‘I had forgotten what it was like to feel like you’re doing a good job’: a longitudinal thematic analysis of teachers’ experiences during the COVID-19 pandemic

ABSTRACT Examining the changes in teachers’ experiences of their job resources, job demands and personal resources can reveal important insights into the profession. Accordingly, semi-structured interviews were conducted with an initial sample of 21 primary and secondary school teachers in England at three time points during COVID-19 (February 2021, July 2021 and July 2022). Four themes were identified in a longitudinal thematic analysis, namely: teachers’ perceptions of disconnection with and disrespect from the government; relationships are central to teacher wellbeing and work engagement; teachers’ preference for the physical school over the online/hybrid school; and reflections on teachers’ increased self-efficacy in using technology and increased resilience to change. Implications for practice and policy are discussed.

Teachers around the world experienced changes in the ways that schools operated during the COVID-19 pandemic, for example, due to school closures, which affected approximately 63 million primary and secondary school teachers (International Task Force on Teachers for Education 2030Education , 2020)).In England, on 20 March 2020, schools closed to pupils except for children of key workers and vulnerable pupils (see Figure 1 for timeline).Teachers continued to experience disruptions in the second year of the pandemic (Kim & Asbury, 2020;Kim, Leary et al., 2021;Kim et al., 2023).A significant date for teachers in England was 4 January 2021, when primary schools fully reopened in 100 councils in England.However, that same evening, the UK Prime Minister announced another strict national lockdown.This was the fifth U-turn in announcements on school partial re-openings and closures in three weeks (Timmins, 2021).At the national level, schools fully reopened on 8 March 2021 and have remained so.However, at the local level, teachers have continued to provide online and hybrid teaching and learning when teachers and/or pupils have been unable to attend in person due to COVID-19 infection.
In the midst of rapid changes in schools -and the world at large -due to COVID-19, teachers continued to look after the welfare of pupils and provide education.For example, teachers and schools checked on the wellbeing, welfare and food situation of pupils and their families (Kim, Dundas et al., 2021;Moss et al., 2020).To continue providing education, they needed to adapt rapidly and upskill themselves to deliver alternative teaching approaches, including socially distanced in-person teaching, online teaching and hybrid teaching.This had consequences for their self-efficacy levels, as illustrated by a US study which found that teachers' self-efficacy levels in 2020 (when they transitioned fully to remote learning) were lower than in studies conducted before the pandemic (Pressley & Ha, 2021).As a result, their already high workloads prior to the pandemic, which were already contributing to teachers' intentions to quit the profession (CooperGibson Research, 2018), affected their mental health and wellbeing (Kim et al., 2022a).Moreover, as a study from the US indicates, the increased complexity of their role and additional stressors at the individual-, classroom-and school leadership levels of their professional lives, also had consequences for their wellbeing (Robinson et al., 2023).For example, a study in Ireland found that COVID-19 related physical (43%) and mental health (67%) consequences were experienced by their sampled teacher participants, including 79% reporting work burnout and 66% reporting low job satisfaction (Minihan et al., 2022).

Job demands and resources
The Job Demands-Resources (JD-R) model proposes that job demands (elements of work which are costly to the individual) contribute to one's experience of burnout, while job resources (elements of work which are beneficial to the individual) contribute to one's engagement in the work.Moreover, job demands and job resources are proposed to buffer the effects of each other (Demerouti et al., 2001).Since the model's initial conceptualisation, personal resources were added, which, like job resources, both contribute to one's work engagement and buffer the effects of job demands on burnout (Bakker & Demerouti, 2014;Bakker et al., 2023).Integrating the JD-R model in the teaching context, the OECD's conceptual framework for teachers' occupational wellbeing (Viac & Fraser, 2020) provided examples of job demands (e.g.physical learning environments, workload) and job resources (e.g.work autonomy, social support).Moreover, previous studies have identified self-efficacy and resilience as commonly examined personal resources for teachers that are relevant to the JD-R model (see Granziera, 2022 for a review).
During the COVID-19 pandemic, teachers were able to identify specific job demands, job resources and personal resources.For example, a longitudinal qualitative study of teachers in England during the first eight months of the pandemic found a general decline in their mental health and wellbeing due to increasing job demands and reduced job resources (Kim et al., 2022a).Moreover, a quantitative study of Italian teachers at the same time found that personal resources (i.e.self-efficacy and resilience) mediated the negative effects of job demands on emotional exhaustion, and that job resources contributed to personal resources, which negated the effects of emotional exhaustion (Manuti et al., 2022).Given the continued challenges that the pandemic brought for teachers, perceptions of the job demands, job resources and personal resources, and their preferences and reflections on these, can be enlightening to understand what teachers value and believe are important in their work and profession.Thus, we will focus on their experiences of relationships, working environment and self-efficacy and resilience, as factors of job resources, job demands and personal resources within the JD-R model.

Narrative identity theory
The narrative identity framework is a useful framework to understand teachers' experiences and their resulting reflection.Narrative Identity Theory (McAdams, 2001;McAdams & McLean, 2013) suggests that individuals develop, internalise and make meaning of their evolving self-stories, which provide them an ongoing sense of unity and purpose.Understanding context is an important part of understanding one's narratives (Edwards & Miller, 2007;Menter, 2008), and teachers' self-stories during COVID-19 are likely to be integrated into their broader life stories, to make sense of their role as teachers amid frequent change.
This framework of capturing teachers' stories of the high, low and turning points of being a teacher during the pandemic is useful for drawing insights about their lives and their work.Previous findings using this framework during the first year of COVID-19 found that teachers' experiences of uncertainty, concerns regarding practicalities and pupils' welfare, the importance of relationships and teachers' professional identity, as well as reflections regarding the profession and the education system at large, were all identified as being important to teacher participants (Kim & Asbury, 2020;Kim, Leary, et al., 2021).
By understanding teachers' experiences as the COVID-19 pandemic progressed, as well as their reflections resulting from these, one can gain new insights into teachers and their work.Specifically, one can identify ways to support teacher wellbeing and teaching effectiveness (Clinton et al., 2018;Madigan & Kim, 2021a), which in turn may improve teacher retention rates (Madigan & Kim, 2021b;Mérida-López et al., 2022;Nguyen et al., 2020).Given that there is a teacher shortage crisis in many countries, including England (Long & Danechi, 2022), understanding teachers' experiences during the pandemic can inform the development of strategies to support teachers' wellbeing, effectiveness and retention as the teaching profession recovers from the effects of the pandemic.
Thus, the current longitudinal study examines the trajectory of the self-stories (high, low and turning-point scenes) of teachers in England since the third national lockdown in 2021 to the end of the second academic year with partial closures and reopenings in July 2021, and the end of the first academic year of full school reopenings in July 2022.Here, we focus on the job resources, job demands and personal resources of the teachers by examining the changes in their experiences -and their resulting reflections -of relationships, working environment and self-efficacy and resilience.

Participants and procedure
The data used in this study was collected as part of a longitudinal project on teachers' experiences during the COVID-19 pandemic (Kim & Asbury, 2020;Kim et al., 2022aKim et al., , 2022b;;Kim, Dundas, et al., 2021;Kim, Leary, et al., 2021) at seven time points (Ts): April 2020 (T1), June 2020 (T2), July (T3), November 2020 (T4), February 2021 (T5), July 2021 (T6) and July 2022 (T7).The interviews lasted 50 minutes on average and examined various aspects of their experiences, including the focus of this study, which was their high, low and turning points.Ethical approval was received from the researchers' university department.Participants were financially compensated for their time.
Participants were interviewed via Zoom by the same researcher at each time point to establish and maintain a good rapport.Participants were reminded of their rights at the beginning of each interview: their right to withdraw at any point and how their anonymised data would be used.Interviews were audio-recorded, transcribed and anonymised before coding and analysis.
The initial sample at T1 consisted of 24 participants with representation from primary and secondary school classroom teachers (CTs) and members of Senior Leadership Teams (SLT).CTs were categorised into early-, mid-and late-career teachers (ECTs, MCTs and LCTs, respectively), depending on their years of teaching experience reported at T1 (≤5, 6-18, ≥ 19, respectively).
The current study examines participants' responses at T5, T6 and T7, which consisted of 47 interviews (Table 1).When presenting quotations, we provide the participant number, school type (primary or secondary), career stage and time point as context (e.g.P1, Primary SLT, T5).

Life story interviews
Individual semi-structured interviews were conducted at each time point.The interview schedule was adapted from Section B of the Life Story Interview (McAdams, 2008) to ask participants to identify and describe three key scenes (a low point, a high point and a turning point) from their experience of being a teacher during the pandemic since they were last interviewed.We asked participants to share as much detail as possible, such as when and where the scene took place, what happened, who was there, and what they were thinking and feeling, as well as to reflect on what their choice of scene might say about them as a teacher.The same wording was used at all seven time points to ensure consistency.

Longitudinal thematic analysis
Analysis was conducted by the second author and was guided by a constructionist framework, which assumes that the data represents the reality of teaching during the COVID-19 pandemic.However, the second author acknowledged how their pre-existing knowledge and experience of being a teacher as well as their experience of the pandemic may have coloured their analysis and interpretation of the data.Therefore, draft and final codes and themes were regularly discussed with the other two authors to ensure they accurately represented the dataset.That is, the process was iterated until all authors agreed with the processes used and with the labelling and description of the findings, with the greatest weight placed on the second author's perspective, given they were most immersed in the data.
It has been recommended that longitudinal approaches to qualitative analysis involve two levels of analysis: cross-sectional and longitudinal (Kirkman et al., 2001;Thomson & Holland, 2003).At both stages, Braun and Clarke's (2021) guidance for reflexive thematic analysis was followed.That is, reflexive coding was used, which represented 'the researcher's interpretations of patterns of meaning across the dataset' (Byrne, 2022(Byrne, , p. 1393)).
During the first phase of analysis, each time point was treated as an independent crosssectional dataset.Each time point was coded sequentially: T5 was coded, then T6 was coded and then T7 was coded.Nevertheless, the researcher acknowledges the impact that coding each time point may have had on the subsequent time points' coding.For example, the researcher coded 'frustration at the government' at T5 and then coded 'disconnect between school and government' at T6.The researcher noted that these codes which depict teachers' perceptions of the government could develop into a potential longitudinal theme.This also supports the claim by Braun and Clarke (2021) that reflexive thematic analysis may not be a linear process and instead may involve shifting back and forth between coding and generating themes.After coding, cross-sectional themes were generated by creating coding maps which grouped the codes.T5 themes included: the influence of social connections on teacher wellbeing; a focus on the social side of teaching; teachers cannot do any more than they are currently doing; and a fractured relationship with the government.T6 themes included: T6 as a transitional time between online and in person; no faith in the government; concern for the whole child; and relationships are central to teaching and its function.T7 themes included: worry for deprived students; tension between knowing pupils need qualifications and knowing they need a broad curriculum; and normalisation of fluidity.
For the second phase of analysis, we initially planned, and attempted to complete, a thematic analysis of T5, T6 and T7's themes to create the longitudinal themes.This would have involved generating longitudinal themes by only treating the cross-sectional themes for each time point as the dataset.However, we abandoned this in favour of using the codes from each time point to create the longitudinal themes.This was done by looking at all the codes across the time points and constructing thematic maps of codes which the researcher saw as demonstrating longitudinal similarities, changes, trends, narratives or experiences of teaching during the COVID-19 pandemic.This was found to be a less rigid approach, which allowed codes which were not represented in a singular time point's theme to be used to construct a longitudinal theme.For example, the following codes from each time point were chosen to construct theme 2: relationships are central to teacher wellbeing and work engagement.The T5 codes were: influence of colleagues on mood/workload; influence of parents; and senior leadership teams managing wellbeing.The T6 codes were: importance of relationships; impact of senior leadership teams on wellbeing (positive); impact of senior leadership teams on wellbeing (negative); and help and impact of COVID-19 on relationship-building.Finally, the T7 codes were: influence of parents and senior leadership teams helping staff.
Once the themes were named and described, the final step was to write these themes into a findings section of this manuscript and further refine the codes used to construct the themes (Braun & Clarke, 2021).Some codes were abandoned during the write-up of themes as, upon reflection, they did not fit sufficiently well with the overarching theme.The multi-stage nature of the process demonstrates how thematic analysis is an iterative process in which each step can influence the final research themes (Braun & Clarke, 2021).

Results
Four longitudinal themes were developed to describe teachers' stories of what it had been like to teach during the COVID-19 pandemic in England between February 2021 and July 2022.The first theme is focused on teachers' perceptions of the government's approach to school operations and learning.The second theme describes how relationships are central to teacher wellbeing and work engagement.The third theme explores teachers' preference for working in a physical school rather than an online/hybrid school.The fourth theme captures teachers' reflections on their self-efficacy in relation to using technology and resilience to change throughout the pandemic.The names of the themes and subthemes can be found in Figure 2.

Theme 1: teachers' perceptions of disconnection with and disrespect from the government
Across all three time points, teachers expressed a feeling of distance from the government, seeing their approach as heavy-handed and 'out of touch' (P8, T6) with the realities of school life.Moreover, they felt they were not being respected as a profession by the government.This theme was generated from teachers' discussions of government decisions and comments across all three time points as they shared their low points and turning points.This section is structured sequentially, starting from T5, progressing to T6 and closing with T7.
At T5 teachers most often drew upon the government's decision to close schools once again to all pupils other than vulnerable pupils and children of key workers in January 2021, which was often cited as a low point at T5.They described how they returned to school on Monday, 4 January 2021, often to plan and prepare for the first day of the teaching term on Tuesday, 5 January.However, later that evening, it was announced to the general population that schools would in fact be closed to most pupils.
Teachers described this announcement as unexpected and they were not in support of the 'last minute' (P6, T5) nature of the decision.P6 felt 'that there was no need for last minute announcements from the government, which led us to having to really quickly put things in place while at the same time not really knowing what we are doing'.
Other teachers described this announcement as 'sudden' (P10, T5) and made at 'such short notice' (P12, T5).P12 recalled thinking, 'surely he [the Prime Minister] wouldn't close schools at 8 pm when children are expected tomorrow, like the next day'.Similarly, two teachers questioned the government's decision with statements such as: 'Why didn't you [the government] just say?You could have saved all this pain and heartache.'(P16, T5) and 'Why would you [the government] choose to do it the day we all come back to school?' (P1, T5) Teachers used this event to exemplify how they felt the teaching profession was 'undervalued' (P16, T5) and '[not] respected' by the government (P2, T5; P6, T5).P4  (T5) drew upon this event to detail their perception of the relationship between the government and the school workforce: It's damaged the relationship of the sector with the government, it is going to take a long time to rebuild that trust.There's only so much you can treat people with a lack of regard and professionalism and that continues to happen: [a] continual revolving pattern of this top-down heavy imposition.
At T6, it appeared that this negative perception of the government continued.Unlike at T5, there appeared to be no specific trigger event for this perception.However, many participants mentioned the announced return of Ofsted inspections (a government body which visits schools and publicly rates their performance), which suggested that this may have been perceived as an indirect trigger event.For instance, P4 (T6) described a polarised view of the government and schools' work during the pandemic: There are people in like Ofsted or the Department for Education who have been at home and are having a really nice comfy time during the rest of the pandemic . . .I would have liked them to lay off their inspectors and put them in schools.
Teachers disagreed with government actions, such as the return of school inspections from September 2021, which they saw as involving a wrongheaded prioritisation of pupils' academic attainment over their mental health and basic needs: The government is still purely focused on getting them ready for reading and writing and back to SATs and all of those things.And I just worry that as a society we are heading for a bit of a crisis point with children's, especially young adults', mental health . . .because of those really important milestone experiences that they've missed out on and are not getting the opportunity to replace.(P2, T6) By T7, teachers did not often mention specific government announcements.Instead, teachers made passing comments about the changes in the government and presented a generally disillusioned view.P2 (T7) stated that 'the local authority and the DfE [Department for Education] didn't really understand that things had changed', which they said had led their headteacher to leave the profession.P17 (T7) referenced having three Education Secretaries in three days in July 2022: 'all this political stuff that has kicked off and the Department for Education just seems very, very unstable at the moment'.This political climate led P17 to 'not feel great, I don't feel reassured'.These comments, although not containing such explicit emotional sentiments as the previous time points, were consistent with earlier interviews, which suggested that teachers felt alienated and disrespected by the government throughout the COVID-19 pandemic, and that this negatively affected how they functioned as teachers.

Theme 2: relationships are central to teacher wellbeing and work engagement
Participants shared the belief that relationships with others are central to their profession, and 'underpin what goes on in the school' (P3, T6).The pandemic 'stresses what we know about teaching, that it is very relationship dependent' (P24, T6).Teachers described relationships with three groups when discussing what it was like being a teacher during the pandemic (i.e.parents, colleagues and pupils), and these groups form the basis of the three sub-themes (Figure 2).A mix of positive and negative experiences were shared regarding all three relationships with colleagues, pupils and parents.

Colleagues as sources of emotional and practical support
Of all of the relationships teachers described, those with colleagues were identified as the most valuable to their wellbeing and performance.This is best seen through P23's (T5) reflection: 'Seeing my colleagues just has to be number one.I mean, I have to confess, I go to work for the social life . . .I work with people that are incredible.'Relationships with other teachers were sources of support for the participants, often provided through informal conversations.These conversations were sometimes about increased workloads as a result of the pandemic and their perception that their teaching was not of high quality compared to pre-pandemic.Interactions with colleagues were often described as turning points, whereby colleagues supported them when they faced adversities through conversing and sharing their experiences.Both P22 (T5) and P13 (T7) discussed how they gained reassurance from their colleagues in relation to their workload and the quality of teaching lessons.P22 (T5) discussed how a conversation about workload during a socially distanced walk with a colleague helped them accept that 'sometimes I need permission that it's ok not to work at that level' and reduce their workload.P23 (T6), who described the 'support from colleagues [as] immense', cited a conversation they had with their colleague which prompted their turning point of feeling positive about the future of holding performances for their Drama students.P13 (T7) described a conversation with a senior leader at their school in which they discussed that the 'lessons . . .are nowhere near as good as I used to teach' because they felt they simply did not have the time.P2 (T7) detailed how conversations with colleagues prompted them to 'reflect on the type of teacher, and leader, I want to be'.
Relationships with SLT provided teachers with validation and practical solutions.For example, P13 (T5) shared that teachers were given more time out of the classroom to complete work during work hours and at T6 the headteacher had discussions with individual teachers about their wellbeing.P11 (T6) described how they wanted to move schools but decided against this after a 'big turning point' conversation with the head teacher about their reasons for wanting to move and career aspirations.P16 (T7) and P3 (T7) both discussed instances in which they felt validated by their SLT, through P16's project on pupil mental health being approved by SLT and P3's multi-academy trust's SLT congratulating the school on their positive Ofsted.
The descriptions of the nature of the relationships with fellow teachers contrasted with that with SLT, perhaps due to the different hierarchical nature of the relationships.That is, the former was described with undertones of friendship while the latter appeared to be more transactional: providing teachers with what they needed to teach in the pandemic (i.e.practical solutions and validation of their teaching).Nevertheless, both of these relationships were identified as sources of positive support for their wellbeing and work engagement.

Teachers care about pupils' relationships
Pupils' experiences were often at the forefront of teachers' minds, showing that pupils were their priority.Teachers described and were often concerned about pupils' development.This was particularly noted in regard to primary school pupils' social skills at T5 versus T6 and secondary school pupils' experiences of transitions and school events at T6 versus T7.In response, their relationship with pupils changed, which both positively and negatively influenced their wellbeing.
Teachers at T5 appeared to be optimistic about primary school pupils' social skills development.Specifically, these teachers described stories of observing pupils interacting online and in school (if vulnerable or a child of a key worker), which brought teachers joy at a time of social distancing measures: [Pupils] just played together and interacted together, and to see some relationships developing that would never have developed before because they were in separate, different friendship groups.(P13, T5) It was just so wonderful to see [pictures sent in of] the parents and the children and their siblings all enjoying [the snow] . . .if we were in school, I would say 'let's scrap phonics for today and get out in the snow'.(P11, T5) This optimism was not present at T6 for primary school teachers; instead, their concern was that they had missed out on opportunities which would have developed their social skills as a result of COVID-19.This ranged from concerns about written communication skills (P12) to public speaking and interpersonal skills (P2).These reflections were grounded in the teachers' stories of having full classes back in the classroom and viewing their progress in person, rather than through online technology.For example, when detailing their low point, P11 expressed concern for their class moving up to the next year group/grade as 'they're not ready for the year one learning and if we could have them for an extra few months in reception, we could work wonders with them'.
Concern for pupils in T6 was also expressed by secondary teachers who indicated their worry about the perceived impact of the pandemic and online learning on their pupils' social skills.Stories of pupils returning with limited social skills were often named as teachers' low points during this time as it conjured negative emotions (e.g.worrying about how to teach pupils with limited social skills: P6, P8, P23).P6 (T6) described the return to teaching a full class in person as a 'battle' when trying to communicate with pupils as 'their answer to everything is "I don't care"'.P23 (T6) recalled 'realising they've really missed out on development . . .they're coming back with very little grasp of sort of social skills'.P8 (T6) agreed and challenged ideas that schools should address 'lost learning'.Instead, P8 (T6) felt that socialisation should be the focus due to 'a lack of interaction with other children' during the pandemic: 'We need to have more of a relaxation on the academics and we need to have more of a look at how to reeducate these children on the basics, you know interacting with each other.' Changing teacher perceptions could be seen between T6 and T7 data, when teachers discussed pupils' experiences of transitions and school events.At T6, teachers expressed their worry about transitioning pupils in Years 1, 6 and 11.The pandemic had meant that 'things that help them grow up' (P2, T6), such as shows and trips, could not go ahead.However, at T7, teachers discussed their delight that these activities had resumed.For example, when discussing a theatre trip, P22 (T7) expressed their realisation that: 'This is what education is about.It's about enriching students.It's about building relationships.' In sum, teachers perceived a change in students' relationships; namely, that pupils' ability to connect and maintain relationships with each other changed.Observing these changes prompted positive (T5 and T7) and negative (T6) emotions for teachers, evidencing teachers' beliefs that their role as teachers is to walk alongside pupils and to promote their academic and social-emotional development.

Parents' feedback and engagement affect teachers
Teachers described scenarios where parents provided positive and negative feedback on their teaching, as well as their efforts to engage with parents.Together, these contributed to teachers' wellbeing and their perceptions of their own effectiveness.
Teachers described how online learning had invited parents, and their comments, into the classroom in new ways, as parents had observed their children's lessons from computer screens.When participants described their high points at T5, five participants described moments when they received praise from parents for their children's online lessons.These instances were named as high points by the participants as they reminded teachers of why they were in the teaching profession.Parents were described as 'thankful' (P10, T5), 'appreciative' (P14, T5) and expressed how their child is 'really lucky to be taught by these committed and enthusiastic staff who know their subject and their child' as paraphrased by P6 (T5).These instances provoked 'really positive' (P6 and P24) emotions for the teachers as they described the praise as 'amazing' (P10) and 'nice [and] lovely' (P14).P6 drew upon parental praise 'as a reminder' of why they do their job.
Negative feedback from parents was also described and identified as a low point across all time points.Four teachers described how they received parental complaints about their teaching and struggled with parental engagement.For example, P13 (T5) described receiving 'an unpleasant email from a parent who basically said, "What you did wasn't good enough, it wasn't as good as we've had [from P13's job share]"' as their low point.P3 (T6) discussed how their school's uptake of online learning was 'too low . . .so we reverted to the paper work packs and then the parents weren't collecting them so we ended up walking around the estate delivering them'.Some teachers expressed how, at either T5, T6 and T7, they felt parents needed their support too.P16 (T7) discussed how their new SLT role was to focus on parental engagement, demonstrating some schools' prioritisation of relationships with parents.Illustrating the role of parents in teachers' perceptions of their role in education, P11 (T5) drew upon a parent's suicide to express their reflection that: Sometimes you need to take that step back and think their education isn't the be all and end all at the moment.These families are suffering enough as it is without pressure from us . . . .See a family as a unit, not just that child in my class that I need to look out for, but to say I need to look out for the whole family as well.
The bi-directional relationship between parents and teachers, though mixed with positive and negative sentiments, can be juxtaposed against teachers' almost entirely negative and strained relationship with the government.At T5, when the government announced partial school closures, P12 described parents as being 'not ready', just like them.P8 (T5) described their delight at the positive praise given by parents to Ofsted when the government asked parents to report any concerns about schools to Ofsted: 'It suddenly made you realise that parents get it, they get what we're trying to do and it's only these idiots, people from the government, who don't get it.'In summary, though the relationship with parents was not always easy, it was mutual and reminded teachers of key elements of their role as teachers beyond the classroom.

Theme 3: teachers' preference for the physical school over the online/hybrid school
Teachers distinguished between two types of schools that operated during the COVID-19 pandemic: the physical in-person school and the online/hybrid school.Teachers expressed a preference for the physical school given their perception of a higher workload attached to online/hybrid education, and the feelings they had about teaching in-person and its effectiveness.

Teacher workload
Across the three time points, teachers reported their perception that online/hybrid education resulted in a greater workload than in-person teaching.At T5, teachers discussed how the sudden move to online learning, as a result of the government's decision to close schools to most pupils, increased their workload.Teachers often contrasted this move to online learning with the first period of lockdown and partial school closures (March 2020), as they felt there were higher expectations of lessons taking place online while they also taught key workers' children and vulnerable pupils in-person.Teaching both online and in-person (i.e.hybrid teaching) resulted in more work for teachers, which led to some describing this moment as their low point.For example, P12 (T5) described this first week of constructing an online/hybrid learning environment as 'the most exhausting week I've ever experienced as a teacher'.Similarly, P10 (T5) described this time as 'extremely stressful'.
At T6, P13 discussed how 'the workload halved overnight' when schools reopened to all pupils.However, hybrid teaching persisted as many teachers discussed how their online learning environments had to be offered if pupils had to isolate due to having COVID-19.Furthermore, in place of external examinations, teachers collected evidence to substantiate a teacher-assessed grade.P8, P9, P16 and P24 described how this increased their workload.For example, P24 (T6) discussed how 'reading an entire year group worth of scripts again and again and again' was 'an awful lot' to do on top of their usual workload.Despite this drop in workload from T5 to T6, workload increased without notice from T6 to T7, due to growing COVID-19 infection rates and subsequent isolation frequencies.
At T7 it appeared as though the need for online/hybrid learning was dwindling as teachers discussed the drop in COVID-19 cases across the 2021/2022 academic year.Many teachers at this time discussed what their schools' online learning environments meant for their future workload.Some discussed how snow days, in which lessons are cancelled as pupils and teachers cannot attend school due to weather conditions, would cease to exist.Instead, they predicted that lessons would move online.For example, P1 (T5) described their turning point as realising schools 'won't' go back to sending a book home and then it coming back in; it will be much more about this kind of live stuff or doing it online'.However, as shown throughout this theme, online learning increased teachers' workloads, which teachers expressed as having a negative impact on their wellbeing.

The feeling of in-person teaching
Teachers drew upon the feeling of the benefits of in-person teaching, compared to online/hybrid teaching, for both pupils and themselves.Teachers discussed how they believed pupils benefit from in-person learning.For example, P23 (T5) believed that 'you get an education by experience', P2 (T6) discussed the need for pupils to have 'sensory experiences' and P16 (T5) described online/hybrid teaching as 'so detached from what it is in the [physical] classroom'.These philosophies were often used when teachers explained why they thought pupils 'had not done very well at all' (P22, T6) while learning online/in a hybrid format.
Teachers discussed how they enjoyed teaching in person as this was what they were trained to do.At T5, when teaching from home, teachers discussed their desire to go back to teaching in person.For example, P8 (T5) believed that online/hybrid teaching was 'really hard work and not particularly rewarding because the rewarding part of the job is being in a classroom with the youngsters and interacting and you just can't do that the same [in the online school]'.
When they returned to in-person lessons with all pupils, self-doubt in their ability to teach online/hybrid lessons was eradicated.For example, P19 (T6) recalled Moreover, they discussed their working environments positively when back in school for T6 and T7.At T6, P23 (T6) described their classroom as 'very calm . . .it's a safe place . . .an oasis of calm'.P3 (T6) discussed how being back in school was better for staff meetings as teachers 'have more confidence asking questions . . .it's just a better way of working'.These feelings continued at T7, evident when P19 expressed that they felt as though being back in school meant that 'everyone [staff and pupils] is having a nice time'.That is, teachers appeared to favour in-person lessons due to the perceived benefit this had for pupils, but also for their own self-efficacy.

Theme 4: reflections on teachers' increased self-efficacy in using technology and increased resilience to change
Many teachers reflected throughout T5, T6 and T7 on their self-efficacy in using technology, and their resilience to change throughout the pandemic, both of which had increased over time and were often discussed as their turning points.

Increased levels of self-efficacy in using technology
Across T5, T6 and T7, many teachers discussed how they felt more confident with their ability to use technology to facilitate learning in comparison to the start of the pandemic.Speaking from an SLT perspective, P3 (T5) felt confident in the school's use of technology to facilitate learning, as they stated 'right now we've got the learning platform and the remote offer right'.They contrasted this practice favourably with paper-based lesson materials they used earlier on in the pandemic.Similarly, P22 (T6) described using technology to facilitate learning as 'second nature' due to their new 'general confidence in technology'.They also reflected upon how such confidence was not found 'last year, [when] it would have been, "Right.I have to annotate it [learning materials] first and just be comfortable with it and then I'll probably have to take four takes of it and then I'll have to watch it."'Barriers to using technology to facilitate teaching seemed to reduce over time.For example, P16 (T7) discussed how they felt more confident as a teacher as they overcame personal psychological barriers during the pandemic.They reflected, 'Actually I can do this [job] . . .I'm really good at getting the kids on board and getting through to them.' Teachers also described how technology was used to facilitate staff training.Specific to this use, differing confidence levels in using technology for this purpose was raised by P24 (T7), who, though confident with using technology for shorter training sessions with teachers, questioned their confidence with using it to facilitate longer training sessions: Once meetings get beyond, I don't know, an hour and a half online, I think it just becomes a bit more challenging . . .so it feels like it's pretty inhibited by technology.When you can't have bits of paper in front of you and sort of collaborate on a sort of shared document visually, it's pretty challenging.
These stories demonstrate how teachers learned to use technology in continuing to teach their pupils and facilitate staff training, which resulted in an increased self-efficacy in using technology.

Increased levels of resilience to change
Teachers discussed how their resilience to the change had grown as a result of the situations the pandemic had created in education.P19 (T5) reflected on how they felt more able to accept and move on from the government's decision to close schools to most pupils than they thought would have been possible earlier in the pandemic: I think if this [government's decision to close schools to most pupils in January 2021] had happened like say six months ago, I'd just be mind blown thinking, 'I don't know what to do.'But because there has been so much change, that's been the only constant.So it's kind of like, 'okay.This is alright.'Others also demonstrated their awareness of how they had developed a greater level of resilience as teachers.When describing a low point, of political instability due to changing Secretary of States for Education in July 2022, P17 (T7) reflected that: 'If that is the biggest pick [of a low point] that I have got over the year, then something has gone right because that is not really a significant thing to be worried about . . .I've probably become more resilient.' Both of these examples demonstrate how teachers were able to cope with structural changes that were out of their control.Furthermore, both participants recognised that this is a new development, which had come about because of their familiarity with change in education throughout the pandemic.However, P23 (T5) questioned whether resilience to change is a result of the pandemic or a feature of the job of a teacher: 'You don't ever become complacent in teaching.There's never that luxury of knowing, "Yeah, everything's sorted.Everything's ok".' Nevertheless, the changes brought about by the pandemic (e.g. the government's partial closure of schools) appeared to have highlighted the resilience needed to be a teacher throughout the pandemic.

Discussion
As schools rapidly adapted to changing requirements to continue educating and looking after the welfare of pupils, teachers experienced a myriad of high, low and turning points, marking changes in their experiences of job resources, job demands and personal resources.To understand changes in their experiences, as well as resulting reflections, throughout this time, we conducted a longitudinal trajectory analysis of teachers' responses between February 2021 (partial school closures due to a third national lockdown), July 2021 (full school reopening; end of the second academic year of partial school closures and reopenings) and July 2022 (full school reopening; end of the first academic year of continued full school reopenings) in England (please see Figure 1 for more details).We found that relationships, the work environment, and self-efficacy and resilience affected teachers' wellbeing and work engagement.We elaborate on the four identified themes (teachers' perceptions of disconnection with and disrespect from the government; relationships are central to teacher wellbeing and work engagement; teachers' preference for the physical school over the online/hybrid school; and reflections on teachers' increased self-efficacy in using technology and increased resilience to change) and how these may indicate ways to support teachers and the teaching profession as we emerge from the pandemic.

Teachers' perceptions of disconnection with and disrespect from the government
Teachers continuously noted feeling separate from and disrespected by the government in their decisions and communications.This was not a surprising finding, given that their perceptions of disconnection and disrespect were discussed throughout 2020 and 2021 (Kim & Asbury, 2020;Kim, Leary, et al., 2021;Kim et al., 2023) and the current finding seemed to indicate that this continued into 2022.One could consider this as a manifestation of what others had noted, which is that the teaching profession has become a directed profession, led by the priorities and policies of the government at the time (Bottery & Wright, 2000).The findings indicate that teachers did not feel that they could contribute to decisions that affected their daily work.In turn, their reduced autonomy would have affected their morale and perceptions of the government, especially given that autonomy is considered one of the basic psychological needs, according to Self-Determination Theory (Deci & Ryan, 2000).In sum, it appears that this disconnect was steady over the pandemic and flared up in response to specific trigger events, which were sources of distress for teachers.

Relationships are central to teacher wellbeing and work engagement
Teaching is a social profession, involving interactions with various groups, including colleagues, pupils and parents.Relationships with the groups that they had the most contact with (i.e.parents, colleagues and pupils) were noted in relation to their impact on their wellbeing and work engagement.
According to various models, including the Job-Demands Resources Model (Bakker & Demerouti, 2014) and the OECD's conceptual framework for teachers' occupational wellbeing (Viac & Fraser, 2020), social support is an important job resource that is helpful for one's mental health and wellbeing.Participants noted that emotional and practical support from colleagues benefited them, which again had been noted in earlier stages of the pandemic with the same sample (Kim & Asbury, 2020;Kim, Leary, et al., 2021), attesting to the value of supportive colleagues, particularly to one's mental health and wellbeing (Kim et al., 2022a).
In contrast to the steadiness of colleague support, parental support was reported to decline over time, which was in line with other studies' findings.Indeed, studies conducted at the initial stages of the pandemic noted families' high appreciation of teachers and improved relationships with teachers (e.g.Bubb & Jones, 2020).The current study indicates that as the demands that parents experienced in their own lives and the expectations they had of school changed over time, so did their perception of and relationship with teachers.Given the benefits of positive school-family relationships on pupils (Hampden Thompson & Galindo, 2017), interventions are examined to increase this (Smith et al., 2022), and restoring the relationship may be a consideration that both families and teachers make as we emerge from the pandemic.
Teachers, as they did in the initial stages of the pandemic (Kim & Asbury, 2020;Kim, Dundas, et al., 2021;Kim, Leary, et al., 2021;Kim et al., 2023), continued to worry about their pupils.Concern for others' wellbeing was noted as negatively impacting their health during the pandemic (Kim et al., 2022a), and it continued to be the case.Indeed, other studies have noted pupils' difficulties in social-emotional development as a result of the pandemic (Hamilton & Gross, 2021) and strategies will need to continue to be proposed and implemented over the longer term.

Teachers' preference for the physical school over the online/hybrid school
Teaching during the pandemic was difficult for all teachers, as individual, school and national circumstances and decisions necessitated the use of different types of teaching modalities (i.e.in-person, hybrid and online) as part of their changing work environments.Prior studies noted the implications of schools moving to online learning at the beginning of the pandemic on teachers' workload (e.g.Kaden, 2020;Phillips et al., 2021).The current study provides a unique insight into teachers' trajectory of adapting to different working environments and their developed preference for one over others.Teaching in person seemed to be preferred by teachers regarding both workload and the feeling of its effectiveness.Switching from one modality to another, especially when required without prior warning, increased teachers' workload.Moreover, providing hybrid teaching, whereby teachers taught some students in person and others online, was particularly difficult for teachers, as noted by teachers in other studies (Bartlett, 2022).In line with current findings for teachers' preference for a physical school, a study found that teachers' self-efficacy was highest for those teaching in person, followed by those teaching in hybrid formats, and lastly by those teaching virtually (Pressley & Ha, 2021).Though discussions are emerging on how to embrace different modes of teaching post the current pandemic and in future pandemics, teachers' preferences should be taken into account as they relate directly to the effectiveness of the teaching.

Reflections on teachers' confidence and resilience
Teachers were confronted with the necessity to rapidly up-skill themselves in using technology to facilitate continued teaching and learning, as well as provide staff training, in whatever format was available at the time (i.e. in person, hybrid and virtual).Despite participants' preference for the physical school, they also noted a general increased confidence in using technology, accompanied by a reduction in the barriers to its use.These are contrary to other study findings reporting lower levels of self-efficacy compared to pre-pandemic times (e.g.Pressley & Ha, 2021).However, the current finding may be due to the timing of the study, where the first data point for this study (i.e.T5) was around 11 months after the first national lockdown.Thus, teachers had time to increase their digital technology skills since then and would presumably increase further over time (i.e.T7).Such a possibility is exemplified by a study finding wherein teachers reported increased levels of selfefficacy in technology application between the beginning and end of the teaching semester in 2020 (Ma et al., 2021).Such behaviour would further be in line with the technology acceptance model (Davis, 1989), and its extended model (Venkatesh & Davis, 2000), which proposes that the factors associated with the perceived usefulness and perceived ease of use predict one's intention to use and thereby use technology.In the context of the pandemic, perceived usefulness was very high and schools did their best to assist teachers in the use of technology through training and the provision of resources, which resulted in teachers experiencing higher self-efficacy in using technology (Baroudi & Shaya, 2022;Cardullo et al., 2021).
Teachers also reported experiencing greater levels of resilience as they survived through the many changes.Teacher resilience, though conceptualised in various ways (see Beltman et al., 2011 for a review), is largely understood as a change-related process during or after adversity resulting in a positive outcome (Ungar, 2021;Van Breda, 2018).Resilience is valuable for teachers as it mediates the impact of job demands (e.g.work overload) and job resources (e.g.school support) on teachers' wellbeing and job performance (Chen & Chi-Kin Lee, 2022).Accepting change that is beyond their control seemed to be helpful in participants' development of resilience.Interventions such as acceptance and commitment therapy hold acceptance as one of its core processes (Hayes et al., 2012), which has been found to be effective in developing resilience in a variety of populations, including teachers (see Kangas-Dick & O'Shaughnessy, 2020 for a review).Indeed, accepting what is beyond one's control is an important component of mindfulness and acceptance and commitment therapy, which have been found to be effective in developing teachers' resilience.
As such, understanding the changes in teachers' experiences -and the resulting reflections -of the job resources, job demands and personal resources in the second and third year of the pandemic in England highlighted what teachers value and believe are important in their work and in the profession.

Implications
Lessons learned from the pandemic are being discussed to improve education as the world moves forward (e.g.OECD, 2020).Contributing to this discussion, teachers, schools and policymakers should consider ways to prepare teachers and schools for unexpected circumstances.At the individual level, for example, training and resources need to be in place to assist teachers for living in a VUCA (volatile, uncertain, complex, ambiguous) world, for example targeting the development of social-emotional competencies like resilience (Hadar et al., 2020).At the school level, contingency plans should be developed that ensure effective communication between colleagues, families and pupils in case of future emergencies (Harmey & Moss, 2021).Specifically, schools should consider how they will support adaptation of the curriculum, and switching between teaching and learning modalities in a way that does not increase teacher workload (Tang, 2023).Such contingency plans would assist teachers' relationships with multiple parties and minimise the impact on their workload, which in turn should assist with their wellbeing and work engagement.At the policy level, structures and systems should be in place where collaborative communication between government and the teaching profession can take place regardless of circumstance, so that teachers can contribute to decisions and implementations that affect their profession and their workplace (UNESCO, 2020).Such preparations will ensure that individuals, schools and educational systems are ready and resilient for the future.

Limitations and future directions
It is important to acknowledge the caveats of survivor bias and its relevance to the data collected.That is, the content of these themes may only be true for the participants who continued participating in the study and may not be necessarily true for those who did not continue (e.g.due to leaving teaching).Within each time point, there were teachers who dropped out of the study.This comment is particularly relevant for the third theme, whereby the stories of developing resilience and confidence are of those teachers who decided to continue with the interviews.It may be that some teachers did not feel that their experience of teaching during the COVID-19 pandemic strengthened their confidence or resilience.This may have been a determining factor in them leaving the profession.For example, P24 (Secondary LCT) in T5 discussed how teaching online was not easy.This may have been a common attitude to have during this time as a teacher.However, in T6, P24 discussed how they were successfully using online learning, thus demonstrating that P24 is a 'survivor'; there may have been teachers which, like P24, struggled with teaching online and decided to leave teaching.Therefore, when discussing the findings, particularly in respect to developing confidence and resilience, it is important to state that this was only true for the participants present, and who decided to continue with this study.As such, given the qualitative nature of the study, together with this caveat bias of a longitudinal study, the findings should not be discussed as those which can be generalised to a wider population of teachers.

Figure 1 .
Figure 1.Timeline of 2020-2022 school events and research data collection.

Figure 2 .
Figure 2. Thematic map with four themes and their subthemes.
thinking . . .oh this is what teaching is like!I remember this!I'm alright at it!I feel calm, I feel like I know what I'm doing . . .I had forgotten what it was like to feel like you're doing a good job.

Table 1 .
Participant groups, gender and time points participated.