Perceived social adversity and online trolling in college students: Depressive symptoms mediation and perceived social support moderation

Introduction

Online trolling refers to the behaviour of individuals deliberately posting and disseminating offensive, inflammatory and destructive information on the internet for the purpose of causing controversy and conflict (Buckels et al., 2014). In contrast to other forms of online aggression, online trolling targets large numbers of internet users who do not know each other, with the purpose of creating chaos for fun or venting negative emotions caused by social pressure (Howard et al., 2019). As a widespread form of online aggression, online trolling is gradually becoming dominant in the online environment, which not only affects the mental health of victims, including increased depressive symptoms, self-harm, and suicidal behaviour (Akhtar & Morrison, 2019; Kapse & Gupta, 2023; Wilson & Seigfried-Spellar, 2023) but also leads to profound negative consequences such as group polarization, the deterioration of the public communication environment, and decreased social trust (Simchon et al., 2022; Goyanes et al., 2021; Tufekci, 2017). As individuals enter adulthood from adolescence, self-presentation becomes especially salient among college students. Their frequent involvement in online activities may, in turn, expose them to a greater likelihood of online trolling. Prior evidence has shown that this behavior is relatively prevalent in this group, with one study finding that 71.2% of college students reported engaging in online trolling at least weekly (Hong & Cheng, 2018).

However, many studies have not fully utilized the person–environment interactions perspective (Herrenkohl, 2013). They often focus on a single, direct association between environmental factors or personal traits and aggressive behavior outcomes, overlooking that online trolling often results from the joint influence of environmental and individual factors (Nitschinsk et al., 2023; Volkmer et al., 2023). In view of the fact that online trolling may lead to serious negative effects, this study chose to focus on how individuals subjectively perceive their social environment, namely the perceived social adversity, to examine its relationship and influence mechanism with online trolling among college students, which has important implications for improving the network environment and maintaining the mental health of college students.

Perceived social adversity and online trolling

Perceived social adversity refers to the extent to which an individual perceives threat or stress in the social environment (Zhang et al., 2017). According to the I³ model, the perceived social adversity may serve as an Instigation for online trolling. When individuals experience long-term social adversity and psychological distress, they often seek a balance in terms of cognition, emotion, and behaviour through aggressive behaviour (Jin et al., 2020). Existing studies have indicated that individuals’ perceptions of long-term and repeated social adversity are associated with aggressive behaviour (Jin et al., 2020; Liu et al., 2021). Furthermore, Individuals who perceive greater social adversity show a significantly greater likelihood of engaging in online aggression (Chen et al., 2024). Some studies have indicated that college students with high perceived social adversity may use online aggression as a way to vent stress and negative emotions (Ybarra & Mitchell, 2004; Hinduja & Patchin, 2014; Huang et al., 2025). Higher perceived social adversity has also been associated with stronger negative emotional experiences and a greater tendency to seek psychological relief (Ding et al., 2018; Chen et al., 2024). Motivated by anonymity and avoidance of accountability, they may create controversy and post inflammatory remarks online to attract attention and interaction, thereby gaining a sense of presence and venting their emotions (Maltby et al., 2016; Wu et al., 2022). We speculate that higher perceived social adversity is associated with a greater likelihood of online trolling among college students.

Mediating role of depressive symptoms

Depressive symptoms refer to an emotional problem characterized by negative emotions such as pain and sadness (Tang & Zhang, 2022). The perceived social adversity may affect college students’ online trolling through depressive symptoms. According to General Strain Theory, when individuals perceive long-term social adversity, they become tired and lose the sense of joy in life, which leads to the emergence of depressive symptoms, and depressive symptoms promotes aggression, thereby relieving the individual’s psychological pressure (Agnew, 1992). A meta-analysis demonstrated that there is a significant positive association between perceived social adversity and depressive symptoms (Jin & Zhu, 2022). Other studies have identified perceived social adversity as a salient antecedent of depressive symptoms in college students (Kokou-Kpolou et al., 2021). Conversely, college students with more severe depressive symptoms are more likely to ruminate and dwell on negative emotions and experiences. With the help of the anonymity of the internet, they can release their pent-up negative emotions by attacking others in an attempt to relieve their inner pain and stress (Hong & Cheng, 2018). Existing findings suggest that depressive symptoms are positively related to cyberbullying (Alrajeh et al., 2021). We speculate that depressive symptoms as an intermediary mechanism linking perceived social adversity to online trolling.

Moderating role of perceived social support

Perceived social support refers to the extent of emotional satisfaction individuals experience when they feel supported and understood (Sarason et al., 1983). According to the buffering hypothesis of the social support mechanism, social support is regarded as a “buffer”. When people face stress and frustration, they can ease their burden through social support and reduce the harm caused by stress to their body and mind (Cohen & Wills, 1985). Similarly, from the perspective of the I3 model, perceived social support may also act as an Inhibition, playing a moderating role between Instigation and aggressive behavior. By providing resources and strengthening self-regulation and coping capacity, it can buffer adversity-related stress, reduce negative appraisals and emotions, and thereby curb aggressive venting (Wang, 2004; Mitchell et al., 1982). Prior findings indicate that perceived social support can buffer the effect of stress on depressive symptoms among college students (Shi, 2021), while also diminishing the association between stress and online aggression (Li et al., 2018). This kind of buffering effect is usually called the “helping-in-need” model, as shown in Figure 1.

Figure 1. “Helping-in-need” model

However, according to previous research, perceived social support may exert different moderating effects under different levels of perceived stress (Stetz et al., 2006). Accordingly, this study also needs to consider the “a-drop-in-the-bucket” model, as shown in Figure 2. The drop-in-the-bucket model suggests that perceived social support may exacerbate the relationship between perceived social adversity and online trolling. Moreover, the stress vulnerability hypothesis supports the drop-in-the-bucket model. In high-pressure environments, individuals’ positive resources lose their buffering effect (Rutter, 2006; Vanderbilt-Adriance & Shaw, 2008; Li et al., 2012). Prior evidence indicates that peer social support may be more beneficial for adolescents in lower-risk contexts (Wang et al., 2019). Therefore, the moderating role of perceived social support warrants further examination in this study.

Figure 2. “Drop-in-the-bucket” model

Theoretical foundations

The I3 model is a theoretical framework designed to explain the factors underlying aggressive behavior. According to the I3 model, an individual’s aggressive behaviour is the result of the combined effects of Instigation, Impellance and Inhibition (Zhang et al., 2021; Finkel, 2014). Instigation refers to external environmental factors that may directly trigger aggressive behavior (e.g., stressful life events, online anonymity) (Finkel, 2014). Impellance refers to individual characteristics or states that increase the likelihood of aggression (e.g., negative emotions, moral disengagement) (Finkel, 2007). In contrast, Inhibition refers to individual or environmental resources that reduce the likelihood of aggression (e.g., social support, collectivism) (Finkel, 2007, 2014). Prior research has typically conceptualized depressive symptoms as a negative emotional state that functions as a key impellance for aggressive behavior, whereas perceived social support plays an important inhibition in online aggression (Finkel, 2011; Yang et al., 2021). Accordingly, guided by the I3 model, the present study conceptualizes perceived social adversity as an Instigation, depressive symptoms as an Impellance, and perceived social support as an Inhibition, to explore the mediating role of depressive symptoms in the relationship between perceived social adversity and online trolling, along with the moderating role of perceived social support.

Goal of the Study

Within the I³ model framework, we developed a moderated mediation model integrating Instigation (perceived social adversity), Impellance (depressive symptoms), and Inhibition (perceived social support) to examine the mediating role of depressive symptoms in the association between perceived social adversity and online trolling, as well as the moderating mechanism of perceived social support, as shown in Figure 3. Specifically, we proposed three hypotheses:

Figure 3. Moderated mediation model. Note. ***p < 0.001.

Hypothesis 1: College students with higher perceived social adversity are more likely to engage in online trolling.

Hypothesis 2: Depressive symptoms will serve as a mediator in the association between perceived social adversity and online trolling. Specifically, higher perceived social adversity is associated with more severe depressive symptoms, which in turn increases the risk of online trolling.

Hypothesis 3: Perceived social support moderates both the direct and indirect pathways linking perceived social adversity to online trolling. Specifically, higher perceived social support weakens these associations.

Method

Participants and procedure

Using a convenience sampling method, this study collected data over a four-month period from March to June 2024. This study recruited 1047 students from universities across China. The participants’ average age was 19.93 ± 1.38 years, and they included 513 male college students (49.00%) and 534 female college students (51.00%), as shown in Table 1.

Measures

Perceived social adversity

The Perceived Chronic Social Adversity Questionnaire (PCSAQ) compiled by Zhang et al. (2017) was used. The Likert 5-point scoring method is used, ranging from 1 (completely disagree) to 5 (completely agree). The higher the score is, the greater the individual’s perceived social adversity is. For example, some of the items include: “I am always being abandoned.” The results of the confirmatory factor analysis (CFA) suggested that the model fit the data well, with TLI = 0.916; CFI = 0.923; SRMR = 0.042; RMSEA = 0.057. Moreover, Cronbach’s α was 0.95.

Depressive symptoms

The Chinese revised version of the Center for Epidemiological Studies Depression Scale (CES-D) compiled by Radloff was used to measure the subjects’ depressive symptoms (Radloff, 1977). The Likert 4-point scoring method is used, ranging from 0 (occasionally or never) to 3 (most of the time or continuously). After the 4 reverse items are rescored, the higher the score is, the greater the degree of depressive symptoms is. For example, some of the items included: “I was bothered by things that usually don’t bother me.” The results of the CFA suggested that the model fit the data well, with TLI = 0.932; CFI = 0.941; SRMR = 0.040; RMSEA = 0.055. Moreover, the Cronbach’s α was 0.86.

Online trolling

The Global Assessment of Internet Trolling Scale (GAIT) compiled by Sest & March was used. This study used the Chinese version translated by Li et al. (2024). A 5-point Likert scoring method, ranging from 1 (strongly disagree) to 5 (strongly agree). Scores that are higher reflect a greater likelihood of engaging in online trolling. For example, some of the items included: “Although some people think my posts/comments are offensive, I think they are funny.” The results of the CFA suggested that the model fit the data well, with TLI = 0.968; CFI = 0.976; SRMR = 0.026; RMSEA = 0.060. Moreover, the Cronbach’s α was 0.77.

Perceived social support

The perceived social support Scale (PSS) compiled by the domestic scholar Jiang Qianjin was used (Jiang, 2001). It uses a 7-point scoring method ranging from 1 (strongly disagree) to 7 (strongly agree). The higher the score is, the greater the ability to perceive social support is. For example, some of the items included: “There is a special person who is around when I am in need.” The results of the CFA suggested that the model fit the data well, with TLI = 0.964; CFI = 0.974; SRMR = 0.031; RMSEA = 0.056. Moreover, the Cronbach’s α was 0.93.

Procedure

The entire research procedures followed the ethical standards of Sichuan Normal University (IRB NO. SCNU-240318). Before completing the questionnaire, all the subjects signed an informed consent statement and were advised that they could withdraw from the study at any time.

Data analysis

All statistical procedures were conducted in IBM SPSS Statistics 26.0. Harman’s single-factor test was used to examine potential common method bias. The results showed that the largest factor accounted for 22.85% of the variance (below 40%), indicating no substantial common method bias. Multicollinearity was examined. The variance inflation factor (VIF) values for all predictors ranged from 1.00 to 1.84, far below the cutoff value of 5, and tolerances were all ≥0.54, suggesting no multicollinearity concerns. Hayes’ PROCESS macro for SPSS was used to test the mediating role of depressive symptoms in the association between perceived social adversity and online trolling while controlling for gender, age, and grade (Model 4). The indirect effect was estimated with bootstrapping, and mediation was considered significant when the 95% confidence interval did not include zero. PROCESS Model 8 was further applied to examine the proposed moderated mediation model.

Results

Descriptive statistics and correlation analysis

Before conducting the relevant analyses, a normality test was performed on each variable. The results showed that the skewness [−0.725, 1.071] and kurtosis [0.504, 0.996] were within the commonly used empirical threshold range (West et al., 1995), and the overall normality assumption was satisfied (see Appendix I). The relevant analysis results showed that perceived social adversity was significantly positively correlated with online trolling (r = 0.351, p < 0.001), supporting Hypothesis1, as shown in Table 2. Perceived social adversity was significantly positively correlated with depressive symptoms (r = 0.526, p < 0.001), perceived social adversity was significantly negatively correlated with perceived social support (r = −0.188, p < 0.001), depressive symptoms were significantly positively correlated with online trolling (r = 0.225, p < 0.001), and perceived social support was significantly negatively correlated with online trolling (r = −0.198, p < 0.001).

Perceived social adversity effects on online trolling

As presented in Table 3, perceived social adversity remained a significant positive predictor of online trolling after gender, age, and grade were controlled for (β = 0.290, p < 0.001), thereby supporting Hypothesis 1.

Mediating effect of depressive symptoms

As shown in the regression analysis results in Table 4, perceived social adversity significantly and positively predicted depressive symptoms (b = 0.340, t = 20.102, p < 0.001, 95% CI = [0.306, 0.373]), and depressive symptoms positively predicted online trolling (b = 0.154, t = 3.410, p < 0.001, 95% CI = [0.065, 0.242]). The results of the mediation effect analysis, as shown in Table 5 revealed a significant total effect (b = 0.325, 95% CI = [0.277, 0.373]) that is, perceived social adversity could significantly and positively predict online trolling. The indirect effect (b = 0.052, 95% CI = [0.016, 0.089]) indicated that depressive symptoms had a significant mediating effect on the relationship between perceived social adversity and online trolling, supporting Hypothesis 2.

Moderating effect of perceived social support

The results revealed that the interaction term between perceived social adversity and perceived social support significantly predicted online trolling (b = 0.056, t = 3.551, p < 0.001), indicating that perceived social support could moderate the direct path between perceived social adversity and online trolling, as shown in Table 6. However, the interaction between perceived social adversity and perceived social support did not significantly predict depressive symptoms, indicating that perceived social support did not moderate the association between perceived social adversity and depressive symptoms. Simple slope analyses further showed that perceived social adversity was positively associated with online trolling at both low and high levels of perceived social support. Specifically, under low perceived social support, perceived social adversity significantly predicted online trolling (b = 0.201, t = 5.786, p < 0.001, 95% CI = [0.133, 0.270]); under high perceived social support, this predictive effect remained significant and was stronger (b = 0.344, t = 9.731, p < 0.001, 95% CI = [0.275, 0.414]). As shown in Figure 4, the effect of perceived social adversity on online trolling varied as a function of perceived social support. At low levels of perceived social adversity, higher perceived social support was associated with less online trolling, suggesting that perceived social support served a protective role in this condition. However, this pattern did not hold when perceived social adversity was high. Under high-adversity conditions, lower perceived social support was associated with lower levels of online trolling, indicating that perceived social support failed to exert the expected buffering effect. Accordingly, the moderating hypothesis 3 was only partially supported.

Figure 4. The moderating role of perceived social support in the relationship between perceived social adversity and online trolling

Discussion

Results revealed a positive association between perceived social adversity and online trolling among college students, providing support for Hypothesis 1. From the perspective of the I3 model, this result suggests that perceived social adversity, as an Instigation, on the one hand, depletes individuals’ coping and self-regulation resources through prolonged stress. According to coping theory (Lazarus, 1984), when faced with stress, individuals deliberately mobilize their coping resources; when such resources are insufficient, individuals may turn to maladaptive coping strategies (Jin et al., 2020). College students with higher perceived social adversity often endure chronic stress from multiple sources, and their emotion regulation and stress-management capacities may gradually deteriorate (Ybarra & Mitchell, 2004; Hinduja & Patchin, 2014; Huang et al., 2025). As coping resources become depleted, they may engage in online trolling. This study further supports the frustration–aggression hypothesis (Dollard et al., 2013). As a form of frustrating experience, higher perceived social adversity is more likely to elicit online aggressive behavior (Chen et al., 2024). Moreover, unlike other forms of online aggression, trolling does not require a specific target; simply posting provocative or divisive remarks can quickly trigger reactions and provide short-term psychological compensation (Cheng et al., 2017), thereby temporarily alleviating psychological distress.

The mediation analysis showed that depressive symptoms mediated the association between perceived social adversity and online trolling, thereby supporting the proposed hypothesis. This finding is consistent with prior research and provides further support for General Strain Theory (Hay & Meldrum, 2010; Agnew, 1992). One possible explanation is that social adversity may evoke feelings of helplessness and frustration in college students, which subsequently give rise to negative emotional responses, with depressive symptoms representing one of the most common outcomes of this process (Jin & Zhu, 2022). Depressive symptoms can also reduce an individual’s ability to regulate emotions, making it difficult for them to cope with stress and negative emotions in a positive way, and as a result, they choose an aggressive coping method (Hong & Cheng, 2018). Online trolling is also an anonymous, low-risk way to express anger and hostility (Wu et al., 2022; Buckels et al., 2014), making it a coping strategy for individuals with depressive symptoms.

Notably, the findings did not support a significant moderating role of perceived social support in the link between perceived social adversity and depressive symptoms. This result may be attributable, first, to the fact that perceived social support reflects a subjective perception that is readily influenced by personal characteristics, psychological status, and cognitive biases (Lakey & Orehek, 2011; Han et al., 2025). Second, depressive symptoms emerge from the interplay of biological, psychological, and social determinants, which may restrict the protective influence of social support (Kendler et al., 2002; Lakey & Orehek, 2011). Furthermore, depressive symptoms are commonly associated with maladaptive cognitions, including self-blame and negative self-evaluation, which may undermine individuals’ willingness or ability to perceive and receive support, thereby attenuating the role of perceived social support (Ciarrochi et al., 2011; Esposito et al., 2024).

However, the present study showed that perceived social support strengthened, rather than buffered, the association between perceived social adversity and online trolling, thereby providing only partial support for Hypothesis 3. One possible interpretation is consistent with the drop-in-the-bucket model: perceived social support was associated with lower levels of online trolling only when perceived social adversity was relatively low, whereas this protective effect disappeared under high levels of adversity. This pattern is also broadly consistent with prior research showing that, when negative stressors become overwhelming, individuals’ self-regulatory capacities may be compromised, limiting the protective role of social support (Rutter, 2006). In this sense, the present findings also accord with the stress vulnerability hypothesis, which posits that positive personal and contextual resources may lose their buffering function in high-stress environments (Vanderbilt-Adriance & Shaw, 2008; Li et al., 2012; Teng et al., 2025). Prior research suggests that under high levels of stress, the protective effects of positive psychological qualities may weaken or even disappear (Fu et al., 2012; Li et al., 2012; George-Levi et al., 2025). For example, adolescents with higher gratitude adapt better under low-stress life events, but under high-stress life events, maladjustment is more likely regardless of whether gratitude is high or low (Li et al., 2012). Similarly, our findings suggest that perceived social support may also operate as a stress vulnerability factor. Although the mechanisms were not directly tested, high perceived social adversity may deplete cognitive resources and reduce individuals’ awareness of available support (Baumeister, 1997), and may also undermine their sense of control, making high levels of support more likely to be perceived as intrusive and thus eliciting psychological reactance. In this context, online trolling may provide a sense of dominance and control that fulfills emotional and self-identity needs, thereby leading to more online trolling (van Baak et al., 2023).

Implication for practice

This study provides some practical insights. First, at the individual level, online trolling should be viewed as an externalizing behavior produced through person–environment interaction. Interventions should enhance self-regulation through positive reappraisal and emotional awareness; perceived social support can serve as a daily resource under low stress, whereas under high stress greater emphasis should be placed on cognitive resource management, a sense of self-control, and self-identity. Secondly, at the school level, regular assessments and crisis screenings are used to identify students at high risk of stress and depressive symptoms early on. At the same time, relevant positive psychology knowledge, such as perceived social support, is integrated into class meetings, ideological and political education classes, and campus culture to enhance psychological resilience and interpersonal trust. Finally, at the societal level, governance should go beyond deleting posts and blocking content by creating healthier channels for emotional expression through rules and content guidance. Platforms can raise the cost of posting aggressive content, curb algorithmic amplification of stigmatizing language, and increase the visibility of prosocial, mutual-aid, and positive-emotion content.

Limitations and future directions

Firstly, this study uses a cross-sectional design and regression-based mediation and moderation analysis, which prevents causal inference and ignores the influence of inter-structural relationships. Future research could use latent variable SEM and longitudinal studies to further examine its robustness and causal mechanisms within a more rigorous framework. Secondly, this study used convenience sampling, resulting in relatively limited sample coverage. Future studies could employ more random sampling methods to improve the robustness of the results. Furthermore, this study’s depressive symptoms measure focused on negative affect, and self-reported assessment of trolling may have been affected by social desirability bias. Future research could adopt a more multidimensional instrument and use real platform posting scenarios to assess online trolling more directly. Finally, perceived social support is highly subjective and susceptible to various influences, such as cultural background. Future research could examine more readily operationalizable psychological resources (e.g., gratitude and mindfulness).

Conclusion

Most prior studies have examined online trolling primarily from an individual-level perspective. Extending this line of work, the present study investigated the association between perceived social adversity and online trolling, together with the mechanisms underlying this association. The findings indicate that perceived social adversity is an important risk factor for online trolling among college students, operating through both a direct pathway and an indirect pathway via depressive symptoms. In addition, perceived social support moderated the direct association between perceived social adversity and online trolling. Specifically, perceived social support was associated with lower levels of online trolling only when perceived social adversity was low; this protective effect was not observed under high levels of perceived social adversity. Accordingly, interventions aimed at reducing online trolling should go beyond platform content regulation and also emphasize the early identification of individuals experiencing elevated stress and depressive symptoms, followed by the provision of appropriate psychological resources and support.

Acknowledgement: We sincerely thank all individuals who took part in this study.

Funding Statement: Project supported by the National Social Science Foundation of China (Grant No. 23BSH143).

Author Contributions: Conceptualization: Siqi Wang, Fang Li; Methodology: Siqi Wang, Fang Li, Biyun Wu; Software: Siqi Wang; Validation: Siqi Wang, Yuedong Qiu; Formal Analysis: Siqi Wang; Investigation: Siqi Wang; Resources: Siqi Wang, Fang Li, Yuedong Qiu, Biyun Wu; Data Curation: Siqi Wang; Writing—Original Draft: Siqi Wang; Writing—Review & Editing: Siqi Wang, Fang Li, Yuedong Qiu; Visualization: None; Supervision: Fang Li, Yuedong Qiu, Biyun Wu; Funding Acquisition: Fang Li. All authors reviewed and approved the final version of the manuscript.

Availability of Data and Materials: The data underlying the findings of this study can be obtained from the first author upon reasonable request.

Ethics Approval: The entire research procedures followed the ethical standards of Sichuan Normal University (IRB NO. SCNU-240318), and all research projects were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards (Date of Application: 18 March 2024, IBR NO. SCNU-240318). Written informed consent was obtained from all the participants prior to the enrollment of this study.

Conflicts of Interest: The authors declare no conflicts of interest.

References

Agnew, R. (1992). Foundation for a general strain theory of crime and delinquency. Criminology, 30(1), 47–88. https://doi.org/10.1111/j.1745-9125.1992.tb01093.x [Google Scholar] [CrossRef]

Akhtar, S., & Morrison, C. M. (2019). The prevalence and impact of online trolling of UK members of parliament. Computers in Human Behavior, 99(1), 322–327. https://doi.org/10.1016/j.chb.2019.05.015 [Google Scholar] [CrossRef]

Alrajeh, S. M., Hassan, H. M., Al-Ahmed, A. S., & Alsayed Hassan, D. (2021). An investigation of the relationship between cyberbullying, cybervictimization and depression symptoms: A cross sectional study among university students in Qatar. PLoS One, 16(12), e0260263. https://doi.org/10.1371/journal.pone.0260263. [Google Scholar] [PubMed] [CrossRef]

Baumeister, R. F. (1997). Esteem threat, self-regulatory breakdown, and emotional distress as factors in self-defeating behavior. Review of General Psychology, 1(2), 145–174. https://doi.org/10.1037//1089-2680.1.2.145 [Google Scholar] [CrossRef]

Buckels, E. E., Trapnell, P. D., & Paulhus, D. L. (2014). Trolls just want to have fun. Personality and Individual Differences, 67, 97–102. https://doi.org/10.1016/j.paid.2014.01.016 [Google Scholar] [CrossRef]

Chen, R., Hu, Y., Shi, H. F., Fang, Y., & Fan, C. Y. (2024). Perceived chronic social adversity and cyberbullying perpetration among adolescents: The mediating role of rumination and moderating role of mindfulness. Frontiers in Psychology, 15, 1376347. https://doi.org/10.3389/fpsyg.2024.1376347. [Google Scholar] [PubMed] [CrossRef]

Cheng, J., Bernstein, M., Danescu-Niculescu-Mizil, C., & Leskovec, J. (2017). Anyone can become a troll: Causes of trolling behavior in online discussions. In: Proceedings of the 2017 ACM Conference on Computer Supported Cooperative Work and Social Computing (pp. 1217–1230); 2017 Feb 25–Mar 1; Portland, OR, USA. [Google Scholar]

Ciarrochi, J., Dean, F. P., & Anderson, S. (2011). Emotional intelligence moderates the relationship between stress and mental health. Personality and Individual Differences, 50(1), 20–25. [Google Scholar]

Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310–357. https://doi.org/10.1037/0033-2909.98.2.310 [Google Scholar] [CrossRef]

Ding, C., Zhang, J., & Yang, D. (2018). A pathway to psychological difficulty: Perceived chronic social adversity and its symptomatic reactions. Frontiers in Psychology, 9, 615. https://doi.org/10.3389/fpsyg.2018.00615. [Google Scholar] [PubMed] [CrossRef]

Dollard, J., Miller, N. E., Doob, L. W., Mowrer, O. H., Sears, R. R. et al. (2013). Frustration and aggression. Oxfordshire, UK: Routledge. [Google Scholar]

Esposito, C. M., Mancini, M., Estradé, A., Rosfort, R., Fusar-Poli, P., et al. (2024). How do depressed people feel perceived by others? A qualitative study from the patient’s perspective. Journal of Affective Disorders Reports, 16(10), 100776. https://doi.org/10.1016/j.jadr.2024.100776 [Google Scholar] [CrossRef]

Finkel, E. J. (2007). Impelling and inhibiting forces in the perpetration of intimate partner violence. Review of General Psychology, 11(2), 193–207. https://doi.org/10.1037/1089-2680.11.2.193 [Google Scholar] [CrossRef]

Finkel, E. J. (2011). I3 Theory: Instigating, impelling, and inhibiting factors in aggression. In: Human aggression and violence: Causes, manifestations, and consequences (pp. 407–421). Guildord, UK: American Psychological Association. [Google Scholar]

Finkel, E. J. (2014). The I3 model: Metatheory, theory, and evidence. In: J. M. Olson, M. P. Zanna (Eds.), Advances in experimental social psychology (Vol. 49, pp. 1–104). Cambridge, MA, USA: Academic Press. [Google Scholar]

Fu, Q., Ye, B., & Wen, Z. (2012). The impact mechanism of stressful life events on adolescents’ subjective well-being. Psychological Development and Education, 28(5), 516–523. [Google Scholar]

George-Levi, S., Faverman, L., Galin-Lonchich, Y., Ben-Gal Dahan, A., & Frei-Landau, R. (2025). Beyond meaning: Hope and secondary trauma in Israeli therapists after the October 7th massacre. Frontiers in Psychology, 16, 1594885. https://doi.org/10.3389/fpsyg.2025.1594885. [Google Scholar] [PubMed] [CrossRef]

Goyanes, M., Borah, P., & De Zúñiga, H. G. (2021). Social media filtering and democracy: Effects of social media news use and uncivil political discussions on social media unfriending. Computers in Human Behavior, 120(3), 106759. https://doi.org/10.1016/j.chb.2021.106759 [Google Scholar] [CrossRef]

Han, H., Yu, X., Li, H., & Sun, N. (2025). The bidirectional relationship between depression and perceived social support among college freshmen: Evidence from longitudinal and daily diary methods. Frontiers in Psychology, 16, 1676105. https://doi.org/10.3389/fpsyg.2025.1676105. [Google Scholar] [PubMed] [CrossRef]

Hay, C., & Meldrum, R. (2010). Bullying victimization and adolescent self-harm: Testing hypotheses from General Strain Theory. Journal of Youth and Adolescence, 39(5), 446–459. https://doi.org/10.1007/s10964-009-9502-0. [Google Scholar] [PubMed] [CrossRef]

Herrenkohl, T. I. (2013). Person-environment interactions and the shaping of resilience. Trauma, Violence, & Abuse, 14(3), 191–194. https://doi.org/10.1177/1524838013491035. [Google Scholar] [PubMed] [CrossRef]

Hinduja, S., & Patchin, J. W. (2014). Cyberbullying identification, prevention and response. Cyberbullying Research Center. https://www.cyberbullying.us. [Google Scholar]

Hong, F. Y., & Cheng, K. T. (2018). Correlation between university students’ online trolling behavior and online trolling victimization forms, current conditions, and personality traits. Telematics and Informatics, 35(2), 397–405. https://doi.org/10.1016/j.tele.2017.12.016 [Google Scholar] [CrossRef]

Howard, K., Zolnierek, K. H., Critz, K., Dailey, S., & Ceballos, N. (2019). An examination of psychosocial factors associated with malicious online trolling behaviors. Personality and Individual Differences, 149(4), 309–314. https://doi.org/10.1016/j.paid.2019.06.020 [Google Scholar] [CrossRef]

Huang, L., Chen, L., & Ma, S. (2025). The relationship between social media fatigue and online trolling behavior among college students: The mediating roles of relative deprivation and hostile attribution bias. Frontiers in Psychology, 15, 1495235. https://doi.org/10.3389/fpsyg.2024.1495235. [Google Scholar] [PubMed] [CrossRef]

Jiang, Q. J. (2001). Perceived social support scale. Chinese Journal of Behavioral Medical Science, 10(10), 41–43. [Google Scholar]

Jin, T., Wu, Y., Zhang, L., Yang, X., Jia, Y., et al. (2020). The effect of perceived chronic social adversity on aggression of college students: The roles of ruminative responses and perceived social support. Psychological Development and Education, 36(4), 414–421. [Google Scholar]

Jin, T., & Zhu, X. (2022). Perceived chronic social adversity, perceived social support, and depression among Chinese college students: Mediating role of self-esteem. American Journal of Applied Psychology, 9(4), 1–8. https://doi.org/10.11648/j.ajap.20221101.11 [Google Scholar] [CrossRef]

Kapse, S. M., & Gupta, S. (2023). Trolling and victims’ perception of trolling: A thematic analysis. In: Cyberfeminism and gender violence in social media (pp. 49–70). Hershey, PA, USA: IGI Global. [Google Scholar]

Kendler, K. S., Gardner, C. O., & Prescott, C. A. (2002). Toward a comprehensive developmental model for major depression in women. American Journal of Psychiatry, 159(7), 1133–1145. https://doi.org/10.1176/appi.ajp.159.7.1133. [Google Scholar] [PubMed] [CrossRef]

Kokou-Kpolou, C. K., Jumageldinov, A., Park, S., Nieuviarts, N., Noorishad, P. G., et al. (2021). Prevalence of depressive symptoms and associated psychosocial risk factors among French University students: The moderating and mediating effects of resilience. Psychiatric Quarterly, 92(2), 443–457. https://doi.org/10.1007/s11126-020-09812-8. [Google Scholar] [PubMed] [CrossRef]

Lakey, B., & Orehek, E. (2011). Relational regulation theory: A new approach to explain the link between perceived support and mental health. Psychological Review, 118(3), 482–495. https://doi.org/10.1037/a0023477. [Google Scholar] [CrossRef]

Lazarus, R. S. (1984). Stress, appraisal, and coping (Vol. 464). Berlin/Heidelberg, Germany: Springer. [Google Scholar]

Li, Y., Li, D., Li, X., Zhou, Y., Sun, W. et al. (2018). Cyber victimization and adolescent depression: The mediating role of psychological insecurity and the moderating role of perceived social support. Children and Youth Services Review, 94, 10–19. https://doi.org/10.1016/j.childyouth.2018.09.027 [Google Scholar] [CrossRef]

Li, F., Tang, X., Ge, X., Yu, M., Wang, S., et al. (2024). Is mindfulness associated with lower online trolling among adolescents? Mediating effects of self-esteem and depression and moderating effect of dark personality traits. Journal of Psychology in Africa, 34(3), 293–298. https://doi.org/10.1080/14330237.2024.2363698 [Google Scholar] [CrossRef]

Li, D., Zhang, W., Li, X., Li, N., & Ye, B. (2012). Gratitude and suicidal ideation and suicide attempts among Chinese adolescents: Direct, mediated, and moderated effects. Journal of Adolescence, 35(1), 55–66. https://doi.org/10.1016/j.adolescence.2011.06.005. [Google Scholar] [PubMed] [CrossRef]

Liu, H., Hao, X., Kong, D., Wang, W., & Yang, J. (2021). Influences of perceived chronic social adversity, psychache and negotiable fate on suicidal risk: A moderated mediation model. Chinese Journal of Clinical Psychology, 30(4), 954–958. [Google Scholar]

Maltby, J., Day, L., Hatcher, R. M., Tazzyman, S., Flowe, H. D. et al. (2016). Implicit theories of online trolling: Evidence that attention-seeking conceptions are associated with increased psychological resilience. British Journal of Psychology, 107(3), 448–466. https://doi.org/10.1111/bjop.12154. [Google Scholar] [PubMed] [CrossRef]

Mitchell, R. E., Billings, A. G., & Moos, R. H. (1982). Social support and well-being: Implications for prevention programs. Journal of Primary Prevention, 3(2), 77–98. https://doi.org/10.1007/BF01324668. [Google Scholar] [PubMed] [CrossRef]

Nitschinsk, L., Tobin, S. J., & Vanman, E. J. (2023). A functionalist approach to online trolling. Frontiers in Psychology, 14, 1211023. https://doi.org/10.3389/fpsyg.2023.1211023. [Google Scholar] [PubMed] [CrossRef]

Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. https://doi.org/10.1177/014662167700100306. [Google Scholar] [CrossRef]

Rutter, M. (2006). Implications of resilience concepts for scientific understanding. Annals of the New York Academy of Sciences, 1094(1), 1–12. https://doi.org/10.1196/annals.1376.002. [Google Scholar] [PubMed] [CrossRef]

Sarason, I. G., Levine, H. M., Basham, R. B., & Sarason, B. R. (1983). Assessing social support: The social support questionnaire. Journal of Personality and Social Psychology, 44(1), 127. https://doi.org/10.1037/0022-3514.44.1.127 [Google Scholar] [CrossRef]

Shi, B. (2021). Perceived social support as a moderator of depression and stress in college students. Social Behavior and Personality: An International Journal, 49(1), 1–9. https://doi.org/10.2224/sbp.9893 [Google Scholar] [CrossRef]

Simchon, A., Brady, W. J., & Van Bavel, J. J. (2022). Troll and divide: The language of online polarization. PNAS Nexus, 1(1), pgac019. https://doi.org/10.1093/pnasnexus/pgac019. [Google Scholar] [PubMed] [CrossRef]

Stetz, T. A., Stetz, M. C., & Bliese, P. D. (2006). The importance of self-efficacy in the moderating effects of social support on stressor-strain relationships. Work & Stress, 20(1), 49–59. https://doi.org/10.1080/02678370600624039 [Google Scholar] [CrossRef]

Tang, J., & Zhang, T. (2022). Causes of the male-female ratio of depression based on the psychosocial factors. Frontiers in Psychology, 13, 1052702. https://doi.org/10.3389/fpsyg.2022.1052702. [Google Scholar] [PubMed] [CrossRef]

Teng, C., Shi, L., Xu, C., & Zhu, J. (2025). Parenting styles and academic burnout in medical students: A moderated mediation model of resilience and stress. Frontiers in Medicine, 12, 1703534. https://doi.org/10.3389/fmed.2025.1703534. [Google Scholar] [PubMed] [CrossRef]

Tufekci, Z. (2017). Twitter and tear gas: The power and fragility of networked protest. New Haven, CT, USA: Yale University Press. [Google Scholar]

van Baak, C., Maher, C. A., Protas, M. E., & Hayes, B. E. (2023). Victims and perpetrators of cyber harassment: The role of power and control and the use of techniques of neutralization. Deviant Behavior, 44(5), 690–707. https://doi.org/10.1080/01639625.2022.2088317 [Google Scholar] [CrossRef]

Vanderbilt-Adriance, E., & Shaw, D. S. (2008). Conceptualizing and re-evaluating resilience across levels of risk, time, and domains of competence. Clinical Child and Family Psychology Review, 11(1), 30. https://doi.org/10.1007/s10567-008-0031-2. [Google Scholar] [PubMed] [CrossRef]

Volkmer, S. A., Gaube, S., Raue, M., & Lermer, E. (2023). Troll story: The dark tetrad and online trolling revisited with a glance at humor. PLoS One, 18(3), e0280271. https://doi.org/10.1371/journal.pone.0280271. [Google Scholar] [PubMed] [CrossRef]

Wang, Y. F. (2004). A review of research on the relationship between social support and physical and mental health. Psychological Science, 27(5), 1175–1177. [Google Scholar]

Wang, Q., Xiao, T., Liu, H., & Hu, W. (2019). The relationship between parental rejection and Internet addiction among left-behind children: A moderated mediation model. Psychological Development and Education, 35(6), 749–758. https://doi.org/10.16187/j.cnki.issn1001-4918.2019.06.12 [Google Scholar] [CrossRef]

West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables: Problems and remedies. In: Hoyle, R. H. (Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 56–75). Thousand Oaks, CA, USA: Sage Publications, Inc. [Google Scholar]

Wilson, N. C., & Seigfried-Spellar, K. C. (2023). Cybervictimization, social, and financial strains influence internet trolling behaviors: A general strain theory perspective. Social Science Computer Review, 41(3), 967–982. https://doi.org/10.1177/08944393211065868 [Google Scholar] [CrossRef]

Wu, B., Zhou, L., Deng, Y., Zhao, J., & Liu, M. (2022). Online disinhibition and online trolling among Chinese college students: The mediation of the dark triad and the moderation of gender. Cyberpsychology, Behavior, and Social Networking, 25(11), 744–751. https://doi.org/10.1089/cyber.2022.0046. [Google Scholar] [PubMed] [CrossRef]

Yang, J., Wang, N., Gao, L., & Wang, X. (2021). Deviant peer affiliation and adolescents’ cyberbullying perpetration: Online disinhibition and perceived social support as moderators. Children and Youth Services Review, 127(4), 106066. https://doi.org/10.1016/j.childyouth.2021.106066 [Google Scholar] [CrossRef]

Ybarra, M. L., & Mitchell, K. J. (2004). Youth engaging in online harassment: Associations with caregiver-child relationships, Internet use, and personal characteristics. Journal of Adolescence, 27(3), 319–336. https://doi.org/10.1016/s0140-1971(04)00039-9 [Google Scholar] [CrossRef]

Zhang, J., Ding, C., Tang, Y., Zhang, C., & Yang, D. (2017). A measure of perceived chronic social adversity: Development and validation. Frontiers in Psychology, 8, 2168. https://doi.org/10.3389/fpsyg.2017.02168. [Google Scholar] [PubMed] [CrossRef]

Zhang, L., Wu, Y., & Jin, T. (2021). The expression mechanism of individual behavior from the perspective of I3 model. Advances in Psychological Science, 29(10), 1878–1886. https://doi.org/10.3724/sp.j.1042.2021.01878 [Google Scholar] [CrossRef]