The Relationship between Social Anxiety and Emotional Intelligence among College Students: Analysis of the Mediating Effect of General Self-Efficacy ()
1. Introduction
The transition to university life represents a critical developmental period, presenting students with a host of new academic and social challenges. Students are required not only to acquire extensive academic knowledge but also to cultivate personal competencies essential for their future success and well-being. An individual’s psychological qualities have a significant influence on their life, and the possession of positive traits can be effective in improving their quality of life. Social anxiety, emotional intelligence, and general self-efficacy play significant roles in the overall development of college students. Therefore, a deeper analysis of these three factors to describe the current state of interpersonal communication among contemporary students and understand the relationships and interactions between them is essential for enhancing and improving students’ quality of life and social environment.
Social anxiety refers to a negative emotion experienced during interpersonal interactions and is a common emotional state (Lacombe et al., 2024). Individuals with social anxiety often feel nervous, fearful, uneasy, hypervigilant, and doubtful of themselves and others, accompanied by feelings of depression. In the 1990s, David Clark and others developed a model of social anxiety, which provides a framework for understanding and explaining anxiety arising in social situations (Clark & Wells, 1995). These authors proposed that social anxiety can be defined and interpreted through specific means, rejecting the notion that it is an elusive phenomenon. This model has been found to explain most social anxiety scenarios. Research indicates that social skills and self-evaluation are two critical factors influencing social anxiety (Lau et al., 2022; Okano & Nomura, 2023). Social skills refer to an individual’s ability to engage in effective interpersonal communication, while self-evaluation pertains to an individual’s assessment of their own traits and abilities, showing a negative correlation with social anxiety. Many international psychologists have reached similar conclusions (Cartwright-Hatton et al., 2003; Koban et al., 2023); suggesting that the primary causes of social anxiety are poor social skills and negative self-evaluation. However, other psychologists hold different views. Studies show that social anxiety is closely linked to negative self-evaluation, self-focused attention, and concerns about others’ impressions, which affect communication and interpersonal functioning (Heerey & Kring, 2007; Tonge et al., 2020). Related studies indicate that the needs for knowledge, self-esteem and self-reliance, and friendship are the three primary needs for students entering university life, highlighting the importance of interpersonal skills during their four years of college (Nuha et al., 2024). As social anxiety is a key factor affecting interpersonal communication abilities, greater attention should be paid to addressing it to meet the basic communication needs of college students.
In the late 1990s, Reuven Bar-On (1997), based on his research, defined emotional intelligence as “an array of personal, emotional, and interpersonal abilities that influence an individual’s capacity to cope with environmental demands and pressures.” He considered emotional intelligence a critical factor in determining an individual’s success in life (Bar-On, 2006), which is related to an individual’s mental health. Bar-On proposed that emotional intelligence consists of five main components: intrapersonal skills, interpersonal skills, adaptability, stress management, and general mood (Di Fabio & Kenny, 2016; Rind et al., 2025). Researchers building on existing literature conducted surveys and factor analysis on collected data. They proposed that emotional intelligence influences an individual’s success or failure in learning, life, and work, representing a non-cognitive psychological ability (Coronado-Maldonado & Benítez-Márquez, 2023; Sharma & Tiwari, 2023; Urquijo et al., 2019). Emotional intelligence encompasses five attributes related to ability: emotional evaluation, perception, adaptation, regulation, and expression (Mayer et al., 2024; Serrat, 2017). These five factors can be further divided into eighteen sub-factors, collectively comprising an individual’s emotional intelligence system (Ching & Kitahara, 2025; Kohli et al., 2025). Additionally, several researchers have approached the concept from the dimensions of object and operation (Bru-Luna et al., 2021; Drigas & Papoutsi, 2018), defining the extension of emotional intelligence by analyzing its components. They suggested that emotional intelligence includes seven aspects, namely, understanding, awareness, evaluation, prediction, experience, expression, and regulation of emotions (Husain et al., 2022). Emotional intelligence has been shown to be correlated to varying degrees with college students’ entrepreneurial activities, interpersonal communication, career decision-making, and social adaptability (Chien-Chi et al., 2020). This highlights the paramount importance of cultivating emotional intelligence among college students (Miezah et al., 2025).
The concept of self-efficacy was introduced by Bandura in the 1970s (Bandura, 1977a). Self-efficacy refers to an individual’s confidence in their ability to handle various pressures in life, study, and work, as well as their belief or perception of successfully completing a task or behavior (Bandura, 2024). When faced with challenges or tasks, the difficulty of tasks chosen by individuals is influenced by their level of self-efficacy (Street et al., 2022). After experiencing failure, individuals with high self-efficacy tend to readjust, regroup, and learn from the experience without becoming disheartened, whereas those with low self-efficacy may become discouraged or even give up (Kerwagen et al., 2024). In situations of success, individuals with high self-efficacy are likely to attribute their accomplishments to their own talents and efforts, while those with low self-efficacy may attribute them to luck or external factors (Caliendo et al., 2023; Glosenberg et al., 2022). Self-efficacy is associated to some extent with the career choices, life circumstances, family environment, academic performance, and practical experience of college students (Zhou et al., 2022). For example, studies in the field of education have shown that teachers with high teaching self-efficacy demonstrate significantly better teaching quality and outcomes compared to those with low self-efficacy (Burić et al., 2024). Additionally, students’ academic self-efficacy is associated with their learning achievements and behaviors; students with higher self-efficacy exhibit greater self-monitoring ability, which has a positive impact on their future goal setting and academic performance (Meng & Zhang, 2023). Self-efficacy holds an important place in the realm of individual cognitive regulation and has received considerable attention across various fields (Bandura, 2024; Gazo & Mahasneh, 2024). For instance, in the workplace, self-efficacy is one of the most effective predictors of job performance. Individuals with high self-efficacy generally perform better, while those with low self-efficacy tend to perform poorly (Alessandri et al., 2024; Pacheco et al., 2023). Moreover, the self-efficacy of organizational managers significantly influences the overall performance of the organization (Kashif & Atta, 2025; Zam et al., 2024).
Previous studies have reported close relationships between emotional intelligence and social anxiety, emotional intelligence and general self-efficacy, and social anxiety and general self-efficacy. However, most studies have focused on correlations between two variables, leaving the underlying mechanisms of these interactions underexplored (Guo & Xu, 2023; Rostami et al., 2024). Specifically, while we know that social anxiety is linked to lower emotional intelligence, the process through which this occurs is not well understood (Cao et al., 2023; Rozen & Aderka, 2023). Bandura’s social cognitive theory suggests that self-beliefs, such as self-efficacy, are central to human functioning. It is plausible that the apprehension and negative self-evaluation characteristic of social anxiety erode an individual’s confidence in their ability to manage situations (i.e., their general self-efficacy), which in turn impairs their capacity to perceive, regulate, and utilize emotions (i.e., their emotional intelligence, Morales-Rodríguez & Pérez-Mármol, 2019; Wang et al., 2022).
While the relationship between social anxiety and emotional intelligence could theoretically be bidirectional, we propose that social anxiety represents the antecedent variable. This is based on the following rationale: First, from a developmental perspective, symptoms of social anxiety often emerge earlier in adolescence (before university entry), whereas emotional intelligence continues to develop through social learning and practice during young adulthood (Mayer et al., 2008). Second, cognitive-behavioral models of social anxiety (Clark & Wells, 1995) suggest that anxiety-driven avoidance behaviors reduce opportunities for emotional learning and the development of social skills, thereby restricting the development of emotional intelligence. Third, longitudinal evidence suggests that anxiety symptoms predict subsequent deficits in skills associated with emotional regulation rather than the reverse (Aldao et al., 2010). Nevertheless, we acknowledge that cross-sectional data cannot definitively establish causality, and future longitudinal research is needed to confirm the directional relationships investigated here.
The present study aims not only to investigate the associations among these three variables but also to examine whether self-efficacy acts as a mediating factor that explains the influence of social anxiety on emotional intelligence. The study seeks to explore whether a mediating model can be constructed to explain the relationships among them. Based on the theoretical framework, the following hypotheses were formulated:
H1: There will be a significant negative correlation between social anxiety and emotional intelligence.
H2: There will be a significant negative correlation between social anxiety and general self-efficacy.
H3: There will be a significant positive correlation between general self-efficacy and emotional intelligence.
H4: General self-efficacy will mediate the relationship between social anxiety and emotional intelligence.
Researching these issues can, on one hand, enrich the findings related to self-efficacy and, on the other hand, enhance the awareness of cultivating corresponding abilities among college students. The findings hold practical significance for helping students manage interpersonal relationships effectively and strengthen their self-efficacy.
2. Subjects and Methods
2.1. Research Subjects
This study employed a convenience sampling method. The subjects were undergraduate students from Guangxi Minzu University, spanning first to fourth years. A total of 505 questionnaires were distributed. After removing incomplete or invalid responses, a final sample of 461 participants was retained for analysis. In terms of grade level, the valid responses included 130 from freshmen, 69 from sophomores, 108 from juniors, and 174 from seniors. In terms of sex, there were 145 male and 316 female respondents. All participants provided informed consent, and the study procedures were conducted in accordance with ethical guidelines for research with human subjects.
2.2. Research Tools
2.2.1. Interaction Anxiousness Scale (IAS)
The IAS was developed by Leary based on clinical experience (Leary & Kowalski, 1993). The scale aims to eliminate the influence of individual behavior and measure an individual’s subjective tendency to feelings of social anxiety. In this study, the total IAS score was used as an objective indicator of interpersonal competence. Chinese studies on the reliability and validity of the scale have shown that although the correlation between item 9 and the total score was r = 0.275 in homogeneity tests, the correlations between other items and the total score all exceeded 0.4, generally meeting survey requirements. The scale showed a Cronbach’s α of 0.81 and a test-retest reliability of 0.78, aligning closely with international findings and indicating its stable performance in measuring social anxiety. The IAS consists of 15 questions and requires the subjects to make a five-level assessment. Lower scores indicate lower levels of social anxiety.
2.2.2. Wong and Law Emotional Intelligence Scale (WLEIS)
In the early 2000s, Law and colleagues developed the self-reporting WLEIS (Law, et al., 2004). This 16-item scale uses a 7-point scoring system, where “0” indicates “strongly disagree” and “6” represents “strongly agree.” Lower scores reflect lower emotional intelligence. The scale shows interactive relationships with the “big five” personality traits and cognitive technical skills, supporting its internal validity. In a sample of university students, the scale achieved a Cronbach’s α of 0.83 (Akbar et al., 2025), with inter-item correlations ranging from 0.12 to 0.59, confirming its predictive validity.
2.2.3. General Self-Efficacy Scale (GSES)
The study utilized a Chinese version of the GSES, originally developed by foreign psychologists and their colleagues, and later adapted by Zhang and Schwarzer (1995). This 10-item instrument employs a 4-point Likert scoring system. Analysis of the reliability and validity of the scale revealed that the correlation coefficients between items and the overall test ranged from 0.60 to 0.77, indicating satisfactory performance of all items. The Cronbach’s α of the scale was found to be 0.87, with a test-retest reliability of 0.83. In terms of validity, the GSES exhibits good predictive validity and internal consistency, supporting its effectiveness in measuring general self-efficacy among university students.
3. Research Results
3.1. Descriptive Statistics and Bivariate Correlations
As hypothesized, social anxiety was significantly and negatively correlated with both emotional intelligence (r = −0.612, p < 0.01) and general self-efficacy (r = −0.0605, p < 0.01). Furthermore, general self-efficacy was significantly and positively correlated with emotional intelligence (r = 0.596, p < 0.01). (Table 1)
Table 1. Presents the means, standard deviations, and Pearson correlation coefficients for the main study variables.
Variable |
M ± SD |
1 |
2 |
3 |
Social Anxiety |
39.86 ± 7.93 |
— |
|
|
General Self-Efficacy |
24.13 ± 6.50 |
−0.605** |
— |
|
Emotional Intelligence |
73.54 ± 19.50 |
−0.612** |
0.596** |
— |
3.2. Analysis of the Mediating Role of General Self-Efficacy between Social Anxiety and Emotional Intelligence
Based on the theoretical assumptions of this study, social anxiety was treated as the independent variable, emotional intelligence as the dependent variable, and general self-efficacy as the mediating variable. The mediating effect was tested using Model 4 in the PROCESS macro in SPSS software and was analyzed by employing Hayes’ Bootstrap method (5000 samples, 2018), resulting in the construction of a mediation model.
The path coefficients among social anxiety, emotional intelligence, and general self-efficacy are shown in Figure 1.
Figure 1. Mediation model.
Table 2. Decomposition of total effect, direct effect, and mediating effect.
|
Effect Size |
se |
LLCI |
ULCI |
Proportion |
Total Effect |
−1.50** |
0.12 |
−1.73 |
−1.28 |
|
Direct Effect |
−0.98** |
0.14 |
−1.24 |
−0.71 |
64.67% |
Mediating Effect |
−0.53** |
0.10 |
−0.73 |
−0.34 |
35.33% |
The results indicate the validity of the constructed mediation model, with general self-efficacy playing a partial mediating role. The total effect of social anxiety on emotional intelligence was significant, with an effect size of −1.50. The direct effect of social anxiety on emotional intelligence was also significant, with an effect size of −0.98. The indirect effect of social anxiety on emotional intelligence through general self-efficacy was significant, with an effect size of −0.53, as the 95% confidence interval did not contain zero. Detailed results are presented in Table 2.
4. Discussion
This study aimed to investigate the relationships among social anxiety, general self-efficacy, and emotional intelligence in college students, with a specific focus on the mediating role of self-efficacy. The findings supported all four of the initial hypotheses and provide a clearer understanding of the psychological mechanisms at play.
Consistent with H1 and previous research, the results revealed a strong negative correlation between social anxiety and emotional intelligence. Many studies have observed that stronger interpersonal skills are associated with higher emotional intelligence (Garcia et al., 2025; Petrovici & Dobrescu, 2014). This phenomenon may result from the frequent lack of effective interpersonal skills and the ability to regulate both their own and others’ emotions in individuals with high social anxiety, leading to lower emotional intelligence. A negative correlation was also found between social anxiety and general self-efficacy, supporting H2 and aligning with earlier findings (Hood et al., 2021; Liao et al., 2023). This suggests that individuals with higher social anxiety tend to have lower confidence in their ability to complete tasks and lower overall self-efficacy (Fatemi et al., 2024; Iancu et al., 2015). Excessive social anxiety may lead to over-sensitivity in daily life, with individuals being inclined toward safe behaviors, resulting in decreased motivation for social interaction and reduced confidence in accomplishing tasks, thereby lowering their self-efficacy (NikMehr et al., 2021). A positive correlation was observed between general self-efficacy and emotional intelligence, in line with H3 and supporting the findings of existing studies (Shubayr & Dailah, 2025). The factors influencing self-efficacy include the individual’s emotional state and level of arousal, indicating that higher general self-efficacy is associated with higher emotional intelligence (Morales-Rodríguez & Pérez-Mármol, 2019; Sun & Lyu, 2022). Moreover, these two traits are likely interrelated, with improvement in one often enhancing the other (Hameli & Ordun, 2022). The most crucial finding of this study was the confirmation of H4: general self-efficacy partially mediated the relationship between social anxiety and emotional intelligence (Guo & Xu, 2023; Rostami et al., 2024). This indicates that social anxiety does not only have a direct negative impact on emotional intelligence but also an indirect one. Specifically, the results suggest that higher levels of social anxiety reduce a student’s belief in their own capabilities (lower self-efficacy), and this diminished self-efficacy, in turn, contributes to lower emotional intelligence. This finding provides an important explanatory mechanism, moving beyond simple correlation to suggest a pathway through which social anxiety exerts its detrimental effects.
Interestingly, no significant differences in social anxiety, self-efficacy, or emotional intelligence were found across ethnic groups, despite collection of data from Guangxi Minzu University, an institution with a highly diverse student body, including Han, Zhuang, Yao, and other ethnic minorities. This finding may reflect the increasingly integrated multicultural nature of higher education in contemporary China, where shared educational experiences, common academic pressures, and peer socialization transcend ethnic boundaries. The psychological constructs examined in this study, namely, social anxiety, self-efficacy, and emotional intelligence, may be influenced more strongly by universal developmental processes and shared educational contexts than by ethnic-specific cultural practices. Additionally, the homogenizing influence of modernization and digital communication may have reduced traditional ethnic differences in psychological functioning among younger generations. However, this null finding should be interpreted cautiously, as the measures used may not have been sufficiently sensitive to capture subtle cultural variations in the expression or experience of these constructs. Future research using culturally adapted instruments or mixed-methods approaches could explore the influence of specific ethnic traditions on coping strategies, help-seeking behaviors, and expression patterns related to social anxiety and emotional competence.
4.1. Theoretical and Practical Implications
Theoretically, these findings contribute to the literature by integrating social anxiety, self-efficacy, and emotional intelligence within a single explanatory model. By demonstrating the mediating role of self-efficacy, this study lends empirical support to the application of social cognitive theory in understanding the interplay between social-emotional difficulties and core self-beliefs. It highlights that cognitive appraisals of one’s own competence are a critical link between feeling anxious in social situations and the ability to manage emotions effectively.
Practically, the results have significant implications for university mental health services and student support programs. Interventions aimed at reducing social anxiety should not only focus on exposure and social skills training but should also incorporate components designed to bolster students’ general self-efficacy. For example, workshops focused on setting achievable goals, celebrating small successes, and reframing failures as learning opportunities could be highly beneficial. By strengthening students’ sense of efficacy, institutions may be able to buffer them from the negative emotional consequences of social anxiety, thereby fostering greater psychological well-being and academic success.
The present finding that general self-efficacy mediates the relationship between social anxiety and emotional intelligence suggests that university mental health interventions should prioritize promoting self-efficacy as a pathway to improving emotional competence among students with social anxiety. Specifically, interventions could incorporate the following evidence-based strategies grounded in Bandura’s (Bandura, 1997b) social cognitive theory:
First, experiences of mastery should be central to interventions. Counselors could design tasks involving graduated exposure in which students can successfully navigate progressively challenging social situations, thereby building confidence through an accumulation of successful experiences. For example, structured programs in social-skill training could begin with one-on-one interactions, progress to small group discussions, and culminate in public speaking or leadership roles. Successful completion of the individual steps reinforces students’ beliefs in their capabilities.
Second, vicarious learning programs could be introduced, involving peer mentoring or group activities where students observe others in similar situations successfully managing social interactions. Watching peers with comparable backgrounds overcome social challenges can enhance observers’ beliefs in their own efficacy. Video modeling or role-playing exercises featuring relatable student models can be particularly effective for socially anxious individuals who might otherwise be reluctant to initiate or engage directly in social interactions.
Third, verbal persuasion requires the training of counselors and educators to provide specific, credible feedback that reinforces students’ capabilities in social and emotional domains. Rather than providing generic praise, the feedback should highlight specific improvements and attribute success to the growing competence of the students rather than to external factors. For instance, “Your ability to recognize and respond to others’ emotions during that group discussion showed real skill” is more efficacy-enhancing than “You did well.”
Fourth, addressing physiological and emotional states requires instruction in techniques for managing anxiety, such as progressive muscle relaxation, diaphragmatic breathing, and cognitive restructuring. These enable students to reinterpret physiological arousal (e.g., increased heart rate) as manageable excitement rather than as evidence of inadequacy or impending failure. Mindfulness-based interventions may also help students develop greater awareness and acceptance of their emotional experiences.
These multi-component interventions targeting self-efficacy may prove more effective than only addressing emotional intelligence skills, as they target underlying barriers and lack of confidence that prevent socially anxious students from applying their emotional knowledge in real-world interactions. The initial reinforcement of students’ confidence in their ability to handle social-emotional challenges can provide a foundation upon which emotional intelligence can develop naturally through practice and experience.
4.2. Limitations and Future Directions
Several limitations of this study should be acknowledged. First, the cross-sectional design does not allow for the establishment of causality. Although the mediation model is based on strong theory, longitudinal research is needed to confirm the directional relationships proposed. Although the present theoretical framework and mediation model propose that social anxiety influences emotional intelligence through self-efficacy, alternative models are also conceivable. For instance, individuals with lower emotional intelligence may struggle to interpret social cues accurately, which may in turn increase social anxiety. Similarly, the relationship between these constructs could be bidirectional or even reciprocal over time. It is possible that the mechanisms underlying these associations may be more complex than suggested by this unidirectional model, with possible feedback loops where low emotional intelligence exacerbates social anxiety, which may in turn further impair the development of emotional skills. Second, the use of a convenience sample from a single university limits the generalizability of the findings to other student populations. Future research should aim to replicate these findings in more diverse samples. Third, all data were collected via self-report questionnaires, which may be subject to common method variance and social desirability bias. Future studies could incorporate behavioral observations or peer-report measures to provide a more comprehensive assessment. Future research could also explore other potential mediators or moderators in the relationship between social anxiety and emotional intelligence, such as coping styles or social support. Additionally, experimental intervention studies that directly manipulate self-efficacy could provide stronger causal evidence for the proposed model.
The gender imbalance in the present sample (68.5% female, 31.5% male) warrants careful consideration when interpreting these findings. While the results revealed significant gender differences, with females reporting higher levels of social anxiety and males demonstrating higher emotional intelligence, the over-representation of female participants may limit the precision of parameter estimates for male students and reduce the generalizability of the findings to more gender-balanced populations. This gender distribution, although reflecting the typical composition of educational psychology and social science programs in Chinese universities, may have inflated the statistical power for detecting effects in females, with insufficient power for males. Consequently, more gender-balanced samples should be investigated before strong conclusions are drawn. It is also possible that the under-representation of male participants may have obscured important gender-specific patterns or moderation effects. Future studies should prioritize the recruitment of more balanced samples through targeted strategies to enable more robust gender comparisons and to ensure that the findings can be generalized to both genders. Gender-stratified analyses or weighting procedures could also be applied to account for unequal sample sizes when examining gender as a moderator.
5. Conclusion
In conclusion, this study confirms the negative relationships between social anxiety and both emotional intelligence and general self-efficacy among college students. More importantly, it provides novel evidence that general self-efficacy serves as a significant mediator in the link between social anxiety and emotional intelligence. These findings underscore the critical role of self-beliefs in the emotional lives of students and suggest that fostering self-efficacy is a key target for interventions aimed at promoting student mental health and interpersonal competence.
Funding Information
Our study is supported by three funding sources with the following grant numbers: This study was funded by the Guangxi Education Science “14th Five-Year Plan” 2023 Special Topics (No. 2023ZJY136), the 2024 Guangxi Higher Education Institutions Undergraduate Ideological and Political Education Theory and Practice Research Projects (No. 2024MSZ019), and the 2024 Guangxi Higher Education Undergraduate Teaching Reform Project (No. 2024JGA175).