Plans of US Parents Regarding School Attendance for Their Children in the Fall of 2020: A National Survey | Infectious Diseases | JAMA Pediatrics | JAMA Network
[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.  Weighted Survey Results
Weighted Survey Results
Figure 2.  Plans to Keep Children Home From School Adjusted by Socioeconomic and Family Characteristics
Plans to Keep Children Home From School Adjusted by Socioeconomic and Family Characteristics

BA indicates bachelor of arts; COVID-19, coronavirus disease 2019; HS, high school, NA, not applicable.

Table 1.  Sample Statistics (N = 730)
Sample Statistics (N = 730)
Table 2.  Associations Between Parent Attitudes and Beliefs, Sociodemographic Characteristics, and Plans for School Attendancea
Associations Between Parent Attitudes and Beliefs, Sociodemographic Characteristics, and Plans for School Attendancea
1.
Hsiang  S, Allen  D, Annan-Phan  S,  et al.  The effect of large-scale anti-contagion policies on the COVID-19 pandemic.   Nature. 2020:1-9. doi:10.1038/s41586-020-2404-8PubMedGoogle Scholar
2.
Flaxman  S, Mishra  S, Gandy  A,  et al; Imperial College COVID-19 Response Team.  Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe.   Nature. 2020:1-8. doi:10.1038/s41586-020-2405-7PubMedGoogle Scholar
3.
Kuhfeld  M, Soland  J, Tarasawa  B, Johnson  A, Ruzek  E, Liu  J. Projecting the potential impacts of COVID-19 school closures on academic achievement. EdWorkingPapers.com. Published online May 4, 2020. Accessed June 5, 2020. https://www.edworkingpapers.com/ai20-226
4.
Brooks  SK, Webster  RK, Smith  LE,  et al.  The psychological impact of quarantine and how to reduce it: rapid review of the evidence.   Lancet. 2020;395(10227):912-920. doi:10.1016/S0140-6736(20)30460-8PubMedGoogle ScholarCrossref
5.
Aviles  AM, Anderson  TR, Davila  ER.  Child and adolescent social-emotional development within the context of school.   Child Adolesc Ment Health. 2006;11(1):32-39. doi:10.1111/j.1475-3588.2005.00365.xGoogle ScholarCrossref
6.
Liu  JJ, Bao  Y, Huang  X, Shi  J, Lu  L.  Mental health considerations for children quarantined because of COVID-19.   Lancet Child Adolesc Health. 2020;4(5):347-349. doi:10.1016/S2352-4642(20)30096-1PubMedGoogle ScholarCrossref
7.
Cluver  L, Lachman  JM, Sherr  L,  et al.  Parenting in a time of COVID-19.   Lancet. 2020;395(10231):e64. doi:10.1016/S0140-6736(20)30736-4PubMedGoogle Scholar
8.
Viner  RM, Russell  SJ, Croker  H,  et al.  School closure and management practices during coronavirus outbreaks including COVID-19: a rapid systematic review.   Lancet Child Adolesc Health. 2020;4(5):397-404. doi:10.1016/S2352-4642(20)30095-XPubMedGoogle ScholarCrossref
9.
Douglas  M, Katikireddi  SV, Taulbut  M, McKee  M, McCartney  G.  Mitigating the wider health effects of covid-19 pandemic response.   BMJ. 2020;369:m1557. doi:10.1136/bmj.m1557PubMedGoogle ScholarCrossref
10.
Dorn  AV, Cooney  RE, Sabin  ML, Sabin  ML.  COVID-19 exacerbating inequalities in the US.   Lancet. 2020;395(10232):1243-1244. doi:10.1016/S0140-6736(20)30893-XPubMedGoogle ScholarCrossref
11.
Van Lancker  W, Parolin  Z.  COVID-19, school closures, and child poverty: a social crisis in the making.   Lancet Public Health. 2020;5(5):e243-e244. doi:10.1016/S2468-2667(20)30084-0PubMedGoogle ScholarCrossref
12.
Armitage  R, Nellums  LB.  Considering inequalities in the school closure response to COVID-19.   Lancet Glob Health. 2020;8(5):e644. doi:10.1016/S2214-109X(20)30116-9PubMedGoogle Scholar
13.
Greenberg-Worisek  AJ, Kurani  S, Finney Rutten  LJ, Blake  KD, Moser  RP, Hesse  BW.  Tracking healthy people 2020 internet, broadband, and mobile device access goals: an update using data from the health information national trends survey.   J Med internet Res. 2019;21(6):e13300. doi:10.2196/13300PubMedGoogle Scholar
14.
Dooley  DG, Bandealy  A, Tschudy  MM.  Low-income children and coronavirus disease 2019 (COVID-19) in the US.   JAMA Pediatr. 2020. doi:10.1001/jamapediatrics.2020.2065PubMedGoogle Scholar
15.
Kossek  EE, Lautsch  BA.  Work–life flexibility for whom? occupational status and work–life inequality in upper, middle, and lower level jobs.   Acad Manag Ann. 2017;12(1):5-36. doi:10.5465/annals.2016.0059Google ScholarCrossref
16.
Deniz Can  D, Ginsburg-Block  M.  Parenting stress and home-based literacy interactions in low-income preschool families.   J Appl Dev Psychol. 2016;46:51-62. doi:10.1016/j.appdev.2016.07.002Google ScholarCrossref
17.
Verdoni  L, Mazza  A, Gervasoni  A,  et al.  An outbreak of severe Kawasaki-like disease at the Italian epicentre of the SARS-CoV-2 epidemic: an observational cohort study.   Lancet. 2020;395(10239):1771-1778. doi:10.1016/S0140-6736(20)31103-XPubMedGoogle ScholarCrossref
18.
Baker  BD. America’s Most Financially Disadvantaged School Districts and How They Got That Way: How State and Local Governance Causes School Funding Disparities. Center for American Progress. Published 2014. Accessed June 16, 2020. https://eric.ed.gov/?id=ED561094
19.
Simpkins  SD, Fredricks  JA, Lin  AR. Families and organized after-school activities for youth. In: APA Handbook of Contemporary Family Psychology: Applications and Broad Impact of Family Psychology, Vol. 2. 2019:235-248. doi:10.1037/0000100-015
20.
Brown  M, Pitt-Catsouphes  M.  A mediational model of workplace flexibility, work–family conflict, and perceived stress among caregivers of older adults.   Community Work Fam. 2016;19(4):379-395. doi:10.1080/13668803.2015.1034656Google ScholarCrossref
21.
The Return of the Multi-Generational Family Household. Pew Res Center’s Soc Demogr Trends Proj. Accessed July 24, 2020. https://www.pewsocialtrends.org/2010/03/18/the-return-of-the-multi-generational-family-household/
22.
Tucker-Seeley  RD, Li  Y, Sorensen  G, Subramanian  SV.  Lifecourse socioeconomic circumstances and multimorbidity among older adults.   BMC Public Health. 2011;11(1):313. doi:10.1186/1471-2458-11-313PubMedGoogle ScholarCrossref
23.
Chandler  J, Rosenzweig  C, Moss  AJ, Robinson  J, Litman  L.  Online panels in social science research: expanding sampling methods beyond Mechanical Turk.   Behav Res Methods. 2019;51(5):2022-2038. doi:10.3758/s13428-019-01273-7PubMedGoogle ScholarCrossref
24.
Callegaro  M, DiSogra  C.  Computing response metrics for online panels.   Public Opin Q. 2008;72(5):1008-1032. doi:10.1093/poq/nfn065Google ScholarCrossref
25.
Standard Definitions - AAPOR. Accessed July 24, 2020. https://www.aapor.org/Standards-Ethics/Standard-Definitions-(1).aspx
26.
Gee  GC, Ro  MJ, Rimoin  AW.  Seven reasons to care about racism and COVID-19 and seven things to do to stop it.   Am J Public Health. 2020;110(7):954-955. doi:10.2105/AJPH.2020.305712Google ScholarCrossref
27.
Raifman  MA, Raifman  JR.  Disparities in the population at risk of severe illness from COVID-19 by race/ethnicity and income.   Am J Prev Med. 2020;59(1):137-139. doi:10.1016/j.amepre.2020.04.003PubMedGoogle ScholarCrossref
28.
The COVID-19 pandemic: a call to action to identify and address racial and ethnic disparities. Europe PMC. Accessed July 24, 2020. https://europepmc.org/article/med/32306369
29.
Taylor  S, Landry  CA, Paluszek  MM, Fergus  TA, McKay  D, Asmundson  GJG.  Development and initial validation of the COVID stress scales.   J Anxiety Disord. 2020;72:102232. doi:10.1016/j.janxdis.2020.102232PubMedGoogle Scholar
30.
Lumley  T.  Analysis of complex survey samples.   J Stat Softw. 2004;9(1):1-19. doi:10.18637/jss.v009.i08Google Scholar
31.
Deville  J-C, Särndal  C-E, Sautory  O.  Generalized raking procedures in survey sampling.   J Am Stat Assoc. 1993;88(423):1013-1020. doi:10.1080/01621459.1993.10476369Google ScholarCrossref
32.
Bureau  UC. Current Population Survey Data Tables. The United States Census Bureau. Accessed June 15, 2020. https://www.census.gov/programs-surveys/cps/data/tables.html
33.
Agresti  A.  Modelling ordered categorical data: recent advances and future challenges.   Stat Med. 1999;18(17-18):2191-2207. doi:10.1002/(SICI)1097-0258(19990915/30)18:17/18<2191::AID-SIM249>3.0.CO;2-MPubMedGoogle ScholarCrossref
34.
Winship  C, Radbill  L.  Sampling weights and regression analysis.   Sociol Methods Res. 1994;23(2):230-257. doi:10.1177/0049124194023002004Google ScholarCrossref
35.
Mplus User’s Guide. Accessed June 15, 2020. https://www.statmodel.com/html_ug.shtml
36.
Benjamini  Y, Hochberg  Y.  Controlling the false discovery rate: a practical and powerful approach to multiple testing.   J R Stat Soc Ser B Methodol. 1995;57(1):289-300. doi:10.1111/j.2517-6161.1995.tb02031.xGoogle Scholar
37.
Murphy  J.  The social and educational outcomes of homeschooling.   Sociol Spectr. 2014;34(3):244-272. doi:10.1080/02732173.2014.895640Google ScholarCrossref
38.
Martin-Chang  S, Levesque  K. Academic Achievement. In:  The Wiley Handbook of Home Education. John Wiley & Sons, Ltd; 2017:121-134. doi:10.1002/9781118926895.ch5
39.
Davies  NG, Klepac  P, Liu  Y, Prem  K, Jit  M, Eggo  RM; CMMID COVID-19 working group.  Age-dependent effects in the transmission and control of COVID-19 epidemics.   Nat Med. 2020:1-7. doi:10.1038/s41591-020-0962-9PubMedGoogle Scholar
40.
Schools Should Prioritize Reopening in Fall 2020, Especially for Grades K-5, While Weighing Risks and Benefits | National Academies. Accessed July 20, 2020. https://www.nationalacademies.org/news/2020/07/schools-should-prioritize-reopening-in-fall-2020-especially-for-grades-k-5-while-weighing-risks-and-benefits
42.
Yank  V, Agarwal  S, Loftus  P, Asch  S, Rehkopf  D.  Crowdsourced health data: comparability to a US national survey, 2013-2015.   Am J Public Health. 2017;107(8):1283-1289. doi:10.2105/AJPH.2017.303824PubMedGoogle ScholarCrossref
43.
Leiner  DJ.  Too fast, too straight, too weird: non-reactive indicators for meaningless data in internet surveys.   Surv Res Methods. 2019;13(3):229-248. doi:10.18148/srm/2019.v13i3.7403Google Scholar
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    1 Comment for this article
    Parameters
    Jeoffry Gordon, MD, MPH | Retired Family Medicine
    The only parameters worth knowing are: what is the test positivity in the community. and what plans does the school system have to deal with a positive case? Otherwise parents concerns are based on emotions, political messaging, and then individual needs and concerns.
    CONFLICT OF INTEREST: None Reported
    Views 13,966
    Citations 0
    Original Investigation
    Impact of Policy on Children
    August 14, 2020

    Plans of US Parents Regarding School Attendance for Their Children in the Fall of 2020: A National Survey

    Author Affiliations
    • 1Center for Child Health, Behavior and Development, Seattle Children’s Research Institute, Seattle, Washington
    • 2Department of Pediatrics, University of Washington, Seattle
    • 3Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle
    JAMA Pediatr. 2020;174(11):1093-1101. doi:10.1001/jamapediatrics.2020.3864
    Key Points

    Question  What do parents plan to do about school attendance in the fall of 2020, and what factors are influencing these plans?

    Findings  In this survey study of 730 US parents of school-aged children, 31% of parents indicated they will probably or definitely keep their child home this fall if schools open for in-person instruction. Factors associated with planning to keep children home were lower household income, not being employed, and whether those employed had a flexible and controllable work schedule.

    Meaning  Outreach to help families make decisions about school attendance should focus on addressing the concerns of families with low income.

    Abstract

    Importance  As schools consider reopening for in-person instruction prior to availability of a coronavirus disease 2019 (COVID-19) vaccine, families may be weighing their priorities regarding school attendance.

    Objective  To characterize the association of planned in-person school attendance during the COVID-19 pandemic with factors, including family socioeconomic characteristics, and parent attitudes and beliefs about their child’s school attendance.

    Design, Setting, and Participants  Cross-sectional survey study. Data were collected from June 2, 2020, to June 5, 2020, weighted to reflect population norms, and analyzed using ordered probit regression. A sample of US parents (of children ages 5-17 years) were recruited using a nonprobability survey panel with stratification by socioeconomic characteristics.

    Main Outcomes and Measures  The main outcome was parent-reported plan to send their child to school or keep their child home, conditional on their school opening for in-person instruction. Additional measures assessed family socioeconomic characteristics, medical vulnerability, worry about COVID-19 and multisystem inflammatory syndrome, confidence in their child’s school, and homeschooling difficulties.

    Results  The sample of 730 parents was balanced by parent sex (53% women) with successful oversampling for Black (28%; n = 201) and Hispanic (27%; n = 200) participants. In estimates weighted to US population norms, 31% (95% CI, 27% to 34%) of participants indicated they would probably or definitely keep their child home this fall, and 49% indicated that they would probably or definitely send their child to school this fall. Factors associated with planning to keep children home included lower income (38% with incomes <$50 000 vs 21% with incomes $100 000-$150 000 per year; difference, 17%; 95% CI, 9% to 26%), being unemployed (40% unemployed vs 26% employed; difference, 14%; 95% CI, 5% to 25%), and having a flexible job (33% with flexible jobs vs 19% with inflexible jobs; difference, 14%; 95% CI, 5% to 30%). Planning to keep children home was also associated with fear of COVID-19 (B = 0.19; P < .001), fear of multisystem inflammatory syndrome (B = 0.12; P = .04), confidence in schools (B = −0.22; P < .001), and challenges of homeschooling (B = −0.12; P = .01). Race and ethnicity were not significantly associated with plans to keep children home.

    Conclusions and Relevance  In this survey study, many parents planned to keep children home in fall 2020. Schools need to act soon to address parental concerns and provide options for what will be available for them should they opt to keep their child home. Structural barriers, such as lack of workplace flexibility and potential school-level inequities in implementation of preventive measures, must be acknowledged and addressed where possible.

    Introduction

    State governments and school districts are grappling with how to safely, effectively, and equitably educate youths during the novel coronavirus disease (COVID-19) pandemic. Recent data suggest that school closures, along with other policies to encourage home isolation, have substantially slowed the spread of COVID-19.1,2 However, such closures have implications for children, parents, and society, including inadequate academic instruction,3 missed opportunities for developmentally important in-person experiences,4-6 parental stress,7 and compromised work productivity.8 Critically, these costs are likely to be disproportionately borne by the most socioeconomically disadvantaged families7,9-12 because they tend to have the least resources for remote learning (ie, access to computers and high-speed internet),11,13 the least support for home learning from their child’s school,14 the least job flexibility,15 and the greatest risk of experiencing other stressors (eg, financial strain, racism, and neighborhood safety) that may compound parenting challenges.16

    As schools consider reopening for in-person instruction prior to availability of a COVID-19 vaccine, families may be weighing their priorities regarding school attendance. This choice will look different for each family as schools will likely vary in the steps they take to mitigate viral transmission, home-based learning will be more difficult for some families than others, and parents will have different levels of worry about COVID-19 and the extremely rare multisystem inflammatory syndrome in children that has been associated with COVID-19.17 Critically, these attitudes and beliefs are likely patterned by family socioeconomic factors. Overcrowded and underfunded schools, more likely to be located in communities characterized by lower-income residents,18 may have more difficulty implementing social distancing and other recommended measures to limit viral transmission while maintaining developmentally appropriate instruction. Yet, families from these communities rely disproportionately on schools for meals and other resources (academic support and counseling) and opportunities (physical activity and extracurricular activities).19 Parents who lack the ability to work from home on a flexible schedule may believe that home-based learning would be more challenging; such job flexibility is more likely among more affluent and educated parents.20 Families with higher COVID-19 mortality risk, particularly those who live with a medically vulnerable person, may be more worried about COVID-19; lower-income families are more likely to live in a multigenerational household21 and to have a family member with a chronic health condition.22

    In a demographically diverse sample of parents of school-aged children in the United States, this study aimed to (1) describe parents’ plans for sending children back to school in fall 2020, (2) characterize the association of planned school enrollment with family socioeconomic characteristics, and (3) understand how parent attitudes and beliefs are associated with socioeconomic characteristics and plans for school attendance. Understanding what factors are associated with parent plans for school attendance can inform how those with a vested stake in this decision, including pediatricians and school districts, approach providing decision support to families.

    Methods
    Sample and Procedure

    Participants were a demographically stratified convenience sample of 730 parents in the United States who have at least 1 child between the ages of 5 and 17 years. Recruitment was facilitated by Prime Panel,23 an online survey recruitment platform that sources participants across multiple nonprobability survey panels applying investigator-determined eligibility criteria and demographic quotas. Consistent with American Association for Public Opinion Research reporting guidelines for survey recruitment using an opt-in nonprobability panel, the participation rate was not reported because the sampling frame was unknown.24,25 Participants accessed the survey online using their own computer or mobile device and received a small monetary incentive for completion. Individuals without computer or internet access were not eligible to participate. Black and Hispanic parents were oversampled relative to 2018 US Census demographic distributions. To meet sample size goals for Black and Hispanic parents, data collection was supplemented by nonprobability survey panel recruitment with increased compensation for Black and Hispanic parents. Race and ethnicity were measured because of emergent evidence and discourse about how due to systemic racism, Black and Hispanic families are being disproportionately negatively affected by COVID-19.26-28 Eligible participants completed an English-language anonymous online survey hosted on the Qualtrics platform, with data collection occurring June 2, 2020, through June 5, 2020. After providing information about family structure, parents with more than 1 child aged 5 to 17 years were instructed to answer questions with reference to the child whose birthday was coming up next. The questionnaire contained 36 questions (eAppendix in the Supplement), which were pilot tested in a convenience sample of parents recruited via Mechanical Turk (n = 200). Research activities were approved by the University of Washington and Seattle Children’s Hospital institutional review boards, and a waiver for documentation of informed consent was granted; potential participants reviewed an information sheet describing the study.

    Measures
    Outcome

    The primary outcome for the study was parent-reported plans for their child’s school attendance in the fall. Parents were asked, “If schools open in the fall, do you plan to keep your child at home (ie, home school) or send them to school?” Response options ranged from definitely keep child home (1) to definitely send child to school (5).

    Parent Attitudes and Beliefs

    Four multi-item scales assessed (1) parent confidence in their child’s school, (2) anticipated difficulty keeping their child home, (3) worry about COVID-19 (as measured by the Danger subscale of the COVID-19 Stress Scale29), and (4) worry about the multisystem inflammatory syndrome in children. Questions are provided (eAppendix in the Supplement).

    Socioeconomic and Family Composition Characteristics

    Parent age, sex, number of other children in their family, number of adults in their home, whether they live with an elderly family member or person with a chronic health condition, and child age and sex were queried. Race/ethnicity was self-reported using US Census categories. Past-year household income was assessed in $10 000 increments. Due to a survey coding error, incomes between $100 000 and $150 000 were included in 1 category. Parent education was assessed in 4 categories (high school or less, some college, bachelor’s degree, or graduate degree). They also reported whether their child qualifies for free or reduced-price lunch at their school and the type of school their child attended (public charter, parochial or religious, or private nonreligious). Parents indicated whether they were employed (full time or part time), and if employed, whether they were considered an “essential worker.”10 They were also asked the extent to which their job allows for work from home (yes, somewhat, or no), and whether they have a flexible work schedule that they can control (yes, somewhat, or no).

    Statistical Analyses

    Because data were oversampled based on race and ethnicity, we developed raking weights using iterative poststratification30,31 to describe parent plans about sending their children back to school in the fall (aim 1). We used census data to generalize to a sample nationally representative in terms of race, ethnicity, and income for households with at least 1 child.32

    To examine differences in plans for school attendance across socioeconomic and family characteristics (aim 2), we used ordered probit regression with all planned variables entered simultaneously. Probit regression makes the assumption that a continuous, latent variable underlies the categorical response options to the survey items and produces coefficient estimates in terms of z scores.33 We controlled for design variables in the probit analysis rather than using weights,34 with all variables centered at their weighted rather than original values.

    To understand the associations between parent attitudes and beliefs, sociodemographic characteristics, and plans for school attendance, we used a series of ordinal probit regression models (aim 3). Plans for school attendance were regressed on all 4 attitude and belief variables (fears about COVID-19, fears about multisystem inflammatory syndrome, confidence in schools, and challenges due to homeschooling), controlling for sociodemographic characteristics. To understand how attitudes and beliefs differed across sociodemographic characteristics, these variables were also regressed on sociodemographic characteristics. Models were estimated using Mplus, version 8.0 (Muthén & Muthén).35

    Multiple Comparisons

    To manage the rate of false positives, we used the Benjamini-Hochberg adjustment, with α criterion of P less than .05 and the false discovery rate set to 5% for all models.36

    Missing Data

    Of the 730 initial participants, 3 participants were dropped owing to missing data. The very small missingness rate is unlikely to have biased the results.

    Results
    Sample Characteristics

    Raw and weighted sample characteristics are presented in Table 1. The unweighted sample included 730 participants and was largely consistent with population norms for households with at least 1 child, with successful oversampling for Black (28%; n = 201) and Hispanic (27%; n = 200) participants and balance across parent sex (53% female; n = 388). Lower-income participants were somewhat overrepresented compared with population norms (37% sample vs 30% in the weighted sample), and higher-income participants were somewhat underrepresented (9% in the sample vs 21% in the weighted sample). After applying the weights, households reported relatively even distributions of grade level for the focal child (range, 19% to 33%), with 71% attending public school. Thirty-four percent of participants reported at least 1 medically vulnerable person living in their household, 57% of those employed said their job could be completed from home, and 44% reported having flexible work hours.

    Survey Results (Aim 1)

    Weighted survey responses for key variables are presented in Figure 1. Notably, only 49% of participants said they would probably or definitely send their child to school if it opened in the fall, with 30% responding they would probably or definitely keep their child home. Forty percent of participants were more than moderately worried that they would catch the virus or that the health care system would not be able to protect their loved ones, and 34% were more than moderately worried about multisystem inflammatory syndrome in children. Confidence in schools was relatively modest, with 29% reporting they were confident or very confident that school could prevent the spread of COVID-19, 35% reporting school could provide enough social interaction for their child while enforcing social distancing, and 45% reporting that schools could meet their child’s academic needs with a modified schedule.

    Socioeconomic Characteristics and Family Composition (Aim 2)

    Factors associated with anticipated school attendance choice are shown in Figure 2. Key factors included income level, employment status, and job flexibility. Controlling for all other model variables, participants with lower incomes reported a higher likelihood of keeping their children home in the fall, with model-adjusted proportions of 38% of those making less than $50 000 per year in the “probably home” or “definitely home” categories vs 21% with incomes of $100 000 to $150 000 (difference, 17%; 95% CI, 9% to 26%). Those with flexible jobs were nearly as likely to keep their children home as those who were not employed (33% vs 40%; difference, 7%; 95% CI, −3% to 17%), whereas just 19% of those whose jobs were not flexible planned to probably or definitely keep their children home (difference from flexible jobs,−17%; 95% CI, −30% to -5%). Parents with children in grades 3 through 5 were more likely to plan to keep their children home than parents of high schoolers. Parents in households with vulnerable people were modestly more likely to plan to keep their children home. Race and ethnicity were not significantly associated with plans to keep children home. Together, family characteristics were associated with a modest proportion of the variability in plans to send children to school (R2 = 16%).

    Parent Attitudes and Beliefs (Aim 3)

    Results from the ordered probit regression models examining parent attitudes and beliefs are shown in Table 2. After controlling for sociodemographic and family composition predictors, planning to keep children home was associated with fear of COVID-19 (B = 0.19; P < .001), fear of multisystem inflammatory syndrome (B = 0.12; P = .04), confidence in schools (B = −0.22; P < .001), and challenges of homeschooling (B = −0.12; P = .01).

    An education level of some college or less was generally associated with less fear about COVID-19 and multisystem inflammatory syndrome (see Table 2 for exact patterning of results). In contrast, having a vulnerable person in the household was associated with more fears of both COVID-19 (B = 0.22; P = .04) and multisystem inflammatory syndrome (B = 0.25; P = .02). Hispanic ethnicity was associated with fear about COVID-19 (B = 0.28; P = .04) but not multisystem inflammatory syndrome, whereas having a younger child was associated with fear of multisystem inflammatory syndrome but not COVID-19 (Table 2).

    Parents anticipated more challenges homeschooling younger children compared with high schoolers (prekindergarten to grade 2, B = 0.56; P < .001; grades 3-5, B = 0.43; P < .001; grades 6-8, B = 0.31; P = .02). Parents who were employed also indicated more expected challenges of homeschooling (B = 0.51; P < .001), and this effect did not vary according to job flexibility. Only education level was associated with confidence in schools, with those with graduate degrees expressing the most confidence. Race was not significantly associated with any of the parent attitude and belief measures. Including attitude and belief variables in the model explained moderately more variation in plans for school attendance (ΔR2 = 11%; total R2 = 27%).

    Discussion

    Nearly one-third of families in this sample plan to definitely or probably keep their child home in fall 2020, highlighting the reality that while policy decisions about opening schools are made at the community and state level, families are separately making their own decisions. Given the potential implications of school attendance for viral transmission, academic progress, social-emotional development, and parent well-being, it is critical that families receive appropriate guidance to make an informed choice that reflects their values, priorities, and specific health risks. It is also critical that structural barriers to families making the choice that works best for them are acknowledged and remedied where possible to minimize worsening disparities. Further, given the reality that many parents are strongly considering keeping their children home, schools and school districts must consider how they can feasibly meet the needs of this potentially sizeable fraction of the population. While many school districts in the US are now planning to start the 2020 school year with fully remote instruction, the present findings will continue to be relevant as instructional options change during the ongoing pandemic.

    That many families plan to keep their child home is not necessarily problematic. Homeschooling is a decision that some families make every year and, if well implemented, can meet a child’s learning and social-emotional needs.37 However, parents may be assuming that their child’s school will provide resources (eg, instructional materials or virtual teacher interaction) for teacher-led learning at home. If families are choosing to keep children home in the absence of adequate resources and support for remote learning from their child’s school, it is critical that they understand the key components of high-quality, developmentally appropriate homeschooling.38 Schools or school districts should clearly communicate to families the level of home learning support that will be provided (if any), along with the competencies and requirements for developmentally appropriate homeschooling in the absence of school support.

    Overall, parents were highly worried about COVID-19 and multisystem inflammatory syndrome, with worry and having a medically vulnerable household member associated with greater likelihood of planning to keep children home. While worry is subjective, it is critical that where there is objective data, this information is clearly communicated to families so that worry is at least relative to accurate probabilities. Extant data suggest that children are half as susceptible to infection compared with adults,39 and multisystem inflammatory syndrome in particular poses very limited risk to most children, suggesting that misperceptions about its prevalence could be addressed within decision support tools while also addressing key risk modifiers, such as having a medically vulnerable household member.

    Few parents were confident that their child’s school would be able to prevent students from spreading COVID-19. Importantly, parent confidence in school practices will likely change as schools finalize and communicate what steps they will be taking to limit viral transmission. Leading organizations, including the American Academy of Pediatrics and the National Academies of Sciences, Engineering, and Medicine, have released age-specific guidance for school reopening.40,41 Further research is needed to understand how parent willingness to send their child to school varies by school adoption of these practices. However, regardless of what practices the school adopts, parent perceptions about whether they will be consistently implemented by their child’s school is likely also an important behavioral determinant. More educated parents in the sample had greater confidence in their child’s school, potentially related to more positive prior experiences with their child’s school meeting student needs in other contexts.

    Critically, having a flexible schedule under one’s own control may be a key structural barrier in families’ abilities to make in-person school attendance decisions that they feel are best for their risk profile. In fact, the effect of job flexibility was a stronger predictor of planning for home-based learning than was having a medically vulnerable person in the household. More than three-quarters (78%) of employed participants reported having at least some flexibility. However, it is possible that a large subset of jobs with temporarily enhanced flexibility return to more traditional schedule demands if there is an option for in-person school attendance, forcing employees to choose between protecting their jobs and minimizing their family’s perceived COVID-19 risk. Notably, job flexibility was distinguished from having a job that can be completed from home, which was not associated with a greater likelihood of home-based learning. Critically, schedule control tends to be more prevalent in senior white-collar jobs,15 meaning that less socioeconomically advantaged families likely will be less able to make the choice about in-person schooling that works best for their family. Corporate, local, or federal policies supporting flexibility after schools reopen may help protect those at greatest health risk, particularly those with socioeconomic disadvantage.

    Limitations

    This study has several limitations. First is the use of a nonprobability sampling frame, meaning that survey respondents may not necessarily generalizable to nonrespondents; samples obtained this way tend to be younger than the average US adult and lack racial/ethnic diversity.42 However, we followed emergent guidance on how to obtain a demographically diverse and high-quality nonprobability sample, stratifying recruitment by socioeconomic characteristics23 and applying attention and data quality checks.43 We also note that this sampling strategy involved parents accessing a survey online. Consequently, participants may be more technologically literate and have greater comfort with remote learning than the average US parent. Second, we did not measure whether the reference child had another parent and the employment status of that parent. Future studies should more holistically understand the family context and how combined parent work demands and flexibility influence decision-making related to school attendance. Third, the term homeschooling was used in the question stem about planned parent behavior. It is possible this wording was interpreted as not subsuming school-facilitated remote learning, in which case the estimate of parent plans to keep their child home may be biased downwards. Fourth, we measured intentions, and the extent to which these will be acted on is unclear. Although many parents reported reluctance to return their children to school, they may in fact do so when presented with a binary choice of returning or homeschooling. Further, we did not ask about school attendance plans under specific schooling options, such as hybrid in-person and distance learning.

    Conclusions

    In spite of these limitations, this study has important implications for parents, schools, and policy makers. Parents should seek counsel from a trusted medical source (eg, their primary health care clinician) regarding the potential risks posed to their child and family by returning to school. There is scope for decision support to help families make this important choice. However, decision support alone is not sufficient for equitable decision-making. Structural barriers, such as lack of workplace flexibility and potential school-level inequities in implementation of health-related policies, must be acknowledged and addressed where possible. Schools need to act soon to allay parental concerns and provide options for what will be available for them should they opt to keep their child home. Policy makers need to ensure that there are adequate resources provided to schools to meet parental expectations regarding personal protective equipment, social distancing, symptom checking, and other steps recommended by organizations such as the US Centers for Disease Control and Prevention. It is expected that school budgets will need to expand rather than contract as might be expected given current fiscal crises. Such steps are critical for all families to be able to equitably act on their preferences regarding school attendance.

    Back to top
    Article Information

    Corresponding Author: Emily Kroshus, ScD, MPH, Seattle Children's Research Institute, 2001 Eighth Ave, Seattle, WA 98105 (ekroshus@u.washington.edu).

    Accepted for Publication: July 30, 2020.

    Published Online: August 14, 2020. doi:10.1001/jamapediatrics.2020.3864

    Author Contributions: Drs Kroshus and Hawrilenko had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: All authors.

    Acquisition, analysis, or interpretation of data: Kroshus, Hawrilenko, Christakis.

    Drafting of the manuscript: Kroshus, Hawrilenko.

    Critical revision of the manuscript for important intellectual content: Hawrilenko, Tandon, Christakis.

    Statistical analysis: Hawrilenko.

    Administrative, technical, or material support: Kroshus.

    Supervision: Christakis.

    Conflict of Interest Disclosures: None reported.

    Disclaimer: Dr Christakis is Editor of JAMA Pediatrics, but he was not involved in any of the decisions regarding review of the manuscript or its acceptance.

    References
    1.
    Hsiang  S, Allen  D, Annan-Phan  S,  et al.  The effect of large-scale anti-contagion policies on the COVID-19 pandemic.   Nature. 2020:1-9. doi:10.1038/s41586-020-2404-8PubMedGoogle Scholar
    2.
    Flaxman  S, Mishra  S, Gandy  A,  et al; Imperial College COVID-19 Response Team.  Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe.   Nature. 2020:1-8. doi:10.1038/s41586-020-2405-7PubMedGoogle Scholar
    3.
    Kuhfeld  M, Soland  J, Tarasawa  B, Johnson  A, Ruzek  E, Liu  J. Projecting the potential impacts of COVID-19 school closures on academic achievement. EdWorkingPapers.com. Published online May 4, 2020. Accessed June 5, 2020. https://www.edworkingpapers.com/ai20-226
    4.
    Brooks  SK, Webster  RK, Smith  LE,  et al.  The psychological impact of quarantine and how to reduce it: rapid review of the evidence.   Lancet. 2020;395(10227):912-920. doi:10.1016/S0140-6736(20)30460-8PubMedGoogle ScholarCrossref
    5.
    Aviles  AM, Anderson  TR, Davila  ER.  Child and adolescent social-emotional development within the context of school.   Child Adolesc Ment Health. 2006;11(1):32-39. doi:10.1111/j.1475-3588.2005.00365.xGoogle ScholarCrossref
    6.
    Liu  JJ, Bao  Y, Huang  X, Shi  J, Lu  L.  Mental health considerations for children quarantined because of COVID-19.   Lancet Child Adolesc Health. 2020;4(5):347-349. doi:10.1016/S2352-4642(20)30096-1PubMedGoogle ScholarCrossref
    7.
    Cluver  L, Lachman  JM, Sherr  L,  et al.  Parenting in a time of COVID-19.   Lancet. 2020;395(10231):e64. doi:10.1016/S0140-6736(20)30736-4PubMedGoogle Scholar
    8.
    Viner  RM, Russell  SJ, Croker  H,  et al.  School closure and management practices during coronavirus outbreaks including COVID-19: a rapid systematic review.   Lancet Child Adolesc Health. 2020;4(5):397-404. doi:10.1016/S2352-4642(20)30095-XPubMedGoogle ScholarCrossref
    9.
    Douglas  M, Katikireddi  SV, Taulbut  M, McKee  M, McCartney  G.  Mitigating the wider health effects of covid-19 pandemic response.   BMJ. 2020;369:m1557. doi:10.1136/bmj.m1557PubMedGoogle ScholarCrossref
    10.
    Dorn  AV, Cooney  RE, Sabin  ML, Sabin  ML.  COVID-19 exacerbating inequalities in the US.   Lancet. 2020;395(10232):1243-1244. doi:10.1016/S0140-6736(20)30893-XPubMedGoogle ScholarCrossref
    11.
    Van Lancker  W, Parolin  Z.  COVID-19, school closures, and child poverty: a social crisis in the making.   Lancet Public Health. 2020;5(5):e243-e244. doi:10.1016/S2468-2667(20)30084-0PubMedGoogle ScholarCrossref
    12.
    Armitage  R, Nellums  LB.  Considering inequalities in the school closure response to COVID-19.   Lancet Glob Health. 2020;8(5):e644. doi:10.1016/S2214-109X(20)30116-9PubMedGoogle Scholar
    13.
    Greenberg-Worisek  AJ, Kurani  S, Finney Rutten  LJ, Blake  KD, Moser  RP, Hesse  BW.  Tracking healthy people 2020 internet, broadband, and mobile device access goals: an update using data from the health information national trends survey.   J Med internet Res. 2019;21(6):e13300. doi:10.2196/13300PubMedGoogle Scholar
    14.
    Dooley  DG, Bandealy  A, Tschudy  MM.  Low-income children and coronavirus disease 2019 (COVID-19) in the US.   JAMA Pediatr. 2020. doi:10.1001/jamapediatrics.2020.2065PubMedGoogle Scholar
    15.
    Kossek  EE, Lautsch  BA.  Work–life flexibility for whom? occupational status and work–life inequality in upper, middle, and lower level jobs.   Acad Manag Ann. 2017;12(1):5-36. doi:10.5465/annals.2016.0059Google ScholarCrossref
    16.
    Deniz Can  D, Ginsburg-Block  M.  Parenting stress and home-based literacy interactions in low-income preschool families.   J Appl Dev Psychol. 2016;46:51-62. doi:10.1016/j.appdev.2016.07.002Google ScholarCrossref
    17.
    Verdoni  L, Mazza  A, Gervasoni  A,  et al.  An outbreak of severe Kawasaki-like disease at the Italian epicentre of the SARS-CoV-2 epidemic: an observational cohort study.   Lancet. 2020;395(10239):1771-1778. doi:10.1016/S0140-6736(20)31103-XPubMedGoogle ScholarCrossref
    18.
    Baker  BD. America’s Most Financially Disadvantaged School Districts and How They Got That Way: How State and Local Governance Causes School Funding Disparities. Center for American Progress. Published 2014. Accessed June 16, 2020. https://eric.ed.gov/?id=ED561094
    19.
    Simpkins  SD, Fredricks  JA, Lin  AR. Families and organized after-school activities for youth. In: APA Handbook of Contemporary Family Psychology: Applications and Broad Impact of Family Psychology, Vol. 2. 2019:235-248. doi:10.1037/0000100-015
    20.
    Brown  M, Pitt-Catsouphes  M.  A mediational model of workplace flexibility, work–family conflict, and perceived stress among caregivers of older adults.   Community Work Fam. 2016;19(4):379-395. doi:10.1080/13668803.2015.1034656Google ScholarCrossref
    21.
    The Return of the Multi-Generational Family Household. Pew Res Center’s Soc Demogr Trends Proj. Accessed July 24, 2020. https://www.pewsocialtrends.org/2010/03/18/the-return-of-the-multi-generational-family-household/
    22.
    Tucker-Seeley  RD, Li  Y, Sorensen  G, Subramanian  SV.  Lifecourse socioeconomic circumstances and multimorbidity among older adults.   BMC Public Health. 2011;11(1):313. doi:10.1186/1471-2458-11-313PubMedGoogle ScholarCrossref
    23.
    Chandler  J, Rosenzweig  C, Moss  AJ, Robinson  J, Litman  L.  Online panels in social science research: expanding sampling methods beyond Mechanical Turk.   Behav Res Methods. 2019;51(5):2022-2038. doi:10.3758/s13428-019-01273-7PubMedGoogle ScholarCrossref
    24.
    Callegaro  M, DiSogra  C.  Computing response metrics for online panels.   Public Opin Q. 2008;72(5):1008-1032. doi:10.1093/poq/nfn065Google ScholarCrossref
    25.
    Standard Definitions - AAPOR. Accessed July 24, 2020. https://www.aapor.org/Standards-Ethics/Standard-Definitions-(1).aspx
    26.
    Gee  GC, Ro  MJ, Rimoin  AW.  Seven reasons to care about racism and COVID-19 and seven things to do to stop it.   Am J Public Health. 2020;110(7):954-955. doi:10.2105/AJPH.2020.305712Google ScholarCrossref
    27.
    Raifman  MA, Raifman  JR.  Disparities in the population at risk of severe illness from COVID-19 by race/ethnicity and income.   Am J Prev Med. 2020;59(1):137-139. doi:10.1016/j.amepre.2020.04.003PubMedGoogle ScholarCrossref
    28.
    The COVID-19 pandemic: a call to action to identify and address racial and ethnic disparities. Europe PMC. Accessed July 24, 2020. https://europepmc.org/article/med/32306369
    29.
    Taylor  S, Landry  CA, Paluszek  MM, Fergus  TA, McKay  D, Asmundson  GJG.  Development and initial validation of the COVID stress scales.   J Anxiety Disord. 2020;72:102232. doi:10.1016/j.janxdis.2020.102232PubMedGoogle Scholar
    30.
    Lumley  T.  Analysis of complex survey samples.   J Stat Softw. 2004;9(1):1-19. doi:10.18637/jss.v009.i08Google Scholar
    31.
    Deville  J-C, Särndal  C-E, Sautory  O.  Generalized raking procedures in survey sampling.   J Am Stat Assoc. 1993;88(423):1013-1020. doi:10.1080/01621459.1993.10476369Google ScholarCrossref
    32.
    Bureau  UC. Current Population Survey Data Tables. The United States Census Bureau. Accessed June 15, 2020. https://www.census.gov/programs-surveys/cps/data/tables.html
    33.
    Agresti  A.  Modelling ordered categorical data: recent advances and future challenges.   Stat Med. 1999;18(17-18):2191-2207. doi:10.1002/(SICI)1097-0258(19990915/30)18:17/18<2191::AID-SIM249>3.0.CO;2-MPubMedGoogle ScholarCrossref
    34.
    Winship  C, Radbill  L.  Sampling weights and regression analysis.   Sociol Methods Res. 1994;23(2):230-257. doi:10.1177/0049124194023002004Google ScholarCrossref
    35.
    Mplus User’s Guide. Accessed June 15, 2020. https://www.statmodel.com/html_ug.shtml
    36.
    Benjamini  Y, Hochberg  Y.  Controlling the false discovery rate: a practical and powerful approach to multiple testing.   J R Stat Soc Ser B Methodol. 1995;57(1):289-300. doi:10.1111/j.2517-6161.1995.tb02031.xGoogle Scholar
    37.
    Murphy  J.  The social and educational outcomes of homeschooling.   Sociol Spectr. 2014;34(3):244-272. doi:10.1080/02732173.2014.895640Google ScholarCrossref
    38.
    Martin-Chang  S, Levesque  K. Academic Achievement. In:  The Wiley Handbook of Home Education. John Wiley & Sons, Ltd; 2017:121-134. doi:10.1002/9781118926895.ch5
    39.
    Davies  NG, Klepac  P, Liu  Y, Prem  K, Jit  M, Eggo  RM; CMMID COVID-19 working group.  Age-dependent effects in the transmission and control of COVID-19 epidemics.   Nat Med. 2020:1-7. doi:10.1038/s41591-020-0962-9PubMedGoogle Scholar
    40.
    Schools Should Prioritize Reopening in Fall 2020, Especially for Grades K-5, While Weighing Risks and Benefits | National Academies. Accessed July 20, 2020. https://www.nationalacademies.org/news/2020/07/schools-should-prioritize-reopening-in-fall-2020-especially-for-grades-k-5-while-weighing-risks-and-benefits
    42.
    Yank  V, Agarwal  S, Loftus  P, Asch  S, Rehkopf  D.  Crowdsourced health data: comparability to a US national survey, 2013-2015.   Am J Public Health. 2017;107(8):1283-1289. doi:10.2105/AJPH.2017.303824PubMedGoogle ScholarCrossref
    43.
    Leiner  DJ.  Too fast, too straight, too weird: non-reactive indicators for meaningless data in internet surveys.   Surv Res Methods. 2019;13(3):229-248. doi:10.18148/srm/2019.v13i3.7403Google Scholar
    ×