From a social comparison perspective, emotions and motivations related to social status may play an important role in understanding aggressive behavior during adolescence. In particular, dispositional envy may motivate adolescents to pursue higher popularity within peer hierarchies, potentially increasing their involvement in aggressive behaviors such as bullying. However, longitudinal evidence examining the processes linking envy, status-related motivations, and aggression remains limited. The present study investigated the longitudinal associations among dispositional envy, need for popularity, and both bullying behavior and cyberbullying behavior in a sample of 707 adolescents (ages 11-17 years; M=13.35, SD=0.93) assessed across three waves. Cross-lagged structural equation modeling was used to examine the temporal relationships among these variables. Results showed that dispositional envy at Time 1 positively predicted need for popularity at Time 2. In turn, need for popularity was associated with higher levels of bullying behavior at Time 3, but not with cyberbullying behavior. The indirect effect linking envy to bullying through need for popularity was small and should therefore be interpreted with caution, and no mediation effect was observed for cyberbullying. Multi-group analyses indicated that the structural relations among the variables were similar for boys and girls. Overall, the findings suggest that dispositional envy may contribute indirectly to bullying behavior through adolescents’ motivation to achieve social status. These results highlight the relevance of status-related motivations in the development of bullying during adolescence and underscore the importance of addressing such socio-motivational dynamics in prevention efforts.
Desde la perspectiva de la comparación social, las emociones y motivaciones relacionadas con el estatus social pueden desempeñar un papel importante en la comprensión de las conductas agresivas durante la adolescencia. En particular, la envidia disposicional puede motivar a los adolescentes a buscar una mayor popularidad dentro de las jerarquías de pares, lo que potencialmente podría incrementar su implicación en comportamientos de bullying. Sin embargo, la evidencia longitudinal que examine estos procesos sigue siendo limitada. El presente estudio ha investigado las asociaciones longitudinales entre la envidia disposicional, la necesidad de popularidad y las conductas de bullying y cyberbullying en una muestra de 707 adolescentes (11-17 años; M=13.35, DT=0.93) evaluados a lo largo de tres olas de medición. Se ha utilizado un modelo de ecuaciones estructurales con efectos cruzados (cross-lagged structural equation modeling) para examinar las relaciones temporales entre las variables. Los resultados han mostrado que la envidia disposicional en el Tiempo 1ha predicho positivamente la necesidad de popularidad en el Tiempo 2. A su vez, la necesidad de popularidad se ha asociado con mayores niveles de conducta de bullying en el Tiempo 3, pero no con cyberbullying. El efecto indirecto que ha vinculado la envidia con el bullying a través de la necesidad de popularidad ha sido pequeño y debe interpretarse con cautela, y no se ha observado un efecto de mediación para el cyberbullying. Los análisis multigrupo han indicado que las relaciones estructurales han sido similares entre chicos y chicas. En conjunto, los hallazgos sugieren que la envidia disposicional puede contribuir indirectamente a la conducta de bullying al incrementar la motivación de los adolescentes por alcanzar estatus social dentro de las jerarquías de pares. Estos resultados destacan el papel de las motivaciones relacionadas con el estatus en el desarrollo del bullying y subrayan la importancia de abordar estas dinámicas en los esfuerzos de prevención.
Traditional (face-to-face) bullying is a deeply harmful pattern of peer aggression that occurs within the school environment. It is commonly characterized by four core elements: (1) intentionality (the perpetrator aims to cause harm), (2) power imbalance (the aggressor holds a real or perceived social, physical, or psychological advantage over the victim), (3) repetition (the behavior is repeatedly enacted over time), and (4) the systematic use of aggressive behaviors (physical, verbal, or relational) intended to intimidate, humiliate, or socially exclude the target (Olweus, 1993). In addition, recent theoretical discussions have emphasized the importance of considering the experience of harm from the victim’s perspective as a relevant criterion in identifying bullying situations (Waseem & Nickerson, 2024). When this form of aggression is carried out using electronic devices, such as social media, instant messaging, email, online gaming, or text messages, it is referred to as cyberbullying (Camacho et al., 2023). Both forms of interpersonal violence (traditional bullying and cyberbullying) have received global recognition as major public health concerns due to their high prevalence among adolescents and the severe physical and psychological consequences they inflict on those affected (Charalampous et al., 2018; Oriol et al., 2019).
The underlying causes of bullying and cyberbullying behavior are complex and multifaceted, arising from intricate interactions among various contributing factors (Cour et al., 2022). Consequently, one of the most crucial aspects in the literature on this form of interpersonal violence involves identifying the factors that best explain aggressive behavior in perpetrators. In recent years, a growing body of research has emphasized the significance of adolescents’ need for social status and popularity within peer groups as a critical driver of bullying and cyberbullying behaviors (Košir et al., 2022; Pouwels et al., 2018). In this regard, adolescence constitutes a pivotal period for identity formation, characterized by heightened motivation to achieve popularity, peer acceptance and social affiliation, coupled with increased sensitivity to social exclusion (Blakemore, 2019; Tomova et al., 2021). These social motivations may compel adolescents to engage in aggressive behaviors as a strategy to maintain or enhance their social standing within the peer hierarchy (Pouwels et al., 2018).
From an evolutionary perspective, social reputation and popularity are regarded as adaptive functions in social environments, serving as key motivators for aggressive behavior in adolescence (Volk et al., 2022). Specifically, evidence suggests that bullies often achieve elevated social status, particularly during early adolescence (e.g., Vaillancourt et al., 2003; Wiertsema et al., 2023). This elevated status can encourage the use of aggression as a strategic tool to gain competitive advantages over peers (Volk et al., 2022; Volk et al., 2014). This drive for social status also appears to extend to virtual environments, where cyberbullying behavior likewise functions as an instrumental strategy that may offer social recognition and dominance (Camacho et al., 2023). From another but complementary point of view, Bandura’s (2002) socio-cognitive theory, a foundational framework for understanding the mechanisms underlying these forms of bullying behavior, also underscores the significance of social status and hierarchical structures in bullying dynamics. According to this theory, behaviors such as bullying are more likely to be repeated when they result in positive reinforcement, such as increased social attention or elevated status within the peer group.
Together, evolutionary and socio-cognitive perspectives provide a compelling account of why bullying and cyberbullying can function as effective status-enhancing strategies in adolescent peer hierarchies and how such behaviors become reinforced over time. However, these approaches offer more limited insight into the psychological processes that initiate adolescents’ engagement in status-seeking aggression. They do not fully explain why, within the same competitive and hierarchical social environments, only some adolescents translate concerns about social status into aggressive behavior, while others pursue alternative, non-aggressive strategies. Addressing this gap requires a closer examination of the affective and motivational signals that arise from social comparison processes and that may channel status-related concerns into specific behavioral responses. In particular, relatively little longitudinal research has examined whether dispositional envy, a socio-affective response emerging from unfavorable social comparisons, may contribute to bullying and cyberbullying behaviors over time. Moreover, limited attention has been paid to whether status-oriented motivations, such as the need for popularity, function as a mechanism through which envy-related perceptions of disadvantage translate into aggressive behavior during adolescence.
Associations between dispositional envy and bullying and cyberbullying behaviorAs previously noted, Bandura’s (2002) socio-cognitive perspective highlights social comparison processes as a critical driver for understanding why adolescents may engage in aggressive behaviors to preserve or enhance their social standing within peer hierarchies. When adolescents perceive a discrepancy between their own social status and that of more popular peers, such unfavourable comparisons may elicit negative self-evaluations and motivate attempts to reduce perceived disadvantage, sometimes through aggressive strategies aimed at status acquisition (Geng et al., 2022; Lange et al., 2018).
One affective disposition that may be particularly relevant in this context is dispositional envy. Contemporary perspectives describe it as a painful, self-focused socio-affective response arising from upward social comparison and the perception of relative inferiority in valued domains (Crusius et al., 2020). From this perspective, envy functions as a psychological signal of threatened social status or access to valued resources, which may activate motivational processes oriented toward restoring one’s relative standing within the peer group (Ramachandran & Jalal, 2017). Such motivational activation may, in turn, reduce empathic concern and increase the likelihood of instrumental behavior aimed at regaining status, including aggressive forms of interaction.
From a social cognitive perspective, emotional states influence motivational processes by shaping individuals’ self-evaluations, outcome expectations, and goal-directed behavior (Bandura, 2002). In the case of envy, upward social comparison may heighten adolescents’ awareness of their relative inferiority in socially valued domains, such as peer status and recognition. This heightened awareness may activate goal-directed efforts aimed at restoring one's relative standing within the peer hierarchy. Given that popularity represents a salient marker of status in adolescent peer groups, such motivational activation may translate into an increased need for popularity, understood as a desire to gain social visibility, influence, and recognition among peers.
A growing body of research suggests that envy may foster aggressive behaviors aimed at overcoming perceived status discrepancies and gaining access to more privileged positions within social hierarchies (Lange & Protasi, 2021; Rengifo & Laham, 2022; Yang & Guo, 2023). Previous studies have distinguished between situational experiences of envy, which are triggered by specific social contexts characterized by limited perceived control over status improvement, and dispositional envy, which reflects a more stable tendency to experience chronic feelings of inferiority toward more advantaged others (Cohen-Charash, 2009; Lange et al., 2018; Smith et al., 1999). Dispositional envy, in particular, has been linked to heightened hostility, resentment, and motivation to reduce perceived disadvantage, which may increase the likelihood of instrumental behaviors directed at undermining superior others. Empirical evidence has associated dispositional envy with deceitful and aggressive behaviors, particularly in adult populations (for a review, see Yu et al., 2018). Emerging research has extended these findings to adolescence, showing that higher levels of dispositional envy are associated with greater involvement in bullying and cyberbullying behaviors (e.g., Geng et al., 2022; Hitokoto & Sawada, 2021; Villanueva-Moya et al., 2023). For instance, adolescents who report elevated envy are more likely to engage in aggression as a means of lowering the social standing of peers perceived as superior (Hitokoto & Sawada, 2021).
In online contexts, the translation of envy into aggressive behavior may be facilitated by structural features such as anonymity, reduced accountability, and the absence of immediate victim feedback, which can attenuate social constraints and disinhibit aggression (Özden-Yıldırım, 2019). Moreover, digital environments may intensify social comparison processes by exposing adolescents to highly curated and idealized representations of peers, potentially amplifying perceptions of inferiority and status threat (Geng et al., 2022).
Although previous research has linked envy to various forms of aggression, including bullying and cyberbullying, the empirical evidence remains limited and predominantly cross-sectional. This methodological limitation restricts our understanding of the temporal dynamics and causal direction between envy and aggressive behavior (Geng et al., 2022; Hitokoto & Sawada, 2021; Lange & Crusius, 2015). Moreover, while theoretical frameworks have suggested that envy may motivate interpersonal harm as a strategy to reduce perceived social disparity, less attention has been paid to the specific motivational goals through which envy is translated into aggressive behavior, particularly during adolescence.
In adolescent peer contexts, status attainment and popularity constitute highly salient social goals (Pouwels et al., 2018). Although prior research has shown that envy is associated with hostility and aggressive tendencies aimed at reducing perceived status disadvantage (Lange & Crusius, 2015; Yu et al., 2018), empirical evidence examining how envy is translated into bullying and cyberbullying remains limited and predominantly cross-sectional (Geng et al., 2022; Hitokoto & Sawada, 2021). More importantly, existing studies have rarely investigated whether envy contributes to aggressive behavior by activating specific status-oriented motivational states, such as an increased need for popularity, particularly during adolescence, a developmental period characterized by heightened sensitivity to peer status and social comparison (Pouwels et al., 2018; Romera et al., 2021). Addressing this gap requires a longitudinal approach capable of capturing how envy-related perceptions of disadvantage evolve into popularity-oriented motivational processes and, subsequently, into aggressive behavioral strategies.
The mediator role of need for popularityAs previously discussed, dispositional envy may signal perceived threats to an adolescent’s social status or access to valued resources within the peer group (Lange & Crusius, 2015). However, such affective signals alone do not explain how feelings of disadvantage are translated into concrete aggressive actions. Identifying the motivational mechanisms that channel envy-related discomfort into specific behavioral strategies is therefore crucial for understanding bullying perpetration during adolescence (Xiang et al., 2019; Yu et al., 2018). In this regard, motivational mechanisms focused on achieving higher status may help to clarify the connection between envy-related experiences and aggressive behavior (Chmielowice‐Szymanski et al., 2024). One such mechanism is the need for popularity, which represents a central status-oriented goal during adolescence and may operate as a bridge between envy-related perceptions of disadvantage and aggressive behavior (Romera et al., 2021).
Consistent with Bandura’s (2002) socio-cognitive framework, affective experiences function as signals that can activate motivational processes and goal-directed behavior rather than as passive responses. In this sense, the drive to achieve popularity may operate as a key motivational mechanism through which envy-related perceptions of disadvantage are translated into aggressive behaviors perceived as effective strategies for elevating one’s status within the peer group hierarchy (Pouwels et al., 2018; Wiertsema et al., 2023). Within this framework, the need for popularity is defined as the desire for social recognition and self-esteem derived from influence and impact within the peer group (Laninga-Wijnen et al., 2019; Pouwels et al., 2018). This need plays a crucial role in the social dynamics of adolescence, particularly in relation to aggressive behavior. Empirical evidence suggests that adolescents with goals oriented toward popularity are more prone to engaging in both physical and relational aggression (Cillessen et al., 2014; Dawes & Xie, 2014; Malamut et al., 2020; Romera et al., 2021). For example, the longitudinal study by Romera et al. (2021), conducted with 3,017 adolescents aged 11 to 16, revealed that those with a high need for popularity exhibited higher levels of moral disengagement and bullying perpetration. Similarly, the study by Laninga-Wijnen et al. (2022), involving a sample of 1,186 adolescents, showed that popularity goals predict bullying behavior, particularly when there is a discrepancy between desired and self-perceived popularity. Adolescents who aspire to be popular but do not perceive themselves as such are more likely to use aggression to climb the social hierarchy. Specifically, adolescents with a strong need for social recognition tend to evaluate the potential benefits and costs of bullying, often deciding to participate when they perceive it will enhance their popularity at minimal risk to themselves (Pan et al., 2020; Pouwels et al., 2019). In some cases, they may even resort to coercive strategies to climb the social hierarchy (Ojanen & Findley-Van Nostrand, 2014).
On the other hand, regarding the link between the need for popularity and this type of online aggression, studies have shown that adolescents with a strong status orientation tend to use digital platforms to display power or publicly ridicule others with high social visibility (Camacho et al., 2023). Specifically, the immediacy, public visibility, and social reinforcement provided by these environments make cyberbullying a functional channel for those seeking to assert dominance or gain social recognition (Nesi et al., 2018). Therefore, virtual settings may facilitate the use of aggression as a strategy for achieving social recognition. However, although popularity motives may similarly shape aggressive behavior in online contexts, the extent to which the need for popularity mediates the link between envy and cyberbullying may depend on contextual features such as anonymity, identifiability, and social feedback, which differentiate digital from face-to-face interactions.
Beyond these motivational mechanisms, it is also important to consider whether the processes linking envy, need for popularity, and aggressive behavior may operate similarly across gender. Adolescence is characterized by important gender differences in socio-emotional functioning and peer dynamics, which may shape how status-related motivations translate into behavior within peer groups. Previous research suggests that boys and girls may differ in their orientation toward social status, peer approval, and the use of aggressive strategies to navigate social hierarchies (Bravo et al., 2022; Yu et al., 2018). These differences raise the possibility that the motivational processes linking envy, popularity goals, and aggressive behavior may not operate identically across gender. Examining whether these associations differ between boys and girls may therefore provide additional insight into the social mechanisms underlying bullying and cyberbullying during adolescence.
The present studyThe present study seeks to advance understanding of the developmental pathways linking dispositional envy to bullying and cyberbullying behavior during adolescence. Specifically, the study aims to examine whether dispositional envy prospectively predicts involvement in aggressive behavior across offline and online contexts, and to identify the motivational processes through which such associations may unfold over time. Drawing on evolutionary and socio-cognitive perspectives, the study focuses on the role of status-oriented motivation, conceptualized as the need for popularity, as a key mechanism through which envy-related perceptions of disadvantage may be translated into aggressive behavioral strategies.
More concretely, the first objective of the study is to analyse the longitudinal associations between dispositional envy and both bullying and cyberbullying behavior across three measurement waves, thereby clarifying the temporal direction and stability of these relationships. The second objective is to test whether the need for popularity functions as a mediating motivational process linking dispositional envy to subsequent engagement in bullying and cyberbullying behavior. Finally, given the distinct structural characteristics of online and offline peer interactions, the study aims to explore whether these longitudinal pathways operate similarly for traditional bullying and cyberbullying, or whether the mediating role of the need for popularity differs across contexts. Additionally, given previous evidence suggesting potential gender differences in peer dynamics and status-related motivations during adolescence, the study also examines whether these longitudinal associations operate similarly for boys and girls.
Based on these objectives, two hypotheses were formulated: (1) Higher levels of dispositional envy at Time 1 were expected to predict higher levels of aggressive behavior over time. Specifically, adolescents reporting higher dispositional envy were expected to show greater involvement in bullying and cyberbullying behaviors at later measurement points; and (2) It was hypothesized that the need for popularity at Time 2 would mediate the longitudinal association between dispositional envy at Time 1 and bullying and cyberbullying behavior at Time 3, such that dispositional envy at Time 1 would predict increases in the need for popularity at Time 2, which in turn would predict greater involvement in aggressive behavior at Time 3. Given previous research indicating gender differences in aggressive behavior during adolescence, the structural relations among dispositional envy, need for popularity, and both forms of bullying behavior were explored across boys and girls.
MethodParticipantsThe study was approved by the Ethics Committee of the Universidad del Desarrollo and was conducted following the ethical standards established for research involving human participants. A convenience sampling strategy was used to recruit participants from six secondary schools located in Ecuador. The inclusion criterion required students to be between 11 and 18 years old. At Wave 1, a total of 971 adolescents participated in the study. Participants ranged in age from 11 to 17 years (M=13.35, SD=0.93), and the sample included 446 males (45.9%) and 525 females (54.1%). At Wave 2, 821 adolescents participated (379 males, 46.2%; 442 females, 53.8%), with a mean age of 13.82 years (SD=0.83). At Wave 3, 794 adolescents were included in the study (374 males, 47.1%; 420 females, 52.9%), with a mean age of 14.24 years (SD=0.91).
Missing data were examined prior to conducting the main analyses. The initial dataset included 1,029 adolescents who participated in at least one wave of the study. Across the three waves of data collection, 711 participants (69.1%) completed all waves, whereas 318 participants (30.9%) had missing data in at least one wave. The proportion of missing data increased across waves, which is typical in longitudinal school-based research. Specifically, 94.2% of participants provided complete data at Wave 1 (5.8% missing), 79.7% at Wave 2 (20.3% missing), and 77.3% at Wave 3 (22.7% missing). To examine the mechanism underlying the missing data, Little’s MCAR test (Little, 2013) was conducted. Results indicated that the missing data pattern was consistent with Missing Completely at Random (MCAR), p=.34, suggesting that the probability of missingness was not systematically related to the observed study variables. To further examine potential attrition bias, baseline differences were tested between participants who completed all three waves (retained group) and those who did not (attrition group). Independent-samples t-tests revealed no significant differences between the groups in the main study variables at Wave 1, including dispositional envy, need for popularity, bullying, and cyberbullying. In addition, no significant association was found between gender and attrition, χ²(2)=0.42, p=.81.
InstrumentsBullying perpetration was assessed using the Illinois Bullying Behavior Scale (Espelage & Holt, 2001), which measures the frequency of aggressive behaviors directed toward other students during the previous 30 days. The instrument was administered in Spanish following previously translated versions used in research with Chilean adolescents (Varela et al., 2019). Participants responded using a Likert-type scale ranging from 1=never to 4=almost always. A sample item is: “I have teased other students”. Internal consistency was good across waves, with Cronbach’s alpha values ranging from .82 to .87 and McDonald’s omega values ranging from .85 to .90 (Wave 1: α=.82, ω=.85; Wave 2: α=.87, ω=.90; Wave 3: α=.87, ω=.90). CFA analyses supported the expected one-factor structure of the scale. At Wave 1, model fit was acceptable (CFI=.981, TLI=.974, RMSEA=.079, SRMR=.066). At Wave 2, the model showed improved fit (CFI=.989, TLI=.985, RMSEA=.063, SRMR=.057). At Wave 3, the model demonstrated good fit (CFI=.994, TLI=.993, RMSEA=.058, SRMR=.051). Standardized factor loadings ranged from .61 to .89 across waves, indicating that the items were strongly associated with the underlying bullying construct.
Cyberbullying perpetration was assessed using a four-item scale adapted from instruments developed by Ybarra et al. (2007) and Low and Espelage (2013), which evaluate the frequency of engagement in aggressive online behaviors, such as spreading rumors or posting threatening messages online. The items were administered in Spanish following versions previously used in studies with Chilean adolescents (Varela et al., 2019). Internal consistency was adequate across waves (Wave 1: α=.73, ω=.77; Wave 2: α=.82, ω=.84; Wave 3: α=.79, ω=.82). Wave-specific confirmatory factor analyses were conducted to assess the factorial validity of the scale. At Wave 1, the model showed acceptable fit (CFI=.970, TLI=.923, RMSEA=.120, SRMR=.031). At Wave 2, model fit improved substantially (CFI=.990, TLI=.970, RMSEA=.090, SRMR=.020). At Wave 3, the model showed excellent fit (CFI=1.000, TLI=1.000, RMSEA=.000, SRMR=.010). Standardized factor loadings ranged from .56 to .85 across waves, supporting the adequacy of the four-item latent structure.
Need for popularity was measured using the need for popularity scale originally developed by Santor et al. (2000) and later validated and used in Spanish adolescent samples (Del Rey et al., 2019). The scale assesses adolescents’ motivation to be perceived as popular and influential within their peer group. It consists of 12 items rated on a Likert scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). Example items include: “I have ignored certain people to be more popular with others”. Internal consistency was high across waves (Wave 1: α=.89, ω=.91; Wave 2: α=.91, ω=.94; Wave 3: α=.92, ω=.94). At Wave 1, model fit was very good (CFI=.995, TLI=.994, RMSEA=.067, SRMR=.048). At Wave 2, the model showed excellent fit (CFI=.997, TLI=.996, RMSEA=.049, SRMR=.041). At Wave 3, model fit remained strong (CFI=.996, TLI=.995, RMSEA=.055, SRMR=.043). Standardized factor loadings ranged from .42 to .90, with most items showing moderate to high saturation on the latent factor.
Dispositional envy was assessed using the Spanish validated version of the Dispositional Envy Scale originally developed by Smith et al. (1999). The instrument consists of eight items designed to measure the general tendency to experience envy as a stable socio-affective disposition arising from unfavorable social comparisons. Participants responded using a five-point Likert scale ranging from 1=strongly disagree to 5=strongly agree. Example items include: “I feel envy every day” and “The bitter truth is that I generally feel inferior to others.” The Spanish version of the scale has shown adequate psychometric properties in Spanish-speaking populations (Reyes Pérez et al., 2023). In the present study, internal consistency was adequate across waves (Wave 1: α=.86, ω=.89; Wave 2: α=.88, ω=.92; Wave 3: α=.88, ω=.92).
Confirmatory factor analyses were conducted separately for each wave. The one-factor structure of the scale showed acceptable fit across time. At Wave 1, model fit was CFI=.987, TLI=.981, RMSEA=.099, and SRMR=.053. At Wave 2, model fit improved (CFI=.992, TLI=.989, RMSEA=.092, SRMR=.046). At Wave 3, the model also showed acceptable fit (CFI=.990, TLI=.986, RMSEA=.094, SRMR=.049). Standardized factor loadings ranged from .68 to .87 across waves. To further evaluate the longitudinal stability of the measurement model, a longitudinal CFA including the four constructs across the three waves simultaneously was estimated. The model showed good overall fit, χ²(4587)=6901.42, p<.001, CFI=.993, TLI=.993, RMSEA=.027 (90% CI [.025, .028]), and SRMR=.049. The pattern of factor loadings remained stable across time, supporting the temporal stability of the measurement structure.
ProcedureThe study was conducted with the authorization and collaboration of school principals. Participation was voluntary, and written informed consent was obtained from both students and their parents prior to data collection. Participants were informed about the objectives of the study, the confidentiality of their responses, and their right to withdraw from the study at any time without consequences. Data collection was conducted in the classroom using paper-and-pencil questionnaires administered by trained members of the research team. Each wave of data collection lasted approximately 30minutes. The presence of researchers during the administration of the questionnaires ensured that participants could ask questions and that the instructions were consistently understood across schools.
Data analysisAll statistical analyses were conducted using R (version 4.3.1). Structural equation models were estimated using the lavaan package (Rosseel, 2012). Missing data were handled using full information maximum likelihood (FIML) estimation, which allows the use of all available data under the assumption that data are missing at random (MAR) and provides unbiased parameter estimates in longitudinal models (Enders, 2010). Prior to the main analyses, patterns of missing data and attrition across waves were examined. Little’s MCAR test indicated that the missing data pattern was consistent with a missing completely at random mechanism. In addition, baseline comparisons between participants who completed all waves and those with missing data were conducted to assess potential attrition bias.
The analyses proceeded in several stages. First, descriptive statistics were computed for all study variables across the three waves, and independent-samples t-tests were conducted to examine gender differences. Second, Pearson correlations were calculated to examine the associations among the variables across time. Third, a longitudinal structural equation model was estimated to test the hypothesized mediation process. Specifically, the model examined whether need for popularity at Time 2 mediated the association between dispositional envy at Time 1 and aggressive behaviors (bullying behavior and cyberbullying behavior) at Time 3. The model included autoregressive paths for each construct across adjacent waves to account for temporal stability and within-time correlations among variables to capture concurrent associations (Kline, 2016; Little, 2013). Indirect effects were estimated using the delta method within the SEM framework (MacKinnon, 2008). Model fit was evaluated using several commonly recommended indices, including the comparative fit index (CFI), Tucker–Lewis index (TLI), root mean square error of approximation (RMSEA) with 90% confidence intervals, and the standardized root mean square residual (SRMR). Following established recommendations, values of CFI and TLI greater than .90 and RMSEA and SRMR values below .08 were considered indicative of acceptable model fit (Hu & Bentler, 1999).
Fourth, multi-group structural equation modeling analyses were conducted to examine whether the structural relationships differed between boys and girls. This approach allows the estimation of the model simultaneously across groups and enables testing whether structural parameters are invariant across gender (Byrne, 2012; Kline, 2016). Finally, a Random Intercept Cross-Lagged Panel Model (RI-CLPM) was estimated as a robustness analysis to distinguish between stable between-person differences and within-person processes over time. Traditional cross-lagged panel models may confound these two sources of variance because cross-lagged paths can reflect stable individual differences rather than intra-individual change processes. The RI-CLPM addresses this limitation by introducing random intercept factors that capture stable trait-like differences between individuals, allowing the cross-lagged paths to represent within-person fluctuations around individuals’ expected levels across time (Hamaker et al., 2015; Mulder & Hamaker, 2021).
ResultsPreliminary analysesDescriptive statistics and correlationsAs preliminary analyses, T-test were conducted to examine descriptive patterns and gender differences in the study variables across the three waves. These analyses provide an initial overview of the data before testing the hypothesized longitudinal mediation model using structural equation modeling (SEM). Descriptive statistics and gender differences for the study variables across the three waves are presented in Table 1. Overall, levels of bullying and cyberbullying behaviors were relatively low across the study period, whereas envy and need for popularity showed moderate levels. Mean levels of envy showed a slight decrease over time (T1: M=1.98, SD=0.80; T2: M=1.91, SD=0.81; T3: M=1.74, SD=0.75). Similarly, need for popularity decreased slightly across waves. Bullying behaviors remained relatively stable across the three waves, whereas cyberbullying showed a noticeable decline from Time 1 (M=2.25, SD=0.78) to Time 3 (M=1.20, SD=0.64). Gender comparisons indicated that boys reported significantly higher levels of bullying and cyberbullying across waves, whereas girls reported slightly higher levels of envy. However, effect sizes were generally small.
Descriptive statistics and gender differences for study variables across the three waves
| Variable | Total M (SD) | Boys M (SD) | Girls M (SD) | t | df | p |
|---|---|---|---|---|---|---|
| Envy T1 | 1.98 (0.80) | 1.84 (0.75) | 2.05 (0.83) | −4.12 (-0.26) | 954.95 | <.001 |
| Envy T2 | 1.91 (0.81) | 1.78 (0.77) | 1.95 (0.83) | −3.09 (-0.22) | 803.93 | .002 |
| Envy T3 | 1.74 (0.75) | 1.68 (0.76) | 1.79 (0.73) | −1.90 (-0.14) | 757.39 | .058 |
| Popularity T1 | 2.30 (1.15) | 2.27 (1.18) | 2.31 (1.10) | −0.59 (-0.04) | 909.01 | .555 |
| Popularity T2 | 2.29 (1.17) | 2.18 (1.12) | 2.35 (1.22) | −2.06 (-0.14) | 805.67 | .039 |
| Popularity T3 | 2.17 (1.14) | 2.13 (1.15) | 2.22 (1.13) | −1.10 (-0.08) | 760.90 | .271 |
| Bullying T1 | 1.56 (0.52) | 1.60 (0.56) | 1.52 (0.48) | 2.20 (0.14) | 868.97 | .029 |
| Bullying T2 | 1.51 (0.55) | 1.63 (0.64) | 1.45 (0.49) | 4.22 (0.30) | 694.83 | <.001 |
| Bullying T3 | 1.53 (0.55) | 1.59 (0.57) | 1.49 (0.54) | 2.49 (0.18) | 746.99 | .013 |
| Cyberbullying T1 | 2.25 (0.78) | 2.34 (0.84) | 2.17 (0.62) | 3.54 (0.24) | 794.72 | <.001 |
| Cyberbullying T2 | 1.19 (0.67) | 1.40 (0.82) | 1.09 (0.53) | 6.09 (0.44) | 617.69 | <.001 |
| Cyberbullying T3 | 1.20 (0.64) | 1.30 (0.79) | 1.12 (0.55) | 3.74 (0.28) | 630.18 | <.001 |
Note. M=mean; SD=standard deviation; t=t statistic; df=degrees of freedom; p=significance level.
Pearson correlations among the study variables across the three waves are presented in Table 2. Overall, the results indicated moderate temporal stability for the main constructs. Envy showed moderate stability across waves (rs= .41-.45), as did need for popularity (rs=.44-.52), bullying (rs=.47-.51), and cyberbullying (rs=.31-.35). Consistent with theoretical expectations, dispositional envy was positively associated with need for popularity across waves. In addition, both dispositional envy and need for popularity were positively related to bullying and cyberbullying behaviors, although these associations were generally small to moderate. Bullying and cyberbullying were moderately correlated, suggesting substantial overlap between these forms of aggressive behavior in adolescence.
Pearson correlations among study variables across the three waves
| Variable | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1. Envy T1 | — | |||||||||||
| 2. Envy T2 | .45*** | — | ||||||||||
| 3. Envy T3 | .41*** | .45*** | — | |||||||||
| 4. Need for Popularity T1 | .55*** | .31*** | .35*** | — | ||||||||
| 5. Need for Popularity T2 | .28*** | .57*** | .33*** | .44*** | — | |||||||
| 6. Need for Popularity T3 | .25*** | .32*** | .56*** | .47*** | .52*** | — | ||||||
| 7. Bullying T1 | .31*** | .19*** | .20*** | .40*** | .28*** | .29*** | — | |||||
| 8. Bullying T2 | .13** | .19*** | .15*** | .20*** | .35*** | .29*** | .47*** | — | ||||
| 9. Bullying T3 | .16*** | .19*** | .26*** | .27*** | .28*** | .41*** | .47*** | .51*** | — | |||
| 10. Cyberbullying T1 | .25*** | .16*** | .18*** | .31*** | .18*** | .21*** | .51*** | .30*** | .25*** | — | ||
| 11. Cyberbullying T2 | .03 | .16*** | .11** | .11** | .27*** | .21*** | .33*** | .58*** | .33*** | .35*** | — | |
| 12. Cyberbullying T3 | .10* | .11** | .16*** | .19*** | .13*** | .23*** | .25*** | .33*** | .48*** | .23*** | .31*** | — |
Note. p < .05*. p < .01**. p < .001***.
A longitudinal mediation model was estimated to examine whether need for popularity at Time 2 mediated the association between dispositional envy at Time 1 and aggressive behaviors (bullying behavior and cyberbullying behavior) at Time 3. The model showed acceptable overall fit, χ²(35)=290.90, p<.001, CFI=.927, TLI=.901, RMSEA=.084, and SRMR=.078. As illustrated in Figure 1 and detailed in Table 3, dispositional envy at Time 1 significantly predicted need for popularity at Time 2 (β=.10, p=.016). In turn, need for popularity at Time 2 significantly predicted bullying behavior at Time 3 (β=.11, p<.001). However, need for popularity did not significantly predict cyberbullying behavior at Time 3 (β=.03, p=.317).
Longitudinal mediation model linking envy, need for popularity, and aggressive behaviors across three waves.
Note. All paths represent standardized coefficients (β). Solid arrows indicate significant paths (p < .05); dashed arrows represent non-significant paths (p ≥ .05). The model includes covariates sex and age controlled for in all endogenous variables. T1=Time 1; T2=Time 2; T3=Time 3.
Longitudinal mediation model predicting bullying and cyberbullying
| Path | B | SE | β | p | 95% CI |
|---|---|---|---|---|---|
| Envy T1 → Need for Popularity T2 | 0.125 | 0.052 | .100 | .016 | [.024, .228] |
| Need for Popularity T2 → Bullying T3 | 0.055 | 0.016 | .109 | <.001 | [.024, .089] |
| Need for Popularity T2 → Cyberbullying T3 | 0.037 | 0.020 | .034 | .317 | [-.012, .074] |
| Envy T1 → Bullying T3 (direct) | 0.018 | 0.020 | .028 | .359 | [-.021, .056] |
| Envy T1 → Cyberbullying T3 (direct) | 0.021 | 0.018 | .031 | .241 | [-.014, .054] |
| Indirect effect (Envy → Popularity → Bullying) | 0.007 | 0.004 | .011 | .047 | [-.001, .019] |
| Indirect effect (Envy → Popularity → Cyberbullying) | 0.003 | 0.003 | .003 | .192 | [-.003, .010] |
Note. B=unstandardized coefficient; SE=standard error; β=standardized coefficient; CI=confidence interval. Indirect effects were estimated within the structural equation modeling framework.
Regarding indirect effects, the mediation pathway linking envy to bullying through need for popularity was small but reached statistical significance (β=.011, p=.047). However, the bootstrap confidence interval included zero, indicating that the magnitude of this indirect effect should be interpreted cautiously. In contrast, the indirect effect linking envy to cyberbullying through need for popularity was not statistically significant (β=.003, p=.192). The direct paths from envy at Time 1 to bullying and cyberbullying at Time 3 were not statistically significant once the mediator and autoregressive effects were included in the model. Overall, the model explained 10% of the variance in need for popularity, 14% of the variance in bullying behavior, and 7% of the variance in cyberbullying behavior at Time 3.
Gender differences in the structural modelMulti-group structural equation modeling analyses were conducted to examine whether the structural relations differed between boys and girls. The constrained model assuming equal structural paths across gender groups showed a similar fit to the unconstrained model, indicating that the longitudinal associations did not significantly differ across gender. Across both groups, dispositional envy at Time 1 significantly predicted need for popularity at Time 2 (B=.13, SE=.052, p=.016). In turn, need for popularity at Time 2 significantly predicted bullying behavior at Time 3 (B=.06, SE=.016, p<.001). In contrast, the path from need for popularity to cyberbullying behavior at Time 3 was not statistically significant (B=.04, SE=.020, p=.317). Similarly, the direct effects from envy at Time 1 to bullying behavior (B=.02, SE=.020, p=.359) and cyberbullying behavior (B=0.021, SE=0.018, p=.241) at Time 3 were not statistically significant. These findings suggest that the longitudinal mediation pathway linking envy, need for popularity, and bullying behavior operates similarly for boys and girls.
Robustness analysisFinally, a Random Intercept Cross-Lagged Panel Model was estimated as a robustness analysis. However, the model showed poor fit to the data (CFI=.889, TLI=.790, RMSEA=.146, SRMR=.063). Given this inadequate fit, the RI-CLPM results were not interpreted further, and the conclusions of the study rely on the standard longitudinal mediation model.
DiscussionThe present study addressed an important gap in the literature by examining the longitudinal relationships between dispositional envy, need for popularity, and aggressive behaviors during adolescence. Although previous research has linked envy to aggressive tendencies, empirical evidence explaining the motivational processes through which envy may translate into bullying and cyberbullying behaviors has remained limited and largely cross-sectional. In this context, the present study investigated whether the need for popularity functions as a motivational pathway linking dispositional envy with aggressive behaviors over time. Overall, the findings provided partial support for the proposed hypotheses and contributed to a better understanding of the socio-motivational mechanisms linking social comparison processes with bullying and cyberbullying behaviors in peer contexts. In line with the conceptualization of envy as a socio-affective signal indicating perceived threats to one's social standing (Crusius et al., 2020), the present findings suggest that adolescents who chronically experience such status alarms may be more likely to engage in behavioral strategies aimed at regulating their position within peer hierarchies.
Regarding the first hypothesis, the results provided partial support. Although dispositional envy at Time 1 was associated with subsequent aggressive behavior, the direct paths from envy to bullying and cyberbullying at Time 3 were not statistically significant once the mediating role of need for popularity and the autoregressive effects were taken into account. Instead, dispositional envy prospectively predicted higher levels of need for popularity at Time 2, which in turn was associated with greater involvement in traditional bullying over time. These findings suggest that envy may contribute to aggressive behavior indirectly by increasing adolescents’ motivation to attain social status within peer hierarchies. These findings are consistent with theoretical perspectives suggesting that envy emerges from unfavorable social comparisons and signals perceived threats to one's social standing (Lange & Crusius, 2015). In adolescence, where peer hierarchies and social evaluation are particularly salient, such feelings of relative disadvantage may motivate behaviors aimed at restoring or enhancing one’s position within the peer group. In this sense, envy may function as an affective signal that motivates strategies oriented toward status regulation.
Interestingly, dispositional envy also showed associations with cyberbullying behavior at later stages, although this relationship appeared less consistent across waves. While envy was related to cyberbullying at later time points, the short-term longitudinal paths were not statistically significant. One possible explanation is that the translation of envy into online aggression may occur more gradually, potentially influenced by repeated social comparisons and prolonged interactions in digital environments (Geng et al., 2022). Previous research has suggested that cyberbullying can serve instrumental functions similar to traditional bullying, such as increasing group visibility, reinforcing social status, or responding to exclusion dynamics within peer networks (Camacho et al., 2023; Wiertsema et al., 2023). However, the weaker short-term association observed in the present study may indicate that the status-related benefits of cyberbullying are less immediate or less visible than those associated with face-to-face aggression. In this sense, envy may require the accumulation of social comparisons before translating into aggressive behaviors in digital environments.
Overall, the observed relationship between dispositional envy and different forms of bullying aligns with previous research indicating that envy emerges from unfavorable social comparisons and may foster hostile responses toward more advantaged individuals (Crusius et al., 2020). Envy has been linked to persistent feelings of inferiority and resentment, particularly when individuals perceive limited opportunities to improve their own status (Cohen-Charash, 2009; Lange et al., 2018). Within adolescent peer contexts, such experiences may increase motivation to regulate one’s social standing. According to Bandura’s (2002) social cognitive theory, adolescents may learn that aggressive behaviors are sometimes associated with social rewards such as attention, influence, or dominance within peer groups. As a result, bullying behaviors may become reinforced in social environments where aggression is tolerated or even valued as a strategy for navigating peer hierarchies (Romera et al., 2021; Volk et al., 2022).
However, robustness analyses using a Random Intercept Cross-Lagged Panel Model suggested that these associations may partly reflect stable between-person differences rather than purely within-person fluctuations over time. One possible interpretation is that dispositional envy represents a relatively stable socio-affective tendency linked to chronic social comparison processes. Individuals who frequently perceive themselves as disadvantaged in comparison with others may consistently experience status-related concerns within peer hierarchies, which may place them at greater risk of engaging in aggressive behaviors across time. Previous research has conceptualized dispositional envy as a trait-like tendency rooted in recurrent upward social comparisons and persistent perceptions of relative inferiority (Crusius et al., 2020; Lange & Crusius, 2015). From this perspective, the association between envy and bullying may partly reflect stable individual differences in how adolescents interpret and respond to social comparison dynamics within peer groups.
Regarding the second hypothesis, the results partially supported the proposed mediational model in which need for popularity acts as a motivational mechanism linking dispositional envy to both forms of bullying behavior. Dispositional envy predicted higher levels of need for popularity over time, which in turn predicted greater involvement in traditional bullying. This finding suggests that status-oriented motivations may operate as an important mechanism through which envy-related perceptions of disadvantage are translated into aggressive strategies aimed at enhancing social recognition within the peer group. In line with Bandura’s (2002) social cognitive framework, envy may function as an affective signal that activates goal-directed motivational processes rather than merely reflecting passive reactions. Adolescents who perceive themselves as disadvantaged within the peer hierarchy may develop stronger motivations to gain social recognition or popularity. The present findings suggest that these status-oriented motivations may encourage the use of bullying as a strategy to restore or enhance one’s position within the peer group. Previous studies have similarly shown that adolescents with strong popularity goals are more likely to engage in physical, verbal, or relational aggression, particularly when there is a discrepancy between desired and perceived popularity (Laninga-Wijnen et al., 2022; Romera et al., 2021).
However, the mediating role of the need for popularity was not supported in the case of cyberbullying. After controlling for the temporal stability of aggressive behaviors through autoregressive effects, the indirect pathway linking envy to cyberbullying through popularity motives was not statistically significant. This finding suggests that the motivational processes underlying online aggression may differ from those involved in traditional bullying. One possible interpretation is that cyberbullying may be influenced by additional psychological processes that were not directly examined in the present study. Previous research has suggested that online aggression may be facilitated by factors such as impulsivity, emotional discharge, or online disinhibition (Barlett et al., 2022; Wang et al., 2022). However, these processes were not directly measured in the present study and should therefore be interpreted cautiously. Future research could examine whether these mechanisms help explain why envy-related status motivations appear to be less strongly associated with cyberbullying than with face-to-face bullying.
Finally, regarding the third hypothesis, the multi-group analyses indicated that the structural relations among dispositional envy, need for popularity, and aggressive behaviors did not significantly differ between boys and girls. This finding suggests that the socio-motivational processes linking envy and aggressive behavior may operate similarly across gender during adolescence. Although previous research has highlighted gender differences in socio-emotional functioning and peer dynamics (Bravo et al., 2022; Yu et al., 2018), the present findings indicate that the underlying mechanisms through which envy-related perceptions of status threat translate into aggressive behavior may be comparable for both boys and girls. In this sense, while boys and girls may differ in the frequency or expression of aggressive behaviors, the motivational processes associated with status regulation and the pursuit of popularity within peer hierarchies may operate in similar ways across gender.
LimitationsThis study presents several limitations that should be considered when interpreting the findings. First, although the longitudinal design allowed for the examination of temporal associations between dispositional envy, need for popularity, and both bullying and cyberbullying behaviors, the reliance on self-report measures may introduce potential biases such as social desirability or shared method variance. Future research could benefit from incorporating multiple informants, such as peer nominations or teacher reports, as well as behavioral or observational measures to obtain a more comprehensive assessment of bullying and cyberbullying dynamics. In addition, participants were recruited from six secondary schools, which may introduce potential clustering effects due to the nested structure of the data. Although the analyses were conducted at the individual level, the relatively small number of schools limited the feasibility of applying multilevel modeling techniques. Future research including a larger number of schools, could examine potential school-level influences on the associations between the variables of the study.
Second, although the study examined gender invariance in the structural relations among dispositional envy, need for popularity, and both forms of aggressive behavior, future research could explore whether these processes vary across different developmental stages of adolescence. Bullying and cyberbullying may serve different social functions depending on the developmental phase, as peer hierarchies, status competition, and the use of digital platforms evolve throughout adolescence. Future studies could therefore examine whether the relationships identified in this study differ between early, middle, and late adolescence.
Finally, the robustness analyses using a Random Intercept Cross-Lagged Panel Model (RI-CLPM) did not show adequate model fit. This suggests that the associations observed in the traditional longitudinal mediation model may partly reflect stable between-person differences rather than purely within-person changes over time. Future research including additional measurement waves and alternative modelling strategies may help to better disentangle the extent to which envy and popularity motivations contribute to changes in bullying and cyberbullying behaviors within individuals across time.
Conclusions and practical implicationsThis study contributes to a better understanding of the socio-motivational mechanisms linking dispositional envy to aggressive behavior during adolescence. The findings showed that dispositional envy was associated with both bullying and cyberbullying behaviors over time, although the association with cyberbullying appeared less consistent across waves. In line with contemporary perspectives that conceptualize envy as a socio-affective signal arising from unfavorable social comparisons, the results suggest that envy may function as an indicator of perceived status threats within peer hierarchies. Furthermore, the need for popularity emerged as a key motivational mechanism linking dispositional envy to traditional bullying. Adolescents who experienced higher levels of envy were more likely to develop stronger motivations for social recognition, which in turn predicted greater involvement in face-to-face bullying. However, this mediating pathway was not observed for cyberbullying, suggesting that online aggression may follow partially different motivational processes.
These findings have practical implications for bullying and cyberbullying prevention. Educational interventions may benefit from addressing social comparison processes and status-related socio-affective experiences such as envy, while promoting emotional awareness and constructive ways of managing perceived status threats within peer groups. Programs aimed at preventing bullying could incorporate strategies that help adolescents recognize and regulate emotions linked to unfavorable social comparisons, as well as foster prosocial ways of gaining peer recognition and belonging. At the same time, cyberbullying prevention initiatives may need to emphasize additional processes such as emotional self-regulation, impulse control, and responsible digital behavior. Given the characteristics of online environments, such as reduced social accountability and greater behavioral disinhibition, educational programs may also encourage reflection on the social and psychological consequences of online aggression, even when such behaviors occur indirectly or under conditions of anonymity.
Ethical approval statementThe present study has been designed and conducted in strict accordance with the ethical principles outlined in the World Medical Association's Declaration of Helsinki, which governs research involving human subjects. The Ethics Committee of the Universidad del Desarrollo approved this study.
Consent statementThe study was conducted with the consent and collaboration of the school principals. Participation was voluntary, and both the students and their parents provided written informed consent. The students and their parents were thoroughly informed of the study's objectives, procedures and confidentiality measures. The research team emphasized that participants could withdraw from the study at any time if they felt uncomfortable.
Funding statementOpen Access funding provided thanks to the CRUE-CSIC agreement with Elsevier.
Data availability statementThe data that support the findings of this study are available on request from the corresponding autor, uponreasonable request. The data are not publicly available due to privacy or ethical restrictions.
CRediT authorship contribution statementCindy López: conceptualization, data curation, investigation, writing – original draft, writing – review & editing. Xavier Oriol: methodology, supervision, validation, visualization, writing – original draft, writing – review & editing. Rafael miranda: methodology, writing – original draft, writing – review & editing. Jorge varela: conceptualization, writing – original draft, writing – review & editing.
The authors confirm that there are no financial interests or personal relationships that could have influenced the work presented in this manuscript.
The authors extend their gratitude to the principals, teachers, and students for their valuable cooperation and contributions to this study. The authors would like to acknowledge the support provided through the CRUE-CSIC agreement with Elsevier, which enabled the Open Access publication of this article.
