The COV-ED Survey: exploring the impact of learning and teaching from home on parent/carers’ and teachers’ mental health and wellbeing during COVID-19 lockdown

An online survey of parents/carers and teachers of adolescents aged 11–15 years in England. The survey consisted of two sections; data of interest were:

The survey consisted of two sections:

Data was collected via the online Qualtrics platform and was downloaded to a secure server at Warwick Clinical Trials Unit (WCTU). Following download, all data on Qualtrics platform was deleted. No personal data was collected or stored on the WCTU database or used for unrelated purposes. Data was only accessible to the research team.

All analyses were undertaken using Stata 16 (StataCorp 2019).

Information on demographics, practicalities of teaching from home and supporting a child with learning from home were presented using basic descriptive statistics. Continuous data was summarised using mean and standard deviation (if normally distributed) and the median and interquartile range (if skewed). Categorical data was reported using the number and percentage of participants in each category. Multiple linear regressions were utilised to explore the effect of demographic variables and practical aspects of teaching from home on the mental wellbeing of teachers, and, similarly, to explore the effect of demographic variables and practical aspects of supporting a child with learning from home on the mental wellbeing of parents/carers. The dependent variable in the regression models was WEMWBS score with demographics and practicalities as independent variables. The normality of WEMWBS score was assessed graphically (through histogram plots). Diagnostic plots were also used to verify that the distributional assumptions of the regression model had been met. To reduce the number of variables to be included in the two regression models, two researchers (CC, NDV) identified the subset of questions that they thought would be most strongly related to mental wellbeing prior to seeing the data. Two questions in the teachers’ survey (Teacher Experience Q10 and Q11, Supplementary Table 1) regarding “feeling supported by the management” were thought to be very closely related to each other, so these were combined for inclusion in the regression models by adding together scores (full details are provided in Supplementary Table 3, Additional File 1). Similary, answers from the teachers’ survey on the two questions about “feeling supported by colleagues” (Teacher Experience Q12 and Q13, Supplementary Table 1) were combined (as described in Supplementary Table 3). Questions where respondents ‘mostly agreed’ or ‘strongly agreed’ were felt to be uninformative and unlikely to explain differences in outcome between participants and so were excluded from the regression models. Owing to small numbers, the last two categories of responses that were recorded using the five-level Likert scale (strongly agree, agree, neutral, disagree, strongly disagree) were combined (except for one in the parents survey, relating to the statement “Working from home at the same time as supporting my child with learning is easy to coordinate” where it was more appropriate to merge the first two categories). Variables considered for inclusion in the teachers regression model were gender, age category (< 45 years, ≥ 45 years), ethnicity (white, minority ethnicities), “having someone to support you with your mental health” (yes/no), “feeling supported by management” (high, moderate, low), “feeling supported by colleagues” (high, moderate, low), “having adequate resources”, “having adequate training”, “having a desk or workspace”, “feeling confident and capable about teaching from home”, “having students that are engaged in online teaching” and “having students who are coping well with online teaching”.

For the parents model, variables considered for inclusion in the regression model were sex, age category (< 45 years, ≥ 45 years), ethnicity (white, minority ethnicities), “having someone else to support their child with learning from home” (none, parent/grandparent/other family member), “having someone to support them with their mental health” (yes/no), “having a structured home schooling day”, “having been provided with adequate home learning resources by their child’s school”, “using other resources for learning from home”, “child having access to their own laptop/computer/tablet/smart phone”, “working from home at the same time”, “finding working from home at the same time as suporting their child with learning easy to co-ordinate”, “having regular contact with their child’s school and/or teacher”, “feeling supported by their childs teacher/school”, “having their child’s school/teacher set regular deadlines”, “feeling capable and confident in supporting their child with learning”, “feeling capable and confident using technology to support their child’s learning”,”having a child that is engaged with and enjoying learning from home””belief that their relationship with their child has improved whilst learning from home” and “belief that their child’s mental health has been affected by learning from home”. Univariable models were fitted and a parsimonious multivariable model built using forward selection. Robust cluster variance estimators were used for the parent/carer model, so that more than one child could be included per parent. Coefficients from the regression models are reported with associated p-values and 95% confidence intervals.

For both groups WEMWBS scores were grouped as follows [41], one standard deviation above the mean or more: above average (60–70); one standard deviation below the mean or more: below average (14–42); remainder: average (43–59). Below average scores indicate a high probability of clinical depression or anxiety.

A total of 501 participants completed the online survey. 138 teachers (Table 1) and 363 parent/carers (Table 2). 21 (15.2%) teachers were not teaching from home and 34 (10%) of parents/carers were not supporting a child with learning from home and were removed from the analysis. Our analysis, therefore, focused on 446 participants (117 teachers and 329 parent/carers). Demographic information is presented in Tables 1 and 2 respectively.

Teacher survey respondents were mostly female (78.6%), of white heritage (86.3%) and working in the West Midlands (84.6%). A large proportion (35%) were also supporting their own child (or children) with learning from home (teacher/parent/carer). Roughly one third of teachers (29.9%) reported mental wellbeing (WEMWEBS) scores that were below the average for the general UK population. This proportion did not differ substantially between those teachers who were also a parent/carer and those who were not.

Parent/carer participants were also predominantly female (85.4%), of white heritage (90%) and from the West Midlands (73%). A greater number (44.7%) reported below average mental wellbeing (as per WEMWBS score) compared with teachers. There was a marked diffference in mental wellbeing between the parents who were teachers and parents who were not, with 46.8% of parents who were not teachers reporting below average mental wellbeing compared to 31.1% of parents who were also teachers.

Overall, we found that teachers had coped well with remote teaching and had adapted to the novel and challenging situation, despite some having responsibility for supporting an average of 100 students, who were teaching remotely for 5 + hours per day and who did not have their own desk/workspace to teach from home. Overall, teachers reported having structured school days, were in regular contact with their students, colleagues and parents/carers and felt they had received adequate resources and training with which to successfully teach online. However, only half of them felt their students were actively engaged in online learning or were coping well. Student engagement in remote learning is a substantial challenge. Government guidelines suggest that engagement may be enhanced by ensuring regular communication with students, using a consistent digital platform for teaching and providing school laptops and/or printed resources for students experiencing difficulties accessing technology [42].

The average WEMWBS score for all teachers in the current sample was 46.27. This is lower than published pre-pandemic national indicators expected for the adult general population (mean = 49.9 [43]). International findings exploring teachers’ mental wellbeing have indeed suggested that teachers mental wellbeing has been negatively affected during the pandemic [44‐46]. However, the cross-sectional nature of such studies mean that it remains difficult to determine the true cause and effect of the pandemic on mental wellbeing or analyse findings over a period of time. Interestingly, however, the mean WEMWBS score for teachers in the current sample was found to be higher than that reported in the 2019 Teacher Wellbeing Index [47] for educational professionals in the West Midlands (mean = 43.6). This suggests that the majority of teachers who participated in our survey found the circumstances of the pandemic, including the challenges of teaching from home, had improved their mental wellbeing. All felt supported by friends or colleagues and almost two thirds felt supported by senior management and 72% by colleagues with regard to their mental wellbeing. These findings are in contrast with the Teacher Wellbeing Index in 2019 which reported that 43% of teachers, at that time, did not believe that their institutions properly supported employees who experienced problems with their mental wellbeing.

Below average mental wellbeing is indicative of mental health issues like depression and anxiety and is typically expected in 15% of the general population [41, 48]. The annual incidence of mental health issues in teachers, however, is 34% [47]. One third of teachers in our survey reported below average mental wellbeing, in keeping with pre-pandemic statistics and in line with those from our own teacher survey in 2017 [28] which showed that, pre-lockdown, teachers reported experiencing high levels of work-related stress, often due to time pressures and excessive workloads. Many failed to seek help due to stigmatic attitudes and fear of negative response by senior management and were failing to cope with extra responsibilities such as student mental wellbeing support. Some had considered leaving the teaching profession altogether, were using alcohol and tobacco to cope with stress, with only a small minority accessing psychological therapies for their problems.

Over half of parent/carers said they had a structured home-schooling day, had been provided with the necessary resources and felt confident and capable in supporting their child’s learning from home. The majority of parents/carers who took part in the survey were supporting only one child with their learning from home and this was often shared with a partner. Shared responsibility may have enabled a more equal distribution of responsibilities within the family, and whilst we cannot conclude this from our data, this may have reduced individual burden and increased parents/carers’ confidence and capability in supporting learning from home. These positive responses may also be a reflection of enhanced quality time spent between a parent/carer and adolescent during lockdown. Indeed, it is believed that the more parents/carers are actively involved in the education of their child, the more likely they are to succeed in the education system [49]. Our findings should be treated with caution, however. Parent/carers schooling more than one child from home, especially single parent/carers without the support of a partner may have a different story to tell with regard to supporting learning from home.

Over one third of parents/carers reported that their child was positively engaged in learning from home and that their relationship with their child had improved during this time. However, the majority were experiencing difficulties with engagement. Encouraging self-motivation is a significant part of the educational process, especially during adolescence, much of which is heightened by a sense of school membership which can enhance motivation [50]. Remote learning practice may have impacted on this sense of membership. Lessons may have become ‘lectures’ and influenced how intrinsically motivated and engaged an adolescent was.

Schools and teachers may have strived hard to maintain engagement and sense of school membership during school closures, for example, by setting regular deadlines, providing ongoing feedback and encouraging active participation during remote teaching sessions. However, families are fundamental in ensuring that these efforts are supported and reinforced at home. Parental engagement is an on-going challenge and schools are continually searching for ways to improve their relationships with families. A range of factors will impact on parental engagement including socio-economic status and education, but this challenge will have become magnified during the pandemic [51]. The level and severity of burden and stress for home-working parents/carers, in particular, will likely have determined level of parental engagement and have varied depending on the individual needs and profiles of families. However, our findings suggest a perceived lack of support for home-working parent/carers to help reduce their burden. Burden may have been reduced significantly if schools had implemented a ‘needs-led’ approach to home learning, and, rather than operating a ‘one size fits all’ strategy, devised ways of working collaboratively with families, to help reduce inequality, support parental mental wellbeing and better understand their needs and tailor support during the pandemic.

Mean WEMWBS score for all parents/carers in the current sample was 44.01 which is markedly lower than published pre-pandemic national indicators for the adult general population (mean = 49.9 [43]). The observed difference is unlikely to be due to self-selection of a high-risk sample because participants tended to have only one child in home schooling and to have partner support. It is more likely due to a combination of factors relating to the significant disruption caused by the current pandemic. Indeed, research suggests a decrease in mental wellbeing when compared to 2019 levels, particularly in young people, women and those living in deprived areas [52]. Parents/carers who were teachers by profession fared better than those who were not. This is likely to reflect the presence of pre-existing strategies, skills and confidence of parent/carer teaching staff or that they were less worried about what a child needed to learn and better able to prioritize learning targets, which may have better prepared them for tackling and implementing learning from home.

Around 60% of the UK population are working from home due to the pandemic [14] and, whilst this could be argued to be preferable during such difficult circumstances, our survey suggests that home-working parents/carers struggled to co-ordinate this with their child’s educational and emotional needs leading to a negative impact on their mental wellbeing. Our findings support previous research which reports a link between level of caregiver burden and poor mental wellbeing [16]. The increase in burden may not only have impacted on parents’ mental wellbeing but also their parenting behaviours [17, 18] and their children’s cognitive and socioemotional development [53‐55].

Poor mental wellbeing was also observed in those parents/carers who had little or no support. Research suggests greater levels of stress in single parent/carers supporting learning from home during the pandemic compared with those living with a partner [56]. Previous studies exploring the impact of past pandemics have also shown a significant association between lack of social support and severity of mental health problems [57, 58]. Social support, whether perceived or actively received, can serve to operate as a powerful buffer, protecting individuals against the harmful impact of stressful situations [59]. There is little doubt that the COVID-19 pandemic has caused immense worry and fear [60] at a time when, paradoxically, social boundaries have been amplified and interpersonal connections and access to social support has been restricted. Problematic and extended social boundaries require innovative solutions. Greater provision and utilisation of digital social support resources, in the absence of face-to-face contact and/or therapeutic care, may have offered much needed succour for families, during this time. However, digital resources are not a ‘cure-all’ and do not address issues such as loneliness and isolation during these challenging times.

Parents/carers whose child had little or no access to electronic devices were also observed to have poorer mental wellbeing. This may be a marker of disadvantage, however, and not solely associated with restricted access. Children and young people from higher income families have been found to receive greater support with remote learning compared to those from poorer families; higher income families devoting 30% more time to home learning and having better access to a range of remote learning resources [61]. Insufficient provision of electronic devices for students has been highlighted by The Children’s Commissioner for England who report that, in August 2020, around 9% of families were struggling with access to technology [62]. The Department for Education is now working in collaboration with several mobile network providers to help ensure that all children and young people have sufficient internet access and, to date, have provided in excess of 1.3 million disadvantaged young people and children with a laptop or tablet to enable them to fully engage with remote learning [63].

COVID related worries, difficulties with remote learning and family conflict during the pandemic have all been suggested to impact on adolescent mental health [8], with greater impact felt by those already experiencing difficulties [9]. 80% of young people with a pre-existing mental health issue believe the pandemic had made their mental health condition worse [10]. Our findings revealed that those parents/carers who felt that their child’s mental health had been affected by learning from home were also more likely to have poor mental health wellbeing. Such child-related concerns are not unexpected. Families are intrinsically motivated to attend to their child’s physical and mental wellbeing but may often lack the capacity, knowledge and awareness to do so [64]. This may lead to increased feelings of powerlessness and high levels of anxiety and stress; parental mental illness is the strongest predictor of child mental illness [65].

Despite our aspiration of recruiting a diverse population, we were unable to translate the survey into other languages due to limited resources. This meant that the majority of respondents were of white heritage, female and from the West Midlands area. The gender and cultural bias found in the present study may be a reflection of that observed in typical survey completion. Whilst women and those of white heritage are more likely to participate in surveys [66‐70] the pandemic has highlighted an important gender divide, with women taking most responsibility for childcare; carrying out an average of two-thirds more per day than men [71]. Survey topic relevance [68], and response burden [72] may have also played a role in determining those who took part, especially during such stressful circumstances. The survey was ‘live’ for a limited period, so our data only reflects a snapshot of the experiences of our particular sample of teachers, parents/carers during lockdown in England and their specific views at that time. Given the small sample size of teachers who responded to the survey, our findings related to teacher mental wellbeing should be treated with caution. Furthermore, there may be bias in our findings; teachers with a particularly positive outlook on teaching from home may have been more likely to respond. Similarly, parents/carers who found supporting their child at home especially difficult may be over-represented in our survey data. It is also possible that our survey may have missed key social and economic wellbeing factors that may have impacted on the experience of supporting teaching and learning from home. Given these limitations, our findings should not be generalised to the wider population and should be interpreted with caution. They are not an indication of the effects of supporting and teaching from home before and after the COVID-19 pandemic.

Nevertheless, the successful recruitment of 329 parents/carers and 117 teachers has provided us with useful preliminary insights into their experiences of teaching and learning from home. Such findings will serve as a foundation for further exploration and research associated with the COVID-19 pandemic.