1. Introduction
On 11 March 2020, the World Health Organization declared the COVID-19 virus a pandemic [
1]. To reduce transmission of the virus, over 90% of countries worldwide implemented school closures, forcing millions of children to continue their education through mandatory homeschooling [
2]. Mandatory homeschooling refers to involuntary at-home learning resulting from mandated school closures, in which schooling responsibility is largely transferred onto parents and caregivers [
3]. The COVID-19 pandemic has had negative impacts on parent and child mental health [
4,
5,
6,
7], which many studies have linked to mandatory homeschooling [
3,
8,
9]. By contrast, a minority of studies have found that the effects of mandatory homeschooling on mental health were non-significant or beneficial [
10,
11].
Chronic (long-term) activation of stress response has been linked to greater adverse mental health impacts in adults and children than acute (short-lasting) stress [
12,
13,
14,
15]. These findings suggest that families engaged in long-term homeschooling may have been at greater risk of adverse mental health effects due to compounding (progressively worsening) effects over time [
12,
16]. However, existing research has not captured compounding mental health effects of mandatory homeschooling over an extended period, which may explain conflicting results across studies. Consequently, it is unknown whether mandatory homeschooling has compounding adverse mental health effects over time or if long-term mandatory homeschoolers are at greater risk.
Alongside the duration of mandatory homeschooling, several other factors may have influenced its effects on parent and child mental health. First, receiving inadequate educational support from schools during mandatary homeschooling may have placed additional academic strain on families, exacerbating mental health harm [
17,
18,
19]. Second, since use of the internet was often necessary for mandatory homeschooling, families with poor internet access may have experienced greater mental health harm [
20,
21]. Finally, pre- and peri-pandemic research has linked parental unemployment to adverse mental health effects in families, which may point to parent unemployment as another risk factor during mandatory homeschooling [
4,
22,
23,
24].
The present study investigates whether children’s cumulative time engaged in mandatory homeschooling has compounding adverse effects on family mental health, looking specifically at parent depression, anxiety, and stress, and child internalizing symptoms (emotional and peer problems) and externalizing symptoms (behavioral and hyperactivity problems). It also explores whether educational supports from schools, child internet access or parent employment status moderate this relationship. Our first aim was to identify which families are at greatest risk of adverse mental health effects during mandatory homeschooling, so they can be given access to the supports they need to recover. Our second aim was to identify educational supports that may mitigate these adverse effects, to assist schools in minimizing mental health harm for families during future mandates.
1.1. Mandatory Homeschooling during COVID-19
During mandatory homeschooling, educational and childcare responsibilities usually fulfilled by schools were shifted onto parents, who often lacked adequate training and resources [
25]. One study of mandated homeschooling during COVID-19 found that children spent 30% of their learning time in contact with their parents, while just under 14% was spent in contact with teachers [
18]. In-person school provides children with academic support, structured routines, and access to mental health and peer supports, which are beneficial to children’s mental health [
26,
27]. However, during the COVID-19 school closures, children were cut off from many of these mental health protections [
19,
27]. Meanwhile, many parents were forced to navigate their newfound schooling responsibilities alongside other stressors, including employment insecurity, financial strain, and social inequities [
4,
7].
1.2. Mandatory Homeschooling and Parent and Child Mental Health
Most research examining mandatory homeschooling during the COVID-19 pandemic has linked it to declines in parent and child mental health. Kishida et al. [
8] found that full school closures the previous week were linked to greater parent and child mental health harm compared to partial school closures, which in turn were linked to greater harm than full in-person learning. Schmidt et al. [
9] found that parents and children experienced poorer affect on mandatory homeschooling days, while Deacon et al. [
1] found that parents who spent more hours per week homeschooling their child due to COVID-19 mandates experienced higher anxiety, depression, and traumatic stress. Interestingly, a minority of studies did not support these findings. DesRoches [
10] found that mandatory homeschooling had non-significant impacts on child mental health, while Monnier [
11] found that more hours per day spent homeschooling were associated with lower externalizing symptoms in children. This lack of consensus on the effects of mandatory homeschooling on child mental health may be explained by how these studies conceptualized and measured mandatory homeschooling.
1.3. Homeschooling Time as a Continuous Measure
The duration of school closures during the COVID-19 pandemic varied by location and school board [
28]. In Canada, school closure duration ranged from a minimum of 50 school days in British Columbia, to a minimum of 135 school days in Ontario [
29]. Past research has studied mandatory homeschooling over short-term periods, often using non-continuous measures that do not capture variations in duration [
3,
8,
9,
10,
11]. For example, some studies have measured mandatory homeschooling as a dichotomous yes/no variables [
9,
10], while others have measured it as the amount of time spent homeschooling per week [
3,
8] or day [
11]. Consequently, these studies are unable to differentiate between families who experienced long-term versus short-term mandatory homeschooling.
Stressors are actual or perceived threats to an organism that provoke physiological, emotional and/or behavioral reactions from the body. These reactions, collectively called the stress response, prepare the body to cope with stressors [
16]. Chronic (long-term) activation of the stress response has been linked to mental health problems in children and adults, including depression, anxiety and behavioral problems, which compound over time [
12,
13,
14,
15].
Based on findings that chronic stress causes compounding mental health harm over time, it is possible that families engaged in long-term mandatory homeschooling experienced greater mental health harm than those engaged in short-term homeschooling. Further, short-term homeschooling may have had positive mental health impacts on some families by providing increased opportunities for family support and removing children from potential school stressors such as bullying [
30,
31]. Because past research used non-continuous measures of mandatory homeschooling, the findings could not capture potential compounding mental health effects over time; this may account for the conflicting results on the effects of mandatory homeschooling. To our knowledge, the present study is the first to measure time spent engaging in mandated homeschooling continuously over an extended period. Consequently, it is unknown whether mandatory homeschooling has compounding negative mental health effects that place long-term homeschoolers at greater risk.
1.4. Educational Support from Schools
Throughout the COVID-19 pandemic, many countries provided minimal national guidance about which educational support should be provided by schools [
32]. Consequently, there was wide variability in the support children received from their schools or governments [
20,
32]. Many families attributed mandatory homeschooling challenges during the COVID-19 pandemic to a lack of educational resources and support from schools [
17,
33], the latter of which was associated with greater parental distress [
34]. Given these findings, it is possible that receiving educational support from school may have benefited mental health during mandatory homeschooling.
Many children received online classes from schools during mandated homeschooling [
20]. Pre-recorded online classes may have provided students with scheduling flexibility and higher-quality presentations, while live online classes may have provided opportunities for live student–teacher interaction and questions [
35,
36]. Live online classes may have been especially beneficial for mental health, given findings that children who interacted more with their teacher found online learning more effective and were more satisfied with their experience [
19]; however, not all students had access to online classes during mandated homeschooling [
18,
20]. It is possible that children struggled more when they did not receive online classes, especially live ones, resulting in greater mental health harm for families during mandatory homeschooling.
Many students also received other digital or physical resources such as assignments, worksheets or videos, which can be conceptualized as home learning packs [
20,
32,
37,
38]. Home learning packs may have likewise contributed to improved mental health during mandatory homeschooling by providing children with a flexible opportunity to supplement and apply knowledge they gained in class. Pre-pandemic research suggests that these opportunities may have improved children’s academic experiences, suggesting that home learning packs may have decreased academic strain on families during mandatory homeschooling [
39,
40]. This study examines live and pre-recorded online classes, as well as home learning packs, as educational supports from schools that may have mitigated adverse mental health effects of mandatory homeschooling.
1.5. Child Internet Connection
While specific school supports varied during mandated homeschooling, most children had an online component to education during the pandemic [
20], making reliable internet connectivity a crucial resource. Unreliable internet connection has been linked to distress and dissatisfaction during mandatory homeschooling [
17,
41], which may suggest that it contributed to families’ mental health harm. Despite its importance, many children struggled to access reliable internet connection during the pandemic in Canada and the United States, with some children relying entirely on public Wi-Fi to complete homework [
21,
42]. Alternatively, it is possible that reliable internet access put children at greater risk of problematic internet use, which increased during the pandemic and has been linked to poor mental health in children [
43,
44]. Together, the prevalence of unreliable internet access and its potential to harm (or benefit) family mental health during mandatory homeschooling prompted us to investigate its effects in the present study.
1.6. Parent Employment
Many parents either lost their jobs during the COVID-19 pandemic or reduced their working hours to care for their children [
45]. Findings are mixed regarding the effects of parent employment on family mental health during COVID-19. Pre-pandemic research has linked self- or spousal unemployment to stress, anxiety and depression in adults [
23,
46,
47]. Some pre-pandemic studies have linked parental unemployment to poorer mental health outcomes for children due to family financial strain and reduced parental well-being [
24], while others have found beneficial impacts on younger children due to increased availability of parent–child quality time [
48]. In line with many pre-pandemic findings, parent unemployment during the pandemic has been associated with increased depression in parents [
5] and increased negative affect in children [
49]. Employed parents, however, reported that balancing the responsibility of their child’s education with work was a major source of stress, and many felt unable to provide their child with adequate school support [
4,
7,
22]. These findings prompted us to explore parent employment status as a factor that may have influenced family mental health during mandatory homeschooling.
1.7. Current Study
The goal of the present study is to assess the mental health impact of children’s time spent engaged in mandated homeschooling during the COVID-19 pandemic for children and parents. We measure mandatory homeschooling continuously, as the percentage of time between September 2020 and 15 February 2021. In line with prior research, we conceptualize child mental health as internalizing symptoms and externalizing symptoms and parent mental health as depression, anxiety, and general stress [
3,
11,
50]. The study further assesses whether this relationship is moderated by parent employment status, child internet access, or educational support from the child’s school, including live online classes, pre-recorded classes, or home learning packs (e.g., worksheets, assignments, videos, etc.).
This study answers recent calls for research into the effects of COVID-19 restrictions on child and parent mental health [
51]. By studying the compounding mental health effects of cumulative mandatory homeschooling over time, we aim to clarify contradictory findings in past research. By further exploring which factors may moderate this relationship, we aim to identify the families at greatest risk of adverse mental health effects during mandatory homeschooling and the educational support mitigating these effects. Our results may help governments and schools connect high-risk families with the support they need to recover from lingering mental health effects of the COVID-19 pandemic and may encourage schools to implement effective educational support in the future. Together, our results may shape how mandatory homeschooling is conceptualized in future research to capture compounding mental health effects. They may also guide policy aimed at minimizing mental health harm when implementing future homeschooling mandates, such as limiting duration. This study assesses the following hypotheses.
Hypothesis 1 (H1). We predict that cumulative time spent homeschooling is positively associated with child internalizing and externalizing symptoms and parent anxiety, depression and stress, such that greater cumulative time spent homeschooling is associated with poorer mental health.
Hypothesis 2a (H2a). We predict that live online class moderate this relationship, such that receiving this educational support will predict decreased mental health harm of cumulative time spent homeschooling for parents and children.
Hypothesis 2b (H2b). We predict that pre-recorded online classes moderate this relationship, such that receiving this educational support will predict decreased mental health harm of cumulative time spent homeschooling for parents and children.
Hypothesis 2c (H2c). We predict that home learning packs moderate this relationship, such that receiving this educational support will predict decreased mental health harm of cumulative time spent homeschooling for parents and children.
Hypothesis 3 (H3). We predict that children’s internet access moderates this relationship, such that unreliable internet access will predict increased mental health harm from cumulative time spent homeschooling for parents and children.
Hypothesis 4 (H4). We predict that parent employment moderates this relationship, such that unemployment of one or both parents will predict greater mental health harm from cumulative time spent homeschooling for parents and children.
4. Discussion
In the present study, we explored compounding adverse mental health effects of mandatory homeschooling for families during the COVID-19 pandemic by assessing parent and child mental health and children’s cumulative time spent homeschooling over several months. We further explored the effects of potential risk and protective factors on this relationship, including educational support from schools, child internet access, and parent employment.
We found that greater cumulative time spent homeschooling was significantly associated with greater child internalizing symptoms and parent depression and anxiety, but not child externalizing symptoms or parent stress, showing partial support for H1. This supports findings from most past research linking mandatory homeschooling to adverse mental health effects for parents and children [
3,
8,
9]. It also extends this research by suggesting that these effects compound over time, placing long-term homeschoolers at greater risk of mental health harm. These findings align with chronic stress research, which shows that long-term exposure to stressors (such as mandatory homeschooling) is harmful for adult and child mental health [
12,
13,
14,
15]. These findings also suggest that measuring mandatory homeschooling continuously over longer periods may be a more effective means of investigating its effects on mental health.
Our results also suggest that the effects of mandated homeschooling may be more salient for some mental health outcomes than others. The non-significant association between cumulative time spent homeschooling and child externalizing symptoms may relate to findings by Monnier et al. [
11], which show that three or more hours spent homeschooling per day improved children’s hyperactivity symptoms. Research has found that daily routines, which could be built around homeschooling tasks, may have had positive impacts on children’s externalizing symptoms [
10,
69]. There may exist a “sweet spot” of time spent homeschooling during school closure mandates that is protective for children’s externalizing symptoms, which would not be captured in the total effects of our mediation models. The non-significant relationship between cumulative time spent homeschooling and parent stress may be explained by our measure of stress, the PSS-4, which captures general stress [
58]. During mandatory homeschooling, parents may have been experiencing forms of stress specific to the COVID-19 pandemic, which may not have been captured by this measure. A stress measure specific to COVID-19, such as the COVID-19 Stress Scale developed by Taylor et al. [
70], may therefore be more appropriate. Deacon et al. [
1] used this scale to find that greater time engaged in mandatory homeschooling was linked to increased COVID-19-related stress in parents, supporting this possibility.
Looking at educational support provided by schools (H2)
, we found that live online classes (H2a) and home learning packs (H2c) but not pre-recorded online classes (H2b) moderated the relationships between cumulative time spent homeschooling and some mental health outcomes. First, we found that cumulative time spent homeschooling was only positively associated with child internalizing symptoms and parent depression and anxiety when children did not receive live online classes. These findings may reflect the importance of live student–teacher interactions for academic success and satisfaction during mandatory homeschooling [
19]. Live classes may have allowed children to depend more on teachers for ongoing academic support, easing the educational strain felt by families when tackling academic and technological challenges at home [
25,
33]. Live online classes also create routine in the child’s day, which has been identified as a protective factor for children’s mental health during COVID-19 [
10,
69]. This routine may have helped parents by offering them structured time to focus on other responsibilities while children are occupied by teachers. By contrast, pre-recorded online classes did not offer live academic support or routine, which may explain their non-significant interaction effects. Instead, children learning from pre-recorded classes may have sought support from parents, adding to the family’s academic strain. The added responsibility of scheduling class time may have also fallen to parents.
We also found that home learning packs protected against the effects of cumulative time spent homeschooling on child internalizing symptoms and parent depression. The relationships between mandatory homeschooling and child internalizing symptoms or parent depression were only significant and positive when children did not receive home learning packs. This suggests that children and parents benefitted when children received educational support such as assignments, worksheets, or videos during mandated homeschooling. For parents, having additional resources to rely on may have helped them instruct their children more effectively while sparing them the additional effort of assembling materials on their own. This may have protected parents by reducing the burden of their newfound educator roles. These resources may also have provided children with opportunities to learn independently and apply their knowledge, which has been linked to improved academic performance [
40]. If home learning packs helped students and parents engage with course content more effectively and overcome academic challenges, this may explain their protective effects on child mental health. Together, these results suggest that educational support for schools, especially live online class and home learning packs, protect families against the adverse mental health impact of mandatory homeschooling.
Our results suggest that reliable internet access is a risk factor for children’s mental health during mandatory homeschooling, contrary to our hypothesis (H3). Cumulative time spent homeschooling was only significantly positively associated with child internalizing symptoms when children had reliable internet access. We hypothesized that reliable internet connection would protect mental health because of its crucial importance for school, work, social connection, entertainment, and health care during the pandemic [
71,
72]. However, excessive internet use before and during COVID-19 has been linked to adverse mental health and academic outcomes in children [
25,
44,
73,
74]. Further, problematic internet use among children, including elevated time online, increased throughout the COVID-19 pandemic [
43,
44]. Therefore, children with reliable internet may have been using the internet excessively or in otherwise problematic ways, which may in turn have interfered with their schooling and had adverse impacts on their mental health. By contrast, when children were distracted by the internet, parents’ mental health may have benefitted from having time to themselves, explaining why adverse effects did not extend to parents.
Finally, our results suggest that having both parents employed is a risk factor for child mental health during mandatory homeschooling, contrary to our hypothesis (H4). There was only a significant positive relationship between cumulative time spent homeschooling and child internalizing and externalizing symptoms when both parents were employed. These findings suggest that cumulative time spent homeschooling only has adverse impacts on child mental health when both parents are employed. This finding may reflect reports that balancing work and homeschooling is a major source of stress for employed parents [
4,
7,
22] and past findings that children show more internalizing and externalizing symptoms when parents experience greater work–family conflict [
75]. It is possible that working parents had less time to support their children academically, which may have caused their children to struggle more with school and consequently experience mental health declines. By contrast, unemployed parents may have had more time and energy to support their children with school. For parents, the unique challenges of being unemployed during a pandemic may have negated the benefits experienced by their children, explaining the non-significant results for parent mental health [
4,
7,
22]. These findings suggest that the effects of unemployment differed in the context of the COVID-19 pandemic, and that families with two working parents may have been at greater risk of mental health harm.
Limitations and Future Directions
Our study includes some limitations. First, our data were collected at a single time point and accounts were retrospective, preventing causal conclusions. Parents and children were mostly white (parent = 69%, children = 67%), parent relationships were mostly heterosexual (94%), and only 14% of families had incomes of USD 50,000 or less. It is important that future research include more diverse samples, given that racial, ethnic, and sexual minorities and people of lower socioeconomic status experienced disproportionate challenges during the COVID-19 pandemic due to inequities [
4,
76]. Given the disproportionate impacts experienced by people of lower socioeconomic status, we ran additional sensitivity tests using family income (adjusted for family size) as a covariate. Our results unchanged, suggesting that the effects of family income on mental health likely do not account for our results; however, future research should continue to include relevant sociodemographic variables as covariates in their models.
Another limitation is that we used binary yes/no measures for educational support from school and did not collect information on its quality or the training provided to families for this support. Families reported that poor-quality educational support and lack of training on how to use it contributed to adverse mandatory homeschooling experiences [
25,
37], suggesting that these factors should be explored in future research.
This study also ran multiple (30 total) tests, which can increase the risk of Type 1 error. A stringent method to protect against Type 1 error with multiplicity testing is the Bonferroni correction, which is calculated by dividing the standard significance threshold of
p < 0.05 by the number of tests [
77]—in our case, 30. With a stringent Bonferroni-corrected alpha of 0.0017 (0.05/30), only one of our findings remained significant, namely the positive relationship between time spent homeschooling and child internalizing symptoms (
p < 0.001). Thus, we can have confidence in the observed positive link of time spent homeschooling with child internalizing symptoms. Other findings reported as statistically significant in the main analyses using traditional significance levels (
p < 0.05) should be interpreted with caution and must be replicated in future studies. Our survey data were also collected at a single time point, which limits our ability to determine causality and temporality [
78]. While our cross-sectional design is a useful first step, further studies using longitudinal designs are needed to replicate our findings and confirm causality [
79]. Our findings therefore represent important early findings.
Our conceptualization of parent employment also presents limitations. We combined “Unemployed” parents with those “Not in the labor force” to account for low sample sizes, based on findings that these employment statuses resembled each other during the pandemic [
60,
61]. However, these categories are usually viewed as distinct from each other, and should also be explored separately in future research to determine whether their effects differ [
80]. To further account for the small sample size of families in which both parents were unemployed (
n = 35), the “Unemployed” category combined families with one and two unemployed parents. This may present a limitation, as experiences of families with two unemployed parents may have differed from those with one employed and one unemployed parent [
81]. For example, the latter families may have benefited from having one parent with stable employment and one parent with extra time to support the child’s homeschooling [
82], which could have influenced our results. To address this limitation, we ran supplementary analyses that assessed the parent employment moderator at three levels: Both Employed, Both Unemployed and Mixed (One Employed, one Unemployed). Results were virtually unchanged, suggesting that combining these groups did not adversely affect our results. However, future studies should still consider exploring potential differences between families with one versus two unemployed parents.
Past research has also found that mothers took on a disproportionate share of childcare responsibilities during the pandemic compared to fathers and may have experienced more adverse mental health effects [
83,
84]. Further, paternal and maternal unemployment may impact child mental health differently [
48]. Considering these findings, future studies should also assess differences in the effects of maternal versus paternal employment on family mental health during COVID-19. Future studies should also explore whether families with multiple school-aged children experienced mandatory homeschooling differently.