Differences between problematic internet and smartphone use and their psychological risk factors in boys and girls: a network analysis

Over the past two decades, increasing access to digital technology has transformed the lives of young people worldwide. Modern adolescents work with the internet and other digital technologies on an intensive daily basis [1]. The expansion and constant accessibility of internet and other technologies have created great opportunities for learning, work, entertainment, and personal exploration and growth [2]. However, intense online technology use may lead to various social and health risks, including reduced sleep quality [3], obesity [4], and reduced academic performance [5]. Concerns regarding excessive and potentially addictive use have been repeatedly expressed [6, 7]. Of the many different forms of excessive technology use, considerable attention has been paid to problematic internet use (PIU) and problematic smartphone use (PSU). Much research has already been devoted to these two phenomena in adolescence [8, 9], but the relationship between them (especially with regard to adolescents) has been studied to a limited extent. Our study fills this gap by comparing the similarities and differences between PIU and PSU in terms of their risk and protective factors (i.e., psychological variables that might affect susceptibility to PIU or PSU). Our focus is specifically on adolescent users, since they are often seen as a particularly vulnerable group in terms of the development of problematic forms of internet and smartphone use [1, 10].

Both PIU and PSU are operationalized as the inability to control one’s use of the medium, which leads to harmful consequences and disruptions in daily functioning [11, 12]. Even when facing these negative consequences, users have a diminished capacity to limit their time spent in the medium, and they are preoccupied with it, even when not online [13]. The term PIU covers a large number of excessive online activities, especially online gaming, social networking sites use (SNS), chatting, video watching, or online shopping [12]. Smartphones are internet-enabled devices providing instant and nearly unlimited access to online activities. In principle, when using a smartphone, the user is almost always simultaneously connected to the internet. Thus, in terms of user patterns, it can be assumed that the two phenomena overlap to a certain extent. On the other hand, the application use may be slightly different, because some social media use (e.g., Instagram, WhatsApp) is optimized for smartphones. The existing literature reports weak to strong correlations between PSU and PIU (e.g., r = 0.21 in Choi et al. [14]; r = 0.40 in Kwon et al. [15]; r = 0.50 in Lachmann et al. [16]; r = 0.64 in Škařupová et al. [17]). Although positive associations have been demonstrated, several studies have pointed out differences in usage patterns, gender, personality traits, and psychological variables between these two types of problematic behavior. While PSU has been found to be especially related (and almost identical) to social media use [18], extreme PIU scores were found to be related to online gaming [19]. In other words, both phenomena rely on somewhat different need satisfaction and anticipated rewards [20]. Specifically, in terms of internet usage, boys were reported to be more prone to addictive use than girls, whereas this pattern was reversed for smartphone use [14, 21]. Furthermore, lower extraversion was associated with higher PIU but was unrelated to PSU, whereas lower openness to experience was linked to higher PSU but not to PIU [22]. These results suggest that while PIU and PSU are related, a significant portion of unexplained variance remains that represents the differences between the constructs.

Both PIU and PSU are often studied in the context of users’ psychological characteristics and susceptibility to developing problematic forms of use. Previous meta-analyses [23‐25] have suggested that the most consistent risk factors stem from the following areas: (1) high impulsivity and attention/hyperactivity disorders; (2) negative emotionality, anxiety, and depressive symptoms; and (3) low self-esteem and self-directedness. On the other hand, resilience and high self-control are often reported as the most important protective factors in terms of PIU and PSU [26, 27].

Impulsivity is frequently linked to addictive behaviors [28]. Internet users with higher impulsivity present executive dysfunction and deficient inhibitory control, which may contribute to problems with online technology use [29]. Together with disrupted self-control, attention problems, aggression, hyperactivity, and impulsivity are part of the construct of externalizing problems [30]. According to previous research, externalizing problems are relatively common in adolescent problematic media users and they were found to be significantly associated with PIU, specifically excessive social media use or internet gaming addiction [31‐34].

In addition to externalizing problems, adolescents may also develop internalizing problems that include affective states, such as anxiety, social withdrawal, depression [35], diminished self-esteem, and feelings of hopelessness—all of which have been identified as risk factors in terms of developing both PIU and PSU [36‐39]. It has been suggested that individuals with negative emotionality may tend to use smartphones or the internet excessively as a coping mechanism to eliminate distress [40, 41].

Internalizing problems in adolescence often go hand-in-hand with social anxiety, withdrawal from peer relationships, lack of social competence, and shyness [42, 43]. Thus, these individuals experience problems with social functioning and social inclusion [44]. Adolescents who need to belong and whose social connections are unsatisfied in real life might tend to fulfill these needs through SNS [45]. Przybylski et al. proposed that fear of missing out (FoMO) may explain these dynamics. Fear of missing out is defined as a pervasive apprehension that others might have more rewarding experiences or acquire useful information that one does not have access to [46]. To not miss something important on the site, people with increased FoMO feel the need to be as often online as possible. They often experience anxiety offline and feel pressured to constantly check for new information [47]. Previous research has demonstrated robust associations between FoMO and problematic social networks use [48], PIU [49], and PSU [50]. A potential explanation to these findings may lie in unmet social needs: it has been demonstrated that adolescents with higher FoMO also tend to have a higher need for popularity/belonging as well as higher social media use intensity [45].

Resilience is one of the strongest protective factors for PIU or PSU. Resilience is a multidimensional construct defined as the ability to adapt positively to life conditions and thrive, even in the face of adversity [51]. A resilient individual can use constructive coping strategies to successfully deal with adverse life events [52]. Resilience has been repeatedly suggested as a protective factor against various psychopathologies and risky behaviors, including internet, smartphone, and social media addiction [53‐56]. Young people face many stressful challenges due to biological, psychological, and social changes. Adolescents with higher resilience have better internal resources to cope with stressful events, which might lead them to become less involved in using the internet to regulate negative emotions [57].

Although a relatively large body of work has examined PIU and PSU use separately, examinations of both are relatively scarce. Based on the literature review, we assume that they are related (e.g., share some of the predictors), but distinct phenomena. Owing to the different usage patterns of boys and girls, we further assumed that the differences would be reflected at the gender level as well. Thus, the aim of this study was to examine what psychological risk and protective factors are shared by PIU and PSU and what factors define the dividing line. With the help of a network analysis, we aim to examine the association separately for boys and girls.

The current study aimed to investigate the associations between adolescents’ PIU, PSU, and related protective and risk factors. Specifically, we examined the extent to which PIU and PSU share the same risk and protective factors, and how these variables are interrelated in boys and girls. The findings showed that PIU and PSU were positively correlated; however, this association was weak in boys and moderate in girls. PIU and PSU showed a roughly similar structure of relationships with other variables—they were both positively associated with psychological risk factors. However, the correlations between PIU and the other variables were significantly stronger than those between the same variables and PSU, with FoMO as the only exception. The relationships studied also differed between boys and girls.

There can be several reasons why the associations between PIU and psychological variables were stronger than those between PSU and psychological factors. One potential reason is that children may have more access to internet-based activities via devices other than smartphones (e.g., PC, tablets, etc.). This may result in more uninterrupted time spent online, as it may be plausible that smartphones prompt interruptive notifications more frequently than, say, tablets. Owing to fewer interruptions, children may have extended their screen time with an activity. It should also be noted that when it comes to assessing internet use, the line between the use of online functionalities of a smartphone may be implicitly included in the evaluation, as PIU may be an umbrella concept covering other online-based problematic behaviors [77, 78]. Another potential explanation could be that internet use may lead to a sense of anonymity when not performed on a smartphone. In other words, one could hypothesize that a smartphone may be associated with reduced online disinhibition [79], because communication with disclosed contacts may create a feeling of lower anonymity. Online anonymity, in turn, may promote lurking behavior—socially passive internet consumption—which has been shown to be associated with reduced mental health and problematic social media use [80, 81].

Based on these results, we cannot claim that the PIU and PSU exhibit the same phenomenon. Their mutual correlation was relatively low and they shared approximately 8.8% of the variance in the total sample. In contrast, our study showed that PIU and PSU had very similar relationships with the psychological variables. They both showed a positive association with fear of missing out, hopelessness, externalizing problems, and internalizing problems, and a negative association with resilience. These variables clearly contributed to the shared variance between PIU and PSU. This is consistent with previous studies that showed that fear of missing out, externalizing problems (i.e., impulsivity, hyperactivity, aggression), and internalizing problems (i.e., various emotional difficulties) could be risk factors for various forms of problematic online behavior [34, 36, 49], whereas resilience is a protective factor in these cases [55, 56]. However, it must be noted that the associations between selected variables and PIU were stronger than their associations with PSU, which is in line with the study by Jeong et al. [82], who found that the risk factors for PIU were different from those of PSU and non-addicted groups. These results raise questions regarding the extent to which PSU is an independent pathological phenomenon. The only variable, whose association with PSU was stronger than with PIU, was the fear of missing out. It was previously found that people who scored high on FoMO had a higher tendency to overuse their smartphones to satisfy their need for constant connectedness [83]. Owing to their portability, smartphones can provide 24/7 internet access allowing users to constantly check what is happening online. At the same time, this permanent connectedness heightens the awareness of possibly missing out on potentially more rewarding experiences, which could fuel FoMO even more [84].

Our findings also indicate interesting differences between girls and boys. With the exception of resilience, girls showed significantly higher values for all measured variables than boys. There are several possible reasons why the associations between PIU, PSU, and psychological variables are generally stronger in girls than boys. It should be noted that similar results have been demonstrated before; specifically, girls tend to spend more time online (and on digital devices) than boys, and problematic digital technology use has also been reported higher in girls than boys [85]. Given that girls place greater importance on social relationships and it also affects their mental well-being more than in boys [86] it is also natural that, for instance, social media usage patterns differ across genders [87]. It has been demonstrated that girls are more affected by online social comparisons [88], which could affect their body image [89]. This could, subsequently, also affect other aspects of mental health.

As for PSU, however, current research states that gender differences are not as evident, although some authors have reported that females are more susceptible to PSU [90, 91]. Recently, it has been found that girls use SNS and other social communication channels much more intensively, such as Facebook Messenger and WhatsApp [92, 93]. These applications are available predominantly on smartphones, which could partially explain why girls exhibit higher PSU values than boys do. In the case of PIU, the latest research predominantly states that boys have higher levels of PIU than girls [e.g., 94, 95], while others are in line with our findings [e.g., 96‐98]. For example, Ha and Hwang [96] found that girls with emotional difficulties had a higher risk of developing internet addiction than boys with the same conditions. As the girls in our sample reported higher levels of emotional problems than boys, it is possible that these factors may have been related to more PIU. Girls also scored significantly higher on all psychological variables, except resilience, for which no difference was found between boys and girls. This result is consistent with previous studies that repeatedly report a higher prevalence for mental health problems in girls compared to boys [99‐101]. During adolescence, girls are more susceptible to specific stressors associated with increased psychological distress and an increased likelihood of mental health problems, such as body dissatisfaction [102], low self-esteem [99], and academic stress and worries about school performance [103]. According to some authors [104, 105] boys may have more difficulties acknowledging and describing their mental health issues and, in comparison to girls, tend to mask or downplay their problems. This may be related to cultural expectations related to gender roles—in many societies, boys are discouraged from showing vulnerability or weakness and, thus, tend to complain less often about their health problems in general [106].

The results of the network analysis also showed that PIU and PSU were not the central nodes in any network. For boys, the node with the highest strength was externalizing problems, whereas for girls, resilience, externalizing problems, and internalizing problems were of comparable importance. In the present study, this could mean that, if the central nodes (e.g., resilience, externalizing problems, internalizing problems) are targeted, they can significantly change the levels of PIU and PSU; thus, they are ideal targets for prevention and treatment. The importance of internalizing and externalizing problems as risk factors for the development of PIU or PSU has been demonstrated previously. For example, internalizing problems, depression, anxiety, and peer-relationship difficulties predicted both PIU and PSU in previous studies [36, 37, 56, 107]. In the case of externalizing problems, impulsivity, aggression, and attention deficit hyperactivity symptoms have been previously identified as risk factors for PIU or PSU [56, 107‐110]. Some studies [101, 111] also suggest that boys have a higher tendency to externalize problems and girls internalize problems, which could explain why internalizing problems are the central node in girls’ networks but not in boys’ networks. At the same time, resilience has been reported to be one of the most important protective factors against the development of PIU or PSU [54, 55].

The results of our study can be used to develop intervention programs to prevent PIU and PSU. As mentioned above, targeting central nodes in a network may lead to improved mental health. As an example, externalizing symptoms were among the nodes with the highest strength in both boys and girls, meaning indicating that targeting these symptoms may lead to improved well-being in children. There are several examples of how externalizing symptoms can be addressed in children. For instance, parent training programs could be useful in teaching how to cope with children’s behavior by communicating clear expectations, providing consistent consequences for misbehavior, and reinforcing positive behaviors [112]. Social skills training may also have beneficial effects on externalizing symptoms [113]. Finally, school-level interventions (e.g., mental health literacy and stigma mitigation) could also be useful in reducing the severity of externalizing symptoms [114].

This study has several limitations. First, because the design was cross-sectional, causal relationships between variables could not be inferred based on the results. Second, the study design only entailed an interindividual perspective. Previous research on individual differences, however, demonstrated that the structure and associations of inter- and intraindividual differences might not necessarily be the same [e.g., 115]; thus, future studies should incorporate an intraindividual perspective next to the interindividual perspective [116]. Third, this study did not include other factors, such as family, in the present analyses. It has previously been demonstrated that children’s family circumstances, such as parental education [117], might be associated with children’s digital device use [118]. Future research should consider intraindividual differences in PIU and PSU within this context. Fourth, the study used self-reported data, which might be prone to response bias such as social desirability or acquiescence. Fifth, since the data came from a complex epidemiological study, it was not always possible to use full-length questionnaires; therefore, shortened versions of most scales were used instead. This could have negatively affected the reliability of the scales. At the same time, both of the PIU and PSU scales are designed to measure generalized internet and smartphone addiction; therefore, they cannot provide information about specific types of internet or smartphone usage behaviors (e.g., social networking, gaming, and online shopping). Despite these limitations, the key advantage of our study was its large and nationally representative sample of adolescents.

PIU and PSU are weakly to moderately related phenomena, yet they are distinct constructs that differ at the psychological level, with psychological risk factors mostly being especially relevant for PIU. Moreover, these phenomena were rather different between boys and girls, with stronger associations between PIU and PSU and psychological risk factors in girls.