Peer pressure and fear of missing out among university students in the social media environment: the influence mechanism of self‑authenticity

Abstract

Background

Fear of Missing Out (FoMO) and Peer Pressure are intertwined among university students and are highly associated with Social Media Addiction. This study explores the potential moderating role of Self-Authenticity in the associations among peer pressure, social media-related factors, and FoMO.

Methods

A cross-sectional survey was administered to 397 Chinese university students. Five variables were measured: Self-Authenticity, Peer Pressure, Social Media Addiction, Online Social Comparison, and FoMO. Structural equation modeling (SEM) and moderation analyses were utilized to examine the structural relationships among these variables.

Results

Social Media Addiction and Online Social Comparison exhibited indirect association patterns between Peer Pressure and FoMO. Self-Authenticity demonstrated significant negative predictive effects on all variables in the model. Additionally, Self-Authenticity negatively moderated the association between Peer Pressure and FoMO (where the direct predictive path became non-significant after its inclusion), while demonstrating positive moderating effects on the associations from Peer Pressure to Social Media Addiction, and from Social Media Addiction to Online Social Comparison.

Conclusion

Self-Authenticity is negatively associated with baseline levels of peer pressure, social media-related risks, and FoMO. Grounded in the S-O-R framework, the study highlights Self-Authenticity as a key individual-level variable that moderates the statistical associations among external pressures, digital behaviors, and psychological states. Given the cross-sectional design, these findings represent structural associations rather than causal mechanisms.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40359-026-04952-1.

Background

In recent years, social media has developed into an indispensable global platform for information dissemination and social interaction. Supported by the widespread use of mobile internet and smartphones, these platforms enable instant transmission and extensive sharing of information. As a highly active group, university students can access massive information and establish social connections almost anytime and anywhere [1]. On the one hand, the rapid development of social media has broadened students’ horizons. It allows easier exposure to diverse cultures and frontier knowledge, and enhances self expression and social participation [2]. On the other hand, social media has also generated a series of negative consequences. Excessive engagement with social platforms may lead to distraction, reduced learning efficiency, and psychological problems such as information anxiety and Social Media Addiction [3–5]. More importantly, highly curated self presentations and idealized life displays by peers on social media can intensify Online Social Comparison and distort self perception. This process may further induce feelings of inferiority or anxiety among university students [6, 7]. Among these psychological outcomes, Fear of Missing Out (FoMO) represents one of the most prominent manifestations [8].

Recent studies have increasingly examined user behavior and psychological outcomes in mobile social media environments. Research has shown that both human and system interactivity can influence users’ continued use of mobile short video applications through mechanisms such as user gratification and platform attachment [9]. Other work suggests that service quality may shape user identification, belongingness, and satisfaction through emotional attachment processes [10]. In addition, multidimensional benefits have been found to contribute to cumulative satisfaction and electronic word-of-mouth engagement in mobile social media contexts [11].

Beyond user engagement, prior studies have also explored the psychological and behavioral consequences associated with digital media use. For example, depressive mood and self-disclosure have been linked to problematic mobile application use and subsequent educational outcomes [12]. Communication overload and cyberbullying behaviors may also influence users’ switching intentions through social cognitive mechanisms [13]. Other research highlights the role of social connectivity and system interactivity in shaping continuance intention in mobile applications [14], as well as the importance of network externalities and perceived gratifications in sustaining platform use [15]. Furthermore, emotional and functional attachment have been identified as key mediating mechanisms linking human–machine interaction with live-streaming addiction [16].

Although a growing body of research has examined the psychological consequences of social media use [17], the psychological mechanisms that may buffer or mitigate these negative effects have received comparatively limited attention. Existing studies often propose general recommendations based on empirical findings, yet the effectiveness of these suggestions in real contexts lacks systematic validation. Based on prior theories and empirical observations, this study proposes that Self-Authenticity may function as an individual trait with potential buffering capacity. This study does not treat Self-Authenticity as a sole or decisive solution to social media related problems. Instead, it conceptualizes Self-Authenticity as a psychological resource with possible moderating and buffering effects, which may extend current understanding of individual differences. To examine this assumption, an exploratory cross sectional quantitative study was designed to address the following two questions.

• A)Whether Peer Pressure is transformed into Online Social Comparison and Social Media Addiction, and ultimately leads to FoMO.
• B)Whether Self-Authenticity can exert a general protective effect on social media related risk factors and FoMO, through direct effects and indirect moderating effects.

Peer pressure and FoMO in the social media context

In contemporary campus settings, Peer Pressure and FoMO are intertwined, and jointly shape university students’ social behaviors and psychological experiences [18]. Peer Pressure refers to the pressure individuals feel to conform to others’ expectations or behavioral norms within a group [19], and this pressure arises not only from face to face interactions, but is also reinforced through likes, comments, and sharing behaviors on social media platforms [20]. Because university students are at a critical stage of seeking identity and social belonging, they are often more susceptible to peer influence [21]. When peers frequently display vibrant social lives, academic achievements, or extracurricular experiences, individuals may perceive implicit or explicit expectations from the group, which generate strong pressure to participate and avoid absence [22]. At the same time, FoMO has become a widespread psychological phenomenon among university students, defined as anxiety and unease about being excluded while others are experiencing meaningful or enjoyable activities [23]. FoMO was first proposed by Przybylski [24], who defined FoMO as “a pervasive apprehension that others might be having rewarding experiences from which one is absent.” In higher education contexts, FoMO is reflected not only in concerns about missing classroom discussions or student organization activities, but also in persistent attention to social media updates, online live streams, or virtual lectures. This tendency drives students to repeatedly refresh information feeds, in order to alleviate exclusion related anxiety triggered by Peer Pressure [25].

More importantly, Peer Pressure and FoMO exhibit a dynamic interactive relationship. Peer modeling effects on social media platforms can intensify individuals’ FoMO, while elevated FoMO further strengthens the desire for peer approval and belonging, making university students more likely to conform to group norms [26]. Empirical evidence indicates that academic related peer influence significantly increases FoMO by heightening students’ attention to others’ academic outcomes and social achievements. FoMO in turn functions between peer influence and student engagement behaviors, and shapes time management and academic investment [27]. This mechanism places university students in a dual pursuit of high academic performance and sustained social media activity. Such competing demands often increase psychological burden, and subsequently elevate risks of anxiety, sleep disturbances, and attentional distraction [28].

As an antecedent variable of social media addiction and online social comparison, peer pressure is theoretically grounded in its role as an external environmental stimulus. According to the S-O-R framework [29], peer pressure represents the direct urgency imposed by group norms on individuals. It compels individuals to rapidly adopt specific behavioral adjustments or cognitive alignment strategies in order to maintain social belongingness or avoid social exclusion [30]. Within this context, addictive behavior can be understood as a compliant behavioral response that enables individuals to satisfy peer expectations and maintain real-time social interaction [31]. Online social comparison (OSC), in contrast, functions as a direct cognitive mechanism through which individuals evaluate their relative position within the group. FoMO is more appropriately conceptualized as an internal anxiety state that emerges alongside such pressure [32]. Therefore, within the S-O-R framework, external peer pressure serves as the primary driving force that initiates subsequent behavioral adaptation and cognitive evaluation processes.

In addition, the present study proposes that addictive behavior constitutes a critical precondition for the emergence of social comparison in the context of passive social media use among university students [33]. Social media addiction reflects a high level of dependence on social networking platforms and frequent platform engagement. This intensive and immersive usage pattern inevitably increases passive exposure to information about others [34]. Under addictive conditions, individuals are more likely to encounter carefully curated and idealized self-presentations from others. Such dense information exposure naturally activates and reinforces social comparison tendencies [35]. In other words, addictive behavior creates the temporal and situational conditions necessary for social comparison to occur. Without frequent engagement with social media platforms, individuals would lack sufficient informational input to sustain comparison processes. Consequently, from both a logical and statistical perspective, excessive information exposure resulting from addictive behavior is likely to precede and dominate the emergence of social comparison.

Self-authenticity

Self-Authenticity refers to the degree to which individuals maintain consistency and integration among inner experiences, behavioral expressions, and social roles. It includes awareness of personal values and emotions, and involves naturally expressing this awareness in behavior and interpersonal relationships [36]. This construct is commonly divided into three core dimensions. Authentic living refers to acting in accordance with one’s inner beliefs and values. Self-alienation indicates a sense of disconnection from one’s true needs or emotions. Accepting external influence reflects excessive conformity to others’ or societal expectations in behavior or decision making [37]. Higher levels of authentic living are positively associated with psychological well being, self esteem, and life satisfaction [38]. In contrast, elevated self-alienation and accepting external influence are often linked to anxiety, depression, and lower subjective well being [39].

In terms of theoretical origins, the concept of Congruence proposed by Carl Rogers in humanistic psychology emphasizes that psychological health depends on the alignment between experience and action [40]. Subsequent scholars developed this idea into more refined multicomponent models. These models suggest that Self-Authenticity is not only a personality trait, but also a fluctuating state experience [41, 42]. Accordingly, Self-Authenticity has been distinguished into state authenticity and trait authenticity [43]. State authenticity refers to the feeling of being real and true to oneself in a specific context or moment. It represents a short term and situational experience [44]. Trait authenticity, also termed dispositional authenticity, describes a general tendency to live authentically across situations. It reflects a long term and stable personality characteristic [45]. This form of authenticity indicates the extent to which individuals remain true to themselves without distortion from external pressures. Self determination theory similarly proposes that deeper authenticity is experienced when behavior aligns with intrinsic motivation. In contrast, externalized motivation tends to undermine feelings of authenticity [46]. In the serial mediation model constructed in this study, Trait Authenticity demonstrates significantly greater theoretical fit and explanatory power compared with State Authenticity. Since peer pressure, social media addiction, and FoMO all represent psychological tendencies or habitual behaviors that develop over extended periods, the moderating variable must exhibit longitudinal stability across contexts to serve as an effective psychological defense mechanism. State Authenticity, in contrast, is more susceptible to fluctuations from transient environmental factors [44], and momentary psychological states are insufficient to explain how individuals resist the long-term influences of complex social media algorithms or habitual social comparison pressures. Therefore, all references to authenticity in the present study specifically refer to trait authenticity.

Although autonomy and authenticity are often closely associated in Self-Determination Theory (SDT) and related psychological literature, the two constructs differ fundamentally in their conceptual meanings. As one of the basic psychological needs, autonomy primarily concerns the sense of volition and self-endorsement underlying behavior. It emphasizes whether individuals perceive their actions as self-directed rather than constrained by external control or internal compulsion. Authenticity, in contrast, focuses more strongly on the congruence between behavior and the true self. It reflects the extent to which individuals remain faithful to their internal values, beliefs, and personality in everyday life [47]. From a theoretical perspective, autonomy often serves as an essential motivational foundation for the experience of authenticity. However, individuals may voluntarily choose to conform to certain social expectations based on autonomous motivation. Although such behavior possesses autonomous characteristics, it does not necessarily represent the expression of trait authenticity.

In recent years, researchers have developed multiple measurement instruments, such as the Authenticity Scale by Wood [37] and the State Authenticity Scale by Lenton [48]. These tools have supported extensive validation of the structure and validity of Self-Authenticity. Empirical evidence further indicates that Self-Authenticity is closely related to mental health and psychological resilience [49, 50]. It has also received increasing attention in fields such as organizational behavior and cross cultural psychology [51, 52].

The buffering role of self-authenticity

Theoretically, Self-Authenticity represents a stable inner state that enables individuals to make choices based on their own values and needs. Conversely, FoMO often reflects ambiguity and instability in self definition, and directly contributes to heightened Online Social Comparison [53]. The emergence of FoMO is typically driven by external standards and excessive attention to others’ lifestyles. When individuals lack Self-Authenticity, they are more likely to become immersed in others’ expectations and prevailing social trends, and experience anxiety and confusion about how they should live [54]. Under such psychological conditions, FoMO functions as a mechanism for avoiding inner emptiness and unease. By contrast, individuals with higher levels of Self-Authenticity are more likely to accept their own choices, boundaries, and pace of life. Even when facing external temptations or social pressure, they are less prone to experience anxiety about missing out. In brief, higher Self-Authenticity is associated with lower reliance on external comparison, and weaker experiences of FoMO.

From a practical perspective, this relationship is evident in various aspects of daily life. For example, within social media environments, many individuals feel that they are falling behind when they see friends attending parties, traveling, or achieving success [24]. When a person lacks Self-Authenticity, they may compel themselves to participate in activities in order to keep up with others and avoid social exclusion. This occurs even when such activities do not align with their interests or values. Over time, this pattern can lead to exhaustion, emptiness, and even self alienation [55]. However, individuals with clear inner orientation and awareness of their genuine desires are more capable of deciding not to participate, without fearing the potential losses of such choices. Their life focus is more centered on experiences that are meaningful to them, rather than on what others are doing.

From the perspective of SDT, self-authenticity reflects a high degree of integration between individual behavior and one’s core values and intrinsic interests. This highly autonomous psychological state fundamentally reshapes the way individuals cognitively process the external social environment [47]. Individuals with high authenticity possess a stable internal locus of evaluation. The maintenance of their self-worth does not depend on external approval or unquestioned conformity to group norms [46]. In social contexts, highly authentic individuals are more likely to regulate their behavior according to their own intentions rather than external rewards or punishments. This tendency substantially reduces the likelihood of experiencing compulsive psychological pressure [37]. Even when objective social norms are present in the external environment, individuals with high authenticity demonstrate stronger psychological filtering abilities. They are more likely to interpret external cues as informational input rather than mandatory commands. As a result, self-authenticity is expected to significantly and negatively predict individuals’ subjective perceptions of peer pressure.

The tension between Self-Authenticity and FoMO reflects a common psychological conflict in modern life. Individuals face the question of whether to live in accordance with the self, or to strive constantly in order to avoid being left behind. From theory to practice, this relationship highlights whether inner self strength is sufficient to resist anxiety induced by information overload and social comparison. We argue that Self-Authenticity can counteract the emergence and development of FoMO, and also represents a key pathway toward personal freedom and psychological calm.

Research hypotheses

Self Determination Theory (SDT) serves as a macro framework for examining human motivation, personality development, and psychological health. Its core logic is based on the satisfaction of three basic psychological needs [56], namely Autonomy, Competence, and Relatedness. This theory views individuals not as passive objects driven by the environment, but as active organisms that integrate the self inwardly and pursue growth outwardly. Such growth is driven by the fulfillment of basic psychological needs [46]. Self-Authenticity can be regarded as the highest psychological manifestation of the need for Autonomy. It represents a high level of congruence between behavior and the inner self. Individuals with high Self-Authenticity tend to engage in social interactions based on intrinsic motivation, rather than due to Peer Pressure or the pursuit of external approval [57]. As a result, they show a natural resistance to highly curated self presentations on social media. This form of psychological independence directly weakens the influence of social media algorithms on individual self esteem.

However, when the sense of autonomy is not satisfied, psychological deprivation may compel individuals to activate compensatory mechanisms. They may seek alternative psychological compensation through social media use. In this context, Online Social Comparison becomes a distorted way of pursuing Competence. Individuals locate their social value by observing others’ lives, and relieve loneliness and alienation through continuous virtual interaction [58]. Such behavior does not stem from authentic self driven motivation. Instead, it reflects passive compensation controlled by a sense of deficiency. This process not only fails to repair psychological gaps, but also draws individuals into a deeper negative feedback loop [6]. Within this pattern of deprivation and compensation, FoMO represents anxiety arising from severe deficits in Relatedness and Autonomy. It also reflects a stress driven attachment to external information [59]. When individuals maintain a high level of Self-Authenticity, they are able to accept their current experiences. This capacity fundamentally interrupts illusory desires for others’ lives. Therefore, Self-Authenticity is not merely a personality trait. It can also satisfy basic psychological needs at their root, promote rejection of blind social comparison and pathological Social Media Addiction, and ultimately foster resistance to FoMO. Social media may then return to its role as a functional life tool.

Based on the above discussion, this study focuses on Chinese university students. It integrates five variables, including Self-Authenticity, Peer Pressure, Social Media Addiction, Online Social Comparison, and FoMO, into a single theoretical model. The study examines potential mediating effects of social media related factors in the association between Peer Pressure and FoMO. It also explores possible direct effects and moderating effects of Self-Authenticity within the overall model. This study aims to construct a chained mediation model that includes moderating and control variables, and it proposes ten theory driven research hypotheses.

• H1. Peer Pressure will significantly and positively predict FoMO.

• H2. Peer Pressure will significantly and positively predict Social Media Addiction.

• H3. Peer Pressure will significantly and positively predict Online Social Comparison.

• H4. Social Media Addiction will significantly and positively predict FoMO.

• H5. Social Media Addiction will significantly and positively predict Online Social Comparison.

• H6. Online Social Comparison will significantly and positively predict FoMO.

• H7. Self-Authenticity significantly negatively moderates the predictive path between peer pressure and social media addiction.

• H8. Self-Authenticity significantly negatively moderates the predictive path between online social comparison and social media addiction.

• H9. Self-Authenticity significantly negatively moderates the predictive path between peer pressure and FoMO.

• H10. After controlling for the mediation and moderation effects in the model, self-authenticity will still significantly and negatively predict the remaining four variables.

Methods

This study employed a cross sectional questionnaire survey to collect quantitative data. The data were processed and examined for reliability and validity using SPSS 29. R Studio was used to construct the structural equation model (SEM) and to test moderating effects, in order to analyze the mechanisms and robustness of relationships among factors.

Preliminary questionnaire design

The questionnaire consisted of six sections, and all scale items adopted a standard five point Likert format. The demographic section collected information on gender, age, and level of study. Scale refinement and modification were guided by the expertise of two quantitative research specialists who held doctoral degrees and had a Chinese cultural background. First, in order to reduce respondent burden and avoid fatigue effects, the researchers aimed to shorten the full questionnaire by approximately ten items. This requirement was provided to the two experts, who independently selected items for deletion. Afterward, their decisions were compared. If both experts agreed that an item should be removed, their judgment was adopted. When disagreements occurred, face to face discussion was conducted to reach consensus. After the translation of the complete questionnaire, the two experts also evaluated and revised the items, resulting in the final Chinese version.

The scale section comprised five subscales. The first was the Self-Authenticity scale, based on the 12 item scale developed by Wood [37]. It was reduced to 11 items in this study, and included three dimensions, namely Self-Alienation, Accepting External Influence, and Authentic Living. This scale assessed trait authenticity, and measured the extent to which individuals act according to their beliefs and values in daily life, as well as whether they feel that they are being themselves. The Self-Alienation and Accepting External Influence dimensions were reverse scored, and reverse coding was applied during data analysis. The second was the Peer Pressure scale, adapted from the 11 item scale developed by Santor [19], and reduced to six items. It primarily assessed perceived peer pressure, especially the extent to which peers influence individuals’ behaviors, attitudes, and choices. The third was the Social Media Addiction scale, using the six item instrument developed by Andreassen [60]. It measured addictive tendencies in social media use, including excessive use, withdrawal responses, increased tolerance, and the degree to which daily life is affected. This scale reflected levels of psychological dependence on social media. The fourth was the Online Social Comparison scale, based on the eight item scale developed by Lee [61], and reduced to six items in this study. It mainly assessed the frequency and tendency of upward social comparison on online social media platforms. The final scale was the FoMO scale, adapted from the ten item scale developed by Przybylski [24], and reduced to seven items. It measured the extent to which individuals worry about missing interesting or important events that others are experiencing. To enhance cultural suitability in the Chinese context, limited modifications were made to the wording, for example replacing the term SNS with a broader description of social media, since SNS is rarely recognized in China. The complete questionnaire contained 41 items in total including attention check items. The deleted items, reasons for deletion, original English scales, and the finalized scales are presented in the Appendix (Appendix 1–3).

Pilot administration and reliability and validity testing

Because modifications were made to the original scales, a pilot survey was first conducted to examine the reliability and validity of the revised instruments. Using an online platform, 200 questionnaires were distributed to Chinese university students through random sampling. A total of 191 valid responses were collected, yielding an effective response rate of 95.5%. The overall Cronbach’s alpha of the questionnaire was 0.925. The KMO value was 0.934. Bartlett’s test of sphericity was significant (p < .001). These results indicate high reliability and suitability for factor analysis.

To examine structural validity, exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted. In the EFA, the first order model of Self-Authenticity was tested first. The results showed a KMO value of 0.899, and Bartlett’s test of sphericity was significant (p < .001), indicating that the data were suitable for factor analysis. Principal axis factoring was applied with three factors fixed for extraction, because the scale is well established and the initial eigenvalue of the third factor was 0.97. Direct oblimin rotation was used and factor alignment was performed. The factor loadings of most items exceeded 0.50, with only one item loading at 0.492. The cumulative variance explained reached 71.7%, indicating that items were well aligned with their theoretical factors.

After excluding Self-Authenticity, a second order model was constructed using the remaining four variables. The KMO value was 0.931, and Bartlett’s test of sphericity was also significant (p < .001), indicating suitability for factor analysis. After removing items with double loadings or misaligned loadings, including two items from the Peer Pressure scale, namely “I’ve felt pressured to have sex, because a lot of people my own age have already had sex” and “I’ve felt pressured to get drunk at parties”, one item from the Social Media Addiction scale, namely “How often during the last year have you felt an urge to use social media more and more?”, and one item from the FoMO scale, namely “It is important that I understand my friends ‘in jokes’”, the same rotation method was applied. All remaining items showed aligned distributions, and all factor loadings exceeded 0.50. The cumulative variance explained was 68.4%, indicating a robust factor structure with clear factor membership. A SEM was then constructed for CFA, and all seven first order factors were tested simultaneously. The results indicated good model fit (χ²/df = 1.683, IFI = 0.922, TLI = 0.912, CFI = 0.921, RMSEA = 0.060), meeting criteria for excellent fit. Overall, the revised measurement instruments demonstrated good structural validity within the new sample and cultural context, and were suitable for subsequent analyses.

After the formal data collection, a complete second order latent variable model was constructed using the full sample, and CFA was conducted again. The results also indicated good model fit (χ²/df = 2.445, IFI = 0.937, TLI = 0.930, CFI = 0.937, RMSEA = 0.060). In addition, the convergent validity of each scale exceeded recommended thresholds in the formal survey data (see Appendix 4), Average Variance Extracted (AVE) > 0.50, Composite Reliability (CR) > 0.70, indicating high internal consistency and stability. Detailed information on CR and AVE is presented in Table 1.

Table 1.

Convergent Validity Information

	AVE	CR
Self-Authenticity	0.576	0.936
Peer Pressure	0.567	0.836
Social Media Addiction	0.657	0.905
Online Social Comparison	0.703	0.934
Fear of Missing Out	0.665	0.922
Over All	0.503	0.814

Ethical considerations and informed consent

This study strictly adhered to ethical standards, and all survey procedures were approved by the Ethics Review Board of Zhejiang University of Science and Technology. Informed consent was obtained from all participants prior to their participation in the study. At the beginning of each questionnaire, participants were informed about the purpose of the research, the voluntary nature of participation, the anonymity of responses, and the confidentiality of the data. Participation was entirely voluntary, and participants were free to withdraw at any time without penalty. The study involved no foreseeable risks, and all data collected were used solely for academic research purposes and not for any commercial or non-research activities. Throughout the research process, participants’ privacy was fully protected, and the anonymity, confidentiality, and security of all data were ensured.

Results

Questionnaire distribution and sample demographic information

This study used random sampling and distributed questionnaires to Chinese university students through an online survey platform (Credamo, available at https://www.credamo.cc). A total of 420 questionnaires were distributed, and 397 valid questionnaires were returned, yielding an effective response rate of 94.5%. Detailed demographic statistics are presented in Table 2.

Table 2.

Demographic Information Statistics Table

		Frequency	Percentage
Gender	Male	140	35.3%
	Female	257	64.7%
Age	18–20	47	11.8%
	21–23	182	45.8%
	24–26	72	18.1%
	> 27	96	24.1%
Education Level	Associate	21	5.3%
	Bachelor	308	77.6%
	Master	61	15.4%
	PhD	7	1.8%

Pearson correlation analysis

To examine linear relationships among variables, Pearson correlation analysis was conducted. The results indicated significant correlations among the five variables, including Self-Authenticity (SA), Peer Pressure (PP), Social Media Addiction (SMA), Online Social Comparison (OSC), and Fear of Missing Out (FoMO). Detailed results are reported in Table 3. These findings indicate that the data are suitable for further structural equation modeling.

Table 3.

Results of Correlation Analysis

	M ± SD	SA	PP	SMA	OSC	FoMO
SA	3.81±0.87	1
PP	2.19±0.94	−0.846**	1
SMA	2.76 ± 1.1	−0.710**	0.689**	1
OSC	3.04 ± 1.2	−0.570**	0.559**	0.650**	1
FoMO	2.37 ± 1.1	−0.761**	0.709**	0.781**	0.640**	1

Note: For p < .05, * * for p < .01, * * * for p < .001, the same applies below

Discriminant validity testing

To test discriminant validity among variables, CFA with competitive model comparisons was conducted. The hypothesized five factor baseline model was compared with three alternative models. The first was a four factor model A, in which Self-Authenticity and Peer Pressure were combined due to their relatively high correlation. The second was a four factor model B, in which Social Media Addiction and FoMO were combined. The third was a single factor model, in which all measurement indicators were loaded onto one latent factor.

The results showed that the five factor baseline model demonstrated the best fit indices, with CFI = 0.908, TLI = 0.900, RMSEA = 0.072, and SRMR = 0.048. In contrast, all alternative models showed poorer fit. The fit indices for four factor model A were CFI = 0.906, TLI = 0.898, RMSEA = 0.073, SRMR = 0.048. The fit indices for four factor model B were CFI = 0.888, TLI = 0.879, RMSEA = 0.080, SRMR = 0.050. The fit indices for the single factor model were CFI = 0.736, TLI = 0.717, RMSEA = 0.122, SRMR = 0.082. Chi square difference tests indicated that the baseline model was significantly superior to four factor model A, with Δχ² = 26.52, Δdf = 4, p < .001, to four factor model B, with Δχ² = 213.88, Δdf = 4, p < .001, and to the single factor model, with Δχ² = 1795.50, Δdf = 10, p < .001. These results indicate that although some variables are correlated, the latent constructs demonstrate good discriminant validity, which supports the five factor structure.

Common method bias testing

This study applied both the single factor CFA approach and the Unmeasured Latent Method Construct (ULMC) approach to examine potential common method bias. The single factor CFA approach requires constructing a model with one latent factor that loads on all measurement items. The results showed poor model fit, with χ²/df = 6.872, RMR = 0.131, GFI = 0.552, RFI = 0.685, RMSEA = 0.122. These results indicate that a single factor cannot adequately explain all measurement items, suggesting that serious common method bias is unlikely.

The ULMC approach was further applied by introducing an unmeasured latent method factor into the original SEM, and allowing all items to load on this factor. The results showed that model fit indices changed very little compared with the original model. Detailed changes in fit indices are reported in Table 4. Therefore, the influence of common method bias on the results can be considered negligible [62]. Taken together, the results from the single factor CFA and the ULMC approaches indicate that this study does not suffer from significant common method bias.

Table 4.

Detailed Changes in Fit Indices

	χ²/df	NFI	IFI	TLI	CFI	RMSEA
Original Model	2.973	0.919	0.945	0.937	0.944	0.071
After adding Latent Method Factor	3.045	0.926	0.949	0.934	0.949	0.072
ΔDifference	0.072	0.007	0.004	− 0.003	0.005	0.001

Multicollinearity test

Correlation analysis indicated significant high correlations among some variables. To ensure regression robustness and exclude potential multicollinearity, this study further calculated variance inflation factors (VIF) [63]. In R Studio, all continuous variables were mean centered. A chained mediation model was then specified and VIF tests were conducted. Results showed that all predictor VIF values were below 5 (Self-Authenticity = 3.98, Peer Pressure = 3.76, Social Media Addiction = 2.54, Online Social Comparison = 1.82). Although some independent variables exhibited relatively strong pairwise correlations, this does not necessarily indicate multicollinearity. Collinearity diagnostics showed that after removing shared variance, each variable retained sufficient unique information at the model level. These variables therefore remained mathematically distinguishable. They did not cause infinite variance inflation in parameter estimates or model distortion [56].

Moderation effects and direct effects testing

This study used R Studio to test moderation effects and direct effects in the model, with a focus on the specific predictive mechanisms of Self-Authenticity. Because the model was relatively complex, the full latent variable interaction model failed to converge. Therefore, this part of the analysis adopted a more robust observed variable path analysis. All five variables were mean centered. The model including Self-Authenticity is shown in Fig. 1, and detailed path information is reported in Table 5 (CFI = 0.948, SRMR = 0.038). Self-Authenticity showed significant moderation effects on three of the six paths in the model, with specific test results reported in Table 6. Simple slope plots for all moderated paths are presented in Figs. 2, 3 and 4. Although the ΔR² values appeared small, the median value of f² in psychology and sociology research is approximately 0.002 [64], and the f² values in this study were well above this benchmark. In addition, McClelland [65] argued that interaction terms typically explain only 1% to 3% of the variance in moderation analyses. Taken together, the moderation effects observed in this study have substantial practical significance. The moderated indirect effect of the model is presented in Table 7.

Fig. 1.

Schematic diagram of moderation model

Table 5.

Details of path analysis (Moderation model)

Peer Pressure → Social Media Addiction	β	S.E.	95%CI	P
	0.398	0.079	[0.273, 0.524]	***
Social Media Addiction → Online Social Comparison	0.373	0.072	[0.238, 0.508]	***
Peer Pressure → Online Social Comparison	0.160	0.084	[0.030, 0.291]	*
Social Media Addiction → FoMO	0.447	0.056	[0.335, 0.559]	***
Online Social Comparison → FoMO	0.198	0.038	[0.118, 0.277]	***
Peer Pressure → FoMO	−0.031	0.077	[−0.166, 0.103]	0.647
Self-Authenticity → FoMO	−0.190	0.089	[−0.335,−0.044]	**
Self-Authenticity → Online Social Comparison	−0.322	0.120	[−0.494,−0.151]	***
Self-Authenticity → Peer Pressure	−0.846	0.030	[−0.881,−0.812]	***
Self-Authenticity → Social Media Addiction	−0.561	0.082	[−0.683,−0.439]	***

Table 6.

Results of moderation effect test

Moderator: Self-Authenticity	β	S.E.	ΔR²	f2	P	95%CI
Peer Pressure → Social Media Addiction (A)	0.267	0.041	0.031	0.071	***	[0.186, 0.348]
Social Media Addiction → Online Social Comparison (B)	0.216	0.083	0.018	0.033	**	[0.069, 0.363]
Peer Pressure → FoMO (C)	−0.202	0.072	0.021	0.078	*	[-0.361, − 0.043]

Fig. 2.

Simple slope analysis plot (Path A)

Fig. 3.

Simple slope analysis plot (Path B)

Fig. 4.

Simple slope analysis plot (Path C)

Table 7.

Summary of conditional indirect effects

Mediation Path	Moderator Level (SA)	β	S.E.	P	95%CI
PP → SMA → FoMO	High (+ 1 SD)	0.301	0.060	***	[0.184, 0.419]
	Mean (0)	0.200	0.036	***	[0.129, 0.272]
	Low (−1 SD)	0.105	0.038	**	[0.031, 0.180]
PP → OSC → FoMO	High (+ 1 SD)	0.025	0.018	0.181	[−0.011, 0.061]
	Mean (0)	0.036	0.017	*	[0.002, 0.070]
	Low (−1 SD)	0.047	0.028	0.090	[−0.007, 0.102]
PP → SMA → OSC → FoMO	High (+ 1 SD)	0.052	0.026	*	[0.002, 0.103]
	Mean (0)	0.033	0.009	***	[0.015, 0.051]
	Low (−1 SD)	0.011	0.007	0.131	[−0.003, 0.024]

In addition, predictive paths were added to the model with Self-Authenticity as the independent variable and the remaining main variables as dependent variables. The data indicated that after controlling for chained mediation relationships and moderation effects among variables, Self-Authenticity still negatively predicted the other four main variables. After the inclusion of Self-Authenticity, the R² values of the four main variables were as follows: FoMO (R² = 0.732); Online Social Comparison (R² = 0.459); Peer Pressure(R² = 0.716); Social Media Addiction (R² = 0.572).

To examine why the direct predictive path from Peer Pressure to FoMO was no longer significant, and to clarify the specific mechanism of Self-Authenticity in this relationship, this study constructed four competing models and compared differences on the path from Peer Pressure to FoMO, with detailed results reported in Table 8. Model A was the basic chained mediation model, which did not consider individual differences in Self-Authenticity. Model B was the control variable model, in which Self-Authenticity was included as a control variable. Model C was the moderation model, in which Self-Authenticity was included as a moderator. Model D was the full effect model, in which Self-Authenticity was included as both a control variable and a moderator. The comparison results show that the loss of significance in Model B is mainly due to Self-Authenticity acting as a strong predictor and competing for explained variance, which indicates that the influence of Peer Pressure substantially weakens after the effect of authenticity is controlled. In Model C, the non-significant main effect reflects structural changes after the interaction term is introduced. Peer Pressure no longer has a simple direct effect on FoMO, but instead becomes a conditional effect that depends entirely on the level of individual authenticity. In addition, the continuous decrease in AIC values confirms that including Self-Authenticity significantly improves model fit and explanatory power.

Table 8.

Comparison result of competing models

	β	S.E.	95%CI	AIC	P
Model A	0.325	0.060	[0.214, 0.457]	2718.9	***
Model B	0.072	0.079	[−0.081, 0.214]	2639.2	0.358
Model C	0.044	0.069	[−0.098, 0.184]	2662.3	0.524
Model D	−0.035	0.078	[−0.193, 0.110]	2580.1	0.650

Given that non normality introduced by interaction terms can undermine traditional fit and testing methods based on normality assumptions [66], this study used likelihood ratio tests and information criteria, including AIC and BIC, to evaluate model performance. First, a baseline model without interaction terms was constructed. Because this model was saturated, fit indices reached optimal values. Subsequently, we constructed a model by incorporating Self-Authenticity into the baseline model as both a moderator and a control variable. The results indicated that the inclusion of this new variable significantly enhanced the model’s explanatory power. The AIC value decreased from 3222.5 to 3163.4, and the BIC value decreased from 3278.3 to 3243.1. The chi square difference test was highly significant (Δχ² = 266.8; Δdf = 6; p < .001), indicating that the introduction of the new moderator substantially improved the model.

Simple slope tests for the moderating effect

To examine the moderating role of self authenticity within the structural equation modeling framework, this study integrated latent variable based simple slope analyses into SEM. Based on the initial model estimation, conditional effects of the endogenous variables across different structural paths were decomposed at the mean level of self authenticity, as well as at high and low levels defined as plus and minus one standard deviation. During model specification, new parameters were defined in the model syntax, so that the slopes of Path A from peer pressure to social media addiction, Path B from social media addiction to online social comparison, and Path C from peer pressure to FoMO, were expressed as functions of the mean centered self authenticity score. A bias corrected bootstrap procedure with 5000 resamples was employed to obtain point estimates, standard errors, and 95% confidence intervals. The results reported in Table 9 illustrate the boundary conditions under which these psychological pathways vary.

Table 9.

Summary of simple slope estimates for hypothesized paths

	β	S.E.	P	95%CI
Path A (-1 SD)	0.260	0.071	***	[0.122, 0.405]
Path A (Mean)	0.478	0.079	***	[0.324, 0.635]
Path A (+ 1 SD)	0.696	0.100	***	[0.499, 0.898]
Path B (-1 SD)	0.176	0.120	0.141	[−0.061, 0.413]
Path B (Mean)	0.386	0.071	***	[0.243, 0.525]
Path B (+ 1 SD)	0.596	0.080	***	[0.422, 0.741]
Path C (-1 SD)	0.119	0.081	0.140	[−0.049, 0.273]
Path C (Mean)	−0.035	0.077	0.647	[−0.190, 0.119]
Path C (+ 1 SD)	−0.190	0.114	0.095	[−0.409, 0.041]

Because the preliminary analyses of Path C were non significant across all three conditional levels, a Johnson Neyman analysis was conducted to further delineate the precise moderating boundaries of self authenticity in the association between peer pressure and FoMO, while controlling for social media addiction and online social comparison. The Johnson Neyman technique identifies regions of significance across the full range of the moderator, thereby determining the critical threshold at which the moderating effect emerges. As shown in Fig. 5, a statistically significant interval from minus 0.7498 to 0.4689 was identified. When the mean centered self authenticity score fell outside this interval, the predictive slope of peer pressure on FoMO reached statistical significance at the p < .05 level. The corresponding simple slope estimates for this path are presented in Table 10.

Fig. 5.

Johnson-Neyman slope gradient for the moderation of self-authenticity

Table 10.

Summary of simple slopes for Path C based on Johnson-Neyman critical values

	β	S.E.	P	95%CI
SA =−1.0000	0.164	0.060	**	[0.047, 0.281]
SA =−0.7498	0.113	0.058	*	[0.000, 0.227]
SA = 0.0000	−0.039	0.061	0.520	[−0.158, 0.080]
SA = 0.4689	−0.134	0.068	*	[−0.269, 0.000]
SA = 1.0000	−0.242	0.081	**	[−0.401,−0.084]

Discussion

The buffering role of self-authenticity

The results of the mediation model explain the mechanism underlying FoMO among university students. Social Media Addiction and Online Social Comparison strongly mediated the relationship between Peer Pressure and FoMO, with the indirect effect comprising 58.8% of the total effect. Social Media Addiction was also found to be significantly positively associated with Online Social Comparison (β = 0.59) and participated in the influence pathway leading to FoMO. The construction of this mediation model strengthened prior conclusions regarding social media use and FoMO, and clarified the source of this influence, namely Social Media Addiction that may be induced by Peer Pressure (β = 0.79). Although the overall indirect effect of social media addiction on FoMO through online social comparison did not reach statistical significance in the bootstrap test (p = .112), the path analysis confirmed that the two core paths forming this mediation process were both statistically significant (β = 0.15, β = 0.11). The non-significant indirect effect is more likely due to the attenuation of effect size after multiplying the two path coefficients, which reduces its statistical power in multiple tests. The role of online social comparison as a psychological bridge linking addiction and anxiety remains theoretically valid, although its statistical contribution to explaining variance in the dependent variable may be smaller than that of other dominant pathways. We also identified significant direct predictive effects and moderation effects of Self-Authenticity in this model. After controlling for mediation and moderation effects, Self-Authenticity still exhibited highly significant negative predictive effects on all four variables in the model (β = − 0.190 / − 0.322 / − 0.846 / − 0.561). This indicates that individuals with higher levels of Self-Authenticity experience substantially lower levels of Peer Pressure, Social Media Addiction, Online Social Comparison, and FoMO.

In the path from Peer Pressure to FoMO, Self-Authenticity also played the expected buffering role (β = -0.202, p < .05). After the inclusion of Self-Authenticity, the direct path coefficient from Peer Pressure to FoMO changed from significant (β = 0.334, p < .001) to nonsignificant (β = -0.031, p = .647). In the model that included moderation effects, the influence of Peer Pressure on FoMO shifted from an average effect describing the full sample to a conditional main effect constrained by the level of the moderator. This coefficient represents the effect of Peer Pressure on FoMO when Self-Authenticity is at its mean level. This change suggests that the direct association between peer pressure and FoMO may be moderated by individuals’ levels of self-authenticity. In combination with the significantly negative interaction term, the findings indicate that self-authenticity exhibits a buffering moderating pattern in the relationship between peer pressure and FoMO. Based on the S-O-R framework, peer pressure can be conceptualized as an external social stimulus, whereas FoMO represents a psychological outcome associated with this stimulus. Self-authenticity, as an individual-level conditional variable, may moderate the strength of the association between external stimuli and psychological outcomes.

The simple slope analysis further demonstrated that the positive association between peer pressure and FoMO gradually weakened as the level of self-authenticity increased. This relationship was primarily significant among individuals with low self-authenticity, whereas it became non-significant at average or high levels of self-authenticity. Consistent with the findings reported above, the direct path coefficient from peer pressure to FoMO also became non-significant after self-authenticity was introduced as a control variable. These results suggest that, within the present model, the statistical predictive effect of peer pressure on FoMO is moderated to some extent by self-authenticity. Although this study provides a theoretical explanation for the non-significant direct path from Peer Pressure to FoMO after introducing Self-Authenticity, we also acknowledge that this result may partly reflect statistical characteristics of multivariate analysis. Given the strong negative correlation between Self-Authenticity and perceived Peer Pressure, the change in path significance may be due to shared variance and possible measurement overlap between the constructs. This statistical effect of variance suppression or redistribution may cause a strong predictor to mask the significance of other related variables.

The amplifying role of self-authenticity

When the dependent variables were social media-related constructs, including social media addiction and online social comparison, self-authenticity demonstrated significant positive moderating effects on these pathways. Higher levels of self-authenticity strengthened the effects of other variables on social media-related outcomes (β = 0.267, β = 0.216). These findings indicate that the statistical associations among the relevant variables became stronger at higher levels of self-authenticity. This pattern was not fully consistent with the original expectations. Therefore, it should be interpreted as an unexpected exploratory finding rather than as evidence of a stable psychological mechanism. Based on the S-O-R framework, this result may reflect differences in how individuals respond to external stimuli [67]. Individuals with higher self-authenticity generally place greater emphasis on consistency in self-expression and the integrity of their self-image. Consequently, when confronted with peer pressure or highly immersive social media environments, they may be more inclined to maintain their sense of self-identity through online expression, interaction, and social feedback [68]. In addition, highly authentic individuals may place stronger importance on preserving a coherent self-concept in social interactions, which could increase their engagement in online self-presentation and feedback-seeking behaviors [69]. The positive moderating effects of self-authenticity on several social media-related pathways may therefore reflect the complexity of authenticity within digital social contexts.

In the pathway from peer pressure to social media addiction, self-authenticity showed a significant positive moderating effect (β = 0.267, p < .001). This finding indicates that the statistical association between peer pressure and social media addiction became stronger at higher levels of self-authenticity. Within the S-O-R framework, peer pressure can be conceptualized as an external social stimulus, whereas social media addiction represents a behavioral response within digital social environments. Self-authenticity, as an individual-level conditional variable, may alter the way university students respond to peer pressure. Individuals with higher self-authenticity may place greater value on consistent self-expression and the integrity of their personal identity. As a result, when experiencing peer pressure, they may be more likely to engage in self-presentation, interaction, and identity expression through social media platforms. The positive moderating effect of self-authenticity in this pathway may therefore reflect a stronger tendency for highly authentic individuals to participate actively in social media contexts. In the pathway from social media addiction to online social comparison, self-authenticity also demonstrated a significant positive moderating effect (β = 0.216, p < .01). At higher levels of self-authenticity, the statistical association between social media addiction and online social comparison became stronger [37]. When highly engaged with social media, individuals with greater self-authenticity may become more attentive to discrepancies between their own online self-image and the online presentations of others. This heightened attention may be associated with more frequent online social comparison processes.

From the perspective of the overall theoretical framework, the positive moderating effects of self-authenticity on social media-related pathways suggest that its role may not follow a purely protective pattern. Based on the S-O-R framework, self-authenticity may function as an individual-level conditional variable that moderates the strength of associations among external stimuli, social media-related behavioral responses, and cognitive responses. On the one hand, self-authenticity was significantly associated with lower levels of peer pressure, social media addiction, online social comparison, and FoMO. On the other hand, it also appeared to exhibit more complex conditional effects within certain social media-related pathways. It should be emphasized that the positive moderating effects of self-authenticity observed in several social media-related pathways should be interpreted cautiously. In addition to potentially reflecting the complexity of authenticity within digital social contexts, these findings may also be influenced by statistical error, shared variance among variables, suppression effects, or overlap in measurement content. Therefore, the present study does not interpret these findings as confirmed psychological mechanisms. Instead, they should be regarded as exploratory association patterns that require further validation in future research.

The overall predictive direction of self-authenticity

From a quantitative perspective, the protective effects of Self-Authenticity in the model were substantially stronger than its compensatory costs. Although positive amplification coefficients of 0.267 and 0.216 appeared in specific moderated paths, the mean negative predictive effect of Self-Authenticity on the remaining variables in the overall model still reached − 0.480. This indicates that Self-Authenticity markedly reduced the baseline levels of Peer Pressure, Social Media Addiction, Online Social Comparison, and FoMO. In statistical terms, individuals with high Self-Authenticity started from much lower initial levels than those with low authenticity. The role of self-authenticity does not appear to follow a purely protective pattern and may instead demonstrate a certain degree of contextual dependence. Based on the S-O-R framework, self-authenticity can be understood as an individual-level conditional variable that may be associated with the ways university students respond to external peer pressure and social media environments.

From the perspective of the total effect, the present study more cautiously conceptualizes self-authenticity as a potential psychological resource associated with lower levels of risk. Higher self-authenticity was generally associated with lower levels of peer pressure, social media addiction, online social comparison, and FoMO. However, the positive moderating effects observed in certain pathways suggest that the role of this variable within social media contexts may be more complex. Because the present study employed a cross-sectional design and several variables demonstrated relatively high intercorrelations, these findings should be further validated through future longitudinal studies, experimental research, or multi-source data.

Conclusion

Based on the S-O-R framework, the present study constructed a moderated chain mediation model to examine the structural associations among peer pressure, social media addiction, online social comparison, self-authenticity, and FoMO among university students. Within this framework, peer pressure was conceptualized as an external social stimulus. Social media addiction and online social comparison were regarded as behavioral and cognitive responses within social media environments, whereas FoMO was understood as a psychological outcome associated with these processes. The findings demonstrated a significant positive association between peer pressure and FoMO. Social media addiction and online social comparison also exhibited indirect association patterns between these variables. Among the mediating pathways, the pathway involving social media addiction showed a relatively stronger statistical effect, whereas the independent indirect effect of online social comparison was comparatively weaker. These results suggest that the mediating pathways did not contribute equally within the model.

Self-authenticity demonstrated a relatively complex pattern of effects in the present model. Self-authenticity significantly and negatively predicted peer pressure, social media addiction, online social comparison, and FoMO. In addition, self-authenticity exhibited a negative moderating effect in the pathway from peer pressure to FoMO. These findings suggest that self-authenticity may represent an important individual-level factor for understanding differences in FoMO under conditions of peer pressure among university students. In several pathways related to social media use, self-authenticity also demonstrated positive moderating effects that were not fully consistent with the original expectations. The close associations among peer pressure, social media addiction, online social comparison, self-authenticity, and FoMO further indicate that the S-O-R framework may provide a relatively clear analytical framework for understanding the relationships among external pressure, individual responses, and psychological outcomes within social media environments. However, because the present study relied on cross-sectional data, all pathway results should be interpreted as statistical associations rather than evidence of causal mechanisms.

Previous studies have similarly suggested that social media addiction, FoMO, anxiety, and social comparison may involve multiple explanatory pathways operating in different directions [70–72]. In addition, the present sample primarily consisted of Chinese university students. Therefore, the findings cannot be directly generalized to other cultural contexts or age groups. Future studies may employ longitudinal designs, experimental approaches, or multi-source data to further examine the temporal ordering, potential causal directions, and stability of the proposed model.

Abbreviations

SA

Self Authenticity

PP

Peer Pressure

SMA

Social Media Addiction

OSC

Online Social Comparison

FoMO

Fear of Missing Out

Authors’ contributions

HL: Conceptualization, Methodology, Data curation, Formal analysis, Writing - original draft, Writing - review & editing.XJ: Conceptualization, Resources, Supervision, Funding acquisition, Visualization, Writing - review & editing.

Funding

This research did not receive funding.

Data availability

The data used in this study are available from the author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study strictly adhered to ethical standards, and all survey procedures were approved by the ethics review board of Zhejiang University of Science and Technology. At the beginning of each questionnaire, participants were provided with information about the purpose of the research, the voluntary nature of participation, the anonymity of responses, and the principle of data confidentiality. Participation was entirely voluntary, and respondents completed the questionnaire based on informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.