.But there is no money for it.I have 15% in this [mentoring role], and I have a very large work capacity.I do not think anyone else would normally get all these things done here within 15%.(Christine, coordinating mentor, School 1) Christine expressed that it would be easier if there was a mentoring team.She admitted that it is mainly because of her personal working capacity that she could fill the role as a coordinating mentor, although she seems frustrated that she was alone in the practice.The lack of financial means to establish mentoring of NQTs was a challenge if the intention was to establish a coherent system of professional development.Such concerns were also mentioned by the coordinating mentor in school 4, which also provided released time for a coordinating mentor.
The group mentoring of NQTs at School 2 was established as a place where the NQTs could meet and talk to the school leadership about pressing issues.This was recently implemented, and according to the assistant principal, the experiences of the social aspect were important: We have chosen to do it after working hours to be able to socialize as well.We start the group meeting with food and talk before moving on to a round of questions about being a teacher.We really appreciate having fun together.(Bredine, Assistant Principal, School 2) At School 2, the assistant principal presents the set-up for mentoring as a social arena.The group mentoring had the character of after-work activity.For NQTs who are already experiencing overwhelming workloads, an after-work meeting may represent a challenge.This aspect was however not brought to the attention of the informants at this school.

Legitimizing the schools' organization of mentoring
Taking care of all the newly employed teachers was considered critical in all the schools.In some of the schools the teacher teams were given both the practical and moral responsibilities to ensure that the newly employed teachers were properly included in the school organization.At School 3, the principal legitimized how the school provided what they called "mentors" by explaining that the responsibility is an included part of being a good colleague: We used to pay mentors3 and expected them to contact the newcomer they were responsible for … But we changed that practice, so we don't pay mentors anymore.We don't want our collaborative team practice to be like that … To be a mentor is a part of being a good colleague.(Olav, Principal, School 3) Olav, the principal, expected the professional community to take collective responsibility for newcomers, which was the legitimization for not paying mentors.To include new colleagues was not considered an assignment for a mentor, so this role was not particularly recognized.All four schools applied several arrangements legitimized as necessary for welcoming and including NQTs and newly employed teachers into the professional community: We have mentoring because it is for both quality assurance and help in [personal and professional] development.I think they are not completely qualified as teachers when they come, and we must provide a program to … take them further on the road to becoming ready to be teachers.The mentor, provide that.(Victor,School Principal,School 4) The mentoring at School 4 was legitimized not only for support but also for quality assurance, hence serving diverging purposes.In school 4 it seemed that the principal entrusted the development and assessment of the new teachers to the coordinating mentor.In school 1, they were cautious about separating these two purposes as illustrated in this quote from the coordinating mentor: I have no leadership responsibility.And that is very important, the principal and I agree on that, because many of these newly employed, and especially the NQTs, they are in temporary positions.And it is not my mandate to discuss any of these with anyone.(Christine, coordinating mentor, School 1) At the same time, the quote from a middle leader at the same school illustrates that the follow-up and mentoring of NQTs is a practice with different and diverging aims in this school as well.
Although, for new employees, it is … you have a six-month probationary period … the mentoring situation should be assessment free, without reporting.But when I [as a leader] enter [the classroom], we can't keep a masquerade; I must know what is going on and whether that is good enough.(Maria, middle manager, School 1) This school leader acknowledges that mentoring should be assessment free, although clearly states she is concerned about NQT assessment, as it is her formal responsibility to provide quality teaching.School leaders are eager to follow up the NQTs themselves, which may be illustrated by how the principal in school 2 legitimized the set-up for NQT mentoring: And then I think it's good for the NQTs to be able to meet the school leaders sometimes.At the same time, we [the leadership] can show … what we want to focus on at our school.It is not given that a colleague with experience agrees with everything we say, so it is good that they can hear it unfiltered directly from us.(Kalmar, Principal, School 2) The arena for group mentoring served as a way for management to meet with the NQTs and have the opportunity to communicate their perspectives from a management level.In contrast, the principal at School 1 legitimized their arrangement for mentoring oppositely: Arranging for mentoring [separated from leadership] can also concern that the NQTs find it difficult to raise issues with the leader for various reasons.At the very least, you should have a management-free room dedicated to mentoring and development.(Oliver, Principal, School 1) Such variance in legitimizing the way of organizing mentoring of NQTs may illustrate some of the challenges in providing mentoring in accordance with government guidelines.
To sum up, our findings show that school leaders and mentors are unaware of policy expectations in policy documents regarding mentoring of NQTs and the organization of mentoring may appear as window dressing.NQT mentoring is perceived as a multitude of practices, and the organization of policy initiated NQT mentoring is experienced as a precarious practice legitimized with diverging aims and purposes indicating tensions in enacting policy expectations regarding mentoring of NQTs.

Discussion
The analysis revealed different perceptions of the policy demand for mentoring NQTs between school leaders and mentors.These differences were illustrated by the school leaders who were not aware of the national policy intention.The mentors had the slightly imprecise impression that NQTs were legally entitled to receive mentoring, 4 and expressed frustration about the lack of resources to expand mentoring practices.The differences in the perceptions of the policy directions can be understood as a challenge when it comes to managing expectations.National intentions and expectations regarding mentoring NQTs may be described as more or less well-defined.Also, the differences in the perceptions from the school leaders and mentors confirmed that a national policy expectation on safeguarding NQTs does not explain "what has to be achieved by whom" (Hopmann, 2008, p. 424).To be realized in practice policy ideas must be made thinkable, calculable and practicable according to Theisens et al. (2016), and steering presupposes a certain alignment between the three.In our findings, it seems that the policy idea of mentoring NQTs has not been realized in accordance with policy intentions, as mentoring seems to be a precarious practice in upper secondary schools characterized by little collaboration, diverging purposes and lack of commitment (with some exceptions, particularly from the mentors).A question thus becomes why policy intentions regarding NQT mentoring is not practiced as intended by educational authorities?4 Mentoring is highly recommended but not compulsory to perform in Norwegian schools.
On the one hand, school leaders may not understand mentoring as a solution or not even recognize mentoring as a preferred measure for taking care of NQTs.On the other hand, the findings illuminated expressions about a lack of resources needed in the local schools to offer the type of mentoring the governing level implied necessary.The school leaders were concerned about the general induction of NQTs or newly employed.However, most were not concerned about mentoring specifically.The informants who did not recognize mentoring as essential had several other ideas or beliefs about what is essential when it comes to welcoming, including, and socializing newcomers in local school practices.The expectations from the national authorities concerning taking care of NQTs mainly centered on mentoring, not other types of supporting arrangements, such as team support or professional group support.This can be regarded as a tension point in the schools' local practices compared to what is expected from educational authorities.A relevant perspective to draw in here concerns the plethora of policy initiatives and requirements that school leaders are expected to pay attention to and to enact upon (Verger et al., 2019).School leaders are found to be creative translators of unclear policy expectations (Hall et al., 2023), however they also have to manage tensions, for instance regarding priorities.For example, one of the principals in our material argued for cutting out half of the mentoring of NQTs that was expected from local authorities.Such tensions between policy and practice challenges the alignment between thinkable, calculable and practicable which Theisens et al. (2016) describe.
We describe NQT mentoring as a precarious practice in schools.By the informants the mentoring of NQTs was described as a practice of its own, not included in the general collaborative practices of the school.Consequently, mentoring arrangements were not part of a "coherent system" of continued professional development from initial teacher education throughout professional life, as described in the intentions of the national education authorities.Research has shown that school leaders' engagement in achieving central policy ambitions was critical (Hobson et al., 2009;Ingersoll & Strong, 2011;Kutsyuruba, 2020).Our analysis implied a lack of sufficient engagement and involvement from school leaders at several levels regarding NQT mentoring.However, even in the schools where the principal had a genuine engagement in mentoring, the mentoring practice appeared precarious, so we ponder whether the Norwegian policy approach concerning mentoring NQTs provides sufficient conditions for establishing good mentoring practices.
A soft governance approach like the Norwegian one, commits school leaders to implement measures like NQT mentoring based on the idea that the school leaders and mentors are professionally motivated to implement measures to improve practice (Maroy & Pons, 2019).Often, such measures are not followed up with sufficient resources and have few accountability mechanisms (Hopmann, 2008); then, there is the risk that what is practiced as mentoring may become small precarious practices depending on the engagement of few actors, or it may appear as window dressing or a "fabrication" of mentoring (Hobson & McIntyre, 2013).One could argue that such limited practices steal time from schools and other practices that the practitioners themselves want to prioritize.Simultaneously, it is of essence to nuance the picture, as the expression "window dressing" may be interpreted as a quite cynical strategy of school leaders to obscure to what extend they comply with policy expectations.Measures of NQT mentoring seem to be well intended in our material, and not just a matter of showing off.Another aspect concerns the complexity of intentions regarding development of practice as a response to policy expectations.One of the school leaders in the material admits that initiation of group mentoring at their school was an "eye-opener" regarding the perspective of the NQTs and gave insight into their needs.In such a way the policy expectation may spark off some reflections which may in turn lead to development of new practices and can thus not be reduced to a practice carried out to "tik-off" some box.
Our findings, however also indicate that the "prescription" offered by the authorities may not serve its purpose, particularly as it is not followed up with economic resources for realization.We suggest that such governing by expectations (Hopmann, 2008) results in substantial differences between how school leaders and mentors legitimize their arrangements at the local school level.The school leaders in our cases definitely agreed that NQT mentoring has the potential to bring about professional development both for NQTs and the whole collegium and they underscored that NQTs and mentors are highly appreciated resources in the professional community.However, NQT attrition, which is an argument for providing mentoring, may also represent an issue when school leaders consider which efforts to invest in regarding human resources (Ingersoll & Strong, 2011).School leaders must consider resources to invest in professional development, and may not agree that mentoring in accordance with government guidelines should be the main approach to building professional development and socializing NQTs?
According to government guidelines, mentoring should be separated from management and assessment (The Norwegian Directorate for Education and Training (TDET), 2018).Several of the school leaders revealed that they think mentoring was for support, control, and relationship building, which may indicate tensions in the approach to the mentoring practice.NQTs frequently only have a temporary position in schools, and they are always employed on probation for the first six months.Assessment of employees is an included part of a school leader's mandate; however, research has shown that mentoring of NQTs may be detrimental if assessment guides the practice (Hobson & Malderez, 2013).The school leaders' legitimization of the mentoring organization as a practice with multiple purposes may thus represent a dilemma between support and control of NQTs.
The power and eventual conflicts in mentoring practices can be regarded as parts of processes of responsiveness.That is, formal leaders have the power to argue for the importance of mentoring.According to Sørhaug (1996), school leaders have the formal power and greatest influence to advance their interests.Simultaneously, mentors have the power to use their knowledge about mentoring.As such, mentors can use their resources in the organizational negotiation on how to arrange for mentoring.Such a perspective finds support in research on mentors as middle leaders with potential to negotiate and design the content of NQT mentoring (Willis et al., 2019).An important question to ask is which mandate mentors are ascribed and whether the property of mentoring is agreed upon by professional actors in and between school and policy levels.

Concluding remarks
The purpose of this study is to contribute to knowledge about school leaders' and mentors' perceptions, experiences, and legitimizations of national policy expectations about the organizing of mentoring of NQTs.Although the four schools organized the mentoring of NQTs differently, the different approaches all had in common that mentoring was not a prioritized activity.A minimum of resources was allocated for organizing mentoring, and the use of resources differed largely between the schools.Also, mentoring was not well integrated as a part of school organizations.There was a need to clarify the core purpose of mentoring NQTs in schools and how it could be integrated with other induction practices.Further research is needed to examine in more detail what characterizes the policy-initiated organization of NQT mentoring in upper secondary schools and policy expectations to schools.Hence, the role of formal school leaders and perspectives of accountability are important issues to investigate.Moreover, research should examine how teacher teams and professional communities collectively respond to the intentions of mentoring in upper secondary schools.The study has some major limitations.The data were collected from only four schools, so the findings will only have an analytical value for generalization.Future studies should utilize even more informants, including mentees.Observation data would also be a relevant design in future studies in the field to gain the overall reliability and validity of the findings.

Table 1 .
Overview of case schools and informants.
Olga: Coordinating mentor All names are pseudonyms, and for anonymity purposes, all principals are presented as male and all middle managers and mentors as female.

Table 2 .
Steps of analysis.