Online Aggressive Behavior, Self-Harm Behavior, and Social Anxiety: The Mediating Effect of Social Network Sites Addictive Tendency and the Moderating Effect of Sex
Article information
Abstract
Objective
Self-harm and online aggressive behavior, recognized as dual-harm behaviors, have become increasingly prevalent among college students, which is strongly related to social anxiety. Yet, the underlying mechanism remains unclear and social network sites (SNS) addictive tendency may have a mediating effect. In addition, the influence of sex requires further clarification.
Methods
A convenient sampling method was employed and 1,608 college students (females=1,245, 77.43%; Meanage=18.95 years) were recruited. Scales measuring social anxiety, online aggressive behavior, self-harm behavior, and SNS addictive tendency were utilized. Structural equation modelling based on Mplus was conducted to testify the mediating effect of SNS addictive tendency and Wald’s χ2 test was employed to clarify sex difference.
Results
Social anxiety is significantly and positively related to online aggressive behavior, self-harm behavior, and SNS addictive tendency. The mediating effect of SNS addictive tendency is tested to be significant (online aggressive behavior: β=0.13, 95% confidence intervals [CIs], 0.077–0.191; self-harm behavior: β=0.05, 95% CIs, 0.000–0.104) and sex could moderate the effect of social anxiety on SNS addictive tendency. The relation between social anxiety and SNS addictive tendency is significantly stronger among males.
Conclusion
The current study suggests that social anxiety could cause a higher risk of both online aggressive and self-harm behavior through SNS addictive tendency, especially among male college students. Therefore, further interventions should target assisting college students to develop interpersonal relations in the real world could be beneficial.
INTRODUCTION
Self-harm and online aggression are serious psychological health issues that have become increasingly prevalent in contemporary society, particularly among college students, with a prevalence of 16.6% [1] and 57% [2]. While the harmful consequences of self-harm and online aggression are well-documented [3], the underlying mechanisms remain unknown, and effective in tervention are lacking. Existing studies suggest that social anxiety (SA) may be a potential risk factor for both self-harm and online aggression [4,5]. Moreover, individuals exposed to repeated setbacks in real-life interpersonal interaction may subsequently turn to online socializing as a way to escape or replace offline interactions. This, in turn, can intensify their tendency toward social network sites (SNS) addiction [6]. Similar to other addictive behaviors, social media addiction is associated with severe negative psychological and social consequences [7].
In sum, although previous studies have explored the relationship between SA, social media addiction, and mental health issues, the mediating role of SNS addiction in the relationship between SA and self-harm/online aggression remains an area that warrants further exploration.
Self-harm
Self-harm, also known as non-suicidal self-injury (NSSI), refers to the intentional act of harming oneself without suicidal intentions, including cutting, punching, burning, scratching bleeding areas, and interfering with wound healing [8]. It has been proven that NSSI is particularly prevalent during adolescence and peaking in early and late adolescence, with college students mainly in this period [9]. Additionally, a meta-analysis conducted in China found that the prevalence of self-harm behaviors among Chinese undergraduates was as high as 16.6% [1]. Evidence has shown that self-harm behaviors have emerged as a major global public health issue that significantly impact individual psychological development [10]. These behaviors are strongly associated with a range of mental problems (e.g., depression, anxiety, and posttraumatic stress disorder) [9] and exhibit greater predictability in relation to suicide compared to other risk factors [11]. Therefore, exploring the psychological mechanisms as well as the risk factors underlying self-harm behaviors among college students is essential for informing the development of targeted intervention and prevention strategies.
Online aggression
Similar to NSSI, aggressive behavior is also considered as one of the common maladaptive behaviors among college students [12]. Empirical evidence has shown that 27.7% of college students exhibited moderate or even higher levels of aggression tendencies [13]. However, with the development and popularization of the internet, interpersonal interactions have increasingly shifted from face-to-face to online. Consequently, aggressive behavior has also taken on new derivative forms, namely online aggression [14].
Online aggression is defined as aggressive behavior executed via a broad range of information and communication technologies, such as social media platforms, email, chat apps, or other messaging tools [15]. Different to traditional aggressive behaviors, aggression occurring in the virtual cyber environment possesses distinct manifestations and characteristics, such as the absence of time and space limitations, the potential for repeated harm, and anonymity [16]. Consequently, online aggressive behaviors have a broader influence and a higher prevalence. Existing evidence indicated that nearly 57% of college students reported initiating online aggression [2]. Moreover, online aggression can seriously influence youth development. A large number of studies have demonstrated that online aggression may result in serious negative impacts on victims, including depression, anxiety, lower academic performance, poor self-esteem, and even suicide [17], as well as the perpetrators [3,18]. Consequently, it is necessary to reveal the antecedents of online aggression, effectively preventing and intervening this behavior.
As for the relationship between self-harm and online aggression, although self-harm and aggression may initially seem distinct, several researches have suggested that self-harm and aggression often co-occur [19]. Recently, the concurrent presence of self-harm behaviors and aggressive behaviors within the same individual has been termed “dual harm.” [20] Individuals manifesting dual-harm behaviors are highly likely to be a high-risk cohort characterized by distinctive attributes and atypical patterns of harmful behaviors [21]. Especially, a hospital-based retrospective study has confirmed that adolescents worrying about maintaining their social connections and reputations are more prone to engage in dual harm [22], indicating the possible predicting effect of SA. However, previous research mainly focused on either self-harm or aggression, neglecting the combination of these two harmful behaviors. Furthermore, very few studies testify to the possible role of SA or clarify the mechanism. To fill the gap, the current study aimed to investigate the underlying mechanisms and risk factors of dual harm behaviors, as well as effective interventions to address this complex issue.
SA, self-harm, and online aggression
SA refers to the negative emotional experiences, including anxiety, nervousness, and shyness, that individuals encounter in social interactions [23]. These emotions often result in significant distress and functional impairments, adversely affecting social activities and interpersonal relationships [24]. On the one hand, according to interpersonal theory [25], thwarted belongingness and perceived burdensomeness are key factors that increase the likelihood of self-harm behaviors. When individuals experiencing SA, they may overly concern about others’ evaluations, increasingly the susceptibility and fear for social interactions [25]. Unable to effectively express and regulate these negative emotions, individuals may turn to maladaptive coping strategies, such as engaging in self-harm behaviors [26]. Several studies have also demonstrated the close relation between SA and self-harm behaviors [24,27]. For instance, in the investigation of the predictors of self-harm behaviors among college students, SA has been identified as the sole significant predictor. 28 Another study involving 2,992 undergraduates also supports this perspective, revealing the significant predictive role of SA in self-injury behaviors [24].
On the other hand, SA is also one of the factors that can be associated with online aggressive behaviors. The general aggression model (GAM) posits that an individual’s behavior is shaped by the combined influence of personal and environmental factors. Inputs from these factors affect cognition, emotion, and arousal, which are then integrated into the processes of assessment and decision-making [29]. Within this framework, SA functions as a dispositional factor that heightens individuals’ sensitivity to perceived interpersonal threats or negative evaluations in online contexts [30]. This heightened internal reactivity may impair emotional regulation, thereby elevating the risk that impulsive reactions, such as posting hostile comments or engaging in retaliatory behavior, escalate into acts of online aggression [31]. Additionally, SA has been empirically shown to significantly predict online aggressive behaviors [32,33]. For instance, in one study, individuals with SA were found to respond aggressively online [32]. Additionally, longitudinal studies have provided further support for this relationship, revealing that higher levels of SA at baseline predict increased online aggression over time [33].
Therefore, we make the Hypothesis 1 and 2 as below:
Hypothesis 1: Social anxiety could be positively associated with online aggressive behaviors.
Hypothesis 2: Social anxiety could be positively associated with self-harm behaviors.
SNS addictive tendency as a mediator
In the age of information, social networks have profoundly transformed the way humans communicate, offering opportunities for instant interaction and extensive connectivity. However, the excessive use of social networks is gradually evolving into a widespread mental health issue, commonly referred to as SNS addiction [34]. As the prodromal stage of internet addiction, internet addictive tendency includes the core symptoms of internet addiction including withdrawal, salience, relapse [35] and its associated psychological and behavioral consequences such as insomnia, loneliness, depressive affect [36]. While SNS addiction is not officially classified as a psychological disorder in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) [37], its symptoms and negative consequences closely resemble those of other behavioral addictions, such as gambling and video game addiction [38]. The development of social network addiction is influenced by a variety of factors. According to the cognitive-behavioral addiction theory model (CBM) [39], these factors can be categorized into distal and proximal influences. Distal influences, such as environmental stressors and psychopathological vulnerabilities, play indirect role in the occurrence of SNS addictive tendency. In contrast, proximal factors (e.g., maladaptive cognitive processes), including dysfunctional beliefs about the self and interpersonal relationships, may directly promote compulsive use of SNS. Individuals with SA often demonstrate cognitive impairments, such as biased attention to negative social cues and persistent negative self-evaluation [40,41]. These dysfunctional cognitions may intensify avoidance of face-to-face interactions and reinforce maladaptive online behaviors that serve as compensatory strategies. Consequently, SA may lead individuals to rely more heavily on social media platforms as perceived safer and more controllable social environments, thereby increasing the risk of developing SNS addiction [7,42,43].
Moreover, emerging evidence has identified SNS addictive tendency as a significant predictor of self-harm and online aggressive behaviors. According to the Problematic Psychosocial Predisposition Model, uncontrollable internet use, especially excessive engagement in online social interaction, can undermine individuals’ socially adaptive functioning, leading to a range of internalizing and externalizing problems, including self-injury and aggressive behaviors [44]. Specifically, continuous access to social networks may heighten exposure to bullying, unfavorable social comparisons, and persistent criticism. These factors are closely linked to elevated levels of anxiety, depression, low self-esteem, and maladaptive emotional regulation strategies [45]. When left unresolved, these negative emotional states may drive individuals to engage in self-injurious behaviors as a means of affect regulation, or to display aggression online as a maladaptive strategy for expressing frustration or reclaiming a sense of social power [46-48].
Given the above, it is reasonable to believe that SA may influence individuals’ self-harm and online aggressive behaviors by enhancing their tendency to engage in SNS addiction. Therefore, within the integrated of theoretical frameworks and empirical evidence, we propose the Hypothesis 3, 4, and 5 as below:
Hypothesis 3: Social anxiety could be positively associated with SNS addictive tendency.
Hypothesis 4: SNS addictive tendency is positively related to self-harm behaviors and could mediate the relation between social anxiety and self-harm behaviors.
Hypothesis 5: SNS addictive tendency is positively related to online aggressive behaviors and could mediate the relation between social anxiety and online aggressive behaviors.
Sex as a moderator
Despite the theoretically established relationship between SA and SNS addictive tendency, not all individuals with high levels of SA will exhibit excessive social network use [49]. Thus, it is essential to explore the factors that may weaken the strength of the relation between SA and SNS addictive tendency. Social role theory (SRT) suggests that individuals’ internalized sex role can significantly influence their responses to social and psychological stress [50]. Sex roles refer to the societal and cultural expectations for male and female behavior, which in turn affect how individuals cope with stressors. Influenced by these social norms, males are generally expected to exhibit independence, decisiveness, and the ability to handle stress without showing vulnerability, whereas females are more likely to depend on interpersonal support and exhibit more emotional reactions [51].
However, this societal expectation for men to suppress emotional vulnerability may discourage them from seeking face-to-face emotional support, thereby prompting them to pursue more controllable and less confrontational alternatives, such as online platforms. As a result, men experiencing high levels of SA may be more likely to engage in SNS use as a compensatory strategy to manage stress and maintain a socially desirable self-image. Supporting this perspective, prior research has demonstrated that socially anxious men are more likely to use SNS for initiating new connections, whereas women tend to use SNS to maintain existing relationships [52]. Further reinforcing this view, a longitudinal research conducted in 2,103 Korean college students showed that men with SA tend to spend more time engaging in online social interactions in order to maintain their self-image [53].
Therefore, based on SRT and empirical evidence, we formulate the Hypothesis 6 that:
Hypothesis 6: The relation between SA and SNS addictive tendency is stronger among males than females.
The present study
In summary, based on the interpersonal theory, GAM, CBM, SRT, and empirical research, the current study aims to investigate the mediating role of SNS addictive tendency and the moderating role of sex in the relation between SA and self-harm behaviors, as well as online aggression. Specifically, the present study proposes the following hypothesized model in Figure 1.
The hypothesized model. SNS, social network sites.
METHODS
Participants and procedures
The data collection process happened in Jiangsu province of China in October 2023 and we used the convenient sampling to recruit college students from four universities. The data collection was based on an online survey platform named “Wenjuanxing” (www.wenjuanxing.com). Before the survey began, participants were presented with informed consent in which the aim and content of the online survey were introduced clearly. All participants were informed that they could freely choose to participate in the survey or not and had the right to quit at any moment. In addition, we used G*power (version 3.1.9.7; Heinrich Heine University) to estimate the needed sample size for verifying the hypothesized model and it is suggested that at least 107 samples are needed with the effect size setting to 0.15.
Among 3,157 college students who received the online survey link, 3,005 of them consented to participate and completed the survey. Noticeably, to prevent any possible negative impacts caused by the survey, participants who reported being at the onset stage for severe mental disorders were not included and provided with the resources for mental health services. Based on participants’ responses to two attention test questions, we identified 1,435 participants as careless responders for not passing the attention test and removed them from the data set. A total of 1,608 participants (females=1,245, 77.43%; Meanage=18.95 years) were included in further data analysis.
All participants and their parents were fully informed about the objectives of the study and provided electronic informed consent. This study was approved by the ethics Committee of Brain Hospital of Nanjing Medical University (reference number: 202305290090).
Measurements
Social anxiety scale for adolescents
The Chinese version of Social anxiety scale for adolescents (SAS-A) was translated and revised by Zhou et al. [54] and its credibility and validity were verified by multiple studies [55,56]. The SAS-A consists of 18 items scored on a 5-point Likert scale (1 “never” to 5 “always”), with higher score indicating a higher level of SA. The 18 items can be divided into three subscales, namely the fear of negative evaluation, social avoidance, and distress in the new situation/with strangers, and social avoidance and distress generally/with acquaintance. In the present study, the SAS-A showed good internal consistency with the Cronbach’s α values of 0.96 for the whole scale, 0.94 for the fear of negative evaluation sub-scale, 0.89 for the social avoidance and distress in the new situation/with strangers subscale, and 0.84 for the social avoidance and distress generally/with acquaintance subscale.
Short version of SNS addictive tendences scale
The short version of SNS addictive tendences scale was developed by Milošević-Đorđević and Žeželj57 and its Chinese version has been verified to demonstrate good reliability and validity [58]. The short version of SNS addictive tendences scale contains 6 items which are scored on a 5-point Likert scale (1 “strongly disagree” to 5 “strongly agree”). A higher score on the scale may indicate a higher level of social network addiction. In the current study, the short version of SNS addictive tendences scale demonstrated good internal consistency with the Cronbach’s α reaching 0.87.
Adolescent online aggressive behavior scale
The Adolescent online aggressive behavior scale (AOABS) was developed and verified by Zhao and Gao [59]. The AOABS consists of 31 items scored on a 4-point scale describing the frequency of the behavior (1 “never” to 4 “always”). The AOABS contains two subscales, namely instrumental aggression and reactive aggression, and each subscale was divided by two factors of overt aggression and relational aggression. Based on the present dataset, the AOABS showed good internal consistency with the Cronbach’s α for the whole scale reaching 0.86, 0.81 for the instrumental aggression subscale, and 0.91 for the reactive aggression subscale.
Self-harm, suicidal ideation, and suicide attempts
Self-harm was assessed using one item from the Juvenile Campus Violence Questionnaire [60]. The item was scored on a 4-point scale (1 “never” to 4 “almost”) and a higher score may indicate higher frequency of having self-harm behaviors.
Besides the aforementioned questionnaires, the demographic information such as sex and age is also measured in the survey.
Data analysis
The data analysis was conducted using R 4.3.1 [61] and Mplus 8.0 (Muthén & Muthén). All variables were standardized before further analysis. The reliability and validity of the measuring instruments were assessed using the reliability analysis and factor analysis. Meanwhile, the Haman’s single-factor test was used in the study to test the existence of the common method bias [62]. The descriptive statistics and Pearson correlations between variables were conducted to examine the relation among online aggressive behaviours, social network addiction, SA, self-harm behaviors, sex, and age.
Then, we examined the mediating effect of social network addiction on the relation between SA-online aggressive behaviors and SA-self-harm behaviors. The indirect effects were tested using the bootstrap approach and were considered significant if the 95% confidence intervals (CIs) did not cover 0. After that, we explore the moderating effect of sex on the mediating process by conducting a multi-group analysis which could testify if there exist significant differences in the coefficients in the model across sexes. In detail, we compare the unconstrained model (allowing freely estimating the coefficients in the model of two sexes) with the constrained model (setting the coefficients in the model of two sexes to be equal). Finally, we used the Wald χ2 test to compare the specific path coefficients between sexes. In the model, adolescents’ age was controlled.
RESULTS
Reliability and validity of the research instruments and the common method bias test
Before we conduct further analysis, we tested the reliability and validity of the scales used in the current study. All instruments employed in the current study showed good reliability and validity with Cronbach’s α >0.80 and Kaiser-Meyer-Olkin values >0.85. Therefore, the data collected by these measuring instruments could be included in further study.
In addition, the result of Haman’s single-factor test shows that the eigenvalues of 10 factors were higher than 1 in unrotated factor analysis, and the first factor accounted for 26.05% of the total variance, which was less than the critical standard of 40% [62].
Descriptive statistics, correlation, and differences on sexes
The mean values, standard deviation, and correlation between variables are shown in Table 1.
Means, standard deviations, and correlations with confidence intervals
As demonstrated in Table 1, all research variables are significantly correlated with each other, which suggests that further analysis of the mediation effect is available. Meanwhile, it is also shown in Table 1 that sex is significantly related to both SA and online aggressive behaviours. Indeed, we conducted a t-test to compare levels of SA, SNS addictive tendences, online aggressive behaviours, and self-harm behaviours among two sex groups. The results are shown in Table 2.
Comparison between two groups
As depicted in Table 2, the level of SA among females is significantly higher than that among males (t=-3.33, p<0.001), while males show more online aggressive behaviours than females (t=3.38, p<0.001).
The mediating effect of social network addiction
The mediation model in which SNS addictive tendence mediates the relation between SA and online aggressive behaviour as well as the relation between SA and self-harm behaviours was conducted. The model fit indices suggest that the model fit well on the current data, with Root Mean Square Error of Approximation equal to 0.049, comparative fit index equal to 0.994 and Tucker-Lewis index (TLI) equal to 0.965. The results of regression indicate that SA could significantly predict online aggressive behaviours (β=0.075, p<0.01), self-harm behaviours (β=0.112, p<0.01), and SNS addictive tendence (β=0.710, p<0.001). Meanwhile, SNS addictive tendence can also significantly predict both online aggressive behaviours (β=0.183, p<0.001) and self-harm behaviours (β=0.073, p<0.001).
Moreover, the indirect effect from SA to online aggressive behaviours was equal to 0.130 (p<0.001) and the indirect effect from SA to self-harm behaviours was equal to 0.052 (p<0.05). Apart from that, the results of bootstrap (times=1,000) also indicate significant indirect between SA and online aggressive behaviours (95% CIs: 0.077–0.191) as well as SA and self-harm behaviours (95% CIs: 0.000–0.104). The results of the mediation analysis are shown in Table 3 and Figure 2 to better illustrate the mediation model.
Regression and mediation analysis, comparing two models with the reverse direction
Results of the mediation model. **indicates p<0.01; ***indicates p<0.001. SNS, social network sites.
The moderating role of sex
Considering that there exist significant differences in the levels of SA and online aggressive behaviours, we further analyzed whether sex plays a moderated role in the mediation model. Hence, the Wald χ2 test was employed in the paths of the mediation model to explore the difference between sexes. Specifically, the difference in model fit index was significant (ΔTLI=0.021), indicating a significant difference between the unconstrained and the constrained model [63]. In other words, there is significant difference between the group models of females and males.
The results of the Wald χ2 test are shown in Table 4. As shown in Table 4, the predicted effect of SA on SNS addictive tendency differed significantly between sexes (χ2=4.585, p<0.05). In other words, sex moderates the relation between SA and SNS addictive tendency. The mediation models of both sexes are depicted in Figure 3, respectively.
Results of Wald chi-square test and comparison between sexes
The mediation model of males and females, respectively. A: Model of males. B: Model of females. Dashed lines indicate insignificant relations between variables. *indicates p<0.05; **indicates p<0.01; ***indicates p<0.001. SNS, social network sites.
DISCUSSION
The current study aims at clarifying the predicting role of SA to self-harm behaviours and online aggressive behaviours, as well as identifying the mediating effect of SNS addictive tendence and the moderating effect of sex. Some findings are worth to be further discussed.
As the results of the descriptive statistics stated, females scored higher on SA scale compared to males, which may indicate that the prevalence and severity of SA could be higher in females than males. This finding aligned with a large number of previous studies as well as DSM-5, stating that SA is more prevalent in females than it is in males, especially during adolescence [64-66]. However, it needs to be mentioned that previous studies with similar findings either were based on the Western population or were conducted years ago [67,68]. Hence, findings in the current study may further supplement the findings of sex differences in the prevalence of SA based on the Chinese population. The higher scores on SA among females could be associated with the heightened emotionality and social sensitivity that is more common in females [69]. As a systematic review suggested, females may experience an earlier puberty compared to males, which could add more risk to the onset of social-emotion disorders [69].
Meanwhile, the descriptive statistics also stated that males tend to have significantly more online aggressive behaviours than females. This finding is consistent with previous studies, which demonstrated that femininity could buffer both hostility and aggressive behaviours [70-72]. This diversity might be related to the different levels of impulsivity among males and females. The connection between impulsivity and aggression has been widely documented and males, in general, could be more impulsive compared to females [73]. Moreover, this difference may also be related to the distinct preferences of online game between male and female. As stated by previous research, males tend to spend significantly a larger amount of time in video game, especially involving in violent video games [74]. Furthermore, the involvement in violent games could positively predict the increase of online aggressive behaviours [75].
In addition, the results of mediation analysis indicate that SNS addictive tendence could mediate the predictive effect of SA on online aggressive behaviours and the predictive effect of SA on self-harm behaviours. Individuals with a higher risk for SA may struggle in the interpersonal interaction in their real life, showing more social avoidance behaviours, which may hinder the satisfaction of their need for relatedness [76]. According to self-determination theory, the need for relatedness is one of the three fundamental mental needs in human being and if remaining unsatisfied, people may strive to make compensation for the unsatisfied need [77]. Hence, people who cannot meet their need for relatedness in real life may turn to social networking sites in search of the compensation, which is in line with the social compensation theory [78]. However, despite the virtual relationship built on the social network could alleviate the loneliness temporarily, it may lead to increasing dependency on social media and may further hurt individuals’ social function as well as the mental health state [79]. On the one hand, more time spent on social media indicates a higher risk of being exposed to passive discussion or hostile arguments, which could directly contribute to more risk of being under attack and having aggressive behaviours. Similarly, Alvarez de Mon et al. [80] found that more screen exposure could lead to higher risk of both conducting and suffering from online aggressive behaviours. More relatively, Lin et al. [81] also identified that the addictive use of social media could predict the increase of online aggressive behaviours.
On the other hand, the possibility of viewing more self-harm content may also rise with more exposure to social media [82]. According to a previous review, being exposed to self-harm content, such as videos or graphs of self-harm may increase the potential of conducting self-harm behaviours by normalizing self-harm [83]. More strikingly, the contagion or copycat effects of self-harm have been documented by abundant studies [84]. In other words, the self-harm behaviour may spread via social media. Indeed, the mediating role of SNS addictive tendency on the predictive effect of SA on self-harm is reasonable. Considering very few studies have identified SA as a risk factor for self-harm behaviour or discussed the mechanism behind it, the current study may help clarify the link between SA to self-harm behaviour.
Additionally, the present research also found that the relation between SA to SNS addictive tendence is significantly tighter between males than females. Partly aligned with the current study, Boursier et al. [85] also suggested that the relation between social appearance anxiety and problematic social media use is stronger among males, especially when males tend to believe that their self-profile could help improve their self-confidence. Similarly, Casale and Fioravanti [86] also observed the unique moderating effect of self-presentation on the relation between SA and problematic internet use among males. Therefore, the stronger relation between SA and SNS addictive tendency among males in the current study could similarly be explained by that if males failed to build self-confidence in the real world, they might turn to the virtual world in pursuit of compensation. Considering previous studies drew similar conclusions based on the Western cultural context, the current study may provide a supplement to this finding from the Chinese cultural context. However, since research comparing the relation between SA and SNS addictive tendency are very limited, future study is needed to verify this hypothesis.
To summarize, the results of the present study fully support the Hypothesis 1–5, suggesting that SNS addictive tendence could mediate the relation between SA and self-harm behaviours as well as SA and online aggressive behaviours. Furthermore, the result partly proved the Hypothesis 6, indicating that there exists significant difference in the relation between SA and SNS addictive tendence across sexes. Indeed, the study heightened the need for dealing with the problematic use of social media to decrease the self-harm behaviour and online aggressive behaviours. More specifically, considering the addiction in social media may come from the unsatisfied relatedness need, intervention that guiding socially anxious individual to build and maintain beneficial interpersonal relationship in real life is in urgent need.
Application implication
Results in the current study declare that SNS addictive tendency mediates the relation between SA and self-harm solely among females. This may indicate that more attention is warranted among females facing a higher risk of SA and showing SNS addictive tendency to prevent the possible self-harm behaviours. Meanwhile, the current study may suggest intervention targeting at equipping individuals with more adaptive social interacting skills could be beneficial in preventing self-harm behaviours.
Limitation
Several limitations in the current study need to be mentioned. First and foremost, the number of males and females in the current study is highly unbalanced. Therefore, the results of the current study need to be explained with caution. Moreover, the definition of problematic social media use or SNS addictive tendency did not reach an agreement. Hence, a more standardized and universal definition of the problematic use of social media is needed. Last but not least, when assessing SNS addictive tendency, more objective indices, such as the time spent on social media could be further included.
Conclusion
The current study used the structural equation model to testify to the mediating effect of SNS addictive tendency on the relation between SA and self-harm behaviours, online aggressive behaviours. By adopting the Wald χ2 test, the study verified the moderating effect of sexes on the mediating effect of SNS addictive tendency. The results demonstrated that SNS addictive tendency plays a significant mediating role in the relation between SA and self-harm behaviours, online aggressive behaviours and sex could moderate the relation between SA and SNS addictive tendency. Our findings suggested that intervention targeted at improving socially anxious individuals’ ability to develop and maintain the interpersonal relationship in real life could help decrease the risk for SNS addiction as well as buffer self-harm behaviours and online aggressive behaviours.
Notes
Availability of Data and Material
The datasets generated or analyzed during the study are not publicly available due to ethical consideration but are available from the corresponding author on reasonable request.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: Huachen Ding, Yaqiu Xu. Data curation: Yaqiu Xu. Formal analysis: Huachen Ding. Investigation: Yaqiu Xu. Methodology: Yaqiu Xu. Supervision: Huachen Ding, Yaqiu Xu. Validation: Huachen Ding. Visualization: Huachen Ding. Writing—original draft: Huachen Ding, Yaqiu Xu. Writing—review & editing: Huachen Ding, Yaqiu Xu.
Funding Statement
This research was supported by the following institutions: High-level hospitals to build the second level talent fund of Nanjing Brain Hospital.
Acknowledgments
None