children

Article
Beyond Screen Time: The Different Longitudinal Relations
between Adolescents’ Smartphone Use Content and Their
Mental Health
Shunsen Huang 1,† , Xiaoxiong Lai 1,† , Yajun Li 2 , Yang Cui 1 and Yun Wang 1, *

1 State Key Laboratory of Cognitive Neuroscience and Learning, Faculty of Psychology, Beijing Normal
University, Beijing 100875, China
2 Guangming Institute of Education Sciences, Shenzhen 518107, China
* Correspondence: wangyun@bnu.edu.cn
† These authors contributed equally to this work.

Abstract: Purpose: Previous studies focusing on the relationship between adolescents’ screen
time and mental health have uncovered contradictory results. By focusing on smartphone use
content (SUC), this study uses specification curve analysis to explore the different effects of
SUCs on mental health-based on longitudinal data. Methods: A total of 2552 adolescents were
surveyed in the first (July 2020) and second year (April 2021). A total of 2049 eligible participants
(average age = 14.39 ± 2.27, female = 1062) are included in the analysis. Participants reported
20 types of content used by them during smartphone use and their mental health (depression,
anxiety, and somatization). Specification curve analysis was used to examine the longitudinal
relationship between SUCs and their mental health. Results: Smartphone use for listening to
music (median β = 0.18, p < 0.001, NSRPD = 25/27, p < 0.05), chatting online (median β = 0.15,
p < 0.001, NSRPD = 24/27, p < 0.05), watching TV (median β = 0.14, p < 0.001, NSRPD = 24/27,
p < 0.05), and playing games (median β = 0.09, p < 0.001, NSRPD = 19/27, p < 0.05) produce high to
medium negative effects on subsequent mental health. Only using smartphones for online courses
exerts no effect on their subsequent mental health (median β = 0.01, p > 0.05, NSRPD = 0/27,
p > 0.05). The left 15 types of smartphone content showed unstable effects on future mental health.
Citation: Huang, S.; Lai, X.; Li, Y.; Depending on the types of content used, these effects ranged from high, medium, and small to
Cui, Y.; Wang, Y. Beyond Screen Time:
none. The relatively descending order of effect on mental health is listening to music, chatting
The Different Longitudinal Relations
online, watching TV, playing games, and types of content (e.g., browsing social media, making
between Adolescents’ Smartphone
payments, reading online novels) with high but unstable effects, types of content with medium
Use Content and Their Mental Health.
Children 2023, 10, 770. https://
(e.g., browsing news and posting/sharing) but unstable effects, types of content (e.g., using the
doi.org/10.3390/children10050770 camera, obtaining life information, and making calls) with small but unstable effects, such as
finishing homework and taking online courses. Conclusions: This study enlightens researchers
Academic Editors: Oana A. David
and policymakers to update their understanding of adolescents’ technology use, especially to
and Marie Danet
adopt a differentiated attitude towards different media use content. As nutritionists often do,
Received: 16 February 2023 a “nutritionally balanced” digital diet for young people should be recommended to the public,
Revised: 7 April 2023 rather than just suggesting limits on the amount of time they can spend using digital media.
Accepted: 17 April 2023
Published: 24 April 2023 Keywords: smartphone use content; adolescent mental health; specification curve analysis; depres-
sion; anxiety; somatization

Copyright: © 2023 by the authors.

Licensee MDPI, Basel, Switzerland.
1. Introduction
This article is an open access article
distributed under the terms and From 2016 to 2022, global smartphone use has increased, with nearly 6.57 billion
conditions of the Creative Commons people worldwide becoming smartphone users in 2022 [1]. According to the 46th Sta-
Attribution (CC BY) license (https:// tistical Report on Internet Development in China, as of April 2020, approximately 169
creativecommons.org/licenses/by/ million Chinese individuals under 19 years old access the Internet via smartphones [2].
4.0/). Through smartphones, people can play online games, chat with friends, watch videos,

Children 2023, 10, 770. https://doi.org/10.3390/children10050770 https://www.mdpi.com/journal/children

Children 2023, 10, 770 2 of 12

access social media, search for information, shop online, and more. These functions
facilitate people’s daily lives. However, academic researchers have debated whether
digital technology improves adolescents’ well being or has a negative or no influence on
their mental health.
The extant research on the topic of the effect of digital technology on adolescents’
mental health or well being has uncovered controversial results. Some researchers have
found that adolescents’ media use may promote their mental health, as Valkenburg
et al. found that receiving positive feedback on social networking sites may improve
adolescents’ mental health [3]. Other researchers have argued that digital media use may
exert a weak effect on adolescents’ mental health [4]. Specification curve analysis (SCA)
has been recently used to explore the relationship between screen time and adolescents’
well being and found little evidence of substantial negative associations, they believe
that these effects are too small to warrant policy change [5]. A host of researchers have
argued that adolescents’ digital media use harms their well being [6,7]. Several recent
longitudinal studies have shown that screen time [8], the frequency of Facebook use [9],
social media use time [10], and internet use time [11] negatively predicted adolescents’
future well being. Even so, many of the existing studies on the relationship between
smartphone/social media use and mental health were cross-sectional and are vulnerable
to reverse causation issues [12].
There exist some gaps in this field that may contribute to these controversial debates.
First, screen time is no longer a useful construct, but it still dominates research and public
discourse, and, thus, changing the conversation to more accurately reflective aspects is
necessary [13], as researchers have argued that it is more likely that what adolescents post
and view online rather than screen time (time children spend on digital media) is associated
with low mental health [14]. For example, more nuanced studies of adolescents’ online
activities have shown that it is not the frequency but the type of social media use that is
associated with their depression [15]. Second, to some extent, these controversial results
may reflect a crisis of reproducibility in psychological science due to selective reporting,
selective analysis, and inadequate description of the conditions necessary to obtain the
results [16]. For example, using the same datasets and the same method but different choices
in selecting independent or dependent variables, researchers found a greater effect than in
previous studies with the same datasets and method [17]. Specification curve analysis (SCA)
can be used to alleviate this problem by exhausting all the potential combinations between
variables to obtain the robustness of interesting relationships between variables [18]. Third,
previous research does not distinguish between the effects of different types of digital
technologies (e.g., smartphones, TVs, computers, and tablets) on adolescents’ mental
health [5]. Different digital technologies with rapidly changing content may have different
effects on adolescents’ mental health. As Twenge argued, some studies were conducted
before smartphones became common and before the significant increase in the usage of
digital media [6]. In addition, the National Study on Internet Use by Minors pointed
out that 93.9% of teenagers in China access the Internet through smartphones, while the
proportions of teenagers accessing the Internet through computers, laptops, TVs, and
tablets are 45%, 31.5%, 56.7%, and 28.9%, respectively [19]. Therefore, it is necessary
to focus on smartphones, the most influential technology in Chinese daily life. Finally,
previous research using SCA based on cross-sectional data could not infer causality. No
studies have applied SCA to longitudinal design in this field. Thus, a longitudinal cohort
design is urged to draw valid inferences about the magnitude of different SUC effects on
adolescents’ mental health.
This study uses specification curve analysis to explore the longitudinal effects of
different SUCs on teens’ mental health. The complementary–interference model [20,21]
argues that smartphone use has many benefits like online communication or entertain-
ment, while using smartphones for accessing valuable information, entertainment, and
communication can not only distract people from giving full attention to friends, family,
and other activities in their social environment but also supplant casual (face-to-face)
Children 2023, 10, 770 3 of 12

social interactions. Both distraction and supplantation will reduce engagement with
potential sources of well being concurrently available in the environment, offset the
potential benefits of smartphones, and decreases people’s well being [20,21]. Therefore,
we assumed that using smartphones for communication (e.g., chatting online), accessing
valuable information (e.g., using search engines), and entertainment (e.g., playing games
or watching TVs) have negative effects on future mental health. Except, adolescents
can also use smartphones for study purposes. The online participation theory of online
learning suggests that online learning is a process of online participation that includes
taking part and maintaining relations with others and collaborating with others [22].
Such participation is usually related to good academic results [22]. Thus, we hypothesize
that using the smartphone for online learning purposes (e.g., using the smartphone for
homework or learning) may be less likely to relate to low mental health. Therefore, the
20 types of SUC should exert different effects on adolescents’ mental health.

2. Materials and Methods

2.1. Participants
2.1.1. Study Design
The present study is a two-wave cohort study, aiming at exploring the longitudinal
effects of different SUCs on adolescents’ mental health. Adolescents were from 55 classes
(4 or 5 random classes per school) of 12 randomly selected high schools in Henan province,
China. With the help of teachers, adolescents with experience using smartphones were
invited. The specific recruitment procedure is that, in the first year (July 2020) and the
second year (April 2021), 2552 adolescents were surveyed, and adolescents and parents were
matched by the given identification numbers and names. Nearly 95% of the adolescents in
selected classes gave their responses.

2.1.2. Recruitment
In the first year, influenced by the COVID-19 pandemic, a link to the online question-
naire was sent to the participants and their parents through WeChat, a popular social media
platform that is easily accessible to adolescents in China. One parent (mother or father)
of each participant reported some demographic information, such as parents’ education,
annual family revenue, and the subjective socioeconomic status (SES) of the family. In the
second year, adolescents completed a questionnaire independently at school.
A total of 2552 parent–adolescent dyads were surveyed, and 149 dyads were ex-
cluded due to incomplete questionnaires (e.g., only providing basic demographic in-
formation (e.g., gender) in adolescents’ questionnaire) (final eligible N = 2403). Then,
2403 parent-adolescent dyads were surveyed in the second year, and 354 dyads with
incomplete questionnaires were excluded. Therefore, the final participants contained
in the later analysis totaled 2049 (average age = 14.39 ± 2.27 years, female = 1062).
Little’s test (MCAR test) showed that the missing pattern (χ2 = 133.815, df = 147, p > 0.05)
belongs to MCAR, and the average missing rate across variables was 0.12%. Thus, the
EM algorithm imputation method [23] was used to process the missing data. See Table 1
for demographic information about the participants.
Table 1. Demographic information of adolescents.

Variables Groups Percentage (%)

City 47%
Residence Township 15.3%
Rural region 37.7%
Yes 91%
Only child
No 9%
<College 90.5%
Mother’s education
≥College 9.5%
Children 2023, 10, 770 4 of 12

Table 1. Cont.

Variables Groups Percentage (%)

<College 87.5%
Father’s education
≥College 11.4%
<50,000 ¥ 1 60.2%
Annual income 50,000–100,000 ¥ 21.6%
>100,000 ¥ 19.2%
1 ¥ = RMB.

2.1.3. Ethical Issues

This study was approved by the IRB of the State Key Laboratory of Cognitive Neuro-
science and Learning, Beijing Normal University, Beijing, China (CNL_A_0003_003; 25 July
2018). All adolescents and their parents gave their consent.

2.2. Measurements
2.2.1. Demographic Information
The demographic information included age, gender, residence, only child, education
level of parents (both mother and father), as well as self-reported social economic
status (SES).

2.2.2. Smartphone Use Content

The SUC was measured by a revised Mobile Phone Use Pattern Scale for Chinese
students [24,25]. The revised questionnaire mainly measures smartphone use content
related to playing games on smartphones, chatting online via smartphones, accessing news,
life-related information via smartphones, etc. This revised questionnaire included 20 items
(α = 0.93), and each item measured the different smartphone use content (e.g., reading
online novels, playing games, listening to music, watching TV, making calls, chatting
online, browsing social media, watching clips, posting/sharing, using dictionaries, taking
online courses, finishing homework, etc.). Detailed information on the items is presented
in Table 2. This questionnaire used a 4-point Likert scale (1 = never, 4 = frequently). Each
item indicates an independent construct of a specific SUC, and one-item measurement
usually has a similar construct to multiple-item measurement [26]. The use scale measures
20 different SUCs.
Table 2. The abbreviation of SUC items.

Abbreviation of SUC Items M ± SD

1. Watching TV, etc. 2.12 ± 0.82
2. Watching clips 2.07 ± 0.85
3. Using utilities 2.21 ± 0.80
4. Using search engines 2.16 ± 0.76
5. Using fitness apps 1.77 ± 0.78
6. Using dictionaries 2.16 ± 0.76
7. Using camera 2.05 ± 0.70
8. Taking online courses 2.28 ± 0.75
9. Reading online novels 1.88 ± 0.82
10. Posting/sharing 1.64 ± 0.71
11. Playing games 2.01 ± 0.83
12. Obtaining life information 2.17 ± 0.83
Children 2023, 10, 770 5 of 12

Table 2. Cont.

Abbreviation of SUC Items M ± SD

13. Making payments 1.91 ± 0.91
14. Making calls 2.16 ± 0.73
15. Listening to music 2.35 ± 0.89
16. Finishing Homework 2.06 ± 0.76
17. Consuming online 1.58 ± 0.72
18. Chatting online 2.30 ± 0.88
19. Browsing social media 2.00 ± 0.83
20. Browsing news 1.99 ± 0.91
Note: JD and Taobao = online shopping apps. WeChat and QQ = online chat apps. Qzone, Weibo, and WeChat
Moments = social media apps. Kuaishou, Volcano video, and TikTok = video-sharing social networking apps.
The full information on SUC items can be found in the work of Huang et al. [25], permission has been obtained
and there is no copyright issue. For example, the full item description for Taking online courses is “Use a mobile
phone to learn online courses” and Posting/sharing is “Post or share pictures, videos, audio, songs, articles, or
applets created by yourself”.

2.2.3. Well Being

Depression. The Center for Epidemiologic Studies Depression Scale (CES-D-10) [27]
was used to measure adolescents’ depression. The CES-D-10 scale contains 10 items
(e.g., “I felt depressed”) and it applies a 4-point Likert scale (1 = “rarely or none of the
time (<1 day)”, 4 = “most or all of the time (5–7 days)”) to rate the items (α = 0.84).
Two items with reversed score were reversed when calculating mean score of the scale.
The CFA of CES-D-10 showed good structural validity (CFI = 0.976, TLI = 0.966, and
RMSEA = 0.073 (90% CI [0.068, 0.079])).
Anxiety. The Generalized Anxiety Disorder-7 scale (GAD-7) [28] was used to measure
adolescents’ anxiety. The GAD-7 includes 7 items (e.g., “Feeling nervous, anxious or on
edge”) and uses a 4-point Likert scale from 1 (“Not at all”) to 4 (“Nearly every day”) to rate
the items (α = 0.96), and the CFA showed good structural validity (CFI = 0.993, TLI = 0.988,
and RMSEA = 0.066 (90% CI [0.057, 0.076])).
Somatization. The Somatic Symptom Scale-8 (SSS-8) [29] was used to measure adoles-
cents’ somatization. The SSS-8 contains 8 items (e.g., “Stomach or bowel problems”). It uses
a 5-point Likert scale from 0 (“Not at all”) to 4 (“Very much”). to rate the items (α = 0.95),
and the CFA showed good structural validity (CFI = 0.987, TLI = 0.978, and RMSEA = 0.074
(90% CI [0.066, 0.082])).
The mean score of each scale was used to represent adolescents’ mental health and
used for later analysis. The reliability and validity of CES-D-10, GAD-7, and SSS-8 were con-
firmed among Chinese adolescents and were widely used by Chinese researchers [30,31].

2.2.4. Covariates
The covariates in this study consisted of adolescents’ gender, age, socioeconomic
status (both composite and perceived), smartphone use time on weekdays and weekends,
annual family revenue, parents’ educational level, and baseline mental health. Of these
covariates, parents only responded to their annual family income, educational level (that
of both the father and the mother), and the SES of the family (perceived SES). Other
covariates were obtained from adolescents. One item was used to measure perceived
SES—“How would you rate your family’s socioeconomic status in this city?” (the item
uses a 5-point Likert scale (1 = “very bad”, 5 = “very good”)). The composite SES was
computed by the average of the standardized scores of educational levels of parents and
annual family income [32]. Smartphone use time on weekdays and weekends ranged
from no use (1) to 7 h or more (7). Finally, to control the auto-regressive effect of the
dependent variables, the baseline mental health (the average value of anxiety, depression,
and somatization) was regarded as a covariate.
Children 2023, 10, 770 6 of 12

2.3. Analytic Procedure

First, SPSS 20.0 was used to preprocess the data, identify missing data, and calculate
the reliability coefficient of each scale. Second, to explore the relationship between SUC
and mental health, and to reduce the subjective bias and arbitrariness of researchers,
we conducted a 3 step specification curve analysis [18]: (1) Defining a set of reason-
able specifications. If the defined specifications are not theoretically or empirically
reasonable, it would cause “truly arbitrary” issues, which might impact the results. To
define all the specifications, the researcher should calculate all non-redundant combi-
nations (specifications) of different independent, dependent, and covariate variables.
In this study, the included variables (e.g., independent variables, dependent variables,
and covariates) are chosen based on previous literature or theories. (see the previous
paragraph regarding measurements). (2) Estimating all reasonable specifications and
describing the outcomes. To estimate all specifications and describe them, a linear
model is used to calculate the regression coefficients across the different specifications,
then all the coefficients were distributed in an ascending curve. (3) Conducting joint
statistical tests. The joint statistics test analyzes whether the actual results were incon-
sistent with the results under the null hypothesis. Consistent with previous practice via
forcing the null on the existing data (the original data), we created the null hypothesis
datasets (the null hypothesis was true). The null dataset consists of the original inde-
pendent variables and covariates but contains newly constructed dependent variables
via forcing the null on original data. Thus, there is no relationship between independent
variables and dependent variables in the null dataset. Then, participants in the null
dataset are randomly drawn with replacement, creating a new SCA model based on the
null datasets (the new SCA model is a null hypothesis model in which there exists no
relationship between independent variables and dependent variables). After 500 times,
the SCA models that tested the effects of SUCs on mental health (which turned out to
have no significant effect) are constructed. Several statistical indicators based on the
original SCA (computed from the original dataset) were compared with the indicators
from the SCA (computed from the null dataset) to find whether significance exists. The
used statistical indicators in this study included the median β (median regression coeffi-
cient) and the number of significant results in the predominant direction (NSRPD) [18].
Although some researchers have used only one of them to determine the significance
of the SCA [33], the prediction effect was robust when both the median regression
coefficient and the NSRPD were significant [18]. For more information on SCA, the
reader is referred to the tutorial [18]. The R package specr was used to estimate and
present the specification curve analysis [34]. All the variables were standardized prior
to SCA. In longitudinal studies considering autoregressive effects, the effect size is
small when β is between 0.03 and 0.07, medium when β is between 0.07 and 0.12, and
large when β is greater than 0.12 [35].

3. Results
3.1. Identification of Specifications and Description of the Estimate
As shown in Figure 1B, the identified specifications total 540 for this curve (identified
specifications = SUCs (twenty choices) × mental health (three choices) × covariate (nine
choices)). The descriptive specification curve displayed the distribution of the estimates
obtained through alternative reasonable specifications and identified which analytical
decisions are the most consequential [18]. Figure 1A presents the estimates and describes
the results from the longitudinal data. A total of 386 specifications obtained positive and
significant regression coefficients, two specifications obtained negative and significant
results, and 152 specifications showed no significance. The results show that the regression
coefficient varies from −0.05 to 0.20.
Children 2023, 10, x FOR PEER REVIEW 7 of 12

Children 2023, 10, 770 7 of 12

results, and 152 specifications showed no significance. The results show that the
regression coefficient varies from −0.05 to 0.20.

Figure 1. The description of the specification-curve analysis. Note. The estimate represents the
Figure 1. Theregression
standardized description of the specification-curve
coefficients, the x-axis representsanalysis. Note.
the number The estimateand
of specifications, represents
the y-axisthe
standardized regression coefficients, the x-axis represents the number of specifications, and the y-
in (B) represents the independent, dependent, and control variables. Red points represent negative
axis in (B) represents the independent, dependent, and control variables. Red points represent
and significant coefficients (significance level α = 0.05). Blue points represent positive and significant
negative and significant coefficients (significance level α = 0.05). Blue points represent positive and
coefficients. Grey points refer to nonsignificant coefficients. The curve in (A) is the distribution of
significant coefficients. Grey points refer to nonsignificant coefficients. The curve in (A) is the
coefficients, of
distribution and the shadedand
coefficients, parts around
the shadedtheparts
curve suggest
around 95%
the confidence
curve suggestinterval.
95% confidence interval.
3.2. Statistical Inference
3.2. Statistical Inference
The statistical inference based on bootstrapped null models is presented in Table 3.
The
In this statistical inference
longitudinal data, only based on bootstrapped
adolescents’ null models
online courses showed is no
presented
predictive in Table
effect 3.
In
on their mental health (median β = 0.01, p > 0.05, NSRPD = 0/27, p > 0.05). Finishingon
this longitudinal data, only adolescents’ online courses showed no predictive effect
their mental via
homework health (median β =also
smartphones 0.01, p > 0.05,
shows NSRPD = effect
a negligible 0/27, p(median
> 0.05). Finishing
β = 0.03,homework
p < 0.05,
via smartphones
NSRPD = 8/27, palso showsWatching
> 0.05). a negligibleTV effect
(median(median β = 0.03,
β = 0.14, p < 0.05,
p < 0.001, NSRPD
NSRPD = 8/27, p
= 24/27,
>p 0.05).
< 0.05),Watching
playing TV
games(median
(medianβ = β0.14, p < 0.001,
= 0.09, NSRPD
p < 0.001, NSRPD = 24/27, p < 0.05),
= 19/27, playing
p < 0.05), games
listening
(median
to musicβ(median
= 0.09, p β< 0.001, p < 0.001,
= 0.18,NSRPD = 19/27,
NSRPD p < 0.05),
= 25/27, p < 0.05),
listening to music (median online
and chatting β = 0.18,
p(median
< 0.001, β = 0.15,=p25/27,
NSRPD < 0.001,
p < NSRPD
0.05), and = 24/27,
chatting p< 0.001)(median
online showedβnegative
= 0.15, p <medium and
0.001, NSRPD
high prediction for future mental health. The left 15 types of content
= 24/27, p < 0.001) showed negative medium and high prediction for future mental health. (e.g., using utilities,
using
The leftcameras,
15 typesobtaining
of contentlife information,
(e.g., making
using utilities, usingcalls, and consuming
cameras, obtaining online) during
life information,
smartphone
making calls,use showed
and significant
consuming online)and negative
during effects foruse
smartphone future mental
showed health only
significant and
with the criterion of median β (β = 0.05 to 0.14, p < 0.05) but showed no
negative effects for future mental health only with the criterion of median β (β = 0.05 to significance with
the standard
0.14, p < 0.05) of
butthe NSRPD
showed no(psignificance
> 0.05). with the standard of the NSRPD (p > 0.05).
Children 2023, 10, 770 8 of 12

Table 3. Results of the SCA bootstrapping tests.

Independent Variables Median Point Estimate (β) NSRPD

Watching TV, etc. 0.14 *** 24/27 *
Watching clips 0.09 *** 21/27
Using utilities 0.10 *** 22/27
Using search engines 0.09 *** 20/27
Using fitness apps 0.10 *** 21/27
Using dictionaries 0.09 *** 21/27
Using cameras 0.07 *** 19/27
Taking online courses 0.01 0/27
Reading online novels 0.15 *** 24/27
Posting/sharing 0.08 *** 20/27
Playing games 0.09 *** 19/27 *
Obtaining life information 0.06 *** 18/27
Making payments 0.14 *** 22/27
Making calls 0.05 *** 14/27
Listening to music 0.18 *** 25/27 *
Finishing homework 0.03 * 8/27
Consuming online 0.10 *** 21/27
Chatting online 0.15 *** 24/27 *
Browsing social media 0.16 *** 24/27
Browsing news 0.11 *** 19/27
Note. * p < 0.05, *** p < 0.001.

The effects of SUC on adolescents’ mental health ranged from none (e.g., participation
in online courses and homework), small (e.g., playing games) to medium (e.g., watching
TV), depending on the SUCs.

4. Discussion
Using SCA, our results show that different types of SUC are differently associated
with depression, anxiety, and somatization in adolescents. Depending on the types
of content used, these effects ranged from high, medium, and small to none. The
relatively descending order of effect on mental health is listening to music, chatting
online, watching TV, playing games, types of content (e.g., browsing social media,
making payments, and reading online novels) with high but unstable effects, types of
content with medium (e.g., browsing news, posting/sharing) but unstable effects, types
of content (e.g., using the camera, obtaining life information, and making calls) with
small but unstable effects, such as finishing homework and taking online courses.
Our longitudinal results support the view that digital technology use is related to low
mental health. However, our results are inconsistent with previous similar research arguing
that adolescents’ screen time has no relationship with their well being [36]. Researchers
believe that the use of content rather than screen time may be related to mental health [14].
Our result is also inconsistent with another study by Orben and Przybylski arguing that
although the effect of technology use is negatively associated with well being, it is too
low to be considered [5]. However, the predicted regression coefficients of smartphone
content in this study are larger than those of the previous research focusing on screen time.
This may be because the current study has focused on smartphones and their content, and
since smartphones have become increasingly popular, the negative effect of smartphone
use can be significantly elevated [6]. The datasets used in previous research [5] were
collected before the wide popularization of smartphones, which may have contributed to
the obtained minus effects (extremely low regression coefficients) [6]. For example, a recent
study has found that longer screen time has strong negative effects on mental health [37].
Experimental research has also found that limiting social media use can decrease loneliness
and depression [38].
We found that entertainment activities (such as playing games, watching TV, and
listening to music) are strongly related to adolescents’ mental health, which is consistent
with previous research [39] and supports our hypothesis. Researchers have explained
Children 2023, 10, 770 9 of 12

that owing to their attractiveness, entertainment activities may contribute to adolescents’

excessive smartphone use [40,41], which may lead to loss of control, disrupt their daily
lives, and negatively affect their mental health [42]. Interpersonal activities (e.g., chatting
online) on smartphones also positively predicted adolescents’ depression, anxiety, and
somatization, which provides more evidence to support studies focusing on social network
use [43]. The pervasive connectivity brought by smartphones has advantages in providing
adolescents with interpersonal communication channels, which are conducive to their
mental health. However, the complementary–interference model also points out that online
chatting and entertainment activities might supplant face-to-face interactions and distract
face-to-face interaction, which will offset the potential advantages of smartphones on
mental health [21].
One interesting finding is that taking online courses through smartphones exerts no
effect on adolescents’ mental health while doing homework is slightly and negatively
related to mental health. Although the results do not fully support our hypothesis,
at least, the negative effect of doing homework via smartphones is negligible. We
suggest that adolescents usually complete their homework alone while collaboration and
interaction with others during online courses may provide potential social support [22],
which may benefit their social adjustment. Furthermore, the left 15 types of smartphone
content (e.g., using search engines, using cameras, and posting/sharing) only show
significance in median β not in NSRPD, which means that although their effect size
is medium or large while such effects may not be very stable. We think this may be
related to the intensity of interference when using different smartphone content and
related contexts. For example, playing games, watching TVs, and chatting online need
plenty of cognitive resources and usually take place in leisure time with real-life others
and need a lot of time to complete [20], which might cause more interference with their
daily sources of well-being. Although obtaining life information, using fitness apps,
making calls, and making payments might interfere with their life, these types of content
exert fewer effects on mental health because of less cognitive resources allocation and
less time needed. As for using smartphones for learning and homework demanding
many cognitive resources, these activities usually take place in an environment with few
interactions with real-life others, which is less likely to disturb their sources of well being.
However, why the effects of the left 15 types of SUC are unstable, and what reasons may
contribute to this, future studies may focus on this.
This study has some potential practical implications. Different types of SUC show dif-
ferent effects on adolescents’ mental health, which can inform the practice of policymakers
and parental media mediation. Policy statements regarding screen-based media exposure
for children and adolescents should carry a distinction between different SUC, rather than
proposing a blanket limit on screen time (as recommended by the American Academy
of Pediatrics). When it comes to adolescents’ smartphone engagement, policymakers or
parents might propose or use different mediation strategies towards different functions
or Apps on smartphones. For example, policymakers can encourage loose screen time on
mental health-friendly smartphone functions (with a negligible and stable effect size), such
as taking online courses and finishing homework; at the same time, advocate restricted
screen time on mental health-harmful smartphone functions (with a high and stable effect
size), such as watching TV, playing games, listening to music, and chatting online. As
nutritionists often do, a “nutritionally balanced” digital diet for young people should be
recommended to the public, rather than just suggesting limits on the amount of time they
can spend using digital media.
This study has some limitations. First, all the measurements are subjective evaluations
and may produce some bias. Second, as Twenge argued that the definition of mental health
or well being should be defined broadly, it may include positive emotions (happiness and
life satisfaction), and indicators of negative emotions (depression and suicide attempts) [6].
In this study, we only measured negative emotions, and, thus, future studies should
include both positive and negative indicators and use SCA to provide more evidence.
Children 2023, 10, 770 10 of 12

Third, the relationship between adolescents’ mental health and smartphone content may
be bidirectional. The present study failed to answer such an issue because our scope
focused on the effect of SUC on adolescents’ mental health. Fourth, although we stress the
smartphone content, we fail to provide an answer as to why some content is associated
with negative mental health, such as using search engines. Future studies should focus on
what is searched by adolescents via smartphones which may provide some explanations,
and more qualitative studies are required in the future. Finally, the effect size for some
SUCs is large, but such effects might not be very stable, more confirmation is needed in the
future (e.g., using a nationally representative sample size).

5. Conclusions
Some SUCs are negatively and longitudinally related to adolescents’ mental health.
Smartphones used for listening to music, chatting online, watching TV, and playing games
are negatively and longitudinally related to mental health in adolescents. Using smart-
phones for online courses is uncorrelated with adolescents’ mental health. The left 15 SUCs
(e.g., using search engines or reading online novels) reveal unstable effects. Researchers and
policymakers should update their understanding of adolescents’ technology use, especially
to adopt a differentiated attitude or different policies towards different media use content
and a more cautious attitude towards the overall screen-time restriction policy.

Author Contributions: Conceptualization, S.H. and X.L.; methodology, S.H. and X.L.; validation,
S.H. and X.L.; formal analysis, S.H.; investigation, S.H., X.L., Y.L. and Y.W.; resources, Y.C.; data
curation, S.H. and X.L.; writing—original draft preparation, S.H. and X.L.; writing—review and
editing, Y.C. and Y.W.; supervision, Y.W.; project administration, Y.W.; funding acquisition, Y.W.; and
S.H. and X.L. have equal contribution to this work. All authors have read and agreed to the published
version of the manuscript.
Funding: This research was funded by the Major Program of the National Social Science Fund of
China, grant number 20&ZD153.
Institutional Review Board Statement: The study was conducted in accordance with the Declaration
of Helsinki and approved by the Institutional Review Board of State Key Laboratory of Cognitive
Neuroscience and Learning, Beijing Normal University (CNL_A_0003_003; 25 July 2018).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author. The data are not publicly available due to ethical restrictions.
Acknowledgments: We would like to thank Shunsen Huang’s colleagues Haojie Chen, Yujian Dai,
and Shuyi Zhen at the State key laboratory of Cognitive Neuroscience and Learning, Beijing Normal
University, as well as Xintong Shan at the Collaborative Innovation Centre of Assessment toward
Basic Education Quality, Beijing Normal University, who provided purely technical help.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Statista Smartphone Users Worldwide 2016–2021. Available online: https://www.statista.com/statistics/330695/number-of-
smartphone-users-worldwide/ (accessed on 20 March 2021).
2. CNNIC. The 46th Statistical Reports on Internet Development in China; CNNIC: Beijing, China, 2020.
3. Valkenburg, P.M.; Peter, J.; Schouten, A.P. Friend Networking Sites and Their Relationship to Adolescents’ Well-Being and Social
Self-Esteem. Cyberpsychol. Behav. 2006, 9, 584–590. [CrossRef] [PubMed]
4. George, M.J.; Jensen, M.R.; Russell, M.A.; Gassman-Pines, A.; Copeland, W.E.; Hoyle, R.H.; Odgers, C.L. Young Adolescents’
Digital Technology Use, Perceived Impairments, and Well-Being in a Representative Sample. J. Pediatr. 2020, 219, 180–187.
[CrossRef]
5. Orben, A.; Przybylski, A.K. The Association between Adolescent Well-Being and Digital Technology Use. Nat. Hum. Behav. 2019,
3, 173–182. [CrossRef] [PubMed]
6. Twenge, J.M. More Time on Technology, Less Happiness? Associations Between Digital-Media Use and Psychological Well-Being.
Curr. Dir. Psychol. Sci. 2019, 28, 372–379. [CrossRef]
5.
5. Orben,
Orben, A.;A.; Przybylski,
Przybylski, A.K. A.K. TheThe Association
Association betweenbetween Adolescent
Adolescent Well-Being
Well-Being and and Digital
Digital Technology
Technology Use. Use. Nat.
Nat. Hum.
Hum. Behav.
Behav.
2019,
2019, 3,
3, 173–182.
173–182. https://doi.org/10.1038/s41562-018-0506-1.
https://doi.org/10.1038/s41562-018-0506-1.
6.
6. Twenge,
Twenge, J.M. J.M. More More Time Time on on Technology,
Technology, Less Less Happiness?
Happiness? Associations
Associations Between
Between Digital-Media
Digital-Media Use Use andand Psychological
Psychological Well- Well-
Being.
Being. Curr.
Curr. Dir. Dir. Psychol.
Psychol. Sci. Sci. 2019,
2019, 28, 28, 372–379.
372–379. https://doi.org/10.1177/0963721419838244.
https://doi.org/10.1177/0963721419838244.
7.
7. Twenge,
Twenge, J.M. J.M. Why Why Increases
Increases in in Adolescent
Adolescent DepressionDepression May May Be Be Linked
Linked to to the
the Technological
Technological Environment.
Environment. Curr. Curr. Opin.
Opin. Psychol.
Psychol.
Children 2023, 10, 770 11 of 12
2018,
2018, 32,
32, 89–94.
89–94. https://doi.org/10.1016/j.copsyc.2019.06.036.
https://doi.org/10.1016/j.copsyc.2019.06.036.
8.
8. Allen,
Allen, M.S.;
M.S.; Vella,Vella, S.A.
S.A. Screen-Based
Screen-Based Sedentary Sedentary Behaviour
Behaviour and and Psychosocial
Psychosocial Well-Being
Well-Being in in Childhood:
Childhood: Cross-Sectional
Cross-Sectional and and
Longitudinal
Longitudinal Associations.
Associations. Ment. Ment. Health Health Phys.
Phys. Act.
Act. 2015,
2015, 9,9, 41–47.
41–47. https://doi.org/10.1016/j.mhpa.2015.10.002.
https://doi.org/10.1016/j.mhpa.2015.10.002.
9.
9.7. Shakya,
Shakya,
Twenge, H.B.;
J.M.Christakis,
H.B.; Christakis,
Why Increases N.A.
N.A.inAssociation
Adolescentof
Association of Facebook
FacebookMay
Depression Use
Use Be with
with Compromised
Compromised
Linked Well-Being:
Well-Being:
to the Technological A
A Longitudinal
Longitudinal
Environment. Curr.Study.
Opin. Am.
Study. Am. J.J.
Psychol.
Epidemiol.
Epidemiol. 2017,
2017, 185,
185,
2018, 32, 89–94. [CrossRef] 203–211.
203–211. https://doi.org/10.1093/aje/kww189.
https://doi.org/10.1093/aje/kww189.
10.
8. Riehm,
10. Allen, K.E.;
Riehm, M.S.;Feder,
K.E.; Vella, K.A.;
Feder, Tormohlen,
Tormohlen, K.N.;
K.A.;Screen-Based
S.A. K.N.; Crum,
Sedentary R.M.;
R.M.; Young,
Crum,Behaviour and A.S.;
Young, A.S.; Green,
Green, K.M.;
Psychosocial K.M.;
Well-Being Pacek,
Pacek,inL.R.;
L.R.; La
La Flair, L.N.;
L.N.; Mojtabai,
Flair, Cross-Sectional
Childhood: Mojtabai, andR.
R.
Associations
Associations
Longitudinal between
between Time
Time Spent
Associations. Ment.Using
Spent HealthSocial
Using Phys. Media
Social Media
Act. and
and9,Internalizing
2015, Internalizing
41–47. [CrossRef] and
and Externalizing
Externalizing ProblemsProblems amongamong US US Youth.
Youth. JAMA
JAMA
9. Psychiatry
Shakya, 2019,
Psychiatry H.B.;76,
2019, 76, 1266–1273.
1266–1273.N.A.
Christakis, https://doi.org/10.1001/jamapsychiatry.2019.2325.
https://doi.org/10.1001/jamapsychiatry.2019.2325.
Association of Facebook Use with Compromised Well-Being: A Longitudinal Study.
11.
11. Schmiedeberg,
Schmiedeberg,
Am. J. Epidemiol. C.; Schröder,
C.;2017,
Schröder, J.J. Leisure
185, 203–211. Activities
Leisure[CrossRef] and
and Life
Activities[PubMed] Life Satisfaction:
Satisfaction: An An Analysis
Analysis with with German
German Panel Panel Data.
Data. Appl.
Appl. Res.
Res. Qual.
Qual.
10. Life
Life 2017, 12,
12, 137–151.
2017,K.E.;
Riehm, 137–151.
Feder, K.A.; https://doi.org/10.1007/s11482-016-9458-7.
https://doi.org/10.1007/s11482-016-9458-7.
Tormohlen, K.N.; Crum, R.M.; Young, A.S.; Green, K.M.; Pacek, L.R.; La Flair, L.N.; Mojtabai, R. Associ-
12.
12. Hartanto,
Hartanto, A.;
A.; Quek,
ations between Quek,Time F.Y.X.;
SpentTng,
F.Y.X.; Tng,
Using G.Y.Q.;SocialYong,
G.Y.Q.; Yong,
Media J.C.
J.C.
andDoes
Does Social
Social Media
Internalizing MediaandUse
Use Increase
Increase Depressive
Externalizing Problems Symptoms?
Depressive Symptoms?
among AA Reverse
US Youth. Reverse
JAMA Causation
Causation
Psychiatry
Perspective.
Perspective. Front.
Front. Psychiatry
2019, 76, 1266–1273. Psychiatry
[CrossRef] 2021,
2021, 12, 12, 641934.
641934. https://doi.org/10.3389/fpsyt.2021.641934.
https://doi.org/10.3389/fpsyt.2021.641934.
13.
11. Odgers,
13. Odgers, C.L.;
C.L.; Jensen,
Schmiedeberg, Jensen, M.R.
M.R. Annual
C.; Schröder, Annual
J. Leisure Research
Research
ActivitiesReview:
Review:
and LifeAdolescent
Adolescent
Satisfaction:Mental
Mental Health
Health in
An Analysis in the Digital
Digital Age:
theGerman
with PanelFacts,
Age: Data. Fears,
Facts, Fears, and Future
andQual.
Appl. Res. FutureLife
Directions.
Directions. J.J. Child
2017, 12, 137–151. Child Psychol.
Psychol.
[CrossRef] Psychiatry
Psychiatry Allied Allied Discip.
Discip. 2020,
2020, 61,
61, 336–348.
336–348. https://doi.org/10.1111/jcpp.13190.
https://doi.org/10.1111/jcpp.13190.
14.
12. George,
14. George,
Hartanto, M.
M.A.; The
The Importance
Importance
Quek, F.Y.X.; Tng, of Social
Social Media
of G.Y.Q.; MediaJ.C.
Yong, Content
Content for
for Teens’
Does Social MediaRisks
Teens’ Risks for
for Self-Harm.
Use Increase Self-Harm.
Depressive J.J. Adolesc.
Adolesc. Health
Symptoms? Health 2019,
2019, 65,
A Reverse 65, 9–10.
9–10.
Causation
https://doi.org/10.1016/j.jadohealth.2019.04.022.
https://doi.org/10.1016/j.jadohealth.2019.04.022.
Perspective. Front. Psychiatry 2021, 12, 641934. [CrossRef]
15.
13. Nesi,
15. Nesi, J.;
J.; Miller,
Odgers, Miller, A.B.;
A.B.; Prinstein,
C.L.; Jensen, Prinstein,
M.R. Annual M.J. Adolescents’
M.J. Research
Adolescents’ Review:Depressive
Depressive
Adolescent Symptoms
Mental and
Symptoms and
Health Subsequent
Subsequent
in the DigitalTechnology-Based
Technology-Based
Age: Facts, Fears, Interpersonal
Interpersonal
and Future
Behaviors:
Behaviors:
Directions. AAJ.Multi-Wave
Multi-Wave
Child Psychol. Study.
Study.
PsychiatryJ.J. Appl. Dev.
Appl.Allied Psychol.
Dev. Discip. 2017,
Psychol.2020,
2017,61, 51, 12–19.
12–19. https://doi.org/10.1016/j.appdev.2017.02.002.
51,336–348. https://doi.org/10.1016/j.appdev.2017.02.002.
[CrossRef]
16.
14. Aarts,
16. Aarts, A.A.;
George,A.A.;M. Anderson,
Anderson,
The Importance J.E.;
J.E.; Anderson,
Anderson,
of Social Media C.J.;
C.J.; Attridge,
Attridge,
Content for P.R.;
P.R.; Attwood,
Attwood,
Teens’ A.; Axt,
Axt, J.;
J.; Babel,
A.; Self-Harm.
Risks for Babel, M.;
M.; Bahník,
J. Adolesc. HealthŠ.;
Bahník, Š.; Baranski,
Baranski,
2019, 65, 9–10.E.; Barnett-
E.;[CrossRef]
Barnett-
Cowan,
Cowan,
[PubMed] M.;
M.; et et al. al. Estimating
Estimating the the Reproducibility
Reproducibility of of Psychological
Psychological Science. Science. ScienceScience 2015,
2015, 349, 349, aac4716.
aac4716.
15. https://doi.org/10.1126/science.aac4716.
https://doi.org/10.1126/science.aac4716.
Nesi, J.; Miller, A.B.; Prinstein, M.J. Adolescents’ Depressive Symptoms and Subsequent Technology-Based Interpersonal
17.
17. Twenge,
Twenge,
Behaviors:J.M.;
J.M.;AHaidt,
Haidt, J.;
J.; Lozano,
Multi-Wave Lozano,
Study. J.;
J.; Cummins,
Cummins,
J. Appl. Dev. K.M.
K.M. Specification
2017, 51,Curve
Specification
Psychol. 12–19.Analysis
Curve Analysis
[CrossRef] Shows
Shows That
That Social
[PubMed] Social Media
Media Use Use Is
Is Linked
Linked to to Poor
Poor
16. Mental
Aarts,Health,
Mental A.A.; Especially
Health, Especially among
Anderson, among
J.E.; Anderson,Girls.
Girls. Acta Psychol.
Acta C.J.;
Psychol. 2022,
2022, 224,
Attridge, 224, 103512.
103512.
P.R.; https://doi.org/10.1016/j.actpsy.2022.103512.
https://doi.org/10.1016/j.actpsy.2022.103512.
Attwood, A.; Axt, J.; Babel, M.; Bahník, Š.; Baranski, E.;
18.
18. Simonsohn,
Simonsohn,
Barnett-Cowan, U.;
U.; M.;Simmons,
Simmons, J.P.;
J.P.; Nelson,
et al. Estimating Nelson, L.D.
L.D. Specification
the ReproducibilitySpecification Curve
Curve Analysis.
of Psychological Analysis. Nat.
Nat. Hum.
Science. Science Hum.
2015, 349,Behav.
Behav. 2020, 4,
4, 1208–1214.
2020,[CrossRef]
aac4716. 1208–1214.
17. https://doi.org/10.1038/s41562-020-0912-z.
https://doi.org/10.1038/s41562-020-0912-z.
Twenge, J.M.; Haidt, J.; Lozano, J.; Cummins, K.M. Specification Curve Analysis Shows That Social Media Use Is Linked to Poor
19.
19. CNNIC.
Mental The
CNNIC. The 2019
Health, 2019 National
Nationalamong
Especially Study
Study Girls. on
on Internet
Internet Use
Use by
Acta Psychol. Minors.
by2022,
Minors.224,Available
Available online:
online: http://www.cnnic.net.cn/n4/2022/0401/c116-
103512. [CrossRef] http://www.cnnic.net.cn/n4/2022/0401/c116-
18. 1117.html
1117.html
Simonsohn, (accessed
(accessed on
on 20
U.; Simmons, 20 May
MayJ.P.; 2021).
2021).
Nelson, L.D. Specification Curve Analysis. Nat. Hum. Behav. 2020, 4, 1208–1214. [CrossRef]
20.
20. Kushlev,
Kushlev,
[PubMed] K.
K. Media
Media Technology
Technology and and Well-Being:
Well-Being: A A Complementarity-Interference
Complementarity-Interference Model. Model. In In Handbook
Handbook of of Well-Being;
Well-Being; Diener,
Diener, E.,E.,
19. Oishi,
Oishi,
CNNIC.S.,
S., Tay,
The L.,
Tay, L.,
2019Eds.;
Eds.; DEF:
DEF: Salt
National Salt
Study Lake,
Lake, UT,
UT, USA,
on Internet USA, Use2018.
2018.
by Minors. Available online: http://www.cnnic.net.cn/n4/2022/0401/c116-1
21.
21. Kushlev,
117.htmlK.;
Kushlev, K.; Dwyer,
Dwyer, R.;
(accessed R.; Dunn,
on Dunn,
20 MayE.W.E.W.
2021). The
The Social
Social Price
Price ofof Constant
Constant Connectivity:
Connectivity: Smartphones
Smartphones Impose Impose Subtle
Subtle Costs
Costs onon Well-Being.
Well-Being.
20. Curr.
Curr. Dir.
Dir. Psychol.
Kushlev, Psychol.
K. Media Sci.
Sci. 2019, 28,
2019,
Technology 28, 347–352.
347–352. https://doi.org/10.1177/0963721419847200.
https://doi.org/10.1177/0963721419847200.
and Well-Being: A Complementarity-Interference Model. In Handbook of Well-Being; Diener, E.,
22.
22. Hrastinski,
Hrastinski,
Oishi, S., Tay, S.
S.L., AEds.;Theory
A DEF: Saltof
Theory ofLake Online
Online
City, UT, Learning
Learning
USA, 2018. as
as Online
Online Participation.
Participation. Comput. Comput. Educ. Educ. 2009, 2009, 52, 52, 78–82.
78–82.
21. https://doi.org/10.1016/j.compedu.2008.06.009.
https://doi.org/10.1016/j.compedu.2008.06.009.
Kushlev, K.; Dwyer, R.; Dunn, E.W. The Social Price of Constant Connectivity: Smartphones Impose Subtle Costs on Well-Being.
23.
23. Schafer,
Schafer,
Curr. Dir. J.L.;
J.L.; Graham,
Psychol.Graham,Sci. 2019,J.W.
J.W.28, Missing
347–352.Data:
Missing Data: Our
Our View
[CrossRef] View of of thethe State
State of of the
the Art. Art. Psychol.
Psychol. Methods
Methods 2002,2002, 7, 7, 147–177.
147–177.
22. https://doi.org/10.1037/1082-989X.7.2.147.
https://doi.org/10.1037/1082-989X.7.2.147.
Hrastinski, S. A Theory of Online Learning as Online Participation. Comput. Educ. 2009, 52, 78–82. [CrossRef]
24.
23. Lai,
24. Lai, X.;
X.; Nie,
Schafer,Nie,J.L.; C.;
C.; Huang,
Huang,J.W.
Graham, S.;
S.; Li,
Li, Y.;
Y.; Xin,
Missing Xin, T.;
T.; Zhang,
Data: Zhang,
Our View C.
C. Effect
of theof
Effect of Growthof theMindset
Growth
State Mindset on
on Mental
Art. Psychol. Mental
Methods Health
2002,Two
Health Two Years
Years Later:
7, 147–177. Later: The
The Role
[CrossRef] Role ofof
24. Smartphone
Lai, X.; Nie,Use.
Smartphone Use. Int.
Int. J.J. Environ.
C.; Huang, Environ.
S.; Li, Y.; Res.
Res. Public
Public
Xin, Health
Health 2022,
T.; Zhang, 2022, 19,
19, 3355.
C. Effect 3355.
of Growth Mindset on Mental Health Two Years Later: The Role of
25.
25. Huang,
Huang, S.;
S.; Lai,
Smartphone Lai,Use.X.;
X.; Zhao,
Zhao,
Int. X.;
X.; Dai,
Dai, X.;
J. Environ. X.; Yao,
Res. Yao,
Public Y.; Zhang,
Y.;Health
Zhang, C.; Wang,
C.; 19,
2022, Wang, 3355.Y. Beyond
Beyond Screen
Y.[CrossRef] Screen Time: Time: Exploring
Exploring the the Associations
Associations between
between
25. Types
Types
Huang,of Smartphone
of S.;
Smartphone
Lai, X.; Zhao, Use
UseX.;Content
Dai, X.; and
Content Yao, Adolescents’
and Y.; Zhang, C.; Social
Adolescents’ Wang, Relationships.
Social Relationships.
Y. Beyond Screen Int. J.J. Environ.
Int.Time: Environ.
ExploringRes. Public
Res.the
Public Health
Health 2022,
Associations 2022, 19,
19, 8940.
between 8940.
Types
https://doi.org/10.3390/ijerph19158940.
https://doi.org/10.3390/ijerph19158940.
of Smartphone Use Content and Adolescents’ Social Relationships. Int. J. Environ. Res. Public Health 2022, 19, 8940. [CrossRef]
26.
26. Gardner,
26. Gardner,D.G.;
Gardner, D.G.;Cummings,
D.G.; Cummings,L.L.;
Cummings, L.L.;Dunham,
L.L.; Dunham,R.B.;
Dunham, R.B.;Pierce,
R.B.; Pierce,J.L.
Pierce, J.L.Single-Item
J.L. Single-Itemversus
Single-Item versusMultiple-Item
versus Multiple-ItemMeasurement
Multiple-Item MeasurementScales:
Measurement Scales:An
Scales: An
AnEmpirical
Empirical
Empirical
Comparison.
Comparison.Educ.
Comparison. Educ.
Educ. Psychol.
Psychol.
Psychol. Meas.
Meas.
Meas. 1998,
1998,
1998,58, 898–915.
58,58,
898–915.
898–915. https://doi.org/10.1177/0013164498058006003.
https://doi.org/10.1177/0013164498058006003.
[CrossRef]
27.
27. Andresen,
27. Andresen,E.M.;
Andresen, E.M.;Malmgren,
E.M.; Malmgren,J.A.;
Malmgren, J.A.; Carter,
J.A.; W.B.;
Carter, W.B.; Patrick,
W.B.;Patrick,
Patrick,D.L.D.L. Screening
D.L.Screening
Screening forfor
for Depression
Depression
Depression in in
in Well
Well
Well Older
OlderOlder Adults:
Adults:
Adults: Evaluation
Evaluation
Evaluation of a ofof aa
Short
Short
FormForm
Short Form
of the of the
the CES-D.
ofCES-D. Am. Am.
CES-D. Am.
J. Prev.J.J. Prev.
Med.Med.
Prev. 1994,1994,
Med. 1994, 10,
10, 77–84.
10, 77–84. 77–84. https://doi.org/10.1016/s0749-3797(18)30622-6.
https://doi.org/10.1016/s0749-3797(18)30622-6.
[CrossRef]
28.
28. Spitzer,
28. Spitzer,R.L.;
Spitzer, R.L.;Kroenke,
R.L.; Kroenke,K.;
Kroenke, K.;Williams,
K.; Williams,J.B.W.;
Williams, J.B.W.;Löwe,
J.B.W.; Löwe,B.
Löwe, B.A
B. ABrief
A BriefMeasure
Brief Measurefor
Measure forAssessing
for Assessing Generalized
Assessing Generalized Anxiety
Anxiety Disorder:
Disorder: TheThe GAD-
GAD-
GAD-7.
7. Arch.
7.Arch. Intern.
Arch.Intern.
Intern.Med. Med. 2006,
Med.2006,
2006,166, 166, 1092–1097.
166,1092–1097.
1092–1097.[CrossRef] https://doi.org/10.1001/archinte.166.10.1092.
https://doi.org/10.1001/archinte.166.10.1092.
[PubMed]
29.
29. Gierk,
29. Gierk,B.;
Gierk, B.;Kohlmann,
B.; Kohlmann,S.;
Kohlmann, S.;S.;Kroenke,
Kroenke,K.;
Kroenke, K.;Spangenberg,
K.; Spangenberg,L.;
Spangenberg, L.;Zenger,
L.; Zenger, M.;
Zenger, M.; Braḧler,
Bra ler, E.;
Braḧler, E.; Loẅe,
Loẅe,
Lo B. The
e, B. Somatic
The Somatic Symptom
SomaticSymptom
SymptomScale-8Scale-8 (SSS-8):
Scale-8(SSS-8):
(SSS-8):A
AABriefBrief
MeasureMeasure
Brief Measure
of Somatic Symptom of
of Somatic
SomaticBurden.Symptom
JAMA Intern. Burden.
Symptom Burden.
Med. 2014, 174, JAMA
JAMA 399–407. Intern.
Intern.
[CrossRef]Med.
Med. 2014,
2014,
[PubMed] 174,
174, 399–407.
399–407.
30. https://doi.org/10.1001/jamainternmed.2013.12179.
https://doi.org/10.1001/jamainternmed.2013.12179.
Zhou, S.J.; Wang, L.L.; Yang, R.; Yang, X.J.; Zhang, L.G.; Guo, Z.C.; Chen, J.C.; Wang, J.Q.; Chen, J.X. Sleep Problems among
Chinese Adolescents and Young Adults during the Coronavirus-2019 Pandemic. Sleep Med. 2020, 74, 39–47. [CrossRef]
31. Yang, L.; Hu, Y.; Silventoinen, K.; Martikainen, P. Childhood Adversity and Depressive Symptoms among Middle-Aged and
Older Chinese: Results from China Health and Retirement Longitudinal Study. Aging Ment. Health 2020, 24, 923–931. [CrossRef]
[PubMed]
32. Li, Y.; Xu, L.; Lv, Y.; Liu, L.; Wang, Y.; Huntsinger, C.S. The Influence of Preschool Rural Migrant Composition on Rural Migrant
Children’s Early Development in China. Int. J. Intercult. Relat. 2015, 49, 121–136. [CrossRef]
33. Hässler, T.; Ullrich, J.; Bernardino, M.; Shnabel, N.; Van Laar, C.; Valdenegro, D.; Sebben, S.; Tropp, L.R.; Visintin, E.P.;
González, R.; et al. A Large-Scale Test of the Link between Intergroup Contact and Support for Social Change. Nat. Hum. Behav.
2020, 4, 380–386. [CrossRef]
34. Masur, P.K.; Scharkow, M. Specr: Statistical Functions for Conducting Specification Curve Analyses. 2020. Available online:
https://masurp.github.io/specr/ (accessed on 20 May 2021).
35. Orth, U.; Meier, L.L.; Bühler, J.L.; Dapp, L.C.; Krauss, S.; Messerli, D.; Robins, R.W. Effect Size Guidelines for Cross-Lagged Effects.
Psychol. Methods 2022. [CrossRef]
Children 2023, 10, 770 12 of 12

36. Orben, A.; Przybylski, A.K. Screens, Teens, and Psychological Well-Being: Evidence from Three Time-Use-Diary Studies.
Psychol. Sci. 2019, 30, 682–696. [CrossRef]
37. Delgado-Floody, P.; Caamaño-Navarrete, F.; Guzmán-Guzmán, I.P.; Jerez-Mayorga, D.; Martínez-Salazar, C.; Álvarez, C. Food
Habits and Screen Time Play a Major Role in the Low Health Related to Quality of Life of Ethnic Ascendant Schoolchildren.
Nutrients 2020, 12, 3489. [CrossRef] [PubMed]
38. Hunt, M.G.; Marx, R.; Lipson, C.; Young, J. No More FOMO: Limiting Social Media Decreases Loneliness and Depression. J. Soc.
Clin. Psychol. 2018, 37, 751–768. [CrossRef]
39. Kim, H.H.S.; Ahn, S.J.G. How Does Neighborhood Quality Moderate the Association between Online Video Game Play and
Depression? A Population-Level Analysis of Korean Students. Cyberpsychol. Behav. Soc. Netw. 2016, 19, 628–634. [CrossRef]
40. Cha, S.S.; Seo, B.K. Smartphone Use and Smartphone Addiction in Middle School Students in Korea: Prevalence, Social
Networking Service, and Game Use. Health Psychol. Open 2018, 5, 2055102918755046. [CrossRef] [PubMed]
41. Park, J.H. Smartphone Use Patterns of Smartphone-Dependent Children. Child Health Nurs. Res. 2020, 26, 47–54. [CrossRef]
42. Elhai, J.D.; Dvorak, R.D.; Levine, J.C.; Hall, B.J. Problematic Smartphone Use: A Conceptual Overview and Systematic Review of
Relations with Anxiety and Depression Psychopathology. J. Affect. Disord. 2017, 207, 251–259. [CrossRef]
43. Mesch, G.S. Parental Mediation, Online Activities, and Cyberbullying. Cyberpsychol. Behav. 2009, 12, 387–393. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.