Child media use during COVID-19: Associations with contextual and social-emotional factors

Abstract

Objective:

To test associations between 1) contextual factors and types of digital media use, and 2) types of digital media use and children’s social-emotional and sleep outcomes during COVID-19.

Methods:

In February-March 2021, 303 parents of elementary schoolers participated in this cross-sectional survey gathering information on: demographics, child school format, contextual factors, duration of types of digital media use, social-emotional outcomes, and sleep. Multivariable regressions examined associations outlined in the objective, adjusting for school format, only child, race/ethnicity and parent: stress, depressive symptoms, education, material hardship.

Results:

Children were 5 to less than 11 years old and spent about 4 hours on screen media daily. In multivariable analyses, remote school format, greater material hardship, Black Indigenous, person of color (BIPOC) child race/ethnicity, lower parenting stress, and parent depressive symptoms were associated with longer duration of various digital media. Longer daily duration of streaming video and video chat were associated with higher prosocial scores, while console games, mobile apps/games, and video-sharing platforms were associated with greater problematic media use (defined as interfering with adaptive functioning). More time on mobile apps/games, video-sharing platforms, and video streaming was linked with shorter sleep.

Conclusion:

Lower parenting stress predicted greater digital media use. Greater digital media use during the pandemic may have enabled parents to focus on other needs. Use of media for social connection predicted greater prosocial behaviors. Engagement-prolonging digital media predicted problematic media use. Pediatric providers may wish to consider family context when addressing digital media use and encourage socially-oriented digital media.

INTRODUCTION

The COVID-19 pandemic prompted school closures globally, with 84% of all children in the United States being affected. These significant changes have led to increased use of digital media by children and families, to sustain connections to school and social supports when other aspects of in-person daily living were reduced. Early in the pandemic, children’s total daily duration of screen media use (outside of educational use) increased from an average of 149 minutes pre-pandemic to 199 minutes per day.¹ More recent parent surveys. have estimated a doubling of the time children spend in screen-based activities.²

Survey-based research across multiple international sites³ suggest the increase in media duration is stemming from increases in gaming, video streaming, and video chat during the pandemic. However, these have not been well-characterized in school-age samples in the U.S. Yet, the U.S. experienced a longer duration and more heterogeneous distribution of school closures as compared with these international sites, which may affect the duration and how children are using screen media.

Additionally, growing evidence suggests that contextual factors in particular may play an even more important role in shaping time spent on digital media during times of stress, such as the COVID-19 pandemic.⁴ As theorized in the Interactional Theory of Childhood Problematic Media Use, distal contextual factors such as a parents’ education level or depressive symptoms may increase the risk of a child engaging in problematic media use.⁵ Exploratory work has tested the this theory ⁵ and identified potential contextual factors during the COVID-19 pandemic, such as older child age¹ and low socioeconomic status,⁶ as contributing to increases in digital media use and problematic media use (PMU). However, these studies did not examine other pandemic-specific contextual factors that may uniquely shape digital media use, including remote learning, parenting stress, and material hardships. Understanding these associations will help clinicians develop more tailored digital media guidance that fits the needs of families in the context of the stressors they are experiencing.

Finally, little research has examined the types of screen media use during the COVID-19 pandemic that are associated with lower child well-being, including social-emotional difficulties, problematic media use, and poor sleep. Many circumstances surrounding the pandemic (e.g., social isolation, school closures) have been associated with the child mental health crisis, ¹¹ in addition to worsened child sleep with later bedtimes and more frequent sleep disturbance.⁷ However, in other work, greater duration of child media use was interestingly associated with better sleep parameters in a sample of children with neurodevelopmental disorders.⁸ This prior work did not examine specific types of digital media that contribute to sleep outcomes, which could be one possible reason for these mixed findings.

Age-inappropriate media content exposure has previously been associated with more child behavioral and sleep difficulties.⁹ Alternatively, use of video chat, texting, or video gaming with friends may be associated with social functioning, and adaptive behavior during the COVID-19 pandemic,¹ and thus might be associated with greater child well-being.¹⁰ However, no studies have identified what types of media use are most strongly associated with children’s social-emotional difficulties, PMU, and sleep problems during the COVID-19 pandemic. Yet, this is important because digital media use is a potential modifiable factor that may bely these outcomes, and different types of digital media carry different implications for children’s well-being.¹¹

Therefore, we aim to examine: 1) the media use habits of 5–10 year old children in winter 2021 of the COVID-19 pandemic; 2) associations of contextual factors such as school format, socioeconomic status (SES), material hardship, and parent stress/depressive symptoms with different types of media use; and 3) associations of different types of media use with child social-emotional difficulties, adaptive social-emotional functioning, PMU, and sleep problems.

METHODS

Study Design

We conducted a survey of parents of elementary school-aged children residing in Michigan during February-March 2021. The survey was deemed exempt from review by the University of Michigan Institutional Review Board. Parents received a $30 gift card for participation.

Participants

We recruited parents by posting study information on the University of Michigan research participant registry, social media, and distribution of study flyers by clinicians, parent-teacher organizations, and non-profit organizations. Interested parents were screened for eligibility criteria: 1) Parent or legal guardian ≥ 18 years of age; 2) child age 5.00–10.99 years; 3) parent lives with child more than half the week; 4) English speaking; 5) Michigan resident. Of 413 parents screened, 313 parents were eligible and provided online informed consent, of which 303 completed >50% of the survey and were included. Parents completed a REDCap survey (average completion time 22 minutes, median 17 minutes).

Parents reported their child’s race/ethnicity (categorized as non-Hispanic white vs. under-represented minorities); educational attainment, household size, income (from which income-to-needs [ITN] ratio was calculated). Child sex was only collected from a subsample of participants due to a database creation error; because of high missingness, this variable is omitted from Table 2 and was not included as a covariate in regression models.

Tables 2a, 2b:

Associations between contextual factors and average daily duration different digital media types. P-values are adjusted using the false discovery rate correction. Confidence intervals are pre-adjustment.

a) Characteristics	Gaming Console		Mobile apps and games		Video sharing platforms
	median [IQR] min/day or Spearman R	β (95% CI)	median [IQR] min/day or Spearman R	β (95% CI)	median [IQR] min/day or Spearman R	β (95% CI)
School format  Remote  Hybrid/in-person	22.5 [0.0;83.6] 2.1 [0.0;27.3]** p=.003	25.0 (10.3, 39.7)** Ref p =.006	41.8 [7.5;90.0] 22.5 [0.0;57.9]* p=.03	8.67 (−5.09, 22.41) Ref p=.33	45.0 [0.0;113.6] 28.9 [0.0;75.0] p=.07	8.54 (−9.43, 26.5) Ref p=.48
Parent Education  HS or some college  4-year college degree +	22.5 [0.0;90.0] 2.1 [0.0;42.3]** p=.004	−6.49 (−34.2, 21.2) Ref p=.75	45.0 [11.8;107.1] 22.5 [ 0.0;58.4]** p=.006	−7.09 (−32.9, 18.7) Ref p=.70	90.0 [41.8;167.1] 11.8 [0.0;61.6]*** p<.001	−10.8 (−44.5, 23.0) Ref p=.65
Hardship score (per 1 point increase)	0.105 p=.15	5.80 (−0.03, 11.6) p=.12	0.189** p=.006	7.04 (1.61, 12.5)* p=.04	0.299*** p<.001	17.4 (10.3, 24.5)*** P<.001
Child race/ethnicity  Non-Hispanic White  Underrepresented Minority	11.8 [0.0;57.9] 7.5 [0.0;57.9] p=.94	5.29 (−10.7, 21.3) Ref p=.64	22.5 [2.1;57.9] 45.0 [5.4;107.1]** p=.006	−19.5 (−34.5, −4.6)* Ref p=.04	22.5 [0.0;90.0] 57.9 [9.6;165.0]** p=.004	−17.1 (−36.6, 2.45) Ref p=.18
PSI-SF score (per 1 point increase)	−0.081 p=.27	−0.94 (−1.73, −0.15) p=.06	−0.058* p=.02	−0.75 (−1.49, −0.15) p=.11	−0.029 p=.71	−0.72 (−1.68, 0.25) p=.25
CES-D score (per 1 point increase)	0.1 p=.17	0.37 (−0.37, 1.14) p=.29	0.195** P=.005	0.73 (0.04, 1.41) p=.10	0.257*** p<.001	1.63 (0.74, 2.54)** p=.003

b) Characteristics	Streaming Video		Video chat		Total Media Duration
	median [IQR] min/day or Spearman R	β (95% CI)	median [IQR] min/day or Spearman R	β (95% CI)	median [IQR] min/day or Spearman R	β (95% CI)
School format  Remote  Hybrid/in-person	67.5 [28.9;132.9] 57.9 [28.9;107.1] p=.38	2.99 (−14.6, 20.6) Ref p=.82	7.0 [0.0;24.1] 0.0 [0.0; 7.5]** p<.001	10.3 (1.35, 19.3) Ref p=.07	279.6 [157.5;423.2] 163.9 [109.8;310.7]*** p<.001	56.2 (5.72, 106.6) Ref p=.08
Parent Education  HS or some college  4-year college degree +	90.0 [32.1;150.0] 57.9 [28.9;107.1]* p=.02	−2.99 (−33.6, 32.6) Ref p=.98	5.4 [0.0;15.5] 0.0 [0.0;17.1] p=.72	7.34 (−9.48, 24.17) Ref p=.18	332.1 [212.1;528.8] 184.3 [106.1;304.3]*** p<.001	1.96 (−92.78, 96.70) Ref p=.98
Hardship score (per 1 point increase)	0.176* p=.01	4.70 (−2.26, 11.7) p=.30	−0.001 p=.99	2.49 (−1.06, 6.03) p=.28	0.304*** p<.001	45.1 (25.1, 65.0)*** P<.001
Child race/ethnicity  Non-Hispanic White  Other	57.9 [28.9;107.1] 90.0 [42.3;150.0]* p=.02	−26.3 (−45.4, −7.15)* Ref p=.03	0.0 [0.0;11.8] 2.1 [0.0;27.3] p=.45	−4.56 (−14.29, 5.18) Ref p=.48	208.9 [111.4;324.6] 308.6 [167.1;520.7]*** p<.001	−88.8 (−143.6,−34.0)** Ref p=.009
PSI-SF score (per 1 point increase)	−0.05 p=.50	−1.12 (−2.07, −0.18) p=.06	−0.021 p=.81	−0.29 (−0.78, 0.18) p=.34	−0.081 p=.27	−4.47 (−7.17, −1.77)** p=.007
CES-D score (per 1 point increase)	0.16* p=.02	1.17 (0.29, 2.05)* p=.04	0.086 p=.24	−0.06 (−0.51, 0.39) p=.89	0.293*** p<.001	3.98 (1.45, 6.50)* p=.01

^*p <.05,

^**p<.01,

^***p<.001 for Spearman correlations or Wilcoxon Rank-Sum tests

^aModels adjusted for all variables in the table, as well as child and parent age, parent marital status (married/partner vs single/separated/divorced), only child (yes v no)

Abbreviations: HS, PSI-SF, CES-D, IQR, aOR, CI, Ref

Survey –Contextual Factors

Socioeconomic status and hardship.

Food insecurity, housing insecurity, and financial stress were assessed with items from the Pediatric Adverse Childhood Experiences and Related Life Events Screener¹² and American Academy of Family Physicians Social Needs Screening Tool.¹³ We calculated a material hardship score by summing whether parents endorsed any food, housing, or financial stress, job loss, or receipt of public assistance during this school year and whether ITN was below 200% of the federal poverty level. Parents also reported demographic information as potential covariates.

School format.

Parents reported the main type of schooling their child received during the 2020–2021 school year, including whether the school is public, private, or charter; mostly in-person (with or without brief periods of remote learning); mostly hybrid; mostly remote; or homeschooling. Remote learning was defined as “accessing the school curriculum, assignments, and learning activities through a computer or electronic device.” For comparisons of school format, we excluded 12 homeschooled participants. Hybrid and in-person learners were included in one group, as there were no substantial differences between groups.

Parent stress and Depressive symptoms.

Parents completed the Parental Stress Index-Short Form (PSI-SF)¹⁴ (α = 0.88), a validated measure of parent stress in which parents rate agreement with items such as “I feel trapped by my responsibilities as a parent” on a 5-point Likert scale. They also completed the Center for Epidemiologic Studies Depression Scale¹⁵ (CES-D, α = 0.92) to assess parent depressive symptoms a widely-used and validated measure that asks parents to rate frequency of items such as “I was bothered by things that usually don’t bother me” from 0 (rarely) to 3 (all of the time).

Survey - Media use:

Parents reported the duration of different types of entertainment media use on weekdays and weekends, in which they were asked to answer: “Please estimate the amount of time your child spends on an average weekday/[weekend], at home, doing the following media activities (do not count the time they are doing schoolwork on a computer or a tablet).” Answer options included: (0; 1–15 minutes; 31–60 minutes; 1–2 hours; 2–3 hours; 3–4 hours; 4–5 hours; 5+ hours; don’t know/prefer not to answer). Types of entertainment use include: console and computer games, live TV, mobile apps and games, podcasts, social media video sites (e.g., YouTube or TikTok), streaming video platforms (e.g., Netflix and Prime Video), text messaging, and videochatting with friends. Questions were adapted from questions frequently used in the CAFÉ consortium, a digital media research group.¹⁶

To calculate a daily duration of each type of media, we assigned the median value for each category (e.g., 45 minutes for the 31–60 minute category). We calculated a weighted daily average for each type of media: [(console_time_wkday*5)+(console_time_wkend*2)]/7. We then summed each media type’s average daily duration to calculate an overall media duration variable; however, this may overestimate true usage since children might use more than two forms of digital media simultaneously.

We created several questions to capture child media use challenges, including: starting social media earlier than the parent intended, stumbling across creepy content, being distracted by media, or being contacted by an adult stranger online. Response options were: yes, no, and don’t know/prefer not to answer.

Survey- Problematic Media Use Measure:

Parents completed the Problematic Media Use Measure-Short Form (PMUM-SF), a validated and reliable¹⁷ 9-item measure of maladaptive screen media behaviors. Items are derived from the criteria for Internet Gaming Disorder from the DSM-5, with questions such as: “It is hard for my child to stop using screen media,” rated on a 5-point Likert scale (1=never to 5=always), and averaged for use as a continuous variable (α = 0.92). The PMUM-SF was used as a social-emotional outcome due to prior associations with psychosocial functioning, conduct difficulties, hyperactivity, and peer relationship difficulties.¹⁷ The PMUM-SF items were derived from exploratory factor analysis. Please see Appendix 1 for the PMUM-SF items.

Survey – Child Behavior and Sleep Outcomes

Child behavior.

Parents completed the Strengths and Difficulties Questionnaire (SDQ),¹⁸ a validated 25-item survey which has a subscale for Prosocial Behaviors (α = 0.80) and a Total Difficulties score (α = 0.83), with prompts such as: “Considerate of other people’s feelings” rated on a 3-point Likert scale (1=not true, 3=certainly true). Scores were examined as continuous variables.

Child sleep.

Childhood sleep questions were adapted with pandemic-specific language from the Childhood Sleep-Wake questionnaire,¹⁹ a reliable and validated survey on sleep, with questions such as “On average, how long does it take your child to fall asleep on a typical night? If you’re not sure, please give your best estimate.” Parents reported the usual time their child goes to bed at night and wakes in the morning on weekdays and weekends, from which we approximated sleep duration, as has been done in previous studies.²⁰ Parents also reported the number of overnight awakenings on a typical night that they were aware of.

Statistical Analysis

Spearman correlations and Wilcoxon rank-sum tests were used for bivariate analyses, which informed covariates to include in multivariable analyses. Linear and logistic regression was used for multivariable analyses. Linearity and homoscedasticity assumptions of the regression models were verified using diagnostic plots. Because of the high number of correlations explored, we included a false discovery rate correction (FDR) to reduce the risk of a Type 1 error. 95% confidence intervals represent comparison-wise confidence of the estimates and the p-values are adjusted for multiple comparisons). Statistical analyses were performed using R version 4.0.5.

RESULTS

As shown in Table 1, parents were 89.4% female, average age 38.4 years (SD = 5.6), 69.9% held a college degree or greater, 86.1% were married or lived with a partner. Children were on average 7.3 years (SD = 1.6). Children were 72.8% white, non-Hispanic; 6.4% Black, non-Hispanic; and 11.1% Hispanic any race. 59.9% of children were mostly virtual or remote, 12.5% were mostly hybrid, and 27.5% were mostly full-time in school receiving face-to-face instruction.

Table 1.

Sociodemographic characteristics of 303 parents and elementary school-aged children enrolled in the present study.

Characteristic	Mean (SD), range or n (%)
PARENT
Parent sex
Female	271 (89.4%)
Male	32 (10.6%)
Parent/guardian age (years)	38.4 (5.6)
Parent education
High school/GED/less than HS	92 (30.1%)
College degree or greater	211 (69.9%)
Parent marital status
Married/live with a partner	260 (86.1%)
Other	42 (13.9%)
Parent occupation/location
Stays at home or unemployed	101 (34.2%)
Works full/part-time outside the home	98 (32.9%)
Works full/part-time from home	98 (32.9%)
CHILD
Child age (years)	7.3 (1.6)
Child grade in school
Young 5s/Kinder-ready	13 (4.3%)
Kindergarten	56 (18.5%)
1st	75 (24.8%)
2nd	52 (17.2%)
3rd	48 (15.8%)
4th	38 (12.5%)
5th	21 (6.9%)
Child race/ethnicity
White, non-Hispanic	217 (72.8%)
Black/African American, non-Hispanic	19 (6.4%)
Hispanic or Latinx, any race	33 (11.1%)
Asian or Pacific Islander	14 (4.7%)
More than one race, non-Hispanic	15 (5.0%)
Only child	52 (17.2%)
School format
Mostly full-time in-school, face-to-face instruction	79 (27.5%)
Mostly hybrid instruction	36 (12.5%)
Mostly virtual/remote instruction	172 (59.9%)
HOUSEHOLD
Income-to-needs ratio	3.3 (1.7)
Hardship score	0.9 (1.4)
PSI-SF	39.6 (9.7)
MEDIA	Average minutes/day (SD)
Console-based video games	39.4 (64.3)
Live TV	20.8 (42.3)
Mobile apps/games	49.8 (61.2)
Podcasts	6.0 (25.0)
Video sharing platforms (e.g., YouTube, TikTok)	65.0 (83.4)
Streaming services (e.g., Disney+, Netflix, Hulu)	85.8 (74.1)
Text message	8.9 (26.3)
Video chat (e.g., FaceTime, Skype, Messenger Kids)	17.5 (35.8)
Total screen media duration	293.2 (248.8)
PMUM Score	2.56 (0.94)
Media environment
Family has access to home WiFi	299 (98.0%)
Child has virtual friends	137 (45.2%)
Child exposure to creepy or inappropriate content	151 (49.5%)
Child has used violent apps	55 (18.2%)
Child used social media younger than planned	95 (31.4%)
During online instruction, child ever distracted with other apps	70 (26.6%)
Child has been contacted by strangers online	11 (3.6%)
Barriers to remote learning
Problems with WiFi	95 (36.0%)
Lack of access to laptop or tablet	15 (5.7%)
Difficulties with chosen technology platform (e.g.: Zoom)	85 (32.2%)
Lack of adult availability for supervision	98 (37.1%)
Difficulties teaching children assigned work	57 (21.6%)
Average sleep duration	10.0 (0.8)

On average, children were spending about 4 hours total on screen media per day, outside of educational uses. As shown in Table 1 and Appendix 2, children were spending most of their time viewing streaming videos on services such as Netflix, on online video sharing platforms such as YouTube or TikTok, and on mobile apps/games (49.8 minutes/day). Almost half of parents reported that their child had come across/were exposed to creepy or inappropriate content online. Almost one-third of children started using social media younger than parents had planned for, 26.6% of children were distracted with other apps during remote learning, and 3.6% of children had been contacted by a stranger online.

Several contextual factors were associated with higher duration of different media types when adjusting for covariates (Table 2). School format being remote compared with in-person was associated with greater time spent using console and computer games (β = 25.0 [10.3, 39.7, p=.006]). Greater material hardship score was associated with more mobile app and game usage (β =7.04 [1.6, 12.5], p=.04), more use of social media video platforms (β =17.4 [10.3, 24.5], p<.001), and total screen media duration (β=45.1 [25.1, 65.0], p<.001). Child non-Hispanic white race/ethnicity compared with minoritized race/ethnicity was associated with a shorter duration of mobile apps and games (β =−19.5 [−34.5, −4.6], p=.04), streaming video platforms (β =−26.3 [−45.4, −7.15], p=.03), and total screen media duration (β =−88.8 [−143.6,−34.0], p=009). Greater parenting stress was associated with less total screen media duration (β =−4.47 [−7.17, −1.77], p=.007). Greater parent depressive symptoms were associated with a small but significant increase in video sharing platforms (β =1.63 [0.74, 2.54], p=.003), streaming video (β =1.17 [0.29, 2.05], p=.04), and total screen media duration (β =3.98 [1.45, 6.50], p=.01). Please see Appendix 3 for additional associations between contextual factors and types of digital media.

Regarding digital media and social-emotional outcomes (Table 3), after adjusting for multiple confounders, greater use of mobile apps and games was associated with higher SDQ total scores (β =0.91 [0.21, 1.61], p=.04). Greater duration of using streaming video platforms (β =0.26 [0.07, 0.46], p=.04), video chat (β =0.57 [0.18, 0.95], p=.02), and total screen media duration (β =0.10 [0.03, 0.16], p=.02) were associated with higher SDQ prosocial behaviors. Console games (β =0.14 [0.03, 0.24], p=.04), mobile apps/games (β =0.26 [0.15, 0.37], p<.001), video-sharing platforms (β =0.24 [0.16, 0.33], p<.001), and total screen media duration (β =0.08 [0.05, 0.11], p<.001) were associated with greater PMUM scores.

Table 3.

Types of digital media use (average duration in hours/day) predicting social-emotional and sleep outcomes. Models are adjusted for school format, parent age, marital status, only child, parenting stress, CES-D, parent education, material hardship, and race/ethnicity.

	PMUM	Prosocial	SDQ total	Sleep duration	Night time awakenings
	β (95% CI)	β (95% CI)	β (95% CI)	β (95% CI)	aOR (95% CI)
Console games	0.14 (0.03, 0.24)* p=.04	0.24 (−0.002, 0.47) p=.12	0.13 (−0.54, 0.79) p=.81	−0.09 (−0.19, 0.004) p=.06	0.79 (0.60, 1.05) p=.20
Mobile apps/games	0.26 (0.15, 0.37)*** p<.001	−0.09 (−0.34, 0.17) p=.62	0.91 (0.21, 1.61)* p=.04	−0.25 (−0.35, −0.15)*** p<.001	1.18 (0.89, 1.57) p=.36
Video sharing platforms	0.24 (0.16, 0.33)*** p<.001	0.17 (−0.03, 0.36) p=.19	0.24 (−0.31, 0.79) p=.51	−0.17 (−0.24, −0.09)*** p<.001	1.09 (0.88, 1.36) p=.53
Streaming Video	0.08 (−0.1, 0.17) p=.18	0.26 (0.07, 0.46)* p=.04	0.19 (−0.37, 0.75) p=.63	−0.11 (−0.19, −0.03)** p=.007	1.40 (1.11, 1.77)* p=.02
Video chat	0.09 (−0.08, 0.26) p=.45	0.57 (0.18, 0.95)* p=.02	−0.75 (−1.83, 0.33) p=.28	−0.15 (−0.30, 0.008) p=.06	1.39 (0.88, 2.20) p=.27
Total Media Duration	0.08 (0.05, 0.11)*** p<.001	0.10 (0.03, 0.16)* p=.02	0.14 (−0.05, 0.34) p=.26	−0.07 (−0.10, −0.05)*** p<.001	1.10 (1.01, 1.20) p=.07

^*p <.05,

^**p<.01,

^***p<.001 for Spearman correlations or Wilcoxon Rank-Sum tests. P-values are adjusted using the false discovery rate correction. Confidence intervals are pre-adjustment.

As shown in Table 3, more time on mobile apps/games (β =−0.25 [−0.35, −0.15], p<.001), video sharing platforms (β =−0.17 [−0.24, −0.09], p<.001), video streaming (β =−0.11 [−0.19, −0.03], p=.007), and total screen media duration (β =−0.07 [−0.10, −0.05], p<.001) were associated with shorter sleep duration. Greater video streaming (aOR =1.40 [1.11, 1.77], p=.02) was associated with higher odds of nighttime awakenings. Please see Appendix 4 for additional types of digital media and social-emotional outcomes.

DISCUSSION

Prior work has found that the COVID-19 pandemic has amplified the stressors that families experience in many ways (e.g., financial stressors, caregiving burden, lack of in-person social supports),²¹ while also increasing children’s digital media use.² Therefore, the COVID-19 pandemic may be an important lens through which to examine the contextual factors that shape children’s digital media experiences.⁵ Several contextual factors shaped how children use digital media during the COVID-19 pandemic. In turn, some of the ways that children used digital media were associated with social-emotional outcomes. Particularly, digital media that was highly gamified or behaviorally reinforcing (e.g., mobile apps and games and video sharing platforms), was associated with greater problematic media use and lower sleep duration during the pandemic.

Prior work has found that attendance in remote schooling is associated with greater duration of digital media use, even outside from educational uses.^1,3 After controlling for the false discovery rate, the association between remote schooling and greater total duration of digital media use was null. However, we did find that remote schooling was associated with greater time spent on console games. This may be related to the fact that gaming consoles are more accessible in the home environment, and would likely not be an option at school for in-person learners.

This study was the first to examine material hardships and associations with types of digital media use during the pandemic, finding strong associations between material hardships and greater time spent on mobile apps and games and video-sharing platforms, possibly due to the lack of access to broadband (e.g., WiFi) and hardware (e.g., desktop computer) that low-income families face.²² Indeed, during the pandemic, prior work has found that children from low-income homes were more likely to use mobile devices for completion of schoolwork, as compared with desktop devices.²² It is also possible that these mobile apps/games and video sharing platforms are sources of free content, which may be more accessible for low-income families; however, these types of media have the disadvantages of less developmentally-appropriate programming and greater commercialized content.⁹

White, non-Hispanic race/ethnicity was associated with less screen media use overall, in addition to a lower duration of live TV, mobile apps, and streaming video. These types of digital media use could be accessed through a mobile device and could potentially be accessed without reliable broadband internet connection. These discrepancies likely reflect a digital divide stemming from systemic divestments in housing, education, and infrastructure for minorities, which has included WiFi access.²³ However, despite access to less WiFi, minority populations are generally exposed to greater duration of screen media due to fewer structural resources to support children’s activities outside of screen media and higher rates of virtual learning.²⁴ Furthermore, higher COVID-19 mortality rates for Black and Hispanic families, may have shaped their decision-making on how many in-person activities they could safely access while balancing the risk of COVID-19. Time spent on digital media could be a safer substitute for these in-person activities.

Interestingly, we found lower parenting stress was associated with greater total child screen media use, which differed from prior findings of associations between greater parenting stress and greater duration of child media use during the COVID-19 pandemic.²⁵ It is possible that the directionality of these associations was starting with children’s use of digital media, which may have provided some relief for parents facing other higher-acuity stressors in their lives during the pandemic. Use of digital media may have enabled parents to focus on such demands, thereby reducing their overall stress.

Notably, there were associations between greater time on streaming apps, greater time spent video chatting, and greater overall duration of digital media use with prosocial skills/peer competence. During the COVID-19 pandemic, it is possible this provided opportunities to connect with peers, or that streaming videos may offer more prosocial options and developmentally-appropriate content, as compared with video sharing platforms such as YouTube.⁹ One limitation is that this work did not assess for the content of what children were streaming or viewing, yet this is an important consideration given prior associations between prosocial content and children’s peer competency.²⁶ That these associations are occurring during the time of the COVID-19 pandemic might represent a moment in time where digital connection is important for facilitating peer relationships, particularly when in-person opportunities are not available. It is also possible that children with greater peer competence seek out more texting and video chatting as a means of meeting their social needs.²⁷

On the other hand, greater use of mobile apps and games was associated with greater child behavioral difficulties (e.g.: difficulties with emotion regulation and peer relational challenges). Prior work has found that children transitioning away from tablet devices may exhibit greater behavioral dysregulation after play.²⁸ More gamified apps may prolong children’s engagement at the detriment of supplanting opportunities for engagement with peers, free play, or other activities that promote behavioral regulation. It is also possible that children with greater behavioral difficulties may gravitate toward mobile apps and games, because they feel greater self-efficacy in this domain, when these gamified apps are providing strong behavioral reinforcements.

Overall, the strongest associations with media types and social-emotional outcomes were with PMU. The most engagement-prolonging types of screen media were associated with PMU (e.g., console games, mobile apps, videosharing platforms). These types of digital media tend to be highly gamified with frequent and strong behavioral reinforcement (e.g. social component, appealing rewards, triggers or prompts to re-engage) that may contribute to more problematic use, consistent with the Interactional Theory of Childhood Problematic Media Use.⁵

Incongruent to prior work in children with neurodevelopmental disorders, duration of digital media use was associated with worse sleep outcomes.²⁹ Mobile apps and games, video sharing platforms, and video streaming were all negatively associated with sleep duration, which may possibly be related to mobile devices being present in the child’s room, in addition to auto-play which may offer endless recommendations and seamless delivery. In addition to the engagement-promoting nature of these types of screen media, there may be more social pressures to lengthen time spent on mobile devices with texting.³⁰ Awakenings were associated with more streaming video content, which may be related to sleep-onset associations – if children are falling asleep with streaming video content, they may have a greater duration of nighttime awakening and difficulty returning to sleep if the sleep environment has changed (e.g.: if video streaming content is turned off during the night). It is also possible that scary media content may contribute to nightmares, contributing to sleep disturbance.

Limitations include the use of a convenience sample, which may limit generalizability; that content of digital media was not assessed; and that child race/ethnicity was dichotomized which may not capture the diversity of experiences between minority groups. Future work may wish to examine longitudinally how children’s digital media use during the pandemic may associate with future digital media use patterns and children’s social-emotional outcomes, and also explore prosocial behaviors on digital media.³¹

CONCLUSION

The COVID-19 pandemic has resulted in substantial challenges for families, which may impact how children and families are engaging with digital media. Notably, our work found associations with parenting stress and digital media duration which may reflect the considerable outside stressors for families that have emerged during this time, and that digital media may serve an adaptive purpose for stress relief during the pandemic. Pediatric providers may wish to recommend ways of engaging over digital media that are prosocial (e.g., video chatting), as well as creating boundaries around more gamified types of digital media, such as mobile apps/games which are more associated with problematic media use and poor sleep duration.

Abbreviations:

COVID-19

Coronavirus disease 2019

BIPOC

Black, Indigenous, People of Color

IT-CPU

Interactional Theory of Childhood Problematic Media Use

ADHD

Attention Deficit Hyperactivity Disorder

PMU

problematic media use

PMUM

Problematic Media Use Measure

SES

socioeconomic status

ITN

Income-to-needs ratio

PSI-SF

Parenting Stress Index- Short Form

CES-D

Center for Epidemiologic Studies Depression Scale

SDQ

Strengths and Difficulties Questionnaire