Finding a way through the fog: school staff experiences of the Covid-19 pandemic

ABSTRACT This exploratory study reports on interviews carried out with 19 staff in UK schools during the early phase of the global Covid-19 pandemic. The focus of this qualitative study was to consider the impact of the pandemic on participants’ mental health and wellbeing, and to identify buffering mechanisms which may have mitigated against adverse effects. Participants were employed in a range of role types (including leaders, teachers and teaching assistants) in different educational settings (primary, secondary and alternative provision), and in different regions of the United Kingdom. A process of thematic analysis identified five key themes from the data set: change and adaptation; loss; impact on wellbeing; risk and protective factors; and opportunities to reflect. Data indicate that staff resilience during this time can be understood as emerging from a nuanced and complex interaction of internal and external factors, and thus conceptualised within a socio-ecological framework.


Introduction
On 20 March 2020 schools, colleges and early years settings in the United Kingdom (UK) were instructed by central government to close for all but the most vulnerable pupils and the children of keyworkers, as part of measures implemented in response to the global Covid-19 pandemic. In addition, all examinations and inspections were suspended. This sudden and extensive disruption resulted in multiple transitions for children and young people, their families and education staff, who adapted to homeworking, remote teaching and ensuring the safety of pupils both at home and in schools. Data from the national charity, Education Support (2020), indicate that aspects of the mental health and wellbeing of education professionals have been adversely affected during this period, although little is understood about the underlying mechanisms.
McConney, 2012). For example, Greenfield's (2015) model of teacher resilience identifies the interaction of internal factors (having a sense of hope, purpose and high self-efficacy) and external factors (supportive and positive school culture, positive relationships with colleagues, supportive social networks and supportive policy contexts). This study investigates how the wellbeing and resilience of education professionals has been influenced by the interaction between multiple individual and contextual variables during the pandemic, and considers the implication of findings for research, policy and practice.

Methods
This qualitative study aimed to explore school staff experiences of the Covid-19 pandemic and identify staff perceptions of the factors influencing their wellbeing, responding to a call for rapid qualitative research to better understand the impact of Covid-19 on mental health (Holmes et al., 2020).

Recruitment and sampling
A call for participants was issued via Twitter on 26 May 2020. Inclusion criteria were: (1) adults employed in UK education settings (including Early Years [EY], Primary, Secondary, Further Education [FE] or Alternative Provision [AP]); and (2) access to an internet connection with Skype software.
Nineteen individuals volunteered participation within three weeks of the call. As the sample included a range of role types from different settings (Table 1), and in order to ensure interviews reflected a similar time period (given the fast moving context of change related to the pandemic), those who made contact after this time were informed that the study had recruited sufficiently. Details of existing confidential support services were provided to all individuals who expressed interest in the study irrespective of their decision regarding participation.
The sample included leaders, teachers and teaching assistants from 12 primary schools, 1 middle school, 5 secondary schools (including 1 specialist setting for children with complex needs) and 1 alternative provision (AP) service. The mean number of years of experience in schools was 13.1.

Ethics
The study was approved by the research ethics committee at the University of Brighton. Ethical guidelines for carrying out Covid-19 related research (Townsend, Nielsen, Allister, & Cassidy, 2020) were incorporated into the study design, alongside online research guidance from the ESRC framework for Research Ethics, the British Psychological Society and the UK Data Service. Recruitment materials emphasised that participation should only be considered if it were unlikely to cause distress or anxiety.
Participants were informed of the study aims, duration and procedure, and were assured of their anonymity and the secure storage of their data. All participants provided informed consent and were made aware of their unconditional right to withdraw. A distress protocol was developed based on an existing model used in previous wellbeing and resilience research (Cocking, Sherriff, Aranda, & Zeeman, 2018). The protocol for this study included short pre-interview discussions in which participants suggested strategies that would help ensure that the experience was beneficial, normalising reflecting, checking or pausing during the interview. A non-directive, person-centred approach (Rogers, 1942) allowed participants to shape the direction of the interviews and demonstrated empathy. Digital recordings and transcripts were password protected on university secure cloud-based storage, and device recordings were deleted after each interview.

Data collection
Nineteen semi-structured interviews were conducted between 3-29 June 2020 using Skype. Existing research highlights the advantages of using Skype to conduct qualitative interviews and suggests digital video communication platforms as a viable alternative to face-to-face methods (Lo Iacono, Symonds, & Brown, 2016).
All interviews were conducted by the first author and lasted between 45-60 minutes. Audio data were digitally recorded on a separate device with the permission of participants to enable verbatim transcription.
The interviews aimed to address the following research questions: The research questions were sent in advance to participants to enable preparation for a reflective and open discussion, aided by additional prompts to enable further depth and ensure relevance of data. A sensitive and responsive approach to the subtly changing dynamics of the interviews (Kvale & Brinkmann, 2009) was informed by Brown and Danaher's (2017) principles of connectivity, humanness and empathy.
All participants checked and approved their full transcript for accuracy and anonymity. To ensure confidentiality, the names of the participants, the schools they worked in and any other people they identified were anonymised during transcription and do not appear in this 1 paper.

Analytical strategy
An seven-stage process was implemented ( Figure 1) informed by Clarke's (2006, 2019) model of thematic analysis. The deliberately iterative, reflexive and collaborative process combined inductive and deductive approaches to analysis. Familiarity with the data was achieved by open reading and re-reading of interview transcripts with the research questions actively in mind and codes were generated inductively, with close attention to each line of text. Concurrently, short reflections, interpretations and questions were recorded alongside specific sections of text. These were later developed into longer memos. 'Focused coding' involved analytical and critical reading of the transcripts to combine or create new codes, which were organised into themes and sub-themes to elucidate 'patterns of shared meaning underpinned or united by a core concept' (Braun & Clarke, 2019, p. 593). Collaborative reflective analysis allowed codes to be compared and the researchers remained open to 'surprises' in the data.
Detailed thematic mapping of findings was produced (see supplementary information), the consistency and contradictions of which were explored in relation to original transcripts. Interpretations of the data were then synthesised to answer the research questions in relation to five categories: adaptation and change; loss; impact on wellbeing; protective and risk factors; and opportunities to reflect and adapt.

Adaptation and change
Participants experienced extensive and sudden change brought about by the pandemic, both in their personal and professional lives. The nature and intensity of work had changed, resulting in a 'different kind of busy' (ST3) both for staff 'shielding' at home and those in school. Tasks included learning to use online teaching software, adapting the curriculum for home learning, communicating with parents, pupils and colleagues, supporting the local community (for example, delivering groceries or protective equipment), developing new procedures based on safety guidance and assuming new roles. Most participants felt that they were as busy or busier than usual, with one participant's school opening 6 days a week, 12 hours a day. This increased workload constituted a contextual risk factor, and a high degree of commitment and flexibility was required to keep schools functioning: [O]ur premises team have been amazing, because all three of them are in the shielding group . . . there's a shift that starts at four o'clock in the morning and finishes at seven to make sure that we're safe for opening. (SL1) The 'merge between work and home' (ST3) resulted in increased vigilance: 'I'm constantly thinking about work . . . I might switch off emails, but I can't switch off my mind' (PL5), and an amalgamation of professional and personal stress: 'this constant feeling of guilt . . . where can you get away from anything?' (ST2). The challenge of juggling personal and professional priorities was particularly acute for staff who were also parents:  I just felt like I was failing my children, because they were just pretty much sitting in front of screens all day while I went 'shh I'm in another meeting, shh I'm talking to another parent' . . . I thrive on being a really successful professional, but I wasn't really feeling like that was happening much either from the shed. (PL2) Participants perceived only fleeting public recognition of the vital role of schools ('it was almost a pivot . . . back to "everyone hates teachers"', (ST2)) and were aggrieved by public misconceptions that schools had closed: [P]oliticians kept talking about schools being closed . . . we are not closed, we are working very differently and we are working non-stop. You know . . . we almost are offering a seven day a week service at the moment, and we have through all the holidays. (SL1) A vicissitudinous process was described, in which participants said they had navigated waves of uncertainty, confusion and feeling overwhelmed: 'it is like a hazy mist . . . and you don't know what the other end looks like' (ST2). Participants found ways to adapt, only to experience rapid further change in their personal or professional lives, including changes to guidance for schools or changing internal perceptions (such as increased acceptance): [I]t seems to have been quite settled until the re-opening was announced and then I think everyone went into like a bit of panic stations . . . after we'd had a few meetings people were back into their routines again . . . it's been a bit of a rollercoaster. (PT2) A sense of frustration was conveyed regarding the timing and communication of central government guidance for schools, alongside what participants considered was limited consultation: '[I]t felt like as a profession we were not listened to. Parents weren't listened to, children were talked about, teachers were seen as a pain' (PT3). Both leaders and teachers suggested that the DfE's guidance lacked clarity, resulting in increased anxiety and workload as repeated revisions arrived after school procedures had already been established: Communication from Government has, I think, antagonised school leaders . . . There seemed to be a clear pattern of making a public announcement at the briefings which would impact directly on education without having told anyone in education beforehand. (PL3) We were getting guidance throughout the holidays, all hours of the night . . . it was all contradictory, it was hardly readable, it was either patronisingly obvious or not really doable.
And it certainly didn't help us to alleviate any anxieties of safety for staff. (PL4) Teachers and leaders identified the challenge of managing uncertainty and loss of control, which conflicted with their professional responsibilities: '[B]eing a little bit of a control freak, situations like this in general just take any sort of element of control away from you, and it's sometimes quite a scary place to be' (APT1). Establishing new routines (both in terms of managing the challenges of home-working and school safety) was considered an important coping strategy: 'You feel that as soon as you enter our building you are in a safe zone, and you feel it, you feel almost in control, in control of the virus in a way' (PTA1). Not all change associated with the pandemic was perceived negatively. Participants regarded reframing the situation to identify positive opportunities as part of adaptation. For staff working at least partially from home, this included spending more time with family, flexible working and being able to prioritise self-care. Staff working in school enjoyed an intense sense of collegiality and those who taught 'key worker bubbles' enjoyed the pastoral focus of small groups and teaching beyond their usual age range or subject: I've got incredibly fond of them . . . . All their little characteristics I'm really picking up on and actually, on one hand I'm really enjoying this low number and being able to see them every day because I'm not primary-trained . . . . I have to say I'm really loving it. (MT1)

Loss
A quarter of participants talked about increased bereavement in their school community, both directly and indirectly related to Covid-19: A child's mum died because she couldn't get her medication for her underlying conditionnot because it wasn't available but because she didn't feel able to go out and get it . . . the next day a child's mum took her own life . . . she was in a pretty difficult relationship with somebody and I think just the whole idea of lockdown proved too much. Since then I have sent out over twenty condolence cards. (SL1) Several interviewees said that they had accessed online training to help them support children and families through loss, trauma or grief. Loss was also experienced in a broader sense. Staff who had been anticipating imminent exams or inspections felt that they had been denied external validation and the sense of professional achievement. Others expressed profound loss in relation to endings or transitions: [Y]ou're not only preparing them to leave, you're preparing yourself to let them go . . . . I hadn't prepared myself to say goodbye . . . when you say it out loud it almost sounds quite strange but [it was] almost like a grieving process. (ST2) The loss of face-to-face interaction was challenging for all participants. Communication between staff was maintained through online meetings, emails and the use of WhatsApp. One participant (employed part-time) felt more included in the staff team than ever, in spite of the loss of face-to-face meetings, as they were able to attend online meetings more regularly. Some participants felt that the increased flexibility as a consequence of using digital communication had facilitated increased collegial support: [W]e've been able to be a lot more honest with each other, a lot closer, and a lot more supportive . . . people have got a bit more time to say: 'well actually, you know, I'm really struggling today'. (ST4) Teaching assistants and also one teacher in alternative provision reported limited (if any) opportunities for direct contact, suggesting that role type may have been linked to increased loss of contact with pupils in some cases. Furthermore, for staff who were in contact with pupils, participants noted the relational limitation of telephone and online interactions: '[S] ome kids will just give you one word answers' (PT1). The loss of face-to-face contact was seen by multiple staff to conflict with professional identity and altruistic motivation: [I]t's become a very, very lonely job. And teaching isn't lonely, teaching is collaborative and fun and engaging and talking and based on strong positive relationships, and the fact that I know my students are struggling and I can't do anything to help them really, really is quite emotive to me because I'm a teacher and I'm there for them. (ST1) Personal sacrifices were made by participants to protect others, including not seeing elderly relatives, being unable to hold new-born grandchildren, having to move to a new house or neglecting personal wellbeing. Participants buffered this loss by focusing on the value of their work or a sense of professional duty: [T]here isn't actually time for reflection about how I feel, other than occasionally I get this overwhelming feeling of 'will I see my parents again?' . . . . But you get up, you go to school and it's fine again, because you know, that's what we're here for and that's what we're meant to be doing. (PTA1) Most participants suggested that a shared sense of loss had led to increased cohesion ('we feel absolutely bonded to each other' [PTA1]). However, a minority warned that resentment, particularly in relation to personal sacrifice, could result in blame being directed at staff who had been working from home:

Impact on wellbeing
Participants had experienced a range of negative emotions including stress, anger, sadness, guilt, shock and fear, peaking for most participants when schools partially closed and when plans to fully re-open were announced (and later postponed). Furthermore, participants associated uncertainty and stress with lack of sleep, loss of appetite, an inability to concentrate, strained personal relationships or poor physical health: I found it really difficult to focus on anything. I couldn't read a book, I couldn't really get anything done, because actually that level of anxiety was intense . . . basically learning how to do a new job . . . plus there was a virus that was out there and killing people, all of that just was the pressure, it was the pressure, that's how it felt. (PT3) Place of work and role type emerged from the data as variables, modifying the impact of the pandemic on staff wellbeing. For example, for those working mostly from home, selfcare strategies and spending more time with family (deemed impossible prior to the pandemic) became part of a new routine, resulting in improved physical and emotional health. However, participants worried that wellbeing strategies employed during 'lockdown' would be difficult to maintain. Home-working was also associated with feeling less useful and more disconnected from professional identity and school community, to the detriment of individual wellbeing. Those mostly working in school reported increased anxiety alongside emotional and physical exhaustion ('we're kind of almost running on empty now' [PTA1]), but felt strongly connected to their professional purpose. For leaders, high work-related anxiety was more or less a continual feature of their experience during the pandemic, resulting from increased workload, high-stakes accountability and feeling inadequately informed and therefore ineffectual. Middle and senior leaders questioned how long they could continue to manage 'putting [their] neck on the line' (PL3), whilst feeling inadequately informed and supported themselves: '[T]he mental work is phenomenal at the moment . . . I couldn't do the job in November/December time if things don't change' (PL5).
Most participants considered social stigma associated with mental health to have been reduced as a result of shared adversity: '[P]eople are not ashamed now, or worried about saying that, you know, they're struggling' (ST4). There was a lack of consensus, however, about the longevity of the prioritisation of wellbeing at school level. Teachers tended to be cautiously optimistic, whereas leaders worried that competing accountability priorities could undermine much needed progress: '[W]hat does catch-up look like or what does remote learning look like . . . unfortunately that will be a pressure that's going to break some of the staff a little bit later on in the year' (PL5).
Related themes of 'masking emotions' and 'running on empty' emerged from the data, as participants described trying to buffer stress from family or colleagues whilst struggling themselves: 'I put a lot of energy and thought into looking after those around me that I'm responsible for and working with, to my own detriment' (ST3). Some participants suggested that silent suffering might later result in staff burnout: [I]t was just almost like painting a mask on. They felt they had to appear in a certain way but were quite vulnerable underneath the surface . . . staff have definitely been affected, but the worry is quite by how much, because I don't know yet. (PL5)

Protective and risk factors
Staff reported a range of individual risk factors influencing their wellbeing including: a need for control; low self-perception of competence; and having existing mental health issues (Figure 2). Limited face-to-face interaction, relentless uncertainty and change, intense workload, competing priorities and the communication of guidance for schools were experienced as the main contextual risk factors ( Figure 3). Multiple contextual protective factors were also identified, including collegial support, participative planning, access to exercise and fresh air and establishing new routines ( Figure 5). Participants recognised the benefit of seeking support, but some noted the difficulty of finding someone to talk to: 'I don't particularly want to bombard or bring that into my family situation . . . I want to try and keep those two separate' (PL1). Others said that their leadership role involved reassuring staff whilst they struggled themselves with limited support: 'So, I fill in a spreadsheet every week of all the staff that I'm responsible for, checking that I've met them. My name isn't on there as somebody to be checking in on me' (ST3). Three headteachers accessed their own external supervision during lockdown, which they said supported their wellbeing by providing a non-judgmental space in which to reflect on professional practice.  Interaction between individual and contextual risk and protective factors was moderated by contextual specificity. For example, leaders reported greater workrelated anxiety in relation to policy communication, and participants who were parents were more likely to cite competing priorities as a source of stress; thus for leaders who we also parents, adversity was compounded. Contextual nuance meant that a protective factor for one person could be a risk factor for another, or even for the same person on a different day. A frequently cited example from the data was social media, which provided networks of information and support, but was also experienced as overwhelming, sometimes exacerbating anxiety. Many examples of reciprocity were provided including the benefits of mentoring, which increased feelings of self-worth and competence: '[H]elping somebody else has helped me . . . it has given me strength and she reflected that it gave her strength as well' (PL3).
Participants working from home were more likely to report feeling disconnected from the school community and their sense of professional purpose, resulting in loneliness and low self-efficacy: '[A] lot of the time you do actually feel quite useless because you're at home' (PT4). Leaders had tried to find manageable and meaningful tasks that helped staff to reconnect:  [W]e've got one person sitting at home who can't be in school because she's over 70, so is in the vulnerable group . . . she's creative, so we've sent stuff to home for her to make friendship bracelets for all the new kids who are coming in. (SL1)

Opportunities to reflect and adapt
Underpinned by a sense that 'the world is a different place now, and we need to reset' (PL1), participants described population-wide disruption as an opportunity away from the frenetic pace of school life to reflect, clarify whole school aims and reinforce shared values: [U]sually just you're on that treadmill and you don't stop running until you get to some holidays . . . we're saying 'you know your pupils best and we're going to create this whole school plan but we need your input'. I think that's contributed to the team continuing to be valued and maintaining high levels of wellbeing. (PL2) Some participants emerged from a period of reflection with a renewed sense of professional purpose ('I think I was a little bit jaded . . . now I've sat back and thought about it I think anything you do for children does make an impact' [PTA3]), and increased collegiality ('we're in this together, we're fighting for the same things . . . rather than who's going to have the highest SAT results and who's outstanding' [PL5]). Others questioned remaining in the profession after taking time to reflect on poor work-life balance, intense emotional demands and perceived lack of appreciation: Increased opportunities were identified for participants to attend training, access research and expand professional networks as a result of increased online content and flexible working. This aided critical reflection, resulting in adapted practice and increased competence. Examples were provided in relation to digital and in-person pedagogical strategies, mental health, inclusion, racial equality and trauma. The cancellation of exams and suspension of school inspection were an opportunity to prioritise wellbeing for pupils and staff and emphasised the need for compassion and differentiated support: [E]verybody will need their hands held in a different way and at a different time . . . they need to know that they've got somewhere to go and that what they've got to say will be respected, listened to and they won't be judged by it. (ST3) Leaders warned that the return of demands such as testing and inspection could make it difficult to capitalise on the holistic approach adopted during partial school closure 'as soon as we get that notification that testing's all back up-and-running . . . teachers will find it really hard not to revert back to that treadmill' (PL2) and 'if your accountability is about scores on the doors in May, it's easy to lose focus of all that important stuff' (PL5). Nevertheless, most staff interviewed for this study remained hopeful that given space and time, schools could harness learning from a time of crisis, leading to meaningful and transformative change.

Discussion
The five key themes outlined in our findings are interrelated. For example, participants' experiences of sudden and excessive change can be considered to constitute a contextual risk factor, whilst opportunities for change brought about during this period of turbulence were also considered by some participants to foster a sense of hope, which can be considered to constitute an internal protective factor. In our discussion, we synthesise data from across all five themes to consider participants' experiences through the lens of a socio-ecological understanding of resilience. In this discussion, the individual and contextual factors that interacted to influence staff resilience during the pandemic are identified with reference to the existing literature and we propose extending Greenfield's (2015) model of teacher resilience to consider the resilience of all school staff, in the context of change.
Consistent with a multi-dimensional conceptualisation of wellbeing (Dodge et al., 2012), participants' wellbeing was influenced by their psychological, social and physical resources. It is reductive to suggest only a negative impact on staff wellbeing, and perhaps more accurate to note that participants found that depending on contextual circumstances, sustaining wellbeing was more difficult in some ways and easier in others during this time. We found that the extent to which participants felt able to adapt in order to sustain or improve their wellbeing was influenced by multiple contextual factors. These included but were not limited to: their capacity to work from home; access to social and professional support; and the changing educational policy context. Thus, in line with previous research (Beltman et al., 2011;Greenfield, 2015), resilience was perceived by participants as more complex and dynamic than a static internal capacity to cope with extreme adversity. Instead, resilience was considered to fluctuate over time and to arise from the interaction between individual factors (e.g. self-efficacy) and changing environmental factors (at school and policy level). Role type and working location appeared to moderate participant exposure to risk and protective factors in complex ways. For example, staff who had worked mostly from home reported improvements in their physical and emotional wellbeing due to increased opportunities for self-care, spending time with family, and professional reflection and development. However, home-working staff (especially parents) also reported increased stress due to a lack of boundaries between professional and personal domains, a reduction in professional self-efficacy and a sense of disconnection. Our data suggest that leaders experienced high workrelated anxiety due to their additional responsibilities to others, which emerging research indicates may also be the case beyond our sample (Allen, Jerrim, & Sims, 2020). The danger of leaders falling through the gaps in terms of staff support mechanisms was also highlighted by our participants.
Echoing existing literature, participants of all role types referred to an internal altruistic motivation or sense of professional purpose as an important protective buffer of adversity (Beltman, 2015;Beltman et al., 2011;Drew & Sosnowski, 2019;Gu & Day, 2011) and linked collegial and leadership support to their own wellbeing (Ainsworth & Oldfield, 2019;Beltman, 2015;Greenfield, 2015;Gu & Day, 2011). Staff of multiple role types reported that participative planning for safety and wellbeing in school had helped them to feel more in control and less anxious, suggesting that distributed leadership and expanded professional networks can help navigate complex change with reciprocal benefit (Littlecott, Moore, Gallagher, & Murphy, 2019;Manning, Brock, & Towers, 2020). Internal values and collegial relationships had increased significance in the context of limited interaction with pupils and disruption to staff-pupil relationships, which have previously been highlighted as strongly protective factors (e.g. Graham, Powell, Thomas, & Anderson, 2017;Milatz, Lüftenegger, & Schober, 2015;Spilt, Koomen, & Thijs, 2011).
Parallels can be drawn between individual and contextual risk factors identified by participants in this study, and those identified in prior research. These include rapid change (e.g. Brown, Ralph, & Brember, 2002), loss of control as a result of increased externally imposed regulations (Ball, 2003;Beltman et al., 2011;Williams-Brown & Jopling, 2020;Winter, 2017), societal cynicism and biased media representation of teachers (e.g. Hester, Bridges, & Rollins, 2020). Due to the pace of change and the scale of uncertainty staff had experienced during the pandemic, these risk factors appeared to have been amplified.
Congruent with Hobson & Maxwell's (2017) research (which extended Self Determination Theory [Deci & Ryan, 2000] in the context of schools), our data suggest that autonomy, relatedness with colleagues and pupils and perceived competence were associated with staff wellbeing, though staff perceived that disruption to their day-to-day role caused by the pandemic had reduced opportunities to meet these innate psychological needs and resulted in a sense of deep loss. Our findings extend existing research that demonstrates that normative transitions within pupils' school careers can significantly influence their wellbeing (Pietarinen, Soini, & Pyhalto, 2010), indicating that disruption to pupils' normative transitions (e.g. the cancellation of 'leaver assemblies' or missed opportunities to say 'goodbye') also impacted detrimentally on the wellbeing of education professionals.
As other empirical research has found (e.g. Drew & Sosnowski, 2019), participants drew on their sense of professional purpose to buffer against emotional exhaustion, stress and personal sacrifice. 'Mattering' has been suggested as a psychological need consisting of 'feeling valued and adding value', a lack of which enhances internal and relational conflict (Prilleltensky, 2020). Participants in our sample were able to connect to a sense of adding value (depending on how closely they were involved in direct communication with parents and pupils, teaching and participative decision making), but felt largely undervalued by central government and in mainstream media. Some participants acknowledged that they did not place sufficient value on their own wellbeing (or masking their own needs), and others warned that feeling undervalued could lead to blame or resentment between staff and mistrust of central government, potentially reducing longterm resilience. As other research has suggested (e.g. Richards, Wilson, Holland, & Haegele, 2020), our data suggest that developing an environment in which school staff both feel that their work adds value, and in which staff feel inherently valued, is likely to improve professional satisfaction and personal wellbeing.
Re-framing uncertainty is thought to help sustain teacher wellbeing during challenging times (e.g. Beabout, 2012;Drew & Sosnowski, 2019), and our data showed that staff had created new routines and were planning for the future, re-framing the crisis as a catalyst for change. Disruption to the education system has presented a unique opportunity to pause and critically reflect, which could lead to valuable insight. A careful balance between harnessing this potential for change whilst minimising unnecessary 'turbulence' for staff (Beabout, 2012) can be fostered through participatory policy development, attentiveness to contextual needs and protected time and space for professional reflection and development.
Recent policy drives (DfE, 2018b) to improve staff wellbeing in schools through reducing workload and developing positive workplace cultures are a step in the right direction. Staff in our sample suggested a range of supportive strategies, including access to professional supervision and creating a whole school culture that prioritises wellbeing. However, some participants anticipated a swift return to a hectic pace and a revival of intense accountability pressures, which they worried could negatively influence wellbeing in the absence of structural support.
The results of this study echo previous findings by Manning et al. (2020) that wellbeing support should be matched to diverse staff needs. Staff wellbeing is not connected simplistically to workload or change, but is influenced by staff perceptions of the value they add through their work, the extent to which they feel valued and supported, and their involvement in the process of change. To ensure support for staff is helpful, this contextual nuance must underpin efforts to improve staff wellbeing at multiple system levels.

A model of school staff resilience in the context of change
In line with Greenfield's (2015) model of teacher resilience, our data suggest that participants' resilience can be conceptualised within a socio-ecological framework. Congruent with Greenfield's model, participants' beliefs influenced their personal wellbeing and capacity to adapt. This included maintaining a sense of hope, vocational purpose and self-efficacy. Greenfield identifies taking positive action as playing a critical role in facilitating resilience. Taking action to help others (including staff, pupils and parents), drawing on strategies to reduce stress (for example, exercise) and utilising increased opportunities for reflection, reframing and professional development enabled some participants to navigate challenges posed by the pandemic. Support from colleagues, interactions with pupils and their families, and supportive personal relationships were significantly protective contextual factors for participants, often building on existing school culture and communication between stakeholders. These supportive relationships buffered the effects of the day-to-day challenges that participants experienced. Challenges included ongoing and relentless uncertainty in participants' personal and professional lives, intense workload and competing priorities. For some, the effects of these challenges resulted in negative emotions, including high levels of work-related anxiety. In line with Greenfield (2015), we found that the extent to which participants were negatively affected by the day-to-day challenges was influenced by broader political, cultural and economic factors. For example, lack of clarity and guidance from the government, alongside insufficient resourcing and funding, was perceived by participants to have created an unsupportive policy context which increased work-related anxiety and resulted in a deterioration of trust between school staff and central government. Some participants also described dissonance between a strong sense of vocational values and negative portrayals of school staff in the media, which they perceived as indicative of a lack of public support.
The experiences of participants (of all role types) can be understood within a socioecological conceptualisation of resilience. We therefore suggest that whilst Greenfield's initial model relates to teacher resilience, the factors included in Greenfield's model are relevant for all school staff and could constitute the basis of a school staff resilience model. Furthermore, staff of different role types appeared to experience these risk and protective factors in different ways. For example, senior leaders experienced high levels of work-related anxiety as a result of relentless uncertainty, loss of professional autonomy and loss of control. They felt disconnected from decision-making processes and often masked their emotions to enable them to more effectively support colleagues, parents and students, in the context of receiving little or no support themselves. We therefore suggest that there may be factors relating specifically to job role which influence participants' resilience, and that job role could therefore be considered an important aspect of the model that helps to determine helpful strategies. For example, leaders experienced high levels of work-related anxiety and lacked access to networks of professional support which negatively impacted on their resilience. We therefore suggest that networks of support for this role type could include access to external professional supervision that is available continuously, not just in times of crisis. Our proposed model is shown in Figure 6.
Change is a continual feature of schools due to their inherent complexity (Keshavarz et al., 2010). However, the Covid-19 pandemic can be understood as a sudden external shock to the system, an event that had significant impact at multiple system levels, exposing and exacerbating existing individual and contextual risk factors associated with poor staff wellbeing. In order to acknowledge the continuous nature of change in the school system, and the importance of understanding staff resilience as a dynamic and malleable concept, we propose an addition to the model that indicates change over time and adaptation in response to changing socio-cultural, historical, political and economic contexts. We have adapted Greenfield's (2015) model to indicate these additional factors ( Figure 6).

Limitations and further research
Data were generated from a relatively small sample and relied on self-reports in the form of in-depth participant interviews. Whilst this exploratory study explored the experiences of staff in a range of role types and across different school settings, the data set represents the perspectives of a small number of self-selecting staff, at a moment in time and within a rapidly changing context. Experiences of the pandemic have been nuanced and personal, and the views of participants may not be generalisable to the heterogenous and complex contexts of other school staff. Despite these limitations, the data provide rich and valuable insights about the nuanced and complex nature of staff resilience in the context of the Covid-19 pandemic, highlighting the interaction of multiple internal and external factors. Due to the dynamic and contextual nature of resilience and the rapidly changing landscape of education, further research is required to establish how initial teacher-training providers, schools and policy makers can prioritise the wellbeing of school staff to foster resilience across the school system in the face of ongoing and future adversity. This should include larger samples of cross-cultural comparison, mixed method and quantitative data collection, including longitudinal studies that examine the long-term impact on staff wellbeing and retention across multiple role types. Furthermore, research suggests that the Covid-19 pandemic has had a disproportionately negative impact for those who, prior to the crisis, were already facing socio-economic disadvantage (e.g. Kirby, 2020;Patel et al., 2020). Thus, whilst beyond the scope of this study, further research is required to explore the moderating influence of existing health and inequalities on school staff experiences.
Key findings of this multiple role-type study were consistent with existing literature on teacher stress and burnout, teacher mental health, teacher wellbeing and teacher resilience. Future research could consider how these concepts overlap and could be applied to or adapted for staff of multiple role types and in the context of adversity associated with system 'events' such as the Covid-19 pandemic.

Conclusion
Significant turbulence caused by Covid-19 has both increased adversity for school staff and offered a unique opportunity to pause and reflect. The immense pace of change necessary to address the global pandemic has resulted in seismic disruption for staff, both in their personal and professional lives. Our data set indicated a range of individual and contextual risk and protective factors which have dynamically interacted in different ways for our participants during this time, and which can be understood within a socio-ecological framework of staff resilience. Contrary to a conceptualisation of resilience as an innate individual capacity (which shifts the responsibility from the structural to the individual member of staff), our data were congruent with existing research (e.g. Beltman, 2015;Beltman et al., 2011;Greenfield, 2015;Gu & Day, 2011;Mansfield et al., 2012) that has highlighted the importance of both internal and external and contextual protective factors as the foundation for staff resilience, thus suggesting the need for structural change as opposed to individual 'responsibilisation' (Hart et al., 2016).
One participant described their experience of this uncertain time as being lost in a 'hazy mist'. To continue their apt metaphor, weather conditions remain unstable and ongoing change in schools is inevitable. As the fog begins to clear, there is an opportunity for lessons to be learned from the crisis to better support staff wellbeing in the future. There are several implications arising from this research. Firstly, school leaders should prioritise developing whole school cultures that prioritise wellbeing for all stakeholders. The global pandemic has resulted in multiple academic, social and psychological transitions for educators, students and families. These transitions are continual, and individuals may therefore also require ongoing additional support to enable them to flourish. Secondly, school leaders may benefit from access to external professional supervision to support individuals to buffer the effects of the challenges they are experiencing in their professional lives. Finally, policy makers should engage in thorough consultation with school staff to ensure that future educational policy effectively prioritises staff wellbeing and preserves existing strengths of the school system, and that guidance for schools is clear and practical to implement.
Note 1 A coding system is used to indicate school setting (P = primary, S = secondary, AP = alternative provision, M = middle school) and role type (L = senior leader, T = teacher, TA = teaching assistant). For example, PL1 = the first primary school senior leader interviewed.