Narcissists Facing Social Media Feedback: Activated Emotions and Subsequent Aggressive and Prosocial Behaviors

ABSTRACT

The appraisal‐oriented Status Pursuit in Narcissism (SPIN) model suggests that narcissists show rivalry‐oriented behavior when they lack social affirmation, but behave prosocially when admired. Aiming to integrate emotional responses into this perspective, this study investigated whether narcissistic rage and pride accounted for narcissists' direct and displaced aggression and prosociality following social feedback. Participants (N = 371) experienced either Low Social Validation or High Social Validation in a simulated social media context. After reporting their emotions, they sent “Dislikes” and “Likes” to either the Same Peers (i.e., direct responses) who had given social feedback or to Different Peers (i.e., displaced responses). Results showed that narcissistic rivalry predicted more narcissistic rage following Low Social Validation, which subsequently predicted more “Dislikes” and fewer “Likes” toward both Same Peers and Different Peers. Narcissistic admiration predicted both more “Dislikes” and more “Likes” via pride across social feedback and peer group conditions. This study suggests that narcissistic rage stemming from narcissistic rivalry accounts for aggression following social failures, while pride stemming from narcissistic admiration accounts for prosociality regardless of social feedback experiences.

1. Introduction

The dynamic nature of social media exposes individuals to frequent shifts in their social standing. While they feel socially affirmed when receiving “Likes”, they may also experience deficits in social validation if the amount of positive attention falls short of expectations or feel socially invalidated when receiving “Dislikes” or negative comments. Narcissistic individuals are particularly vigilant and tend to show heightened emotional and behavioral reactivity to status‐related cues (Grapsas et al. 2022; Kroencke et al. 2023). Narcissism is characterized by a grandiose yet fragile self‐view, along with a motivation to pursue social status (Grapsas et al. 2020). Self‐regulatory processing theories of narcissism (Morf and Rhodewalt 2001) propose that both cognitive and affective processes contribute to narcissistic individuals' subsequent interpersonal behaviors following social feedback. Previous studies have directly examined the mediating role of cognitive factors, such as attention‐seeking, in the associations between narcissism and interpersonal behaviors (Hawk et al. 2019; Wang et al. 2023). However, prior experimental research addressing affective components of the self‐regulatory processes has primarily focused on emotional reactivity itself without explicitly examining the connection between emotional reactivity and subsequent behaviors. This study focused on narcissistic individuals' aggressive and prosocial responses to social media feedback, examining whether and how emotions mediate the relations between narcissism and these behaviors.

1.1. Narcissism and Behavioral Responses to Social Feedback

Self‐regulatory processing theories characterize narcissistic individuals' aggressive and prosocial behaviors as outcomes of the interplay between environmental factors and psychological processes (Morf and Rhodewalt 2001). Aggressive behavior refers to “behavior directed toward the goal of harming or injuring another living being who is motivated to avoid such treatment” (Baron and Richardson 1994, p. 7), while prosocial behavior refers to “actions for the benefit of others” (Penner et al. 2005, p. 366). Many studies have emphasized narcissistic individuals' problematic responses to social frustration, finding that insufficient social validation can trigger aggression (for a review, see Kjærvik and Bushman 2021). There are far fewer studies examining whether and how their prosocial behaviors change according to social feedback. Extending prior self‐regulatory processing theories (Morf and Rhodewalt 2001), the Status Pursuit in Narcissism (SPIN) model suggests that narcissistic reactivity can also manifest as prosocial behaviors in response to social validation.

The SPIN model proposes that narcissistic individuals consciously appraise the extent to which the environment has granted (or potentially will grant) social validation based on the feedback that they have received, and strategically respond in either a rivalry‐oriented or admiration‐oriented manner in both ongoing and new social interactions (Grapsas et al. 2020). When narcissistic individuals do not receive expected positive feedback, they view evaluators and competitors as hindrances to their status pursuit, resorting to aggression to undermine others and protect their own social standing (i.e., the rivalry‐oriented pathway). If narcissistic individuals obtain the validation to which they feel entitled, however, they might respond in a socially desirable manner and engage in prosocial behaviors (i.e., the admiration‐oriented pathway). These prosocial behaviors might not solely be driven by a desire to reciprocate positive feedback, but also function as a strategic means to enhance their reputation and extend their influence beyond the initial interaction to new social contexts and a broader audience. The SPIN model explains why narcissistic individuals exhibit consistent patterns of status‐seeking behaviors over time (i.e., at the trait level), while also accounting for situational variations in these behaviors. On the one hand, narcissistic individuals exhibit a consistent tendency to seek status through admiration‐oriented and rivalry‐oriented behaviors, shaping trait‐level aggression and prosociality. On the other hand, whether these chronically accessible admiration‐oriented and rivalry‐oriented behaviors are activated in any given moment depends on available social cues and reinforcements.

Earlier research has divided narcissism ¹ into the facets of narcissistic rivalry and narcissistic admiration, with different dimensions potentially associated with different self‐regulatory patterns (Back et al. 2013). Narcissistic rivalry reflects maladaptive aspects of narcissism, such as antagonism, disagreeableness, coldness, and vulnerability, while narcissistic admiration embodies agentic traits, including extraversion, assertiveness, and self‐esteem stability. Accordingly, this study examined the SPIN model using a multidimensional narcissism scale and tested the extent to which narcissistic rivalry and narcissistic admiration respectively account for aggressive and prosocial responses to social feedback.

1.2. Emotional Reactivity Accounts of Narcissists' Aggressive and Prosocial Behaviors

The SPIN model primarily focuses on narcissists' cognitive appraisals of social status, without sufficiently addressing the emotional processes proposed by broader self‐regulatory processing theories. Recent perspectives on affective contingencies of narcissism propose that narcissistic individuals' emotional systems are especially reactive to status‐related social feedback (Grapsas et al. 2022; Kroencke et al. 2023). This emotional reactivity has a motivational function, leading to behaviors aimed at either self‐protection from unpleasant experiences or enhancing pleasant experiences (Dufner et al. 2024; Grapsas et al. 2022). Following suggestions to measure narcissistic individuals' status‐specific emotional reactions (Dufner et al. 2024), this study explored whether narcissistic rage and pride can account for the rivalry‐ and admiration‐oriented behavioral pathways outlined in the SPIN model.

Narcissistic rage is an emotional response to narcissistic injury that involves anger and shame (Kohut 1972; Krizan and Johar 2015). The absence of positive feedback can trigger negative self‐appraisals and elicit shame‐related feelings, including inferiority, humiliation, and embarrassment. Simultaneously, narcissistic individuals experience anger towards others as a means of shifting attention from their painful recognition of a diminished self (Kohut 1972; Thomaes et al. 2008). A combination of shame and anger thus reflects narcissistic individuals' emotional responses when self‐promotional motives are frustrated (Krizan and Johar 2015). As an approach‐oriented emotion, narcissistic rage often drives individuals to defend themselves in aversive ways. Cross‐sectional studies have found that trait hostility and rage proneness mediated the association between narcissism and trait reactive aggression (Krizan and Johar 2015; Li 2023). An experimental study found that narcissism predicted the co‐occurrence of hostile emotions and aggressive behaviors following physical provocation, though it did not investigate mediation (Krizan and Johar 2015). These findings offer preliminary support that narcissistic rage might at least partially account for the rivalry‐oriented pathway in the SPIN model.

Although prior research has emphasized negative aspects of narcissistic emotional reactivity, individuals might also experience heightened positive emotionality following social affirmation. Narcissistic individuals tend to attribute their success to their own personality and competence, which can boost grandiose self‐views and make them susceptible to experiencing pride (Tracy et al. 2011). Prior research using Experience Sampling Methods (ESM) found that individuals scoring higher in narcissistic admiration experience more pride and joy when they perceive respect from others (Kroencke et al. 2023). Pride stemming from ego‐boosting experiences can foster a favorable perspective toward social interactions and potentially inspire prosocial behaviors (Tracy and Robins 2007). Therefore, the admiration‐oriented pathway in the SPIN model might be a pride‐driven process.

To our knowledge, this is the first experimental study to directly examine the mediating role of emotions in associations between narcissism and behavioral responses to social feedback. Doing so can connect narcissists' intrapersonal affective processes and interpersonal behaviors, thereby elaborating on whether emotional elements can be integrated into the SPIN model.

1.3. Study Overview

This experimental research manipulated the valance of social feedback that participants received, to address the potential causal effects of social feedback on narcissistic individuals' emotional and behavioral responses. We expected to find that narcissistic rivalry would be positively associated with rage‐driven aggression in low social validation situations, while narcissistic admiration would be positively associated with pride‐driven prosocial behavior in high social validation situations. Aggressive and prosocial behavior were respectively operationalized in this study as the number of “Dislikes” and the number of “Likes” participants gave to their online peers (Lutz and Schneider 2020). Prior research has interpreted giving “Dislikes” as a form of direct negative attention, undermining the recipients' feelings of social regard (Lutz and Schneider 2020). In contrast, qualitative research has demonstrated that social media users perceive giving “Likes” as carrying prosocial meaning and as a tool for expressing social support, positively reinforcing individuals' sense of social value (Sumner et al. 2018). Additionally, this study experimentally manipulated the peer groups participants encountered after receiving social feedback, aiming to examine whether individuals' behavioral reactions to those who provided social feedback would also be displaced onto an innocent third party. We compared these effects on direct and displaced behaviors without a priori hypothesis, since findings regarding displaced responses are relatively limited and inconclusive.

2. Methods

2.1. Participants

Our goal was to obtain 0.80 power to detect a small effect size of 0.05 for the three‐way interaction effect between narcissism, social feedback, and peer group, at the standard of α = 0.05 error probability, which required at least N = 347. To ensure the acquisition of adequate high‐quality data, 476 undergraduate students from 30 Chinese provinces were recruited via Sojump (www.wjx.com), a professional data collection company in China (Del Ponte et al. 2024). We excluded 89 (18.3%) participants answering incorrectly to either or both manipulation check questions (see Supporting Information S1: Materials) and 16 cases containing outliers (defined as scores of three or more box lengths outside of the Box‐plot). The final sample consisted of 371 participants (53.6% female; M _age = 20.39, SD = 1.22). All participants reported being active on at least two (M = 6.57, SD = 2.21) social media platforms, such as WeChat (98.9%), QQ (93.3%), TikTok (80.9%), Bilibili (74.1%), Weibo (64.2%), and RED (63.6%). On average, participants reported spending 4.21 (SD = 2.22) hours daily on social media.

2.2. Procedure

This study was approved by the Survey and Behavioral Research Ethics Committee of The Chinese University of Hong Kong (SBRE‐22‐0884). An online link was sent to participants by Sojump, where participants provided informed consent and completed all tasks. Demographic information and narcissism were measured before the online experiment.

The experimental procedure was adapted from the Chinese version (Wang et al. 2020) of the Social Media Ostracism Paradigm (Wolf et al. 2015). Participants engaged in two rounds of simulated social media interactions with 11 online peers (actually fictitious profiles) through sending and receiving “Likes” and “Dislikes” on their posts. Participants were informed at the beginning of the experiment that “likes will increase the receiver's popularity,” and “dislikes will decrease the receiver's popularity”. This instruction brought participants' sending of “Likes” and “Dislikes” in line with common definitions of prosocial and aggressive behaviors, respectively (Baron and Richardson 1994; Penner et al. 2005). Additionally, the fictitious profiles described mundane, neutral activities (e.g., enjoying hotpot), which minimized the likelihood that “Dislikes” were motivated by factors unrelated to aggression, such as disagreement or moral disapproval of inappropriate content. Supporting Information S1: Materials provide details on the experimental design rationale and complete experimental materials.

This study used a 2 (social feedback conditions: Low vs. High Social Validation) × 2 (peer group conditions: Same vs. Different Peers) between‐participants design. In the first online interaction round, participants were randomly assigned either to the High Social Validation condition where they received nine “Likes”, putting them into second place, or to the Low Social Validation condition where they received two “Likes” (i.e., nine peers not sending “Likes” to them), putting them in eleventh place (i.e., second to last). In the second interaction round, participants were randomly assigned to interact with the same peers from the first round (i.e., Same Peers condition) or to interact with 11 new peers who were not from the first‐round group (i.e., Different Peers condition). The number of “Dislikes” and “Likes” participants gave in this round indicated aggressive and prosocial responses to social feedback (Lutz and Schneider 2020).

2.3. Measures

All measures were translated into Chinese and back‐translated by two bilingual speakers for accuracy checking. The Supporting Information S1: Materials provide complete scales and CFA results.

2.3.1. Narcissism

Narcissism was measured by the Narcissistic Admiration and Rivalry Questionnaire (Back et al. 2013), which has been translated and used with Chinese samples (Li et al. 2022). Participants responded to nine items for narcissistic admiration (α = 0.82) and nine items for narcissistic rivalry (α = 0.85) on a 6‐point scale from 1 = not agree at all to 6 = agree completely.

2.3.2. Affect

Narcissistic rage and pride scales were used to measure participants' emotional reactivity toward social feedback. Narcissistic rage (Kohut 1972; Krizan and Johar 2015; α = 0.93) was measured by six anger items (Twenge and Campbell 2003; α = 0.91) and seven shame items (Liu and Giner‐Sorolla 2023; α = 0.91). The pride scale contained seven items (Tracy and Robins 2007; α = 0.96). Participants used a 5‐point scale (from 1 = not at all to 5 = extremely) to report the extent to which these items aligned with their feelings upon seeing the number of “Likes” and the popularity rank. Mean scores of anger, shame, and pride were calculated separately. A latent construct of narcissistic rage was created with shame and anger as its two observed variables.

2.3.3. Aggressive and Prosocial Behavior

Aggressive and prosocial behaviors were measured by the number of “Dislikes” (i.e., aggression) and the number of “Likes” (i.e., prosociality) that participants gave to their online peers after seeing the “Likes” they had received (Lutz and Schneider 2020). Supporting Information S1: Materials provide additional analyses to examine the notion that sending “Dislikes” and “Likes” are valid proxies for aggressive and prosocial behavior, respectively.

2.4. Analytic Strategy

A multigroup analysis based on the moderated mediation model was examined in Mplus v.8.3 (Muthén and Muthén 2017), with the dummy variable regarding peer group manipulation (0 = Same Peers condition, 1 = Different Peers condition) input as the grouping variable. Narcissism variables, social feedback, and affect variables were standardized using z‐scores. The numbers of “Dislikes” and “Likes” were specified as count variables. The model was controlled for gender. ² Significance of direct and indirect effects were evaluated using 95% CI based on 5000 bootstrap resamples, given the skewed distribution of some variables. Since the model included count variables, AIC and BIC were used to evaluate the model fit, with lower AIC and lower BIC being preferred. ΔAIC and ΔBIC of 7–10 indicate sufficient support for the model with the lower AIC or BIC, while values greater than 10 indicate strong support (Anderson and Burnham 2004). See the Supporting Information S1: Materials for further details about the analytic plan.

3. Result

A series of 2 × 2 factorial ANOVAs suggest a successful random assignment of participants and a successful manipulation of social feedback (see Supporting Information S1: Materials). Table 1 shows descriptive statistics and correlations of variables.

Table 1.

Means, standard deviations, range, and bivariate correlations (N = 371).

	1	2	3	4	5	6	7
1. NARQ_NR	—
2. NARQ_NA	−0.00	—
3. Anger	0.24***	0.08	—
4. Shame	0.21***	−0.07	0.68***	—
5. Pride	−0.08	0.25***	−0.23***	−0.40***	—
6. Dislikes	0.18***	0.14**	0.32***	0.31***	−0.04	—
7. Likes	−0.16**	0.04	−0.18***	0.22***	0.21***	−0.31***	—
Mean	3.13	3.82	1.38	1.68	2.79	0.81	6.55
SD	0.95	0.79	0.61	0.81	1.25	1.43	2.96
Range	1.00–5.56	1.56–5.67	1.00–4.17	1.00–4.86	1.00–5.00	0–8	0–11

Note: *p < 0.05, **p < 0.01, ***p < 0.001.

Abbreviations: NARQ_NA, narcissistic admiration measured by NARQ; NARQ_NR, narcissistic rivalry measured by NARQ.

The fully constrained model (AIC = 12865.071, BIC = 13150.954) showed better model fit than the freely estimated model (AIC = 12873.782, BIC = 13331.978), ΔAIC = 8.7111, ΔBIC = 181.024, suggesting there were no differences between participants assigned to the Same Peers condition and those assigned to the Different Peers condition in terms of emotions and the numbers of “Likes” or “Dislikes” given. Therefore, we constrained all paths to be equal across conditions, see Figure 1 and Table 2 for model results. Unstandardized estimates are reported.

Figure 1.

Fully constrained model examining narcissism, narcissistic rage, pride, and the number of “Dislikes” and “Likes”, moderated by social feedback, controlling for gender (N = 371). Note: “Social Feedback” is the dummy‐coded variable for the manipulation regarding how many “Likes” that participants received from peers (0 = Low Social Validation condition, n = 192; 1 = High Social Validation condition, n = 179). All paths were constrained across the same peers condition and the different peers condition. This figure only shows significant predicting paths that were evaluated by 95% CI. Unstandardized estimates are reported. The exogenous variables were specified to be correlated, and only narcissistic rivalry and social feedback was significantly correlated (b = 0.099, 95% CI [−0.197, −0.002]). The two endogenous variables (i.e., the number of “Likes” and the number of “Dislikes”) were not modeled to be correlated, as Mplus does not allow covariances between count variables. The model was controlled for gender and the two endogenous variables were controlled for the number of “Likes” participants gave before the experimental manipulation.

Table 2.

Direct and indirect effects on narcissistic rage, pride, and the numbers of “Dislikes” and “Likes” (N = 371).

Direct and indirect effect	Narcissistic rage		Pride		Number of “Dislikes”		Number of “Likes”
	B	95% CI	B	95% CI	B	95% CI	B	95% CI
Control variables
Gender	−0.110	[−0.195, −0.028]	0.058	[−0.015, 0.130]	−0.215	[−0.432, 0.151]	0.029	[−0.015, 0.070]
“Likes” before manipulation					−0.073	[−0.135, −0.001]	0.103	[0.084, 0.120]
Direct effects
Social feedback (SF)	−0.401	[−0.495, −0.315]	0.677	[0.603, 0.749]	−0.017	[−0.373, 1.113]	−0.033	[−0.131, 0.027]
Narcissistic rivalry (NR)	0.186	[0.097, 0.276]	0.001	[−0.079, 0.077]	0.041	[−0.499, 0.296]	−0.008	[−0.049, 0.041]
Narcissistic admiration (NA)	−0.012	[−0.102, 0.074]	0.227	[0.157, 0.295]	0.176	[−0.046, 0.436]	−0.008	[−0.048, 0.033]
NR × SF	−0.170	[−0.257, −0.083]	−0.043	[−0.119, 0.034]	−0.012	[−0.273, 0.483]	−0.013	[−0.060, 0.027]
NA × SF	−0.049	[−0.135, 0.033]	0.006	[−0.063, 0.072]	0.026	[−0.186, 0.272]	0.020	[−0.015, 0.056]
Narcissistic rage					0.745	[0.284, 4.447]	−0.118	[−0.388, −0.034]
Pride					0.222	[−0.099, 0.629]	0.064	[0.007, 0.126]
Indirect effects
SF → narcissistic rage →					−0.299	[−1.841, −0.103]	0.047	[0.013, 0.159]
SF → pride →					0.151	[−0.069, 0.424]	0.043	[0.005, 0.086]
NR → narcissistic rage →					0.139	[0.049, 0.864]	−0.022	[−0.078, −0.006]
NR → pride →					0.000	[−0.029, 0.030]	0.000	[−0.005, 0.006]
NA → narcissistic rage →					−0.009	[−0.197, 0.086]	0.001	[−0.010, 0.022]
NA → pride →					0.050	[−0.019, 0.150]	0.014	[0.002, 0.031]
NR × SF → narcissistic rage →					−0.127	[−0.780, −0.044]	0.020	[0.005, 0.070]
NR × SF → pride →					−0.010	[−0.060, 0.008]	−0.003	[−0.011, 0.001]
NA × SF → narcissistic rage →					−0.036	[−0.305, 0.030]	0.006	[−0.004, 0.031]
NA × SF → pride →					0.001	[−0.018, 0.034]	0.000	[−0.004, 0.006]
Conditional indirect effects
NR → narcissistic rage →
Low social validation					0.261	[0.094, 1.605]	−0.041	[−0.145, −0.012]
High social validation					0.007	[−0.061, 0.146]	−0.001	[−0.016, 0.007]

Note: “Social Feedback” is coded as 0 = Low Social Validation condition, 1 = High Social Validation condition. Gender is coded as 0 = male, 1 = female. Unstandardized estimates are reported, with significant effects in bold. “Likes” before Manipulation refers to the number of “Likes” that participants gave to online fictitious peers in the first round, before receiving any social feedback.

Social feedback predicted narcissistic rage ³ (b = −0.401, 95% CI = [−0.495, −0.315]), with the Low Social Validation condition reporting more narcissistic rage than the High Social Validation condition. Narcissistic rage, in turn, predicted more “Dislikes” (b = 0.745, 95% CI = [0.284, 4.447]) and fewer “Likes” (b = −0.118, 95% CI = [−0.388, −0.034]). Indirect paths from social feedback to “Dislikes” and “Likes” via narcissistic rage were significant (b = −0.299, 95% CI = [−1.841, −0.103]; b = 0.047, 95% CI = [0.013, 0.159], respectively). Social feedback predicted pride (b = 0.677, 95% CI = [0.603, 0.749]), with the High Social Validation condition reporting more pride than the Low Social Validation condition. Pride, in turn, predicted more “Likes” (b = 0.064, 95% CI = [0.007, 0.126]), but did not predict “Dislikes” (b = 0.222, 95% CI = [−0.099, 0.629]). The indirect path from social feedback to “Likes” via pride was significant (b = 0.043, 95% CI = [0.005, 0.086]).

Narcissistic rivalry positively predicted narcissistic rage (b = 0.186, 95% CI = [0.097, 0.276]), which in turn predicted more “Dislikes” and fewer “Likes”. Indirect paths from narcissistic rivalry to “Dislikes” and “Likes” via narcissistic rage were significant (b = 0.139, 95% CI = [0.049, 0.864]; b = −0.022, 95% CI = [−0.078, −0.006], respectively). The interaction between narcissistic rivalry and social feedback predicted narcissistic rage (b = −0.170, 95% CI = [−0.257, −0.083]), but not pride (b = −0.043, 95% CI = [−0.119, 0.034]). Simple slopes analysis (Figure 2) showed that narcissistic rivalry positively predicted narcissistic rage in the Low Social Validation condition, but not the High Social Validation condition. Indirect paths from narcissistic rivalry to “Dislikes” (b = 0.261, 95% CI = [0.094, 1.605]) and “Likes” (b = −0.041, 95% CI = [−0.145, −0.012]) via narcissistic rage were significant in the Low Social Validation condition.

Figure 2.

Narcissistic rage and narcissistic rivalry in different social feedback conditions (N = 371).

Narcissistic admiration positively predicted pride (b = 0.227, 95% CI = [0.157, 0.295]), which in turn predicted more “Likes”. The indirect path from narcissistic admiration to “Likes” via pride was significant (b = 0.014, 95% CI = [0.002, 0.031]). The interaction between narcissistic admiration and social feedback did not predict narcissistic rage (b = −0.049, 95% CI = [−0.135, 0.033]), pride (b = 0.006, 95% CI = [−0.063, 0.072]), “Dislikes” (b = 0.026, 95% CI = [−0.186, 0.272]), or “Likes” (b = 0.020, 95% CI = [−0.015, 0.056]).

4. Discussion

Situational constraints or affordances can trigger narcissistic individuals' self‐regulatory processes aimed at reducing feelings of inadequacy or boosting feelings of superiority. Focusing primarily on narcissistic individuals' appraisal of social status, the SPIN model suggests that the behavioral outcomes of these self‐regulatory processes can manifest as rivalry‐oriented aggression in response to status frustration, or admiration‐oriented prosociality in response to social validation (Grapsas et al. 2020). However, the SPIN model does not adequately address connections between narcissistic individuals' emotions and their interpersonal behaviors, proposed by prior self‐regulatory processing perspectives (Morf and Rhodewalt 2001). To our knowledge, this experimental study is the first to simultaneously include narcissistic individuals' emotions and behaviors in a single model and to directly whether emotions mediate associations between narcissism and behavioral responses to social evaluation. Integrating emotional elements into the SPIN model (Grapsas et al. 2020), the present research found that narcissistic rage accounted for associations between narcissistic rivalry and both direct and displaced aggressive behavior following social failure (i.e., the rivalry‐oriented pathway). Additionally, pride stemming from narcissistic admiration predicted more prosocial behavior (i.e., the admiration‐oriented pathway), regardless of the valence of social feedback.

4.1. Rivalry‐Oriented Pathway and Narcissistic Rage

The SPIN model suggests that narcissistic individuals engage in rivalry‐oriented aggression when they do not receive the positive feedback they expect. Based on the self‐regulatory processing theories (Morf and Rhodewalt 2001) and perspectives on narcissistic rage (Kohut 1972; Krizan and Johar 2015), we expected that narcissistic rage triggered by the absence of social validation would contribute to rivalry‐oriented aggressive behavior. Our hypotheses were supported; although both narcissistic rivalry and narcissistic admiration are associated with dominance‐based strategies to gain status (Zeigler‐Hill et al. 2019), the fact that social feedback only interacted with narcissistic rivalry to predict narcissistic rage and subsequent “Dislikes” suggests that narcissistic rivalry is more likely than narcissistic admiration to predict narcissistic rage and rage‐driven aggression following social invalidation. This finding converges with prior research on the nomological networks of narcissistic admiration and narcissistic rivalry suggesting that “a certain vulnerability and emotional instability is unique to narcissistic rivalry” (Back et al. 2013, p. 1023). Individuals scoring higher on narcissistic rivalry tend to base their self‐concepts on external validation, experience greater self‐esteem challenges in daily life (Geukes et al. 2017), and are prone to hostile biases and behaviors in social interactions (Li 2023). Our experimental study provides novel empirical support for this notion.

Narcissistic individuals also exhibited rage‐driven aggression in the condition where they entered a new social interaction context. This finding aligns with a meta‐analysis linking narcissism to displaced reactive aggression (Kjærvik and Bushman 2021). These aggressive reactions appear to serve instrumental purposes, such as regaining social status or regulating intense emotions like narcissistic rage. Individuals with narcissistic traits may be more inclined to adopt aggressive approaches to protect their perceived social standing, a tendency that carries over to novel social exchanges (Grapsas et al. 2020). By exerting indiscriminate aggression upon others, narcissistic individuals can externalize their narcissistic rage and recover from the aversive affective state (Krizan and Johar 2015). Further experiments can explore how these aggressive behaviors affect their perceived social status and to what extent such rage‐driven aggression actually alleviates narcissistic individuals' distress.

4.2. Admiration‐Oriented Pathway and Pride

Results did not support the hypothesis that social validation would heighten narcissistic individuals' feelings of pride. Instead, narcissistic admiration predicted pride across different social feedback conditions. One potential explanation is that feelings of superiority buffer individuals high in narcissistic admiration from intense emotional fluctuations (Geukes et al. 2017). Just as they are less likely to experience sharp drops in pride when faced with social threats, they might similarly be less likely to experience strong increases in pride when obtaining validation. Alternatively, our social validation manipulation (nine “Likes” from 11 peers and ranking second) may not have been sufficient to satisfy the unrealistic expectations of narcissistic individuals. Highly narcissistic individuals might only experience increased pride when ranking first in a social hierarchy. Further experiments should explore other manipulations.

Narcissistic individuals' pride‐driven prosocial behaviors were independent of social feedback and peer group conditions. This partially supports the SPIN model's admiration‐oriented pathway, linking narcissism to prosocial behavior. However, our results contradict the model's suggestion that social validation enhances narcissistic prosociality. One potential explanation is the lack of explicit information about whether online peers could identify the senders of “Likes” in this simulated online interaction, which might have reduced participants' view of sending “Likes” as a self‐promotional strategy. The appraisal processes of narcissistic individuals, as described in the SPIN model, encompass not only the appraisal of feedback already received but also the evaluation of potential future feedback. Narcissistic individuals might be less motivated to act prosocially if they believe others cannot connect these behaviors back to them. Future studies can investigate how narcissistic individuals' anticipation of future affordances interacts with feedback to predict subsequent admiration‐oriented behaviors.

4.3. Practical Implications

This study indicates that narcissistic individuals experience narcissistic rage and enact direct and displaced aggression when they do not receive sufficient validation. These processes can be exacerbated in social media contexts, since these platforms provide users with feasible avenues to track and immediately respond to those who made them feel invalidated. Additionally, social media is a dynamic environment where users frequently switch between interactions and encounter different people, enabling individuals to carry over their narcissistic rage states and displace aggression onto others. Previous suggestions for reducing narcissistic individuals' cyberbullying have mainly focused on cognitive processes, such as instructing them to consider the long‐term consequences of aggression (Wang et al. 2023). Our study suggests that practitioners and educators should also address narcissistic individuals' emotion regulation, providing strategies for managing emotional responses to status‐pursuit frustration. Furthermore, while narcissistic admiration was directly associated with aggressive behaviors, it also positively predicted prosocial behaviors through pride. Practitioners and educators can focus on boosting the positive emotions that narcissistic individuals experience from social validation and steering their interpersonal behaviors in a positive direction.

4.4. Strengths, Limitations, and Future Directions

This study has a number of strengths. First, by manipulating the number of “Likes” participants received, we investigated the causal effect of social feedback on individuals' reactions. Second, we adapted a recent experimental paradigm (Lutz and Schneider 2020) to measure aggression and prosocial based on the number of “Likes” and “Dislikes” participants gave others – behaviors that are common in real‐world settings and thus have relatively higher ecological validity. Third, by directly testing the mediating role of emotion in the relation between narcissism and behavioral outcomes for the first time, our study provides novel information regarding potential emotional mechanisms underlying narcissistic individuals' behavioral responses to social feedback. Finally, the SPIN model was originally developed based on research conducted in Western Europe and North America; this study represents one of the first examinations of SPIN in an Asian context.

This study also has some important limitations. First, narcissistic individuals' self‐reported emotions are susceptible to self‐enhancement biases. Future studies should replicate these results using fEMG‐ or observational measures of emotions. Second, our artificial social media context may not reflect the complexity and variability of real‐life situations. Future studies can examine the generalizability of our findings by tracking individuals' authentic social media experiences across multiple platforms. Third, our study only focused on one particular type of social feedback, namely receiving many or few “Likes” from unfamiliar peers. Social feedback can be differentiated based on several criteria, such as its public/private nature, the source of the feedback (e.g., the degree of closeness with the provider), and the content being evaluated (e.g., whether the posts include self‐disclosures). Future research can conduct more nuanced examinations of the potential impacts different forms of social feedback have on narcissistic individuals' reactions. Fourth, we designed the experimental protocol to reduce alternative interpretations of “Likes” and “Dislikes” and to strengthen their alignment with previous conceptualizations of aggressive and prosocial behaviors; nevertheless, future studies would benefit from comparisons of liking and disliking behaviors with well‐established, state‐based measures of aggression and prosocial tendencies. Fifth, the correlation between narcissistic admiration and narcissistic rivalry in our sample was low; a notable proportion of participants scored high on rivalry but low on admiration, while very few participants scored low on both dimensions. Future research should examine whether narcissistic admiration and rivalry consistently function as coexisting facets of narcissism or whether they can manifest as distinct traits, with some individuals scoring high on one dimension but not the other. Investigating these possibilities across diverse samples would provide valuable insights into the structure and coherence of narcissistic traits. Finally, the strength of the associations between narcissism and aggressive or prosocial reactions to social feedback may vary across cultures. Future research could further explore cultural variations in the activated SPIN processes using experimental cross‐cultural comparisons.

5. Conclusion

This study investigated how social feedback triggers changes in narcissistic rage and pride among individuals with narcissistic features, and how these emotions contribute to subsequent aggressive and prosocial behavior in a simulated social media context. Individuals scoring higher on narcissistic rivalry experienced more narcissistic rage when they experienced insufficient social validation, and this narcissistic rage accounted for subsequent direct and displaced aggression (i.e., sending more “Dislikes” to peers). Individuals scoring higher on narcissistic admiration engaged in more pride‐driven prosocial behavior (i.e., sending more “Likes” to peers), regardless of the social feedback that they had received. Our study identifies key personal traits and intrapersonal mechanisms to potentially target in future interventions aimed at constructing harmonious social media relationships.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplemental Materials 2025‐03‐21.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.