Social anxiety, the fear of negative evaluation in social situations, is a common concern among college students and might significantly affect their academic and social experiences (Rapee and Heimberg, 1997). In China, research by Zhao and Dai (2016) indicated that Chinese college students exhibited significantly higher average levels of social anxiety. A study involving 820 participants revealed that 29.65% of college students had experienced high levels of social anxiety (Zhang et al., 2020). Among Chinese college students, social anxiety manifested uniquely through avoidance of group discussions and reluctance to seek professional help (Zha et al., 2023). These behaviors exacerbated academic underachievement. While mobile social networks allowed for interpersonal communication without the immediate pressure of in-person interactions, the fear of being judged online might have exacerbated social anxiety (Hofmann, 2007; She et al., 2023). Given its pervasive impact, identifying modifiable risk factors is critical for intervention.

In the digital era, mobile social networks are integrated into college students’ social lives as an essential component, providing platforms that enable communication, offer entertainment and allow for self-expression (Andreassen et al., 2017; Shannon et al., 2022; Tandoc et al., 2015). Although these technologies facilitate connection and engagement, they bring about risks, particularly regarding problematic usage. Problematic mobile social network usage is typified by an inordinate preoccupation with mobile social networks, advancing to a level where it affects an individual’s daily activities and undermines their wellbeing (González-Nuevo et al., 2023; Wang et al., 2024; Young, 1998; Young et al., 2017). This phenomenon is of particular concern among college students.

There is increasing attention on the problematic mobile social network usage among college students, concentrating on exploring the psychological mechanisms underlying such behavior. The association between problematic mobile social network usage and various negative outcomes have been widely reported, including increased anxiety, depression and academic impairment (Hussain and Griffiths, 2021; Peng et al., 2006; Yigiter et al., 2024). problematic mobile social network usage has been identified as a predictor of social anxiety (Caplan, 2002; Weidman et al., 2012; Lee and Stapinski, 2012). A study involving 1,082 Chinese college students found that problematic mobile social network usage was significantly positively correlated with online social anxiety, and that problematic mobile social network usage could significantly predict online social anxiety (Liu et al., 2022). While previous studies have explored direct links between social media use and anxiety, few have examined the internal psychological processes.

Loneliness, conceptualized as a subjective sense of social isolation, may play a pivotal role in the relationship between problematic mobile social network usage and social anxiety. According to Social Compensation Theory (Caplan, 2005), individuals experiencing social difficulties may engage in excessive mobile social network usage as a compensatory strategy. However, this heightened online interaction often displaces face-to-face communication, diminishes the quality of real-world social connections, and consequently intensifies feelings of loneliness (Lee and Lee, 2023; O’Day and Heimberg, 2021; Wang and Zeng, 2024). The Cognitive-Behavioral Model of Social Anxiety (Rapee and Heimberg, 1997) posits that elevated loneliness reinforces negative self-perceptions and maladaptive cognitions (e.g., “others will reject me”), leading to increased avoidance behaviors and exacerbated social anxiety (Lim et al., 2016; Federman et al., 2024). Therefore, we propose that loneliness may mediate the relationship between problematic mobile social network usage and social anxiety.

However, empirical findings on the mediating role of loneliness between problematic mobile social network usage and social anxiety are inconsistent. While some studies reported significant indirect effects of problematic mobile social network usage on social anxiety through loneliness (Liu et al., 2022; She et al., 2023), replication of these results has not been achieved in other research (Kim et al., 2019). This inconsistency suggests the potential influence of moderating variables, e.g., self-concept clarity. Self-concept clarity, defined as the extent to which an individual’s self-beliefs are clear, internally consistent and temporally stable (Campbell et al., 1996), has been examined in various studies. Current reports present divergent perspectives regarding the role of self-concept clarity. On one hand, individuals with high self-concept clarity possess more stable and coherent self-views, typically demonstrating greater resilience against stressors and potentially increased resistance to the appeal of problematic mobile social network usage (Lodi-Smith and Roberts, 2010; Parise et al., 2019), owing to their reduced reliance on external validation (Lodi-Smith and Roberts, 2010). Conversely, emerging evidence suggests that when mobile social network usage becomes problematic, individuals with high self-concept clarity might be more sensitive to the dissonance between their well-defined self-perception and the superficial nature of online interactions, potentially resulting in stronger experiences of loneliness (Pang et al., 2024). Considering these contrasting possibilities, this study aims to investigate whether self-concept clarity moderates the effect of problematic usage on loneliness.

Grounded in the Cognitive-Behavioral Model of Social Anxiety (Rapee and Heimberg, 1997) and Social Compensation Theory (Caplan, 2005), this study establishes a moderated mediation model. This model addresses two primary questions: (1) how loneliness transmits the effect of problematic mobile social network usage on social anxiety, and (2) whether self-concept clarity moderates this mediation process by altering the strength of the pathway from problematic mobile social network usage to loneliness.

The present study investigated the complex relationships among problematic mobile social network usage, social anxiety, loneliness and self-concept clarity among college students, as well as a moderated mediation model in which loneliness mediates the relationship between problematic mobile social network usage and social anxiety, with self-concept clarity moderating this mediation process.

Both loneliness and social anxiety were positively associated with problematic mobile social network usage. However, a negative correlation was detected between problematic mobile social network usage and self-concept clarity. These findings were consistent with previous research, suggesting that individuals suffering from higher levels of social anxiety and loneliness were more likely to engage in problematic mobile social network usage as a coping mechanism (Caplan, 2002; Ge et al., 2025; O’Day and Heimberg, 2021; Weidman et al., 2012). Both loneliness and social anxiety were positively associated with problematic mobile social network usage, while self-concept clarity was negatively correlated with such usage. This aligns with previous research, suggesting that individuals with unclear self-concepts may rely more on external validation from social media, leading to problematic usage patterns (Lodi-Smith and Roberts, 2010).

The mediating role of loneliness in the relationship between problematic mobile social network usage and social anxiety has been confirmed through mediation analysis. Consistent with the Cognitive-Behavioral Model (Rapee and Heimberg, 1997), the results indicated that loneliness, as a form of negative self-perception, exacerbated social anxiety. Loneliness significantly mediated the relationship, accounting for a substantial proportion of 18.50% of the total effect. The observation suggested that college students have problems using mobile social networks, which might increase loneliness and exacerbate their social anxiety. The finding was consistent with the Social Compensation Theory (Caplan, 2002), which suggested that individuals who had difficulties in real-world social interactions tend to over-rely on mobile social networks for compensation. Problematic mobile social network usage reduces offline socializing, weakens genuine emotional connections, and thereby intensifies loneliness, ultimately leading to increased social anxiety.

The moderation analysis results supported our research hypothesis, indicating that self-concept clarity plays a moderating role in the relationship between problematic mobile social network usage and loneliness. Specifically, compared to individuals with low self-concept clarity, problematic mobile social network usage had a more pronounced impact on loneliness among those with high self-concept clarity. This finding aligns with studies suggesting that individuals with high self-concept clarity have a heightened awareness regarding the gap between their social needs and actual social experiences (Pang et al., 2024). When these individuals engage in excessive use of mobile social networks, they become more acutely aware of the negative impact of excessive social media use on their well-being and recognize that online interactions fail to fulfill their deeper social needs, thereby experiencing more intense feelings of loneliness (Lewandowski et al., 2010). For example, a student with a clear self-view as “sociable” might feel lonelier if problematic scrolling through social media highlights a lack of deep, meaningful interactions compared to their self-perception.

This finding appears inconsistent with previous research demonstrating that self-concept clarity generally buffers against negative psychological outcomes (Lodi-Smith and Roberts, 2010; Wong et al., 2019). The discrepancy could be explained by the distinctive nature of problematic mobile social network usage. While prior studies have primarily examined general stressors or normal social media use, in the specific context of problematic usage characterized by loss of control and preoccupation, the self-awareness of individuals with high self-concept clarity may paradoxically heighten their sensitivity to the negative impact of their usage behaviors on social well-being, thereby strengthening its association with loneliness.

These findings hold significant implications for exploring the psychological mechanisms underlying social anxiety among college students. The study emphasizes that individuals’ problematic mobile social network usage and loneliness serve as effective intervention entry points for addressing social anxiety (Du et al., 2024). Intervention strategies such as enhancing social skills, strengthening social support, improving interpersonal interactions, and adjusting negative social cognition can effectively alleviate individuals’ loneliness (Masi et al., 2011). Additionally, fostering self-concept clarity may help college students cope with the complexity of social anxiety in a healthier manner.

The current study used a cross-sectional design method. Longitudinal research will be needed to establish the temporal sequence of the observed relationships and to further explore the potential reciprocal effects among problematic mobile social network usage, loneliness, social anxiety, and self-concept clarity.

This study revealed significant insights into the relationship between problematic mobile social network usage and social anxiety among college students, highlighting the mediating role of loneliness and the moderating role of self-concept clarity. The findings indicated that loneliness acts as a mediator, linking excessive use of mobile social networks to increased levels of social anxiety. Additionally, the present study observed the moderating role of self-concept in the relationship.

Educational institutions are encouraged to implement programs aimed at promoting responsible use of mobile social networks. Workshops and seminars can be organized to educate students about the potential negative consequences of excessive use, including increased loneliness and social anxiety. These could include cognitive restructuring sessions to challenge beliefs. Additionally, peer mentoring programs could facilitate offline social skill practice, encouraging students to engage in real-world social activities and hobbies—reducing reliance on mobile social networks for interaction. To further reinforce stable self-perception, institutions might introduce self-concept journaling (Wu et al., 2023), where daily entries help students cultivate healthier self-awareness.