View in EDS HTML Full Text PDF Full Text

Network Structure of the Links between Extracurricular Time-Use and Delinquent Behaviors: Moving Forward and Beyond Linear Relations

Saved in:
Bibliographic Details
Title: Network Structure of the Links between Extracurricular Time-Use and Delinquent Behaviors: Moving Forward and Beyond Linear Relations
Language: English
Authors: Xia, Yiwei (ORCID 0000-0001-7360-732X), Ma, Zhihao (ORCID 0000-0001-7983-4136)
Source: Child Development. 2023 94(6):1697-1712.
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 16
Publication Date: 2023
Document Type: Journal Articles
Reports - Research
Descriptors: Foreign Countries, Students, Extracurricular Activities, Time on Task, Delinquency, Delinquency Prevention, Student Behavior, Behavior Problems, Smoking, Drinking, Negative Reinforcement, School Schedules, Weekend Programs, Risk, At Risk Students
Geographic Terms: China
DOI: 10.1111/cdev.13953
ISSN: 0009-3920
1467-8624
Abstract: Using psychological network analysis, this study explored the heterogeneity of the network structure between extracurricular time-use and delinquency using a nationally representative longitudinal survey of at-school students in China (N = 10,279, 47.3% female, average age 13.6, 91.2% Han ethnicity). The results are threefold: First, time stimulation of activities occurs on weekdays, while time displacement and stimulation occur on weekends. Second, delinquent behaviors are positively correlated, forming a problem behavior syndrome. Smoking or drinking is the central delinquent behavior. Third, negative consequences of specific time-use behaviors are more likely to occur on weekends than on weekdays, and time-use behavior may function differently on weekdays versus weekends. Among them, going to coffeenets or game-centers serves as the bridge with the highest potential of triggering delinquency.
Abstractor: As Provided
Entry Date: 2023
Accession Number: EJ1403845
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwEv1GCUVpxzyPak_2iUoo8KAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDAOvIJ-72Q-VZvbNkQIBEICBmxRi5cKt1gnV4f6RqK_GuG68DdJT0Sh8ieVrWfQhV7oNrrxiai82mcVThQR6hT2FFsCBUEYolsVF6-bkF9dwYnE4A_8RbLqdjHpKS85tPMAUBg9oaqCgQaCrnDEqN2e-8O0kKgHAtEtXbGuUargo9WvGc8Usr-jHlq3_nokCRGWw4a5UZqvAgNhHi8cyjbyxD6go9SZIEn0bo7YZ
Text:
  Availability: 1
  Value: <anid>AN0173972090;cdv01nov.23;2023Dec05.05:34;v2.2.500</anid> <title id="AN0173972090-1">Network structure of the links between extracurricular time‐use and delinquent behaviors: Moving forward and beyond linear relations </title> <p>Using psychological network analysis, this study explored the heterogeneity of the network structure between extracurricular time‐use and delinquency using a nationally representative longitudinal survey of at‐school students in China (N = 10,279, 47.3% female, average age 13.6, 91.2% Han ethnicity). The results are threefold: First, time stimulation of activities occurs on weekdays, while time displacement and stimulation occur on weekends. Second, delinquent behaviors are positively correlated, forming a problem behavior syndrome. Smoking or drinking is the central delinquent behavior. Third, negative consequences of specific time‐use behaviors are more likely to occur on weekends than on weekdays, and time‐use behavior may function differently on weekdays versus weekends. Among them, going to coffeenets or game‐centers serves as the bridge with the highest potential of triggering delinquency.</p> <p></p> <ulist> <item> Abbreviations</item> <p></p> <item> BGGM Bayesian Gaussian graph model</item> <p></p> <item> CEPS China Education Panel Survey</item> <p></p> <item> GGM Gaussian graphical model</item> <p></p> <item> PBT problem behavior theory</item> <p></p> <item> PNA psychological network analysis</item> <p></p> <item> SDH structured days hypothesis</item> <p></p> <item> SNA social network analysis</item> </ulist> <p>The relation between extracurricular time‐use and delinquent behaviors has been discussed extensively in the literature. Control theorists suggest that time spent on extracurricular activities, especially conventional activities like doing homework, reading, and working, that generate or result in social bonds could inhibit delinquent behaviors (Hirschi, [<reflink idref="bib21" id="ref1">21</reflink>]). Opportunity theories, such as routine activity and lifestyle theories, imply that time spent on extracurricular activities in the absence of authority figures and in the presence of deviant peers increases the likelihood of delinquency (Cohen & Felson, [<reflink idref="bib8" id="ref2">8</reflink>]; Hindelang et al., [<reflink idref="bib20" id="ref3">20</reflink>]). This is because these activities may increase the situational motivation for delinquent behavior and create an environment conducive to delinquency (Osgood et al., [<reflink idref="bib41" id="ref4">41</reflink>]; Wikström et al., [<reflink idref="bib47" id="ref5">47</reflink>]). Additionally, when these activities lack clear‐stated structures, they are referred to as <emph>unstructured socializing</emph> and are associated with higher delinquency levels (e.g., Hoeben & Weerman, [<reflink idref="bib22" id="ref6">22</reflink>]; Osgood et al., [<reflink idref="bib41" id="ref7">41</reflink>]), as they offer opportunities for delinquency, reinforce group pressure, increase the tolerance for delinquency, and increase the exposure to delinquent peers (Hoeben & Weerman, [<reflink idref="bib22" id="ref8">22</reflink>]).</p> <p>Although different theorists interpret the function of extracurricular time‐use differently, they all agree that involvement in different activities may play distinct roles in delinquency. Thus, a differential involvement perspective is required to understand the relation between extracurricular activities and delinquency (Wong, [<reflink idref="bib53" id="ref9">53</reflink>]). For instance, empirical studies have shown that spending time on school‐ or family‐based activities, like doing homework (e.g., Farrell et al., [<reflink idref="bib15" id="ref10">15</reflink>]), attending cram school or classes (e.g., Tanioka & Glaser, [<reflink idref="bib44" id="ref11">44</reflink>]), doing housework (e.g., Barnes et al., [<reflink idref="bib2" id="ref12">2</reflink>]), and extracurricular reading (e.g., Lushin et al., [<reflink idref="bib33" id="ref13">33</reflink>]) are negatively associated with delinquency. The reasons are straightforward: these activities generally foster social bonds (like attachment to school or family) and reduce the opportunities for delinquency because of the presence of parents or teachers, absence of delinquent peers, and a structure with clear goals. Alternatively, the association between other activities and delinquency remains unclear. For example, some studies found that playing sports may increase the odds of delinquency (e.g., Martins et al., [<reflink idref="bib35" id="ref14">35</reflink>]), especially when the players have developed "jock" identities (Miller et al., [<reflink idref="bib36" id="ref15">36</reflink>]). However, several other studies found no direct relation between sports and delinquency (e.g., Henson et al., [<reflink idref="bib19" id="ref16">19</reflink>]). Two other controversial activities are watching TV and playing video games; although studies have shown that both activities may be directly linked to delinquency (e.g., Henson et al., [<reflink idref="bib19" id="ref17">19</reflink>]), the content of the activities is a major determinant (Rasmussen et al., [<reflink idref="bib43" id="ref18">43</reflink>]).</p> <p>Previous studies outline the critical role of different extracurricular time‐use behaviors in fostering delinquency, which is essential to understand the effectiveness of extracurricular activities in curbing delinquency, considering that the rate of delinquency increases during after‐school hours (Gottfredson et al., [<reflink idref="bib18" id="ref19">18</reflink>]). However, the relation between time‐use among different types of extracurricular activities and the co‐occurrence of different types of delinquency is underexplored. For instance, extracurricular activities may compete with or stimulate each other, resulting in a heterogeneity of extracurricular time‐use patterns. Similarly, as the problem behavior theory (PBT) suggests, diverse types of delinquency tend to co‐occur, showing a syndrome of problem behaviors. How do the mechanisms within extracurricular time‐use and diverse delinquencies affect the activity‐delinquency nexus? Previous studies have yet to provide an answer. Moreover, according to the structured days hypothesis (SDH), extracurricular activities may function differently on weekdays and weekends. It is also unclear how the activity‐delinquency nexus varies between weekdays and weekends.</p> <p>To answer the research questions, the current research adopted a novel methodology—psychological network analysis (PNA)—to explore the complex nexus. Traditional approaches such as factor analysis and structural equation models assume that observed variables originate from underlying latent variables and thus they explore the relations between latent variables. These assumptions imply that external factors can only influence the observed variables through the latent construct (Borsboom, [<reflink idref="bib3" id="ref20">3</reflink>]). PNA is a recently developed approach to exploring and mapping the variable‐level correlation networks of all the observed variables (Borsboom, [<reflink idref="bib3" id="ref21">3</reflink>]) without assuming the existence of latent variables. With this approach, PNA provides an alternative understanding of the psychological construct—a complex system constituted by connections and interactions among multiple observed variables (Borsboom, [<reflink idref="bib3" id="ref22">3</reflink>]). A certain observed variable could be directly linked to external factors and directly (or indirectly) linked to any other observed variables. Based on this theoretical standpoint, PNA provides several new values to explore the complex nexus compared with traditional approaches. These include identifying a central construct among relevant variables (Borsboom, [<reflink idref="bib3" id="ref23">3</reflink>]; Jones et al., [<reflink idref="bib29" id="ref24">29</reflink>]), estimating network comparisons among multiple networks (Epskamp & Fried, [<reflink idref="bib13" id="ref25">13</reflink>]; Williams et al., [<reflink idref="bib50" id="ref26">50</reflink>]), and identifying a path from a certain external factor to a certain psychological variable (Liu & Ma, [<reflink idref="bib31" id="ref27">31</reflink>]). PNA can effectively assist in investigating the network structure of the links between extracurricular time‐use and delinquent behaviors and simultaneously identify the relations among diverse behaviors. The result yielded by PNA could considerably extend the current understanding of the linear relation between extracurricular time‐use and delinquent behaviors.</p> <hd id="AN0173972090-2">LITERATURE REVIEW</hd> <p></p> <hd id="AN0173972090-3">Interactions among extracurricular time‐use: Time displacement, time stimulation, and their c...</hd> <p>Time‐use is not a specific scientific terminology, but rather a general term that describes the pattern in which individuals spend their personal time on certain activities. This term has been widely studied as a critical determinant of individual well‐being among diverse populations (e.g., Bu et al., [<reflink idref="bib6" id="ref28">6</reflink>]; Ma et al., [<reflink idref="bib34" id="ref29">34</reflink>]). In the adolescent population in particular, the outcome of extracurricular time‐use is widely discussed (e.g., Hu & Mu, [<reflink idref="bib24" id="ref30">24</reflink>]; Hunt & McKay, [<reflink idref="bib26" id="ref31">26</reflink>]).</p> <p>Previous studies have provided valuable information on the intercorrelation between extracurricular time‐use behaviors, but the literature differs in their findings. Some scholars have proposed time displacement and time stimulation hypotheses to explain the interaction mechanism between two time‐use behaviors. Time displacement describes a zero‐sum pattern of time‐use. Under the time displacement hypothesis, time spent on one activity is traded for time spent on other activities. The nature of time‐use behaviors is displaced, and focused time‐use behavior is the external shock that alters peoples' previous social life (Putnam, [<reflink idref="bib42" id="ref32">42</reflink>]). For instance, Putnam ([<reflink idref="bib42" id="ref33">42</reflink>]) indicated that time spent on electronic media isolated people and decreased public engagement. Thus, the displacement objects under such a hypothesis are obvious: time‐use on media and on social activities may compete. Several adolescent samples also revealed similar findings (e.g., Twenge et al., [<reflink idref="bib45" id="ref34">45</reflink>]). Time stimulation, however, describes an opposite pattern of time‐use: time spent on one activity can stimulate time spent on other activities. For example, several adolescent and youth studies revealed that time‐use on certain online technologies intensifies individual social interaction with their peers (Lee, [<reflink idref="bib30" id="ref35">30</reflink>]; Valkenburg & Peter, [<reflink idref="bib46" id="ref36">46</reflink>]) and leads to positive well‐being (Valkenburg & Peter, [<reflink idref="bib46" id="ref37">46</reflink>]). Although scholars believe that the context and functions of media consumption are crucial factors in interpreting the outcomes of media use (e.g., Moses et al., [<reflink idref="bib39" id="ref38">39</reflink>]), the interrelations among diverse time‐use behaviors may vary for different scenarios, complicating the role of time‐use on media. However, linking the time displacement or stimulation phenomenon to delinquent outcomes is untenable without further evidence.</p> <p>Previous studies have generally adopted the linear regression approach to investigate time displacement and time stimulation hypotheses. However, this approach is methodologically and theoretically inadequate to explore the complex interaction of time‐use behaviors. From a methodological perspective, previous studies mainly focused on the association between two time‐use behaviors or the effects of multiple time‐use behaviors on another behavior, which may run the risk of omitted variable bias (Williams et al., [<reflink idref="bib51" id="ref39">51</reflink>]) since the associations between time‐use behaviors are multivariate in nature. Considering the theoretical framework, time displacement does not stand in opposition to time stimulation. Given that human beings have 24 h in one natural day, there is enough time for them to choose consecutive time‐use behaviors from a large list of activities. Thus, time displacement and time stimulation may occur simultaneously when studying more than two time‐use behaviors. Unfortunately, the traditional linear regression approach cannot address these flaws.</p> <p>Psychological network analysis provides a potential way to address the above issues, especially when multiple behaviors are involved. Under PNA, behaviors are explained as constituents in a complex system (Mkhitaryan et al., [<reflink idref="bib37" id="ref40">37</reflink>]), and interactions among behaviors are interpreted by network attributions (Borsboom et al., [<reflink idref="bib4" id="ref41">4</reflink>]). While most previous PNA studies purely focused on relations among diverse psychological constructs or psychiatric symptoms, Ma et al. ([<reflink idref="bib34" id="ref42">34</reflink>]) have made an initial effort to employ PNA to identify interplays among different time‐use behaviors and their links to depression. They proposed that the pattern of time‐use behaviors could be hypothesized as a dynamic network where the relation between two time‐use behaviors was estimated after controlling for other time‐use behaviors. Their study revealed heterogeneous patterns of diverse time‐use behaviors, suggesting that time displacement and time stimulation occurred simultaneously (Ma et al., [<reflink idref="bib34" id="ref43">34</reflink>]). Similarly, PNA can be used to understand the interactions between different extracurricular time‐use behaviors among adolescents.</p> <hd id="AN0173972090-4">The co‐occurrence of delinquent behaviors: PBT</hd> <p>Researchers have long been aware of the co‐occurrence of different types of delinquent behavior. In the 1970s, Jessor and Jessor ([<reflink idref="bib28" id="ref44">28</reflink>]) found high intercorrelations among behaviors, such as smoking, alcohol use, marijuana use, and sexual activity. Thus, the authors proposed PBT to interpret the co‐occurrence phenomenon. The PBT suggests that the intercorrelation among behaviors results from a single factor called the problem behavior syndrome, which describes a general tendency to engage in problem or risky behavior. To date, the PBT has gained support from many studies (see Donovan & Jessor, [<reflink idref="bib12" id="ref45">12</reflink>]; Jessor, [<reflink idref="bib27" id="ref46">27</reflink>]). It has also been extended to interpret a wide range of problem behaviors, including gambling (Yakovenko & Hodgins, [<reflink idref="bib54" id="ref47">54</reflink>]), delinquency (Dembo et al., [<reflink idref="bib10" id="ref48">10</reflink>]; Willoughby et al., [<reflink idref="bib52" id="ref49">52</reflink>]), and risky behavior (Dong & Ding, [<reflink idref="bib11" id="ref50">11</reflink>]).</p> <p>Recent studies have questioned the "single factor" assumption of the PBT and suggested that considering more than one latent factor may provide a better fit for many delinquent behaviors. Applying factor analysis, Willoughby et al. ([<reflink idref="bib52" id="ref51">52</reflink>]) found that a three‐latent‐factor model, namely problem behaviors (primarily substance use and sexual activity), aggression, and delinquency were a better fit than a single‐latent‐variable model. Similarly, other studies have also yielded a higher variance explained by multiple‐latent‐variable models (e.g., Dembo et al., [<reflink idref="bib10" id="ref52">10</reflink>]). Notwithstanding the dispute between one‐latent and multiple‐latent variable models, delinquent behaviors are highly intercorrelated with each other in nearly all the studies, even for studies which adopted multiple‐latent‐factor models, the latent variables were also highly correlated with each other. Perhaps just as Willoughby and colleagues stated that, "Given the substantial intercorrelations among these three factors, results also support the possibility of a global factor accounting for the associations among the three latent factors" (Willoughby et al., [<reflink idref="bib52" id="ref53">52</reflink>], p. 1032).</p> <p>The failure to reach a consensus on a relevant number of latent factors alludes to the elusiveness of seeking common causes. In other words, the co‐occurrence of delinquency could be a reflection of a "problem of living" featured with "a syndromic constellation of symptoms that hang together empirically" (Borsboom, [<reflink idref="bib3" id="ref54">3</reflink>], p. 5) rather than a "disease" that arises from a common pathogenic pathway. Thus, it is worthwhile to re‐evaluate the co‐occurrence of delinquency through the lens of PNA by depicting the variable‐level network structure among all the delinquent behaviors (Borsboom, [<reflink idref="bib3" id="ref55">3</reflink>]). In fact, the network approach conveys the genuine meaning of PBT, just as Jessor, the founder of PBT, states, "A key contribution of Problem Behavior Theory ... has been to demonstrate the embeddedness of health‐related behaviors in a larger explanatory network of psychosocial and behavioral variables" (Jessor, [<reflink idref="bib27" id="ref56">27</reflink>], p. 248). The <emph>network approach</emph> enables us to explore the network structure of diverse delinquencies, such as identifying a central delinquency that is highly correlated with other delinquent behaviors, finding sub‐communities of delinquent behaviors, and discovering heterogeneous pathways which trigger specific types of delinquent behavior. For instance, de Vries et al. ([<reflink idref="bib9" id="ref57">9</reflink>]) applied PNA to explore the co‐occurrence of adolescent dating‐abuse experiences and victimization. Their findings revealed clusters of co‐occurring victimization and perpetration incidents and concluded that a diverse range of victimization experiences did not typically co‐occur with perpetration.</p> <hd id="AN0173972090-5">Weekday and weekend activities: SDH</hd> <p>Research has shown that the function of extracurricular activities differs significantly between weekdays and weekends (Brazendale et al., [<reflink idref="bib5" id="ref58">5</reflink>]). The underlying difference, according to SDH proposed by Brazendale et al. ([<reflink idref="bib5" id="ref59">5</reflink>]), is that structured activities characterized by "a pre‐planned, segmented, and adult‐supervised compulsory environment" (Brazendale et al., [<reflink idref="bib5" id="ref60">5</reflink>], p. 2) are more likely to occur during weekdays than on weekends. Interestingly, the pre‐planned and adult‐supervised features defined by SDH coincide with most parts of the definition of <emph>unstructured socializing</emph> proposed by criminologists (Hoeben & Weerman, [<reflink idref="bib22" id="ref61">22</reflink>]; Osgood et al., [<reflink idref="bib41" id="ref62">41</reflink>]), except for the presence of peers. However, weekend activities increase the possibility of hanging out with friends or peers. Therefore, it is reasonable to hypothesize a distinct pattern between the activity‐delinquency nexus on weekdays and weekends. Unfortunately, literature has yet to explore the difference.</p> <p>In addition, the nature of structuredness between weekdays and weekends could also influence the frequency and nature of time displacement or stimulation among extracurricular time‐use. However, few studies have attempted to compare the nature and extent of time‐use between weekdays and weekends.</p> <p>The unique features of PNA enable us to explore and compare the heterogeneous structure between extracurricular time‐use and delinquency during weekdays and weekends. After separately estimating the network structure between extracurricular time‐use and delinquency during weekdays and weekends, PNA could compare the edges of both networks (Epskamp & Fried, [<reflink idref="bib13" id="ref63">13</reflink>]; Williams et al., [<reflink idref="bib50" id="ref64">50</reflink>]) and yield nuanced conclusions on the differing network structures between weekdays and weekends, while traditional methods may not achieve the same depth and distinctiveness.</p> <hd id="AN0173972090-6">THE CURRENT STUDY</hd> <p></p> <hd id="AN0173972090-7">Research questions</hd> <p>This study used the China Education Panel Survey (CEPS), a nationally representative longitudinal study of at‐school students, to explore the interactions between extracurricular time‐use and juvenile delinquency. Specifically, this study proposed the following research questions: (<reflink idref="bib1" id="ref65">1</reflink>) What is the pattern of time displacement or stimulation among diverse extracurricular time‐use behaviors? What are the differences in weekdays and weekends patterns? (<reflink idref="bib2" id="ref66">2</reflink>) What is the network structure of delinquent behaviors? What is the central delinquent behavior with the highest potential to be linked to other delinquent behaviors? (<reflink idref="bib3" id="ref67">3</reflink>) What are the triggering pathways from diverse extracurricular time‐use behaviors to different delinquent behaviors during weekdays and weekends? Which time‐use behavior presents the highest potential for triggering juvenile delinquency? Do the patterns differ between weekdays and weekends? Given these research questions, the current study was mainly exploratory in nature.</p> <p>In the current study, PNA could be used to answer all the above research questions. First, PNA could estimate the variable‐level network structure of time‐use behaviors and delinquency between weekdays and weekends, thus simultaneously exploring time‐use behavior interactions, delinquent behavior co‐occurrence, and the interplay between time‐use behavior and delinquency. Second, the network comparison technique in PNA could estimate the edge difference of the networks between weekdays and weekends. Finally, centrality indices of each variable were calculated to identify the central delinquent behavior and the bridge time‐use behaviors that have the highest potential of triggering other behaviors.</p> <hd id="AN0173972090-8">Methodological framework</hd> <p>The PNA technique has several advantages over traditional linear regression when addressing the current research questions. In this section, the methodological framework of PNA is presented in Figure 1 to further illustrate the difference. Figure 1a shows the linear relation approach of identifying the link between two behaviors. Such a linear connection may be biased as confounders were not controlled for (see Figure 1b). In the current study, considering that both time‐use and delinquent behaviors present internal interaction and co‐occurrence mechanisms, interactions among time‐use on different extracurricular activities and the co‐occurrence of delinquent behaviors could be hypothesized as two variable communities—time‐use community and delinquent behaviors community (see Figure 1c for a hypothetical network structure). The "community" is theoretically equivalent to "latent" in the traditional approach. The inter‐community linkages could be identified as the pathways between time‐use and delinquent behaviors. Under the assumption of Figure 1c, the linear regression combined with a traditional latent variable approach may not effectively tackle tasks such as identifying a detailed pathway from a certain time‐use behavior to a certain delinquent behavior.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CDV/01nov23/cdev13953-fig-0001.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cdev13953-fig-0001.jpg" title="1 Relations among diverse behaviors. (a) The relation between two behaviors (B1 and B2). (b) The relation between two behaviors (B1 and B2) may be affected by a third confounder behavior (B3). (c) Relations among diverse time use behaviors and diverse delinquent behaviors. (d) The conceptual model being investigated in the current study. In (c, d), black edges denote intra‐community relations while red edges denote inter‐community relations. Moreover, blue nodes (T4 and D1) denote bridge behaviors if all edges present an equal weight." /> </p> <p></p> <p>The complex network structure shown in Figure 1c requires a novel method to answer the research questions. PNA enables researchers to investigate the network structure of variables. It has been widely adopted in previous literature to (<reflink idref="bib1" id="ref68">1</reflink>) explore the comorbidity of mental disorders (e.g., Jones et al., [<reflink idref="bib29" id="ref69">29</reflink>]), co‐occurrence of behaviors (e.g., de Vries et al., [<reflink idref="bib9" id="ref70">9</reflink>]), and personal mental reactions toward external events (e.g., Fried et al., [<reflink idref="bib17" id="ref71">17</reflink>]); (<reflink idref="bib2" id="ref72">2</reflink>) identify the central symptom or behavior within a certain network (e.g., Liu & Ma, [<reflink idref="bib31" id="ref73">31</reflink>]); (<reflink idref="bib3" id="ref74">3</reflink>) identify the bridge symptom or behavior of one network community which has a huge potential for triggering other symptoms or behaviors within different network communities (e.g., Jones et al., [<reflink idref="bib29" id="ref75">29</reflink>]); and (<reflink idref="bib4" id="ref76">4</reflink>) compare the edges of different networks (e.g., Williams, [<reflink idref="bib48" id="ref77">48</reflink>]).</p> <p>The latter three applications are especially useful in the current study to gauge the central delinquent behavior, identify the bridge time‐use behavior, and compare the network structures of variables between weekdays and weekends. To better understand the conceptual difference between bridge and central variables, the hypothetical network in Figure 1c is taken as an example (assuming all edges are equal). T5 is the central time‐use behavior within the time‐use community. However, its potential to trigger delinquent behaviors is lower than T4, which establishes three links with delinquent behaviors. T4 is the bridge time‐use behavior of the time‐use community. Correspondingly, D1 is the bridge behavior within the delinquent behaviors community since it presents the highest number of linkages with time‐use behaviors.</p> <p>Figure 1d presents a modified version of Figure 1c. Since the current study employed a longitudinal dataset, the directions from time‐use behaviors to delinquent behaviors could be investigated.</p> <hd id="AN0173972090-10">METHOD</hd> <p></p> <hd id="AN0173972090-11">Data</hd> <p>The study used CEPS baseline and follow‐up surveys conducted by the National Survey Research Center at Renmin University of China in the 2013–2014 academic year. The baseline survey of CEPS randomly selected a school‐based, nationally representative sample of 19,487 seventh (<emph>N</emph> = 10,279) and ninth graders (<emph>N</emph> = 9208) nested in 438 classrooms at 112 schools in 28 county‐level units in China. In the 2014–2015 academic year, the CEPS accomplished its Wave Two follow‐up survey with a follow‐up rate of 91.5%. Recruitment procedure, detailed sampling design, and ethical approval information of CEPS can be found at its official website (<ulink href="http://ceps.ruc.edu.cn/English/Documentation/Manual.htm">http://ceps.ruc.edu.cn/English/Documentation/Manual.htm</ulink>).</p> <p>The CEPS collected data on the demographic and socioeconomic characteristics of the respondents. In addition, the CEPS contains rich information on the time spent on various activities and delinquent behaviors. To the best of our knowledge, CEPS is the only national survey, with a sufficiently large sample of middle‐school adolescents, to have collected data on children's time‐use and delinquent behaviors.</p> <p>The offline survey of Wave Two of CEPS was only conducted for the seventh grade. Therefore, the current study only adopted the responses of seventh graders in the baseline survey for further analysis. After merging the baseline with Wave Two data, the final sample size was 10,279. The imputation method was applied to handle item missingness, which occurred in two forms: item non‐response and invalid item response.</p> <p>Of these participants, 47.3% were female, 52.4% reported having rural hukou, and 43.0% lived in a single‐child family. Participants' mean age was 13.6 years, and most were of Han ethnicity (91.2%) and local citizens (79.8%). Most participants reported that their family socioeconomic status was in the middle‐income bracket (74.8%).</p> <hd id="AN0173972090-12">Measures</hd> <p></p> <hd id="AN0173972090-13">Independent variables: Extracurricular time‐use behaviors (baseline)</hd> <p>Extracurricular time‐use behaviors during weekdays and weekends were measured separately at the baseline. Participants were asked a series of questions about the average number of hours and minutes spent per day on different activities in the past five weekdays and the weekends. The activities were "doing homework," "attending cram school or classes," "doing sports," "extracurricular reading," "watching TV," "surfing the Internet or playing video games," and "doing housework."</p> <p>These seven time‐use behaviors are also prevalent and commonly adopted by adolescents in other countries and regions. Although the types of behaviors may be similar, the allocation of time for behaviors may vary in the current sample. In most East Asian countries, including China, the majority of adolescents' daytime activities are assigned to formal schooling during weekdays. In addition to that, much of the adolescents' extracurricular time is spent on attending cram school lessons and doing homework assigned by school teachers, parents, and private tutors (Hu & Mu, [<reflink idref="bib24" id="ref78">24</reflink>]; Nishino & Larson, [<reflink idref="bib40" id="ref79">40</reflink>]). As a result, in the current sample, the amount of available time is extremely limited for most adolescents in China, making time‐displacement more likely to occur, especially on weekdays.</p> <p>The final responses were coded in hours (minutes were transformed to hours, if any). When the response to a certain activity exceeded 24, the items were considered invalid responses. In addition, when the total number of hours for all the activities (weekdays and weekends) exceeded 24, responses to these activities were all coded as invalid. Although the retrospective measures method with stylized questions is subject to recall bias, studies suggest measuring activities (such as doing homework, watching TV) repeatedly and frequently in a recent and brief reference period (e.g., past five weekdays and the weekends) may help mitigate the bias (Hu & Mu, [<reflink idref="bib24" id="ref80">24</reflink>]).</p> <hd id="AN0173972090-14">Outcome variables: Delinquent behaviors (Wave Two)</hd> <p>The measures of delinquent behaviors were used in the follow‐up survey. Participants were asked eight separate questions regarding the frequency of certain delinquent behaviors they undertook in the past year, including "swearing," "quarreling," "fighting," "bully," "truancy," "plagiarism or cheating," "smoking or drinking," and "going to coffeenets or game centers."</p> <p>These eight behaviors are considered as delinquent behavior in other counties and regions. The only exception is "going to coffeenets or game centers." In China, the government has imposed strict restrictions on minors' entry into Internet cafés (or coffeenets) to protect adolescents from a series of negative consequences caused by Internet addiction (Bao, [<reflink idref="bib1" id="ref81">1</reflink>], pp. 169–171). Thus, going to coffeenets or game centers is also considered as a typical delinquent behavior in the Chinese context (Liu et al., [<reflink idref="bib32" id="ref82">32</reflink>]).</p> <p>Response choices ranged across a 5‐point scale: 1 (never), 2 (occasionally), 3 (sometimes), 4 (often), 5 (always). Each question was treated separately as the frequency of a certain type of delinquent behavior for further analysis.</p> <hd id="AN0173972090-15">Control variables (baseline)</hd> <p>Several demographic and socioeconomic factors were considered as control variables. Sex was coded as female or male, age was measured by years, and ethnicity was coded as Han ethnicity (the majority ethnicity in China) or minorities. Further, Hukou status was coded as rural or urban, and the migrant status was coded as migrant or local. Moreover, to measure whether a participant was living in a single‐child family, participants without siblings were coded as yes; otherwise, as no. Additionally, the self‐rated family socioeconomic status was measured by asking participants, "how was your family economic status before you attended primary school?" Response choices ranged across a 5‐point scale (from 1 = "very poor" to 5 = "very rich") and "not clear." Responses with a "not clear" option were coded as missing.</p> <hd id="AN0173972090-16">Data analysis</hd> <p>First, a descriptive analysis was conducted to obtain an outline of the participants' characteristics. Paired <emph>t</emph>‐tests were conducted to test whether extracurricular time‐use behaviors during weekdays differed from those during weekends.</p> <p>Second, to investigate the time displacement and stimulation, co‐occurrence of delinquent behavior, and their interconnection relations, the Gaussian graphical model (GGM) was applied to estimate the inter‐variable correlation network. In the GGM, all key variables were denoted as nodes, and partial correlations among variables were denoted as edges (thicker edges meant higher correlations). Two network models were estimated: weekdays and weekends. In each network model, variables related to extracurricular time‐use behaviors and delinquent behaviors were categorized as two separate communities so that the connections within each community represented time displacement or stimulation and co‐occurrence of delinquency, respectively. In contrast, connections across communities represented the association between extracurricular time‐use behaviors and delinquent behaviors.</p> <p>According to the literature, the GGM can be estimated with both regularized and non‐regularized algorithms (Borsboom et al., [<reflink idref="bib4" id="ref83">4</reflink>]). Regularized algorithms were adopted in several PNA literature, due to the advantage of penalizing coefficients (Foygel & Drton, [<reflink idref="bib16" id="ref84">16</reflink>]) and then making the edge selection (Epskamp & Fried, [<reflink idref="bib13" id="ref85">13</reflink>]). This feature is useful in high dimensional settings—the number of variables (<emph>p</emph>) reaches or exceeds the number of observations (<emph>n</emph>). However, literature suggests that when encountering low dimensional settings (<emph>p</emph> ≪ <emph>n</emph>), non‐regularized algorithms could be considered as alternatives (Borsboom et al., [<reflink idref="bib4" id="ref86">4</reflink>]; Williams et al., [<reflink idref="bib50" id="ref87">50</reflink>]). Additionally, regularized algorithms have serious limitations to estimate edges' standard errors or confidence intervals (Borsboom et al., [<reflink idref="bib4" id="ref88">4</reflink>]), and non‐regularized algorithms, especially newly developed Bayesian approaches, could overcome such limitations (Williams, [<reflink idref="bib48" id="ref89">48</reflink>]; Williams et al., [<reflink idref="bib50" id="ref90">50</reflink>]). Thus, in the current study, the Bayesian Gaussian graph model (BGGM) was adopted to estimate the PNA.</p> <p>Third, edge differences between the two network models were further estimated to show how time displacement and stimulation, co‐occurrence of delinquent behavior, and their interconnection relations differ on weekdays and weekends.</p> <p>Finally, centrality analysis was conducted to identify which delinquent behavior was the most central and which time‐use behavior had the highest potential for triggering delinquent behaviors. Two node centrality measures were adopted: strength centrality and bridge strength centrality.</p> <p>Despite the similarity of calculations of the measures, centrality indices in PNA were not theoretically equivalent to centrality indices used in social network analysis (SNA), due to the fundamentally different meaning of edges in PNA and SNA. In SNA, an edge usually reflects whether and to what degree two nodes establish relations or connections. In this sense, edges in SNA reflect the <emph>distance</emph> between two nodes, while in PNA, an edge is a conditional partial relation coefficient between two variables, which reflects the <emph>strength</emph> of the relation between two variables.</p> <p>Therefore, degree centrality conveys a different meaning in PNA compared to SNA. In SNA, degree centrality is usually calculated based on the number of relationships (e.g., friendship, cooperation relationship, etc.) with other people (Faris et al., [<reflink idref="bib14" id="ref91">14</reflink>]; Huitsing et al., [<reflink idref="bib25" id="ref92">25</reflink>]), reflecting individuals' popularity in their friendship networks. In PNA, the degree centrality of a certain node is calculated by summing all absolute edge weights (partial correlations) that nodes were directly linked to (Epskamp & Fried, [<reflink idref="bib13" id="ref93">13</reflink>]). Thus, in PNA literature, degree centrality is often called "strength centrality" which reflect the ability of a certain node to predict other nodes in the network (Borsboom et al., [<reflink idref="bib4" id="ref94">4</reflink>]; Epskamp & Fried, [<reflink idref="bib13" id="ref95">13</reflink>]).</p> <p>Similarly, bridge strength centrality is not equal to betweenness centrality and closeness centrality in SNA. In SNA literature, betweenness and closeness centrality represent the "bridge" and "geographic center" in the social network (Faris et al., [<reflink idref="bib14" id="ref96">14</reflink>]), while in the PNA, since all edges were partial correlations rather than social distances, both indices of centrality cannot be interpreted identically as they are in SNA. In fact, leading scholars in PNA criticize the abuse of betweenness and closeness centrality in PNA literature by pointing out the failure to provide a reasonable interpretation of these two indices (Borsboom et al., [<reflink idref="bib4" id="ref97">4</reflink>]). Nevertheless, the idea of "bridge," meaning the potential for linking multiple communities (Burt, [<reflink idref="bib7" id="ref98">7</reflink>]), can be applied in PNA. In PNA, the "bridge" can be interpreted in terms of multiple variable communities linked via the bridge variable(s)—the variable(s) within one variable community that present the highest (or higher) potential link to variables in other communities (Borsboom, [<reflink idref="bib3" id="ref99">3</reflink>]). Based on the new definition of "bridge" in PNA, Jones et al. ([<reflink idref="bib29" id="ref100">29</reflink>]) proposed a bridge strength to measure one variable's bridge capacity by summing the absolute edge weights that nodes were directly linked to outside its community.</p> <p>The BGGMs were estimated using Bayesian statistics. A novel matrix‐F prior distribution with a scale of 0.50 and the posterior samples of 5000 (Williams, [<reflink idref="bib48" id="ref101">48</reflink>]) was set for the estimator. All missing responses were imputed with the respective posterior predictive distribution via the Gibbs sampling scheme (Hoff, [<reflink idref="bib23" id="ref102">23</reflink>]). To obtain more concise results, only posterior distributions of estimated partial correlations significant at the 99% credible interval were retained for further discussion. Further, all demographic and socioeconomic variables were considered control variables when estimating the edges. All analyses were conducted using R, and the BGGMs were estimated via the R package <emph>BGGM</emph> (Williams & Mulder, [<reflink idref="bib49" id="ref103">49</reflink>]). The detailed code used in the current study (including data cleaning, descriptive analysis, and BGGM estimation) is presented on the Open Science Framework Platform (https://osf.io/qrxv6/).</p> <hd id="AN0173972090-17">RESULTS</hd> <p></p> <hd id="AN0173972090-18">Descriptive results</hd> <p>Descriptive statistics for all study variables, including the number of valid responses, missing numbers, and minimum and maximum values, are presented in Table 1. Means and standard deviations are reported for continuous variables, and the counts and percentages are reported for categorical variables. Further, the paired <emph>t</emph>‐test results for the differences between extracurricular time‐use behaviors during weekdays and weekends among validated participants are presented in Table 2.</p> <p>1 TABLE Descriptive statistics.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Variables</th><th align="left"><italic>N</italic> (NA<xref ref-type="fn" rid="tfn1" />)</th><th align="left"><italic>M</italic> (SD)</th><th align="left">Count (%<xref ref-type="fn" rid="tfn2" />)</th><th align="left">Min</th><th align="left">Max</th></tr></thead><tbody valign="top"><tr><td align="left">Time use during weekdays</td></tr><tr><td>Doing homework</td><td>10,079 (200)</td><td>0.630 (1.171)</td><td>—</td><td>0</td><td>20.333</td></tr><tr><td>Attending cram school/classes</td><td>10,079 (200)</td><td>0.333 (0.991)</td><td>—</td><td>0</td><td>16</td></tr><tr><td>Doing sports</td><td>10,079 (200)</td><td>0.799 (1.121)</td><td>—</td><td>0</td><td>22.183</td></tr><tr><td>Extracurricular reading</td><td>10,079 (200)</td><td>1.050 (1.181)</td><td>—</td><td>0</td><td>20</td></tr><tr><td>Watching TV</td><td>10,079 (200)</td><td>0.954 (1.483)</td><td>—</td><td>0</td><td>20</td></tr><tr><td>Surfing the Internet or playing video games</td><td>10,079 (200)</td><td>0.619 (1.363)</td><td>—</td><td>0</td><td>24</td></tr><tr><td>Doing housework</td><td>10,079 (200)</td><td>0.984 (1.447)</td><td>—</td><td>0</td><td>19.583</td></tr><tr><td align="left">Time use during weekends</td></tr><tr><td>Doing homework</td><td>10,031 (248)</td><td>0.835 (1.351)</td><td>—</td><td>0</td><td>20</td></tr><tr><td>Attending cram school or classes</td><td>10,031 (248)</td><td>0.732 (1.504)</td><td>—</td><td>0</td><td>13</td></tr><tr><td>Doing sports</td><td>10,031 (248)</td><td>1.000 (1.308)</td><td>—</td><td>0</td><td>12.5</td></tr><tr><td>Extracurricular reading</td><td>10,031 (248)</td><td>1.332 (1.337)</td><td>—</td><td>0</td><td>14</td></tr><tr><td>Watching TV</td><td>10,031 (248)</td><td>1.727 (1.827)</td><td>—</td><td>0</td><td>24</td></tr><tr><td>Surfing the Internet or playing video games</td><td>10,031 (248)</td><td>1.313 (1.982)</td><td>—</td><td>0</td><td>24</td></tr><tr><td>Doing housework</td><td>10,031 (248)</td><td>1.397 (1.623)</td><td>—</td><td>0</td><td>24</td></tr><tr><td align="left">Delinquent behaviors</td></tr><tr><td>Swearing</td><td>9,413 (866)</td><td>2.218 (0.981)</td><td>—</td><td>1</td><td>5</td></tr><tr><td>Quarreling</td><td>9,406 (873)</td><td>1.836 (0.880)</td><td>—</td><td>1</td><td>5</td></tr><tr><td>Fighting</td><td>9,393 (886)</td><td>1.343 (0.700)</td><td>—</td><td>1</td><td>5</td></tr><tr><td>Bully</td><td>9,404 (875)</td><td>1.152 (0.520)</td><td>—</td><td>1</td><td>5</td></tr><tr><td>Truancy</td><td>9,411 (868)</td><td>1.096 (0.440)</td><td>—</td><td>1</td><td>5</td></tr><tr><td>Plagiarism or cheating</td><td>9,393 (886)</td><td>1.493 (0.788)</td><td>—</td><td>1</td><td>5</td></tr><tr><td>Smoking or drinking</td><td>9,413 (866)</td><td>1.110 (0.490)</td><td>—</td><td>1</td><td>5</td></tr><tr><td>Going to coffeenets or game centers</td><td>9,418 (861)</td><td>1.208 (0.644)</td><td>—</td><td>1</td><td>5</td></tr><tr><td align="left">Control variables</td></tr><tr><td>Sex (ref. male)</td><td>10,083 (196)</td><td>—</td><td>4773 (47.3%)</td><td>0</td><td>1</td></tr><tr><td>Age</td><td>10,045 (234)</td><td>13.578 (0.734)</td><td>—</td><td>12</td><td>18</td></tr><tr><td>Ethnicity (ref. minorities)</td><td>10,244 (35)</td><td>—</td><td>9342 (91.2%)</td><td>0</td><td>1</td></tr><tr><td>Rural hukou (ref. no)</td><td>9,626 (653)</td><td>—</td><td>5043 (52.4%)</td><td>0</td><td>1</td></tr><tr><td>Local citizen (ref. no)</td><td>10,022 (257)</td><td>—</td><td>8002 (79.8%)</td><td>0</td><td>1</td></tr><tr><td>Single kid (ref. no)</td><td>10,277 (2)</td><td>—</td><td>4415 (43.0%)</td><td>0</td><td>1</td></tr><tr><td>Self‐rated family socioeconomic status<xref ref-type="fn" rid="tfn3" /></td><td>9,527 (752)</td><td>2.975 (0.582)</td><td /><td>1</td><td>5</td></tr><tr><td>Very poor</td><td /><td /><td>179 (1.9%)</td><td /><td /></tr><tr><td>Poor</td><td /><td /><td>1102 (11.6%)</td><td /><td /></tr><tr><td>Middle</td><td /><td /><td>7125 (74.8%)</td><td /><td /></tr><tr><td>Rich</td><td /><td /><td>1023 (10.7%)</td><td /><td /></tr><tr><td>Very rich</td><td /><td /><td>98 (1.0%)</td><td /><td /></tr></tbody></table> </ephtml> </p> <p>1 a NA refers to the count of missing responses.</p> <ulist> <item>2 b All the percentages of categorical options were based on the valid responses.</item> <item>3 c The self‐rated family socioeconomic status was reported by two approaches: <emph>M</emph> (SD) and the percentages of each option.</item> <item>2 TABLE Paired t ‐test results for the differences between time use during weekdays and time use during weekends (N  = 9956).</item> </ulist> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Variables</th><th align="left"><italic>M</italic> (SD)</th><th align="left"><italic>M</italic> (SD)</th><th align="left"><italic>t</italic></th><th align="left"><italic>p</italic></th><th align="left">Difference (during weekends − during weekdays)</th><th align="left">Cohen's <italic>d</italic></th></tr><tr><th align="left">During weekends</th><th align="left">During weekdays</th><th align="left">[95% CI]</th><th align="left">[95% CI]</th></tr></thead><tbody valign="top"><tr><td>Doing homework</td><td>0.620 (1.136)</td><td>0.826 (1.316)</td><td>18.234</td><td><.001</td><td>0.206 [0.184, 0.228]</td><td>.183 [.163,.202]</td></tr><tr><td>Attending cram school or classes</td><td>0.331 (0.988)</td><td>0.730 (1.501)</td><td>28.281</td><td><.001</td><td>0.399 [0.371, 0.426]</td><td>.283 [.264,.303]</td></tr><tr><td>Doing sports</td><td>0.789 (1.083)</td><td>0.997 (1.303)</td><td>17.262</td><td><.001</td><td>0.209 [0.185, 0.232]</td><td>.173 [.153,.193]</td></tr><tr><td>Extracurricular reading</td><td>1.038 (1.145)</td><td>1.330 (1.334)</td><td>23.811</td><td><.001</td><td>0.292 [0.268, 0.316]</td><td>.239 [.219,.258]</td></tr><tr><td>Watching TV</td><td>0.942 (1.457)</td><td>1.717 (1.807)</td><td>44.008</td><td><.001</td><td>0.775 [0.740, 0.809]</td><td>.441 [.421,.461]</td></tr><tr><td>Surfing the Internet or playing video games</td><td>0.609 (1.343)</td><td>1.305 (1.953)</td><td>40.750</td><td><.001</td><td>0.695 [0.662, 0.729]</td><td>.408 [.389,.428]</td></tr><tr><td>Doing housework</td><td>0.965 (1.402)</td><td>1.389 (1.610)</td><td>30.474</td><td><.001</td><td>0.424 [0.397, 0.451]</td><td>.305 [.286,.325]</td></tr></tbody></table> </ephtml> </p> <p>As seen in Table 1, among the eight delinquent behaviors, swearing was the most prevalent (<emph>M</emph> = 2.218, SD = 0.981), while other delinquent behaviors were at the minimum value, suggesting that delinquent behavior was not prevalent among the respondents. Among the seven extracurricular time‐use behaviors, extracurricular reading ranked first during weekdays, while watching TV ranked first during weekends. Results of the paired <emph>t</emph>‐test revealed that respondents spend more time on all seven behaviors on the weekends than on the weekdays. Results of Cohen's d revealed that differences in attending cram school and classes, extracurricular reading, and watching TV yielded small effect sizes, while the difference in surfing the Internet or playing video games yielded a middle effect size.</p> <hd id="AN0173972090-19">Results of PNA</hd> <p>Figure 2 presents the results of the BGGM of the relations between extracurricular time‐use behaviors and delinquent behaviors during weekdays (Figure 2a) and weekends (Figure 2b). Taking into consideration that the network models estimated in the current study controlled the demographic and socioeconomic variables, detailed results of regression models for each node within two networks are reported in Tables S1 and S2. Also, coefficient of each edge are presented in Tables S3 and S4. Figure 3 presents the edge differences in the internal linkage among different extracurricular time‐use behaviors on weekdays and weekends, and the detailed coefficients of each edge's difference are reported in Table S5. Figure 4 presents the node strength and node bridge strength of two network models. Detailed information on the value of node strength and node bridge strength on weekdays and weekends can be found in Tables S6–S9.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CDV/01nov23/cdev13953-fig-0002.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cdev13953-fig-0002.jpg" title="2 Network estimations for relations among time use and delinquent behaviors. Blue edges denote positive partial correlations and red edges denote negative partial correlations after controlling other variables. Thicker edges denote more higher partial correlation between two nodes after controlling other variables. Only partial correlations with 99% credible intervals lower or higher than zero were summarized in the network estimations. (a) Relations among time use and delinquent behaviors on weekdays. (b) Relations among time use and delinquent behaviors on weekends." /> </p> <p></p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CDV/01nov23/cdev13953-fig-0003.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cdev13953-fig-0003.jpg" title="3 Edge differences between the network estimation of relations among time use and delinquent behaviors on weekdays and the network estimation of relations among time use and delinquent behaviors on weekends. The error bars correspond to 99% credible intervals while the results highlighted by light blue denote the edge differences were statistically significant. The reference edges were derived from the network estimations for relations among time use and delinquent behaviors during weekends. The edge difference with the 99% credible intervals lower than 0 denote the edge within the network results during weekdays was weaker than that within the network results during weekends while The edge difference with the 99% credible intervals higher than 0 denote the edge within the network results during weekdays was stronger than that within the network results during weekends. Numbers of each node could be identified as following: 1 ("Doing homework"); 2 ("Attending cram school or classes"); 3 ("Doing sports"); 4 ("Extracurricular reading"); 5 ("Watching TV"); 6 ("Surfing the Internet or playing video games"); 7 ("Doing housework"); 8 ("Swearing"); 9 ("Quarreling"); 10 ("Fighting"); 11 ("Bully"); 12 ("Truancy"); 13 ("Plagiarism or Cheating"); 14 ("Smoking or drinking"); 15 ("Going to coffeenets or game centers")." /> </p> <p></p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CDV/01nov23/cdev13953-fig-0004.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="cdev13953-fig-0004.jpg" title="4 Results for node strength and node bridge strength. Numbers of each node could be identified as following: 1 ("Doing homework"); 2 ("Attending cram school or classes"); 3 ("Doing sports"); 4 ("Extracurricular reading"); 5 ("Watching TV"); 6 ("Surfing the Internet or playing video games"); 7 ("Doing housework"); 8 ("Swearing"); 9 ("Quarreling"); 10 ("Fighting"); 11 ("Bully"); 12 ("Truancy"); 13 ("Plagiarism or Cheating"); 14 ("Smoking or drinking"); 15 ("Going to coffeenets or game centers"). (a, c) Node strength and node bridge strength for the network estimation of relationships among time use and delinquent behaviors during weekdays, respectively. (b, d) Node strength and node bridge strength for the network estimation of relationships among time use and delinquent behaviors during weekends, respectively." /> </p> <p></p> <hd id="AN0173972090-23">Time displacement and stimulation patterns and the differences between weekdays and weekends</hd> <p>As shown in Figure 2, both time displacement and time stimulation phenomena among different time‐use behaviors occurred on weekends, but only time displacement occurred on weekdays. On the weekends, participants who spent more time watching TV usually spent less time doing homework and attending cram school or classes. Doing housework during weekends also decreased the time‐used on attending cram school or classes. However, these three pairs of negative associations did not reach statistical significance on weekdays.</p> <p>Figure 3 presents eight significant edge differences between weekdays and weekends. Among them, seven were internal linkages among time‐use behaviors. First, attending cram school or classes and extracurricular reading were positively correlated on weekdays, while such a link was interrupted on weekends. Second, the positive relation between extracurricular reading and doing housework was stronger on weekdays than on weekends. Third, the linkage between attending cram school or classes and surfing the Internet or playing video games and the linkage between doing housework and surfing the Internet or playing video games were positive on weekdays but not significant on weekends. Fourth, the link between watching TV and doing housework, and the positive association between watching TV and surfing the Internet or playing video games on weekdays were stronger than on weekends.</p> <hd id="AN0173972090-24">Network structure of delinquency and central delinquent behavior</hd> <p>Figure 2 also reveals that eight delinquent behaviors were significantly intercorrelated after controlling for other variables. Two edges, the linkages between swearing and truancy and between quarreling and smoking or drinking, were negative, and the other 26 were positive. Also, it was observed that smoking or drinking was at the center of all the delinquent behaviors, suggesting that smoking or drinking could be the central behavior connecting other delinquency. Further centrality analysis was required to confirm the role of smoking or drinking in the delinquent behavior network.</p> <p>Figure 4a,b shows that smoking or drinking had the highest strength value, indicating it is the most central delinquent behavior among all other types of delinquent behaviors. Except for the negative linkage with quarreling, the other six delinquent behaviors were all positively related to smoking or drinking.</p> <hd id="AN0173972090-25">The relations between extracurricular time‐use behavior and delinquent behavior</hd> <p>Figure 2a presents the relations between extracurricular time‐use behavior and delinquent behavior on weekdays. The results showed that time‐use on extracurricular reading inhibited fighting but increased the risk of smoking or drinking. Further, time‐use doing housework decreased the likelihood of swearing at others. Moreover, time‐use on watching TV inhibited smoking or drinking but increased the risk of plagiarism or cheating in academic activities. Time spent on surfing the Internet or playing video games was positively linked to truancy and going to coffeenets or game centers.</p> <p>Figure 2b presents the relations between extracurricular time‐use behavior and delinquent behavior during weekends. Time spent on extracurricular reading decreased the likelihood of plagiarism or cheating on academic activities. Participants who spent more time doing sports had a higher risk of fighting with others. Further, with more time spent watching TV, participants had higher risks of swearing at others and plagiarism or cheating on their academic activities. Additionally, time spent surfing the Internet or playing video games increased the risk of four delinquent behaviors (swearing, truancy, plagiarism or cheating, and going to coffeenets or game centers).</p> <p>Figure 4c,d presents the extracurricular time‐use behavior on surfing the Internet or playing video games which had the highest value of bridge strength, suggesting it may have the highest potential for triggering delinquent behaviors.</p> <p>Figure 3 reveals only one edge related to relations between extracurricular time‐use behavior and delinquent behavior, namely the edge linked to watching TV and swearing, was positively and significantly correlated on weekends but not significant on weekdays.</p> <hd id="AN0173972090-26">DISCUSSION</hd> <p>The current study aimed to explore the heterogeneity of the relation between extracurricular time‐use behaviors and juvenile delinquency using a nationally representative longitudinal survey of school students in China. Specifically, the current study focused on exploring the variable‐level network structure of different extracurricular time‐use behaviors, delinquent behaviors, and the association of specific extracurricular time‐use behaviors with specific delinquent behaviors during weekdays and weekends. Clearly, the traditional literature, which focuses on linear relations, may be inadequate for exploring such complex mechanisms. Therefore, the current study adopted PNA, a recently developed approach that explores and displays the variable‐level correlation network of all observed variables, to elucidate the nuanced relation between time‐use behaviors and juvenile delinquency.</p> <p>The results of the current study suggested that time displacement and stimulation occurred on weekdays while only time displacement occurred on weekends. Moreover, the current study observed a strong intercorrelation among varieties of delinquent behaviors, implying the possible existence of problem behavior syndrome. More importantly, taking advantage of the PNA, the current study found that smoking or drinking were central delinquent behaviors that have the potential to trigger other delinquent behaviors, which may shed light on future prevention strategies. Lastly, negative consequences of specific time‐use behaviors were more likely to occur on weekends than on weekdays. Moreover, the same time‐use behavior may function differently on weekdays and weekends. Both of which support the hypothesis of <emph>unstructured socializing</emph>.</p> <p>Considering the debate between time displacement and time stimulation hypotheses, the current study employed PNA to provide a global view of the time‐use patterns—these hypotheses could be both right but under different circumstances. On weekdays, the time stimulation phenomenon was more likely to occur, as the current study found time‐use behaviors are all positively correlated on weekdays. On weekdays, adolescents' extracurricular activities are planned and embedded in a relatively short period of time, which does not lead to significant competition among various types of extracurricular activities. However, on the weekends, due to the amount of available time and lack of structuredness, adolescents could flexibly arrange their activities, and some pairs of behaviors exhibited negative correlations, suggesting time displacement occurred on the weekends. The results of edge difference also indicated that the effect of positive correlations between some pairs of time‐use behaviors attenuated or even disappeared on the weekends.</p> <p>The difference between weekdays and weekends could also be explained by the nature of structuredness between extracurricular time‐use activities on weekdays and weekends. As suggested by SDH, "a pre‐planned, segmented, and adult‐supervised compulsory environment" is more likely to occur on weekdays than on weekends. Further, the paired <emph>t</emph>‐test supported the prediction of SDH. The results showed that the amount of time spent on all seven behaviors was significantly larger on weekends than on weekdays, and typical unstructured behavior like surfing the Internet or playing video games had the largest effect. Therefore, as the structured environment often allows little discretion for adolescents to allocate their time for extracurricular time‐use activities, time displacement may be less likely to occur on weekdays than on weekends.</p> <p>In line with PBT, the findings of the current study suggested that most delinquent behaviors are positively correlated with each other, forming a "network of problem behaviors." More crucially, centrality analysis suggests that smoking or drinking has the largest strength value, suggesting it could be the central delinquent behavior with the potential to trigger other delinquent behaviors. The role of smoking or drinking could be understood through Moffitt's assumption of the "maturity gap." As suggested by Moffitt ([<reflink idref="bib38" id="ref104">38</reflink>]), delinquency is "social mimicry" of the antisocial style of life‐course‐persistent youth. Crimes that serve to meet adolescents' lust for acknowledgment and privilege are especially prevalent among these adolescent‐limited offenders (p. 691). In this sense, the central role of smoking or drinking, which could be considered typical of adult‐privileged behavior, could serve as a sign of "gateway delinquency" that links or triggers other delinquent behaviors.</p> <p> <emph>Unstructured socializing</emph> suggests that time spent on activities featuring lack of structure, absence of authority figures, and presence of deviant peers increases the likelihood of delinquency. In the current study, two pieces of evidence were presented: On the one hand, negative consequences of specific extracurricular time‐use behaviors were more likely to occur on the weekends than on the weekdays. For example, doing sports was positively associated with fighting on the weekends and surfing the Internet or playing video games was positively associated with more delinquent behaviors on weekends than on weekdays. On the other hand, the same behaviors may also function differently on weekdays and weekends. For example, watching TV was negatively associated with smoking or drinking on weekdays but not on weekends. The above‐mentioned evidence could also be explained by the difference in the activities' structuredness on weekdays versus weekends. On the weekends, watching TV, surfing the Internet or playing video games, and doing sports are less structured, so they have a higher potential of triggering delinquent behavior. The only unexpected finding was the positive association between extracurricular reading and smoking or drinking. This may be attributed to different reading content on weekdays versus weekends. It could also imply a reversed effect of time‐use behavior under extreme structured time. Further research is needed to explore the side effects of extracurricular reading.</p> <p>The current study has three methodological limitations due to the nature of secondary data analysis. First, data from two waves were merged to identify causal effects from extracurricular time‐use to delinquent behavior. However, the key variables within specific communities were measured at the same time. Thus, it is hard to make causal inferences on community intercorrelations. Second, the measurement of extracurricular time‐use activities did not indicate whether the activities were accompanied by certain peers or the absence of authority figures. Thus, the nature of the activities may not be thoroughly investigated and discussed. Another limitation is the use of stylized questions to measure extracurricular time‐use activities, leading to the possibility of contributing missingness in the time‐use variables (e.g., recall bias) and social desirability in the responses (e.g., more reports of socially desirable activities and fewer reports of undesirable ones).</p> <p>To overcome the above three limitations, it is suggested that future studies adopt a longitudinal design with more waves to identify directions of time displacement and stimulation and delinquent behavior diffusion. Moreover, compared with the 1‐year time interval used in the current study, more frequent time intervals between each wave (e.g., 1 month or even 1 week) should be employed to investigate temporal causalities. Furthermore, future studies are expected to develop more contextual measures of extracurricular time‐use activities. Lastly, the missingness of stylized questions can be addressed by the imputation method adopted in the current study. The social desirability issue of stylized questions may call for a more sophisticated method of measurement, such as a time‐diary, in future studies.</p> <p>In addition, the current study provided several important directions (testable hypotheses) for future research. First, the generalizability of the current study should be tested, especially whether the role of smoking or drinking as a central delinquent behavior and the role of surfing the Internet or playing video games as a core risk factor for delinquent behavior are worth exploring in other countries and regions. Specifically, the unique features in China, and perhaps other East Asian countries, provide a unique opportunity to examine the time displacement of adolescents. A lower degree of time displacement is likely to be observed in western countries given the relatively lower amount of time‐use on after‐school academic activities, which require further comparative study to verify an optional time‐use pattern and its consequences. Moreover, although going to coffeenets or game centers is a unique delinquent behavior in the Chinese context, it can be understood as a status offense in the Western context. Therefore, when generalizing the findings of this study, this behavior can be understood as a status offense behavior such as going to a bar or running away from home. Of course, future empirical studies are needed to test whether this substitution is reasonable. Second, the relations between extracurricular time‐use and delinquent behavior may vary across developmental stages according to Moffitt's ([<reflink idref="bib38" id="ref105">38</reflink>]) "maturity gap" hypothesis. Future studies could also examine the trajectories of time‐use activities, delinquent behaviors, and their relations. Third, the current study found a significant difference in patterns between weekdays and weekends, leading to support for the SDH. It is still worth investigating whether the difference can be observed on a larger time scale (e.g., behavior patterns during school vs. during vacation).</p> <p>The results of the current study have important policy implications for preventing delinquency, especially the findings on the existence of time displacement during weekends and the central role of drinking or smoking among all delinquent behaviors. On the one hand, as time displacement is more likely to occur on weekends, time‐use activities that may trigger delinquent behavior may be crowded out or substituted by other time‐use activities. For example, watching TV, which is positively associated with swearing and plagiarism or cheating, could be crowded out or substituted by academic activities such as doing homework and attending cram school or classes during the weekend. Thus, policies that increase the availability of academically related activities (e.g., publicly financed study tours) during weekends could reduce the consequences of watching TV.</p> <p>On the other hand, since drinking or smoking is the most central delinquent behavior, it may serve as a precursor to other (serious) delinquent behaviors. The implications for preventing adolescents from using alcohol and cigarettes could be a direct and effective approach to breaking the problem behavior syndrome. However, as suggested by the PNA, drinking or smoking is not linked to time‐use activities on weekends. Instead, two pathways exist: "doing sports—fighting—drinking or smoking" and "surfing the Internet or playing video games—truancy and going to coffeenets or game center—drinking or smoking." Both pathways could be sabotaged by the presence of appropriate authority figures (such as parents, teachers, etc.) and strengthening the structuredness of the activities (such as organizing sport or e‐sport competitions). Moreover, given that the bridge strength of "surfing the Internet or playing video games" is the highest among multiple time‐use behaviors in both weekday and weekend networks, a priority approach for policymakers and social workers is to block the latter pathway.</p> <p>Taken together, the results of the current study highlight the complexity of the activity‐delinquency nexus, which may once again suggest the inadequacy of using traditional methods to understand the nexus. The complexity requires sophisticated methods that move forward and beyond linear relations. With the new development of network science, network‐based approaches such as PNA allow researchers to embrace this complexity and have the potential to stimulate both theoretical and empirical efforts in future studies.</p> <hd id="AN0173972090-27">ACKNOWLEDGMENTS</hd> <p>This study was supported by the Major Project of the National Social Science Fund of China (Grant No. 19ZDA324) and the Guanghua Talent Project of Southwestern University of Finance and Economics.</p> <hd id="AN0173972090-28">CONFLICT OF INTEREST STATEMENT</hd> <p>The authors declare no competing interests.</p> <hd id="AN0173972090-29">DATA AVAILABILITY STATEMENT</hd> <p>Data are available at the following URL: <ulink href="http://ceps.ruc.edu.cn/English/Home.htm">http://ceps.ruc.edu.cn/English/Home.htm</ulink>. Code is available at the following URL: https://osf.io/qrxv6/. There are no additional materials necessary to attempt to replicate the findings. The analyses presented here were not preregistered.</p> <hd id="AN0173972090-30">ETHICS STATEMENT</hd> <p>This is a secondary analysis of a national survey with ethical approval. No ethical approval was sought for this study.</p> <p>GRAPH: Appendix S1.</p> <ref id="AN0173972090-31"> <title> Footnotes </title> <blist> <bibl id="bib1" idref="ref65" type="bt">1</bibl> <bibtext> This article was accepted under the Editorship of Glenn I. Roisman.</bibtext> </blist> </ref> <ref id="AN0173972090-32"> <title> REFERENCES </title> <blist> <bibtext> Bao, W.‐N. (2017). Delinquent youth in a transforming China. Springer International Publishing. https://doi.org/10.1007/978‐3‐319‐63727‐3</bibtext> </blist> <blist> <bibl id="bib2" idref="ref12" type="bt">2</bibl> <bibtext> Barnes, G. M., Hoffman, J. H., Welte, J. W., Farrell, M. P., & Dintcheff, B. A. (2007). Adolescents' time use: Effects on substance use, delinquency and sexual activity. Journal of Youth and Adolescence, 36 (5), 697 – 710. https://doi.org/10.1007/s10964‐006‐9075‐0</bibtext> </blist> <blist> <bibl id="bib3" idref="ref20" type="bt">3</bibl> <bibtext> Borsboom, D. (2017). A network theory of mental disorders. World Psychiatry, 16 (1), 5 – 13. https://doi.org/10.1002/wps.20375</bibtext> </blist> <blist> <bibl id="bib4" idref="ref41" type="bt">4</bibl> <bibtext> Borsboom, D., Deserno, M. K., Rhemtulla, M., Epskamp, S., Fried, E. I., McNally, R. J., Robinaugh, D. J., Perugini, M., Dalege, J., Costantini, G., Isvoranu, A.‐M., Wysocki, A. C., van Borkulo, C. D., van Bork, R., & Waldorp, L. J. (2021). Network analysis of multivariate data in psychological science. Nature Reviews Methods Primers, 1 (1), 1. https://doi.org/10.1038/s43586‐021‐00055‐w</bibtext> </blist> <blist> <bibl id="bib5" idref="ref58" type="bt">5</bibl> <bibtext> Brazendale, K., Beets, M. W., Weaver, R. G., Pate, R. R., Turner‐McGrievy, G. M., Kaczynski, A. T., Chandler, J. L., Bohnert, A., & von Hippel, P. T. (2017). Understanding differences between summer vs. school obesogenic behaviors of children: The structured days hypothesis. International Journal of Behavioral Nutrition and Physical Activity, 14, 100. https://doi.org/10.1186/s12966‐017‐0555‐2</bibtext> </blist> <blist> <bibl id="bib6" idref="ref28" type="bt">6</bibl> <bibtext> Bu, F., Steptoe, A., Mak, H. W., & Fancourt, D. (2021). Time use and mental health in UK adults during an 11‐week COVID‐19 lockdown: A panel analysis. The British Journal of Psychiatry: The Journal of Mental Science, 219 (4), 551 – 556. https://doi.org/10.1192/bjp.2021.44</bibtext> </blist> <blist> <bibl id="bib7" idref="ref98" type="bt">7</bibl> <bibtext> Burt, R. S. (2012). Network‐related personality and the agency question: Multirole evidence from a virtual world. American Journal of Sociology, 118 (3), 543 – 591. https://doi.org/10.1086/667856</bibtext> </blist> <blist> <bibl id="bib8" idref="ref2" type="bt">8</bibl> <bibtext> Cohen, L. E., & Felson, M. (1979). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44 (4), 588 – 608. https://doi.org/10.2307/2094589</bibtext> </blist> <blist> <bibl id="bib9" idref="ref57" type="bt">9</bibl> <bibtext> de Vries, I., Abeyta, S., Lockwood, S., Cuevas, C. A., & Rothman, E. (2022). A network approach to examining co‐occurring victimization and perpetration in dating abuse among a nationally representative sample of US adolescents. Journal of Adolescent Health, 70, 934 – 941. https://doi.org/10.1016/j.jadohealth.2021.12.026</bibtext> </blist> <blist> <bibtext> Dembo, R., Wareham, J., Schmeidler, J., & Winters, K. C. (2016). Longitudinal effects of a second‐order multi‐problem factor of sexual risk, marijuana use, and delinquency on future arrest among truant youths. Journal of Child & Adolescent Substance Abuse, 25 (6), 557 – 574. https://doi.org/10.1080/1067828X.2016.1153554</bibtext> </blist> <blist> <bibtext> Dong, Y., & Ding, C. (2012). Adolescent risk behaviors: Studying typical and atypical individuals via multidimensional scaling profile analysis. Journal of Adolescence, 35 (1), 197 – 205. https://doi.org/10.1016/j.adolescence.2011.02.006</bibtext> </blist> <blist> <bibtext> Donovan, J. E., & Jessor, R. (2016). Problem behavior theory and the problem behavior syndrome. In I. R. Jessor (Ed.), The origins and development of problem behavior theory (pp. 89 – 109). Springer International Publishing. https://doi.org/10.1007/978‐3‐319‐40886‐6_6</bibtext> </blist> <blist> <bibtext> Epskamp, S., & Fried, E. I. (2018). A tutorial on regularized partial correlation networks. Psychological Methods, 23 (4), 617 – 634. https://doi.org/10.1037/met0000167</bibtext> </blist> <blist> <bibtext> Faris, R., Felmlee, D., & McMillan, C. (2020). With friends like these: Aggression from Amity and Equivalence. American Journal of Sociology, 126 (3), 673 – 713. https://doi.org/10.1086/712972</bibtext> </blist> <blist> <bibtext> Farrell, A. D., Henry, D. B., Mays, S. A., & Schoeny, M. E. (2011). Parents as moderators of the impact of school norms and peer influences on aggression in middle school students. Child Development, 82 (1), 146 – 161. https://doi.org/10.1111/j.1467‐8624.2010.01546.x</bibtext> </blist> <blist> <bibtext> Foygel, R., & Drton, M. (2010). Extended Bayesian information criteria for Gaussian graphical models. In J. Lafferty, C. Williams, J. Shawe‐Taylor, R. Zemel, & A. Culotta (Eds.), Advances in neural information processing systems (Vol. 23, pp. 1 – 14). Curran Associates, Inc.</bibtext> </blist> <blist> <bibtext> Fried, E. I., Papanikolaou, F., & Epskamp, S. (2022). Mental health and social contact during the COVID‐19 pandemic: An ecological momentary assessment study. Clinical Psychological Science, 10 (2), 340 – 354. https://doi.org/10.1177/21677026211017839</bibtext> </blist> <blist> <bibtext> Gottfredson, D. C., Gottfredson, G. D., & Weisman, S. A. (2001). Timing of delinquent behavior and its implications for after‐school programs. Criminology & Public Policy, 1 (1), 61 – 86. https://doi.org/10.1111/j.1745‐9133.2001.tb00077.x</bibtext> </blist> <blist> <bibtext> Henson, B., Wilcox, P., Reyns, B. W., & Cullen, F. T. (2010). Gender, adolescent lifestyles, and violent victimization: Implications for routine activity theory. Victims & Offenders, 5 (4), 303 – 328. https://doi.org/10.1080/15564886.2010.509651</bibtext> </blist> <blist> <bibtext> Hindelang, M. J., Gottfredson, M. R., & Garofalo, J. (1978). Victims of personal crime: An empirical foundation for a theory of personal victimization. Ballinger.</bibtext> </blist> <blist> <bibtext> Hirschi, T. (1969). Causes of delinquency. Transaction Publishers.</bibtext> </blist> <blist> <bibtext> Hoeben, E. M., & Weerman, F. M. (2016). Why is involvement in unstructured socializing related to adolescent delinquency? Criminology, 54 (2), 242 – 281. https://doi.org/10.1111/1745‐9125.12105</bibtext> </blist> <blist> <bibtext> Hoff, P. D. (2009). A first course in Bayesian statistical methods. Springer. https://doi.org/10.1007/978‐0‐387‐92407‐6</bibtext> </blist> <blist> <bibtext> Hu, S., & Mu, Z. (2020). Some time is better spent than other time: Chinese Adolescents' time use and developmental outcomes. Child Indicators Research, 13 (5), 1739 – 1765. https://doi.org/10.1007/s12187‐020‐09718‐8</bibtext> </blist> <blist> <bibtext> Huitsing, G., Veenstra, R., Sainio, M., & Salmivalli, C. (2012). "It must be me" or "it could be them?": The impact of the social network position of bullies and victims on victims' adjustment. Social Networks, 34 (4), 379 – 386. https://doi.org/10.1016/j.socnet.2010.07.002</bibtext> </blist> <blist> <bibtext> Hunt, E., & McKay, E. A. (2015). What can be learned from adolescent time diary research. Journal of Adolescent Health, 56 (3), 259 – 266. https://doi.org/10.1016/j.jadohealth.2014.11.007</bibtext> </blist> <blist> <bibtext> Jessor, R. (2013). Problem behavior theory: A half century of research on adolescent behavior and development. In R. M. Lerner, A. C. Petersen, R. K. Silbereisen, & J. Brooks‐Gunn (Eds.), The developmental science of adolescence (1st ed., pp. 253 – 270). Psychology Press.</bibtext> </blist> <blist> <bibtext> Jessor, R., & Jessor, S. L. (1977). Problem behavior and psychosocial development: A longitudinal study of youth. Academic Press. <ulink href="http://lib.ugent.be/catalog/rug01:000002543">http://lib.ugent.be/catalog/rug01:000002543</ulink></bibtext> </blist> <blist> <bibtext> Jones, P. J., Ma, R., & McNally, R. J. (2021). Bridge centrality: A network approach to understanding comorbidity. Multivariate Behavioral Research, 56 (2), 353 – 367. https://doi.org/10.1080/00273171.2019.1614898</bibtext> </blist> <blist> <bibtext> Lee, S. J. (2009). Online communication and adolescent social ties: Who benefits more from Internet use? Journal of Computer‐Mediated Communication, 14 (3), 509 – 531. https://doi.org/10.1111/j.1083‐6101.2009.01451.x</bibtext> </blist> <blist> <bibtext> Liu, N., & Ma, Z. (2022). Psychiatric reactions among the non‐exposed population who viewed disaster‐related short videos: Evidence from the 2021 Henan floods. Journal of Psychiatric Research, 150, 21 – 33. https://doi.org/10.1016/j.jpsychires.2022.03.036</bibtext> </blist> <blist> <bibtext> Liu, T.‐H., Li, S. D., Zhang, X., & Xia, Y. (2019). The Spillover mechanisms linking family conflicts and juvenile delinquency among Chinese adolescents. International Journal of Offender Therapy and Comparative Criminology. https://doi.org/10.1177/0306624X19842057</bibtext> </blist> <blist> <bibtext> Lushin, V., Jaccard, J., Ivaniushina, V., & Alexandrov, D. (2017). Vocational education paths, youth activities, and underage drinking in Russia: How early does the trouble start? International Journal of Drug Policy, 45, 48 – 55. https://doi.org/10.1016/j.drugpo.2017.05.035</bibtext> </blist> <blist> <bibtext> Ma, Z., Zhao, F., Wang, Y., Liu, T., & Chao, N. (2022). Network analysis of time use and depressive symptoms among emerging adults: Findings from the Guizhou population health cohort study. Frontiers in Psychiatry, 13, 809745. https://doi.org/10.3389/fpsyt.2022.809745</bibtext> </blist> <blist> <bibtext> Martins, P. C., Oliveira, V. H., Mendes, S. M., & Fernández‐Pacheco, G. (2021). After‐school time use of urban adolescents: Effects on achievement, problem behaviors, and happiness. Journal of Leisure Research, 52 (3), 286 – 306. https://doi.org/10.1080/00222216.2020.1785977</bibtext> </blist> <blist> <bibtext> Miller, K. E., Melnick, M. J., Barnes, G. M., Sabo, D., & Farrell, M. P. (2007). Athletic involvement and adolescent delinquency. Journal of Youth and Adolescence, 36 (5), 711 – 723. https://doi.org/10.1007/s10964‐006‐9123‐9</bibtext> </blist> <blist> <bibtext> Mkhitaryan, S., Crutzen, R., Steenaart, E., & de Vries, N. K. (2019). Network approach in health behavior research: How can we explore new questions? Health Psychology and Behavioral Medicine, 7 (1), 362 – 384. https://doi.org/10.1080/21642850.2019.1682587</bibtext> </blist> <blist> <bibtext> Moffitt, T. E. (1993). Adolescence‐limited and life‐course‐persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100 (4), 674 – 701. https://doi.org/10.1037/0033‐295X.100.4.674</bibtext> </blist> <blist> <bibtext> Moses, A. M., Jennings, N. A., Bennett, C. M., Cregan, G. E., Maras, E. Q., & Saritoprak, S. N. (2019). Time, access, context and beliefs: A survey of electronic media in early childhood settings. Computers in the Schools, 36 (4), 274 – 292. https://doi.org/10.1080/07380569.2019.1675033</bibtext> </blist> <blist> <bibtext> Nishino, H. J., & Larson, R. (2003). Japanese Adolescents' free time: Juku, Bukatsu, and governmental efforts to create more meaningful leisure. New Directions for Child and Adolescent Development, 2003 (99), 23 – 36. https://doi.org/10.1002/cd.64</bibtext> </blist> <blist> <bibtext> Osgood, D. W., Wilson, J. K., O'Malley, P. M., Bachman, J. G., & Johnston, L. D. (1996). Routine activities and individual deviant behavior. American Sociological Review, 61 (4), 635. https://doi.org/10.2307/2096397</bibtext> </blist> <blist> <bibtext> Putnam, R. D. (2000). Bowling alone: America's declining social capital. In L. Crothers & C. Lockhart (Eds.), Culture and politics: A reader (pp. 223 – 234). Palgrave Macmillan. https://doi.org/10.1007/978‐1‐349‐62965‐7_12</bibtext> </blist> <blist> <bibtext> Rasmussen, E. E., Strouse, G. A., Colwell, M. J., Russo Johnson, C., Holiday, S., Brady, K., Flores, I., Troseth, G., Wright, H. D., Densley, R. L., & Norman, M. S. (2019). Promoting preschoolers' emotional competence through prosocial TV and mobile app use. Media Psychology, 22 (1), 1 – 22. https://doi.org/10.1080/15213269.2018.1476890</bibtext> </blist> <blist> <bibtext> Tanioka, I., & Glaser, D. (1991). School uniforms, routine activities, and the social control of delinquency in Japan. Youth & Society, 23 (1), 50 – 75. https://doi.org/10.1177/0044118X91023001003</bibtext> </blist> <blist> <bibtext> Twenge, J. M., Spitzberg, B. H., & Campbell, W. K. (2019). Less in‐person social interaction with peers among U.S. adolescents in the 21st century and links to loneliness. Journal of Social and Personal Relationships, 36 (6), 1892 – 1913. https://doi.org/10.1177/0265407519836170</bibtext> </blist> <blist> <bibtext> Valkenburg, P. M., & Peter, J. (2007). Online communication and adolescent well‐being: Testing the stimulation versus the displacement hypothesis. Journal of Computer‐Mediated Communication, 12 (4), 1169 – 1182. https://doi.org/10.1111/j.1083‐6101.2007.00368.x</bibtext> </blist> <blist> <bibtext> Wikström, P.‐O. H., Oberwittler, D., Treiber, K., & Hardie, B. (2012). Breaking rules: The social and situational dynamics of young people's urban crime. Oxford University Press.</bibtext> </blist> <blist> <bibtext> Williams, D. R. (2021). Bayesian estimation for Gaussian graphical models: Structure learning, predictability, and network comparisons. Multivariate Behavioral Research, 56 (2), 336 – 352. https://doi.org/10.1080/00273171.2021.1894412</bibtext> </blist> <blist> <bibtext> Williams, D. R., & Mulder, J. (2020). BGGM: Bayesian Gaussian graphical models in R. Journal of Open Source Software, 5 (51), 2111. https://doi.org/10.21105/joss.02111</bibtext> </blist> <blist> <bibtext> Williams, D. R., Rast, P., Pericchi, L. R., & Mulder, J. (2020). Comparing Gaussian graphical models with the posterior predictive distribution and Bayesian model selection. Psychological Methods, 25 (5), 653 – 672. https://doi.org/10.1037/met0000254</bibtext> </blist> <blist> <bibtext> Williams, T. C., Bach, C. C., Matthiesen, N. B., Henriksen, T. B., & Gagliardi, L. (2018). Directed acyclic graphs: A tool for causal studies in paediatrics. Pediatric Research, 84 (4), 4. https://doi.org/10.1038/s41390‐018‐0071‐3</bibtext> </blist> <blist> <bibtext> Willoughby, T., Chalmers, H., & Busseri, M. A. (2004). Where is the syndrome? Examining co‐occurrence among multiple problem behaviors in adolescence. Journal of Consulting and Clinical Psychology, 72 (6), 1022 – 1037. https://doi.org/10.1037/0022‐006X.72.6.1022</bibtext> </blist> <blist> <bibtext> Wong, S. K. (2005). The effects of adolescent activities on delinquency: A differential involvement approach. Journal of Youth and Adolescence, 34 (4), 321 – 333. https://doi.org/10.1007/s10964‐005‐5755‐4</bibtext> </blist> <blist> <bibtext> Yakovenko, I., & Hodgins, D. C. (2018). A scoping review of co‐morbidity in individuals with disordered gambling. International Gambling Studies, 18 (1), 143 – 172. https://doi.org/10.1080/14459795.2017.1364400</bibtext> </blist> </ref> <aug> <p>By Yiwei Xia and Zhihao Ma</p> <p>Reported by Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib21" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib20" firstref="ref3"></nolink> <nolink nlid="nl3" bibid="bib41" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib47" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib22" firstref="ref6"></nolink> <nolink nlid="nl6" bibid="bib53" firstref="ref9"></nolink> <nolink nlid="nl7" bibid="bib15" firstref="ref10"></nolink> <nolink nlid="nl8" bibid="bib44" firstref="ref11"></nolink> <nolink nlid="nl9" bibid="bib33" firstref="ref13"></nolink> <nolink nlid="nl10" bibid="bib35" firstref="ref14"></nolink> <nolink nlid="nl11" bibid="bib36" firstref="ref15"></nolink> <nolink nlid="nl12" bibid="bib19" firstref="ref16"></nolink> <nolink nlid="nl13" bibid="bib43" firstref="ref18"></nolink> <nolink nlid="nl14" bibid="bib18" firstref="ref19"></nolink> <nolink nlid="nl15" bibid="bib29" firstref="ref24"></nolink> <nolink nlid="nl16" bibid="bib13" firstref="ref25"></nolink> <nolink nlid="nl17" bibid="bib50" firstref="ref26"></nolink> <nolink nlid="nl18" bibid="bib31" firstref="ref27"></nolink> <nolink nlid="nl19" bibid="bib34" firstref="ref29"></nolink> <nolink nlid="nl20" bibid="bib24" firstref="ref30"></nolink> <nolink nlid="nl21" bibid="bib26" firstref="ref31"></nolink> <nolink nlid="nl22" bibid="bib42" firstref="ref32"></nolink> <nolink nlid="nl23" bibid="bib45" firstref="ref34"></nolink> <nolink nlid="nl24" bibid="bib30" firstref="ref35"></nolink> <nolink nlid="nl25" bibid="bib46" firstref="ref36"></nolink> <nolink nlid="nl26" bibid="bib39" firstref="ref38"></nolink> <nolink nlid="nl27" bibid="bib51" firstref="ref39"></nolink> <nolink nlid="nl28" bibid="bib37" firstref="ref40"></nolink> <nolink nlid="nl29" bibid="bib28" firstref="ref44"></nolink> <nolink nlid="nl30" bibid="bib12" firstref="ref45"></nolink> <nolink nlid="nl31" bibid="bib27" firstref="ref46"></nolink> <nolink nlid="nl32" bibid="bib54" firstref="ref47"></nolink> <nolink nlid="nl33" bibid="bib10" firstref="ref48"></nolink> <nolink nlid="nl34" bibid="bib52" firstref="ref49"></nolink> <nolink nlid="nl35" bibid="bib11" firstref="ref50"></nolink> <nolink nlid="nl36" bibid="bib17" firstref="ref71"></nolink> <nolink nlid="nl37" bibid="bib48" firstref="ref77"></nolink> <nolink nlid="nl38" bibid="bib40" firstref="ref79"></nolink> <nolink nlid="nl39" bibid="bib32" firstref="ref82"></nolink> <nolink nlid="nl40" bibid="bib16" firstref="ref84"></nolink> <nolink nlid="nl41" bibid="bib14" firstref="ref91"></nolink> <nolink nlid="nl42" bibid="bib25" firstref="ref92"></nolink> <nolink nlid="nl43" bibid="bib23" firstref="ref102"></nolink> <nolink nlid="nl44" bibid="bib49" firstref="ref103"></nolink> <nolink nlid="nl45" bibid="bib38" firstref="ref104"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1403845
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Network Structure of the Links between Extracurricular Time-Use and Delinquent Behaviors: Moving Forward and Beyond Linear Relations
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Xia%2C+Yiwei%22">Xia, Yiwei</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-7360-732X">0000-0001-7360-732X</externalLink>)<br /><searchLink fieldCode="AR" term="%22Ma%2C+Zhihao%22">Ma, Zhihao</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-7983-4136">0000-0001-7983-4136</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Child+Development%22"><i>Child Development</i></searchLink>. 2023 94(6):1697-1712.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 16
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2023
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22Students%22">Students</searchLink><br /><searchLink fieldCode="DE" term="%22Extracurricular+Activities%22">Extracurricular Activities</searchLink><br /><searchLink fieldCode="DE" term="%22Time+on+Task%22">Time on Task</searchLink><br /><searchLink fieldCode="DE" term="%22Delinquency%22">Delinquency</searchLink><br /><searchLink fieldCode="DE" term="%22Delinquency+Prevention%22">Delinquency Prevention</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Behavior%22">Student Behavior</searchLink><br /><searchLink fieldCode="DE" term="%22Behavior+Problems%22">Behavior Problems</searchLink><br /><searchLink fieldCode="DE" term="%22Smoking%22">Smoking</searchLink><br /><searchLink fieldCode="DE" term="%22Drinking%22">Drinking</searchLink><br /><searchLink fieldCode="DE" term="%22Negative+Reinforcement%22">Negative Reinforcement</searchLink><br /><searchLink fieldCode="DE" term="%22School+Schedules%22">School Schedules</searchLink><br /><searchLink fieldCode="DE" term="%22Weekend+Programs%22">Weekend Programs</searchLink><br /><searchLink fieldCode="DE" term="%22Risk%22">Risk</searchLink><br /><searchLink fieldCode="DE" term="%22At+Risk+Students%22">At Risk Students</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22China%22">China</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1111/cdev.13953
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0009-3920<br />1467-8624
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Using psychological network analysis, this study explored the heterogeneity of the network structure between extracurricular time-use and delinquency using a nationally representative longitudinal survey of at-school students in China (N = 10,279, 47.3% female, average age 13.6, 91.2% Han ethnicity). The results are threefold: First, time stimulation of activities occurs on weekdays, while time displacement and stimulation occur on weekends. Second, delinquent behaviors are positively correlated, forming a problem behavior syndrome. Smoking or drinking is the central delinquent behavior. Third, negative consequences of specific time-use behaviors are more likely to occur on weekends than on weekdays, and time-use behavior may function differently on weekdays versus weekends. Among them, going to coffeenets or game-centers serves as the bridge with the highest potential of triggering delinquency.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2023
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1403845
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1403845
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1111/cdev.13953
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 16
        StartPage: 1697
    Subjects:
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: Students
        Type: general
      – SubjectFull: Extracurricular Activities
        Type: general
      – SubjectFull: Time on Task
        Type: general
      – SubjectFull: Delinquency
        Type: general
      – SubjectFull: Delinquency Prevention
        Type: general
      – SubjectFull: Student Behavior
        Type: general
      – SubjectFull: Behavior Problems
        Type: general
      – SubjectFull: Smoking
        Type: general
      – SubjectFull: Drinking
        Type: general
      – SubjectFull: Negative Reinforcement
        Type: general
      – SubjectFull: School Schedules
        Type: general
      – SubjectFull: Weekend Programs
        Type: general
      – SubjectFull: Risk
        Type: general
      – SubjectFull: At Risk Students
        Type: general
      – SubjectFull: China
        Type: general
    Titles:
      – TitleFull: Network Structure of the Links between Extracurricular Time-Use and Delinquent Behaviors: Moving Forward and Beyond Linear Relations
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Xia, Yiwei
      – PersonEntity:
          Name:
            NameFull: Ma, Zhihao
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2023
          Identifiers:
            – Type: issn-print
              Value: 0009-3920
            – Type: issn-electronic
              Value: 1467-8624
          Numbering:
            – Type: volume
              Value: 94
            – Type: issue
              Value: 6
          Titles:
            – TitleFull: Child Development
              Type: main
ResultId 1