Coping with the ‘new (ab)normal’ in school: an EMA study of youth coping with the return to in-person education during the COVID-19 pandemic

Abstract

Many concerns exist about potential long-term psychosocial impacts of the COVID-19 pandemic on young people. While the school has been identified as having a vital role in psychological recovery post-disaster more generally, it is unclear as yet how young people have adapted to the return to in-person education. This paper reports on the preliminary findings from an intensive Ecological Momentary Analysis exploration of the affective wellbeing, experiences and coping of 82 Irish second-level students. The participants were found to experience more positive than negative events, and to have moderate-high levels of positive affect and perceived coping during the 7-day period of monitoring. However, the findings also suggest that some students, particularly those with pre-existing psychological difficulties, may be in need of additional targeted support. Accordingly, it is recommended that in the short to medium-term, second-level schools should strive to prioritise the psychological recovery and resilience of students such as through an emphasis on re-establishing and consolidating a sense of student connectedness.

Keywords:

• COVID-19
• EMA
• coping
• wellbeing
• youth

1. Introduction

1.1. Psychosocial impacts of COVID-19: Insights from the general disaster literature and the COVID-19 literature

Despite being less likely to experience debilitating symptomatic illness relative to adults during the COVID-19 pandemic (Viner et al. 2021), children and young people have been particularly vulnerable to indirect impacts of the pandemic. These have included dramatic changes in daily routine due to the closure of school buildings (Prime, Wade, and Browne 2020), a loss of social engagement and social support (Loades et al. 2020), the postponement or loss of milestone life events (Efuribe et al. 2020), and in some cases, complicated grief due to unexpected bereavement (Weinstock et al. 2021). Considering the myriad of disruptions to the predictability, structure and perceived safety of daily life, many questions have been raised about the social, emotional and mental health consequences of the pandemic for children and young people (Hsieh et al. 2021). The study described in this paper sought to investigate some of these questions. The overarching aim was to explore real-time experiences of coping and affective wellbeing in young people in the Republic of Ireland during the return to in-person education using an innovative Ecological Momentary Assessment (EMA) methodology. Empirical and theoretical literature relevant to the study is reviewed in this section, while Sections Two, Three and Four, respectively, discuss the study’s methodology, key findings and conclusions.

Previous psychological research exploring the impacts of disasters (e.g. natural disasters, acts of mass violence etc.) on children and young people offers a useful starting point for reflection on the impact of the pandemic, whilst recognising that these events do not directly map on to the current pandemic situation (Engzell, Frey, and Verhagen 2021). This research has identified children and youth as a highly vulnerable group during and after disasters due to dependence on adults, reliance on the home environment, and developmentally-determined difficulties with initiating help-seeking behaviour (Peek 2008). While elevated psychological impairment is common initially following a disaster, symptoms tend to decline over time, with a minority rather than a majority of youth (typically around 30%) manifesting chronic problems (Bonanno et al. 2010). Nonetheless, a prevalence rate of up to 30% for chronic psychological impairment is concerning since it can translate into a large number of affected youth requiring clinical intervention (Dyregrov, Yule, and Olff 2018). Peek (2008) has proposed that there is a complex array of factors at individual, family and community levels, which may either buffer or exacerbate the impacts of a disaster. Candidate factors include: the child’s characteristics; the child’s coping skills; the coping assistance received at home and at school; the intensity of exposure to the disaster; family factors; and community wellbeing (e.g. Fothergill 2017; Pfefferbaum et al. 2015). In general, the assumption is that psychological impacts on children and youth are compounded by pre-disaster family contexts where there are more vulnerabilities, fewer resources and fewer protective factors to buffer against the cascading effects of disasters (Fothergill 2017; Vandana 2019).

Although research is still in its nascent stage, associations have been found between COVID-19 stressors and adverse short-term mental health outcomes in children and young people including heightened levels of psychological distress connected to social isolation (e.g. Vogel et al. 2021; Shovestul et al. 2020), and increased symptoms of anxiety and depression (e.g. Xie et al. 2020). Furthermore, predictions about differential psychological impacts of disasters have been borne out, with socioeconomic status (SES) having emerged as a significant moderator of immediate psychological responses during the acute phases of the pandemic among children and young people (e.g. Ravens-Sieberer et al. 2021). While it is unknown whether long-term psychological impairment will result from the disruptions and losses of the pandemic (Thorell et al. 2021), it is anticipated that this scenario will be more likely for children from certain groups; including children from disadvantaged communities, children in alternative care arrangements, children from unstable home environments, children experiencing poverty, children with disabilities, and migrant, asylum-seeking and refugee children (Larkins et al. 2020)

1.2. The vital role of the school in psychological recovery

Beyond the family context, the school has been identified as having a unique and vital role in psychological recovery from disasters for young people. It has been hypothesised that this function is mediated through the restoration of daily routines (Gibbs et al. 2015), the facilitation and promotion of social connectedness, and the provision of coping assistance, either informally through supportive relationships, or formally through emotional or behavioural health interventions (Vandana 2019). However, a current concern in relation to youth mental health and wellbeing is how young people are adapting to the return to in-person education following the reopening of school buildings. The transition to what Crick (2021) terms as the ‘new (ab)normal’ has been constructed as a challenging one for all young people (Guessoum et al. 2020), due to the concurrent demands of detaching from the routines of homeschooling, catching up with academic progress, adhering to the new COVID-19 safety protocols, and becoming comfortable with daily peer interactions once again; all under the elevated risk of infection in congregated settings (Fegert et al. 2020). However, particular concerns have been raised for students from marginalised groups, who may experience heightened challenges in readjusting to in-person education (Mohan et al. 2021). Lower levels of attendance have been expected for these students, partly because they tended to experience lower levels of engagement in emergency remote learning, and also because marginalised groups were more likely to experience negative economic and health impacts as a result of the pandemic (OECD 2020). However, little research to date has examined the adjustment and coping of young people, either from marginalised or non-marginalised groups, during the transition to in-person education following the reopening of school buildings (Qu et al. 2021).

1.3. Conceptualising coping among young people

There is a multiplicity of theoretical frameworks germane to the construct of coping. One of the classic distinctions in the psychological literature on coping is between problem-focused coping, aiming towards managing or alleviating the challenge causing the distress, and emotion-focused coping, aiming towards regulating the emotional responses to the challenge (Lazarus and Folkman 1984). This conceptualisation was subsequently extended to incorporate the elements of engagement versus disengagement coping, where one orients oneself either towards or away from the challenge or the emotions associated with it (Compas et al. 2001). Another guiding theoretical model for this study was that of Zimmer-Gembeck and Skinner (2016), which conceptualises coping as an adaptive intra- and inter-personal phenomenon that operates across three different time scales; namely, the developmental scale (where processes become reorganised over time due to reciprocal developmental changes), the episodic scale (coping that unfolds over days and months), and the real-time scale (Skinner and Zimmer-Gembeck 2007, 137). Real-time coping, according to Skinner and Zimmer-Gembeck (2009, 6) involves the activation of interactive regulatory processes that enable young people to ‘mobilise, modulate, manage, and coordinate the own behaviour, emotions and attention’ in response to internal or external stressors. Both episodic, but predominantly real-time coping were explored in this study.

Traditionally, coping has featured prominently in research with young people due to consistently observed associations between ways of coping and symptoms of psychopathology (Compas et al. 2017). Moreover, it has been recognised that adolescence is a critical period for coping, with significant developmental shifts being observed as to the centrality of social partners, the use of cognitive reappraisal, the availability of overt behavioural strategies (e.g. avoidance, distraction), and the capacity to use strategies flexibly (Zimmer-Gembeck and Skinner 2011). However, research expressly exploring the coping strategies utilised by young people in response to the manifold stressors of COVID-19 is still emergent. A noteworthy example is the co-produced multi-methods study of Dewa et al. (2021) which examined the coping strategies of adolescents in the UK between the ages of 16 and 24 during the acute phase of the pandemic through online surveys and semi-structured interviews. Eating-related, acceptance, and distraction coping strategies were found to be the most commonly employed strategies. Coping factors significantly associated with poor mental health, which was evident in approximately 30% of the sample, and was particularly prevalent among participants identifying as being of Black/Black-British ethnicity, were behavioral disengagement, self-blame, and substance use strategies. In another COVID-19 youth coping study, Domínguez-Álvarez et al. (2020) found that among Spanish children and young people, disengagement coping predicted higher emotional and behavioural difficulties. In contrast, engagement coping (e.g. acceptance, positive thinking) was associated with the greater psychological adjustment (Domínguez-Álvarez et al. 2020).

1.4. The current study

Coping has traditionally been assessed using retrospective measures. However, it has been widely acknowledged that responses on retrospective measures are subject to a multitude of cognitive biases including the effects of prior beliefs, and the influence of recent and salient events (Doherty, Balaskas, and Doherty 2020). An alternative to the retrospective report approach is EMA which has been found to be associated with mitigation of retrospective bias, increased longitudinal (temporal) validity, and increased ecological validity (Shiffman, Stone, and Hufford 2008). EMA aims to explore unfolding human experience by asking participants to report on current or very recent states and situations through a series of brief, repeated assessments. It has been used successfully with children and adolescents (from the age of 7 upwards) across different gender, ethnic, nationality and socioeconomic groups (Heron et al. 2017).

This EMA study is the fourth phase of a multi-phasal multi-methods exploration of young people’s educational experiences during and after the acute phases of the pandemic in the Republic of Ireland. The central objective of this study was to explore the adjustment and coping of young people during the period of transition back to in-person education; a pandemic-related temporal context for youth which remains under-researched. Within this framework, a specific goal was to adopt a balanced approach to youth functioning, informed by recent academic commentary on the perils of focusing exclusively on psychological distress and impairment to the exclusion of positive outcomes such as stress-related growth and resilience processes (Dvorsky, Breaux, and Becker 2021; Waters and Arslan 2021). Accordingly, both positive and negative daily events and coping processes were explicitly investigated. Arising from these objectives, the research questions framing the study were as follows:

1. What is the affective experience, and what types of stressors and uplifts are Irish second-level students encountering during the educational transition phase?

2. What strategies are Irish young people using to cope with stressors during this phase of the pandemic?

This paper reports on preliminary normative (rather than ipsative) findings on affective wellbeing and coping from phase 1 analysis of the very extensive EMA dataset (2380 sampling moments) so as to disseminate findings that might inform interventions to support young people’s re-engagement with in-person education.

2. Method

2.1. Study recruitment and final sample

Full ethical approval for the study was obtained from the research ethics committee of the relevant institution. The inclusion criteria for the study were: a young person of secondary school-going age (typically 12–19 years); currently in, or having been in an educational placement during the preceding 12-month period (mainstream, home tuition or homeschooling); and having access to an android or IOS phone with internet access. Students were excluded from participation in the study if they had a diagnosis of a psychological disorder, and were currently receiving therapeutic intervention, and had high scores on a baseline measure of psychological adjustment. A variety of different community-based channels (i.e. organisations that work with second-level students from marginalised communities, and second-level schools with a designated disadvantaged status; DEIS schools) were mobilised in an attempt to recruit a geographically-dispersed convenience sample. Participation in the study was completely voluntary, and both parental consent and student assent were obtained.

A 30% EMA compliance threshold resulted in a final sample of 82 adolescents [Mage = 15.79 years (SD = 1.8), 12–18 years, 51.8% female gender identity, 44.6% male gender identity] across the Republic of Ireland. The majority of the participants (67.1%) were in the upper level of secondary school [i.e. Transition Year (TY), 5th and 6^th year). Most reported White Irish ethnicity (85.5%), and most had been born and had always lived in Ireland (71.1%). Most participants reported that English was the primary language at home (85.5%), and that they were attending an English-medium school (73.5%). The majority lived in a home situation with two parents (73.5%) and one other child. Despite concerted attempts to recruit participants from marginalised backgrounds, only a small minority of participants (18.1%) were from homes where parents had a medical card, which is a commonly used indicator of socioeconomic disadvantage in Ireland. Six percent of the sample had a diagnosis of a psychological disorder or were having a diagnosis explored at the time of the study. Parent reports indicated that 6% of participants were receiving mental health support at the time of the study, and 10.8% were not receiving support but had done so in the past.

2.2. EMA protocol

The study used a signal-contingent variable-interval sampling scheme (c.f. van Roekel, Keijsers, and Chung 2019) where participants received prompts at four different times of the day, aiming to capture maximum variability in settings and activities. Measurement burden was a prominent consideration, and in line with other research conducted during the COVID-19 period (e.g. Fried, Papanikolaou, and Epskamp 2021) a timeframe of 7 days was adopted for data collection. All assessments (across baseline, EMA, and exit survey phases) were comprised of a mix of items that had either been abstracted from pre-existing validated scales, developed for the study, or adapted from items used in other EMA youth coping studies. Compliance monitoring and follow-up with participants (at day 3 and day 5 of the study), and the use of a monetary inventive (a 50 euro gift voucher upon attaining a 75% assessment completion level) were implemented to boost compliance. Responses to the exit survey indicated that the young people largely found participation in the study to be positive, (Mean rating = 71.68, SD = 19.44, out of a maximum possible score of 100 on a scale from ‘very negative’ to ‘very positive’).

2.3. Procedure

Data collection took place in May 2021, at which stage the participants had been back in physical school for approximately 6 weeks. All data collection instruments were developed through, and housed on, the Redcap platform.¹ After parental consent was received, online training was provided before the links to the young person baseline survey and the parent/guardian baseline surveys were distributed contemporaneously. The daily EMA assessments were initiated approximately one week after the completion of the baseline surveys. Text messages containing assessment links were delivered to participants’ personal mobiles at random time points within four time intervals during the day (7.30 to 9am; 12-1pm; 5 to 6pm; and 8 to 9pm). Participants were asked to respond to each signal immediately, unless they were unable to use their mobiles in class/school. A single reminder text was automatically issued 15 min after the signal in the case of non-response. On day 8, participants were asked to complete an online exit survey involving reflections on their experiences over the monitoring period.

2.4. Study measures

The baseline measure focused on sociodemographic factors, psychological wellbeing, perceived social support, and coping. Current psychological wellbeing was assessed using the World Health Organisation Well-Being Index (WHO-5; WHO 1998), and parent and youth versions of the Strengths and Difficulties Questionnaire (SDQ; Goodman 1997). Perceived social support was measured using the Social Provisions Scale Child Version (Cutrona 1989; Dolan 2006), whereas baseline coping was explored using the Coping Strategies Inventory, Short Form (CSI-SF; Addison et al. 2007), which assesses engagement and disengagement in relation to both problem-focused and emotion-focused strategies.

The EMA measures centred on learning engagement, perceived social support, momentary affect, experience of and coping with positive and negative events, and perceived control. Momentary perceived social support was measured using a current time-framed version of the SPS scale, while the Shortened Positive and Negative Affect Schedule for Children (Ebesutani et al. 2012) was used to explore momentary positive affect (PA) and negative affect (NA). Participants’ experiences of recent positive and negative events were explored using three items on occurrence, intensity, and associated affect (c.f.Uink et al. 2018), together with one open-ended item on responses to the event. Perceived control was measured through a single item which was included in the final of the EMA daily assessments.

A short exit survey was administered to participants the day after the final EMA assessments. This asked participants to reflect on the preceding 7-day EMA period with respect to overall perceived support, types and sources of support received, and overall perceived coping. The exit survey also measured factors that could potentially influence compliance and response validity such as effort involved in responding to the daily text messages, and perceived comprehensibility of the questions used (van Roekel, Keijsers, and Chung 2019). Responses to these items indicated good levels of understanding of the EMA questions, M = 80.79, SD = 19.64 (out of a maximum possible score of 100), and an acceptable level of perceived effort of response, M = 61.61, SD = 23.30 (out of a maximum possible score of 100).

2.5. Analysis

Descriptive statistics and inferential statistics [e.g. independent samples T-tests, Pearson Product Moment Correlations, Analysis of Variance (ANOVA), and Multiple Regressions, using IBM SPSS Version 25], were utilised to examine the normative quantitative data and to examine differences across sub-groups. It should be noted that only significant effects for subgroup analyses are reported in the results section. The qualitative responses to the open-ended questions were analysed using a two-step process, depending on the nature of the question (i.e. nominal data items or unstructured text). A content analysis identified the type and frequency of response items (e.g. source of support received, negative/positive events experienced, reactions to negative/positive events). A thematic analysis was then used to group the responses into categories and to develop overarching themes.

3. Results

3.1. Baseline levels of psychological wellbeing

The mean score for psychological wellbeing, as indexed by the WHO-5, was just above the midpoint of the scale, at 12.98 (SD = 4.56) [out of a possible maximum of 25]. The majority of the participants were classified the ‘close to average’ range for psychological adjustment, based on scores for total difficulties on the self-report SDQ (70.7%) and the parent-report SDQ (62.2%).

3.2. Daily affective experience

Mean ratings for mood were calculated for each day so as to account for participant non-response, and the mean scores were summed. As shown in Table 1, the highest ratings were for ‘happy’ and ‘relaxed’, whereas the lowest ratings were for ‘angry’ and ‘afraid’. A statistically significant difference was found between total mean PA and total mean NA for the whole sample, t(61) = 12.51, p < .001, d = 2.50, with higher ratings for total mean PA (M = 19.57, SD = 3.75) than for total mean NA (M = 10.92, SD = 3.14). The mean PA score for the sample was slightly above the mid-point (17.5) of the scale.

Table 1. Whole sample means (+ SD) for affect indicators across the 7-day EMA monitoring period.

Variable	Sample Mean	SD	Observed range*
Total mean sad score	10.51	3.14	7-18.83
Total mean angry score	9.43	2.47	7-16.83
Total mean happy score	22.42	4.67	7–31
Total mean worried score	13.60	5.46	7-29.75
Total mean excited score	15.07	4.99	7–21
Total mean afraid score	9.88	3.99	7-19.50
Total mean relaxed score	21.08	5.24	10.5-33.08
Total mean PA	19.57	3.75	9.25-18.56
Total mean NA	10.92	3.14	7-20.75

*The possible range for each mood indicator was 0-35.

A multiple linear regression was conducted to examine whether sociodemographic variables and baseline levels of wellbeing, coping and social support explained a significant proportion of the variance in total mean PA. It was found that a model including quantity of sibling support, B = .47, p = .002, psychological wellbeing (as measured by the WHO-5), B = .25 p = .013, and 5th year in school (dummy coded variable), B = −2.19, p = .018, significantly predicted total mean PA, F (3, 59) = 11.58, p < .001, explaining 35% of the variance. Higher levels of quantity of sibling support and higher levels of psychological wellbeing were associated with higher levels of total mean PA. However, being in 5th year (relative to other school years) was associated with lower levels of total mean PA. Regression analyses indicated that the only significant predictors of total mean NA were self-reported total difficulties (measured using the SDQ), B = .23, p = .01, being in 3^rd year in school, B = 4.63, p = .004, and being in TY at school, B = −2.14, p = .011, F (3, 57) = 12.29, p < .001, explaining 36% of the variance. Higher levels of self-reported mean NA were associated with higher levels of self-reported total difficulties and being in 3^rd year, whereas lower levels of mean NA were associated with being in TY at school.

A significant difference was found between occurrences of positive events and negative events over the monitoring period, t(61) = 6.32, p < .001, d = .91, with the mean number of occurrences of positive events (M = 1.29, SD = .99) being higher than that of negative events (M = 0.55, SD = .58). As shown in Table 2, qualitative content analysis indicated that the top five most frequently referenced positive events were face-to-face interactions with friends, outdoors/sports/extracurricular activities, family time, school success, and special foods/treats. Face-to-face interactions with friends often also involved concurrent participation in outdoor activities/sports or concurrent sharing of special foods/treats. Responses to positive events typically involved interactional responses such as expressing appreciation or emotional reactions such as feeling happy or relaxed. A strong theme in the open-ended responses was the welcomed sense of returning to familiar, and strongly missed, routines, environments, interactions and experiences:

Being back in the band room for drum lesson. Some parts of it I struggles with, but it was great to be back somewhere that I haven't been since before Covid. It was something familiar that I missed (Participant 54).

I got to go to the beach with my friends for the first time this year (Participant 77).

Table 2. Types of positive events mentioned by participants across the 7-day monitoring period. (N = 78 responses)

Type of Positive Event	Frequency of Mentions
Physical/face-to-face interaction with friends	49
Outdoors/sports/extracurricular activities	46
Family time	41
School success	31
Special food/treats	30
School-related events	22
Activities alone	22
Skipping school	10
Compliments	9
Other	9
Online contact with friends	8
Money/gifts	8
Graduation/birthday	7
Applying for job	7
Receiving mail	6
Shopping	5
Haircut	5
Pet	4
School tour	4
Driving test	3

3.3. Coping with negative events

The baseline measure of episodic coping (i.e. the CSI-SF) indicated that a variety of different coping strategies were habitually utilised by the participating young people. However, as shown in Table 3, disengagement strategies, particularly emotional disengagement strategies, were more frequently used than engagement strategies, t(80) = −2.90, p = .005, d = .05. No significant sociodemographic subgroup differences were found for baseline coping engagement. However, for baseline coping disengagement, ANOVAs revealed significant main effects for clinical diagnostic status, F(2,64) = 4.50, p = .015, partial η² = .12, reflecting a significant difference between students with queried diagnoses and no diagnoses, mean difference = 9.27, p = .012, and for clinical mental health support, F(2,64) = 3.25, p = .045, partial η² = .09, reflecting a significant difference between students receiving support at the time of the study and students who had never received support, mean difference = −9.27, p = .012. Baseline coping disengagement was higher for students with queried diagnoses than no diagnoses, and for students receiving mental health support at the time of the study than those who had never received mental health support.

Table 3. Descriptives for coping strategies habitually used by the participating young people.

	Mean	SD	Range	Max. score*
Baseline coping engagement	18.90	3.79	11–30	32
Baseline coping disengagement	21.06	4.57	11–32	32
Baseline problem engagement	9.92	2.09	5–15	16
Baseline problem disengagement `	`9.93	2.95	4–16	16
Baseline emotional engagement	8.97	2.66	4–16	16
Baseline emotional disengagement	11.13	2.67	6–16	16

As shown in Table 4, qualitative analysis of real-time EMA responses indicated that the top five most frequently experienced negative events during the monitoring period were academic stressors, peer interaction challenges, conflict with parents, physical illness/medical intervention, and technology problems. A strong theme in the academic stressor responses was anxiety around exam performance; either anticipatory anxiety related to an upcoming in-school exam or the future state exams, or anxiety stemming from poorer than expected performance in a recent in-school exam. Another prominent theme in this category of responses was concerns about self-regulated learning and management of school workload. Academic concerns were also interwoven within multiple other categories including technology problems, and conflict with parents. Coping with academic stressors most typically involved either problem engagement strategies such as problem solving (e.g. ‘I tried my best and after I just tried to forget about it and try study harder’; Participant 67) or emotional engagement strategies such as venting emotions (e.g. ‘Cried’; Participant 79).

Table 4. Types of negative events referenced by participants across the 7-day monitoring period. (N = 125 responses)

Type of Negative Event	Frequency of References
Academic stressors	29
Peer interaction challenges	15
Conflict with parents	12
Physical illness/medical intervention	10
Technology problems	9
Negative events related to hobbies/sports	7
Commuting stressors	6
Parental discord	5
Negative subjective wellbeing	4
Conflict with siblings	4
Conflict with other adults in the community	3
Issues related to current/future employment	3
Appearance-related concerns	3
Food and weight-related factors	3
Household/family responsibilities	3
Pet-related negative events	3
Sleep-related factors	2
Events related to bereavement	2
Conflict with others (unspecified)	1
Cancelled plans (unspecified)	1

In relation to peer interaction stressors (the second most frequently cited negative event), participants spoke about a range of negative experiences with other young people including betrayals of trust, arguments, loneliness, exclusion, dismissal/invalidation, and ambiguity:

Received a text from someone (who I don't know personally) accusing me not being supportive to someone else (who they do not know personally) randomly for no reason which left me confused. (Participant 63).

In contrast to coping with academic stressors, the most prominent approach to coping with peer interaction stressors was problem disengagement strategies such as problem avoidance: ‘walked away’ (Participant 21); ‘nothing’ (Participant 56); ‘ignored it and her’ (Participant 42).

Another coping-related measure was an item in the exit survey which asked participants to retrospectively rate their coping over the 7-day EMA monitoring period. The mean rating was 60.03 (SD = 21.35) out of a possible maximum rating of 100. ANOVAs exploring subgroup differences revealed a significant main effect for clinical mental health support status, F (2, 53) = 5.26, p = .008, partial η² = .17. This reflected the fact that self-reported coping was lower for students who were receiving mental health support at the time of the study and those who were not at that time and had never received mental health support, mean difference = −30.12, p = .011. Regression analysis indicated that a model including school year significantly predicted perceived coping, F (6, 49) = 4.43, p = .001, explaining 27.2% of the variance, after controlling for clinical mental health status (dummy coded variable). Being in 5th year, B = −14.59, p = .043, and being in 6^th year (marginally significant effect; B = −14.33, p = .058) were negative predictors of perceived coping.

4. Discussion and conclusion

An intensive EMA methodology was utilised in this study to explore the real-time wellbeing and coping of young people in the Republic of Ireland during the transitional period of the return to in-person education during the 2020–2021 academic year. It was found that trends in affective wellbeing were generally quite positive for the participating young people. Firstly, PA (which was found to be at moderate levels) was more prominent than NA, with ‘happy’ and ‘relaxed’ being the strongest feelings experienced by the participants across the week of monitoring. Research has indicated that levels of pleasurable emotions are salient because over time, they may moderate stress and coping appraisals, reduce stress reactivity, and expedite stress recovery (Pressman, Jenkins, and Moskowitz 2019, 637). Modelling analyses indicated that a small number of baseline and demographic variables were predictive of average affect. Specifically, higher levels of mean PA were predicted by higher levels of baseline psychological wellbeing, and higher levels of perceived quantity of sibling support. The latter finding suggests that perceived habitual support from siblings may be an important protective factor in relation to affective experience for the young people who participated in the study. Indeed, whilst peer relationships are most often the focus of research during the period of adolescence, sibling relationships also appear to have significant implications for youth adjustment (Kim et al. 2007; Soli, McHale, and Feinberg 2009). Perhaps spending more time at home during the preceding lockdown period improved sibling relationships, making sibling support a more important protective factor for the participating young people (Mantovani et al. 2021). In addition, higher levels of mean NA were predicted by lower levels of baseline (self-reported) psychological adjustment. Further, relative to other year groups, students in 3^rd year were found to be more susceptible to higher levels of NA, students in TY were more likely to experience lower NA, and students in 5th year were found to be more susceptible to lower levels of PA. These differential findings for PA and NA are not surprising considering that both constructs are statistically separable; that is, PA is believed to be more than simply the absence of NA and vice versa (Pressman, Jenkins, and Moskowitz 2019).

A specific objective of the research was to examine young people’s coping during the return to in-person education. Baseline results indicated that disengagement strategies, particularly emotional disengagement strategies, featured more prominently than engagement strategies in self-reports of typical coping behaviours, cohering with research with youth conducted during the COVID period (e.g. Dewa et al. 2021; Domínguez-Álvarez et al. 2020). Disengagement coping was particularly high among participants with queried mental health diagnoses and for participants currently receiving mental health support. Disengagement coping has traditionally been regarded as maladaptive due to associations with poorer psychological outcomes over the long-term relative to engagement coping (Compas et al. 2017). Notwithstanding, it has been proposed that the use of disengagement coping, especially self-distraction, may actually be adaptive in highly uncontrollable situations/contexts (Zimmer-Gembeck and Skinner 2016; Fear et al. 2009), and indeed, some research has identified this type of coping as a protective factor for distress among youth during the pandemic (e.g. Hsieh et al. 2021). However, a continuing reliance on disengagement coping strategies in situations that are no longer uncontrollable or unpredictable may be problematic for longer-term mental health and wellbeing during the non-acute phases of the pandemic.

The real-time coping data indicated that the most frequent negative events experienced by the participants were academic stressors, most notably, anxiety around performance on in-school tests, and anticipatory anxiety in relation to the future state exams. This aligns with a large body of research pointing to the substantial stress triggered by high-stakes testing (Pascoe, Hetrick, and Parker 2020). Another prominent category of negative event, albeit appreciably less frequent, was stressors in peer interactions or peer relationships. The participating young people tended to cope with peer challenges through maladaptive (that is, maladaptive in predictable and controllable situations) problem-focused disengagement strategies such as problem avoidance. Contrastingly, there were trends of a predominance of coping engagement strategies in response to academic stressors. These findings align with previous empirical evidence of differential patterns of coping strategies in response to family stressors, peer stressors, and school stressors (Herres 2015). It is important to note that less adaptive emotional engagement strategies such as externalising emotions made up a large part of participants’ coping engagement responses to academic stressors. Whilst these strategies do represent an approach orientation to coping, they are less effective than problem-focused coping strategies for long-term youth wellbeing and coping (Evans et al. 2015). Looking back over the EMA week, participants rated their coping to be at a moderate level (as indicated by a score of 60 out of a maximum possible 100). There were subgroup differences in perceived coping, with lower levels being reported by students who were receiving mental health support at the time of the study. In addition, self-reported coping was lower among 5th year and 6^th year students, even after controlling for mental health support status. However, since only a single-item measure of perceived coping was utilised (as part of the exit survey), a conservative approach to interpretation and extrapolation is needed here.

The findings of this study should be interpreted in the context of methodological limitations including the measures used, the sample reached, and the study timeline. First, whilst an attempt was made to use validated measures, some of the EMA daily measures were necessarily limited (e.g. the single-item perceived coping measure) due to the imperative of minimising measurement burden in the context of repeated assessment. Thus, it is advisable that future research utilise a differentiated and multi-faceted measure of perceived coping so as to facilitate a more nuanced and fine-grained analysis of perceptions of coping, that would also allow for a comparison of perceived versus enacted real-time coping. Second, despite concerted efforts at both community and school levels to recruit diverse participants (particularly those from marginalised communities), the final self-selected sample was very ethnically and socioeconomically homogenous, and only a minority of participants in the final sample would be classified as ‘marginalised’ using traditional SES indicators. Accumulating research evidence suggests that the academic progress and potentially, the psychological wellbeing, of students from marginalised backgrounds have been disproportionately impacted by the pandemic and its associated counter-measures (e.g. Dewa et al. 2021; Engzell, Frey, and Verhagen 2021). Consequently, it is strongly recommended that future research examining student re-engagement following the return to in-person schooling should specifically target educationally at-risk students (Russomanno, Patterson, and Tree 2019). Third, a non-probability convenience sampling strategy was utilised which inevitably limits the generalisability of the findings to the general population of second-level students in Ireland. Finally, whilst our study presents a window into young people’s ‘life as it is lived’ (Hektner, Schmidt, and Csikszentmihalyi 2007, 8) during the transitional period of the return to in-person education, additional follow-ups would allow for the examination of continuity or discontinuity in coping and wellbeing further into the readjustment period.

The significance of the study findings lies predominantly in what they suggest for supporting students during the transitional period following the pandemic-related disruptions to education. The participants as a whole were found to have relatively good levels of affective wellbeing and perceived coping, aligning with emerging evidence of psychological resilience to the adversities of the pandemic among young people (e.g. Beames et al. 2021), and corresponding with findings from the disaster literature that recovery is the normative pattern among children and young people (Galatzer-Levy, Huang, and Bonanno 2018). However, a pattern of reliance on less adaptive coping strategies such as problem avoidance or emotional venting was observed across all participants, most notably in response to peer stressors, and to a lesser extent, to academic stressors, which were found to be the most prominent negative events for the participating young people. This suggests that second-level students, particularly those in the final years of secondary school, may benefit from more specific support in relation to understanding and adaptively coping with the stress stemming from workload demands and exam-related anxiety, which is a particularly prominent feature of their everyday lives (Wuthrich, Jagiello, and Azzi 2020). This support could take the form of increased awareness of stress by school personnel, stress-focused wellbeing classes, and more psychosocial support through ‘mental health check-ins’, as articulated by one of the participants in this study. This need may be particularly acute for students in exam or exam-adjacent years, as suggested by the present findings of lower levels of affective wellbeing and perceived coping among 3^rd, 5th and 6th year students. It is our view that additional supports for some students, especially educationally at-risk students, are critical in an academic year which will see the reinstatement of traditional state examinations following a hiatus of two years; the anticipation and experience of which will likely function as a significant stressor for young people. We recognise that the provision of academic supports to students who have been most affected by the disruption to in-person education is a current policy concern, as evidenced by the fast-tracking of national-level funding to primary and second-level schools through the CLASS (COVID-19 Learning and Support Scheme). In this context, it is hoped that the findings reported in this paper may be instructive to second-level school leaders and educators in developing and implementing CLASS supports during the current and next academic years.

Furthermore, the findings of this study suggest that another cohort of students may be in need of specific additional support due to pre-existing vulnerabilities that could compound the challenges of returning to in-person second-level education. In this study, students with pre-existing psychological adjustment difficulties, students with mental health diagnoses (queried or confirmed), and students who were receiving clinical mental health support were more likely to experience lower levels of affective wellbeing, to have lower perceptions of coping, and to report higher levels of disengagement coping, which as previously mooted, may only be adaptive and protective in short-term, highly uncontrollable situations. Accordingly, it is recommended that schools should strive to prioritise the reestablishment of a sense of psychological safety and wellbeing, with a particular focus on students with mental health needs. In the words of one of the participating young people: ‘emotional support should be provided because mental health is more important then [sic] homework’.

A useful framework for supporting students’ readjustment to in-person education is Hobfoll et al. (2007) model of post-disaster psychosocial support, and indeed this has formed the basis of guidance to schools from the National Educational Psychological Service (NEPS) in Ireland during the period of the pandemic. We believe that particular attention should be focused on the ‘sense of connectedness’ element of this framework, because as highlighted by Hobfoll et al. (2007), social connectedness ‘provides opportunities for a range of social support activities including practical problem solving, emotional understanding and acceptance, normalisation of reactions and experiences, and mutual instruction about coping’ (296). A compelling argument arising from this framework is that social interaction with peers should be explicitly planned for and facilitated so as to maximise psychosocial recovery. A commitment in this domain could take the form of planning and resourcing structured support for meaningful social interaction through organised clubs during break-times. Another option could be to increase the quantity (and quality) of peer learning and active learning activities within classes. Overall, it is recommended that in the short to medium-term, second-level schools should strive to prioritise the psychological recovery and resilience of their students through an emphasis on re-establishing and consolidating a sense of student connectedness.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Irish Research Council/Health Research Boad: [Grant Number COV19-2020-037].

Notes on contributors

Niamh Flynn

Dr Niamh Flynn is an Educational Psychologist, and lecturer in educational psychology, inclusive education, and research methods in the School of Education at NUI Galway. Her research interests are focused on inclusive teaching, social-emotional learning, and student and teacher wellbeing.

Clíona Murray

Dr Clíona Murray is a lecturer at the School of Education, NUI Galway, where she teaches sociology of education and research methods. Her research interests include education policy studies, social inclusion, education for sustainable development, and narrative inquiry.

Cormac Forkan

Dr Cormac Forkan is a lecturer in the School of Political Science and Sociology, and Senior Researcher in the UNESCO Child and Family Research Centre at NUI Galway. His research interests include youth work and adolescent development, youth social support, and use of real-time data collection methods with young people.

Carmen Kealy

Dr Carmen Kealy is a Postdoctoral Researcher within the UNESCO Child and Family Research Centre at NUI Galway. She holds a Doctorate in Sociology and a BA in Political Science, Sociology and Psychology from NUI Galway. Her research interests include parenting and migration but also the investigation of inequality experienced by marginalised groups.

Notes

1 RedCap is a secure web-based application for building and managing online surveys and databases. It was originally developed at Vanderbilt University and the RedCap Consortium is now composed of thousands of active institutional and not-for-profit organisational partners (https://www.project-redcap.org)