The frequency of reports of abusive violence among US service members deployed to post-9/11 operations and the psychosocial associations of participating in such behaviors are largely unexamined. We examined reports of participation in and observation of abusive violence behaviors in a sample of deployed veterans (N = 1345) and explored their correlates, which were contrasted with reports of exposure to combat stressors, using Emotional Processing Theory as a framework for variable selection. Forty percent of the sample (46% of men, 35% of women) endorsed observing at least one abusive violence behavior and fourteen percent of the sample (23% of men, 6% of women) reported participation in at least one abusive violence behavior. The nomological networks of abusive violence participation and observation were highly similar, but distinct from combat exposure. Abusive violence participation and observation were comparable to combat exposure in their associations with externalizing problems (e.g., alcohol and substance use), but each had consistently smaller associations with internalizing distress (e.g., depression, PTSD symptoms). Reports of exposure to abusive violence, particularly via observation, appear to be relatively common among deployed veterans with varying severities of internalizing distress, and these unique experiences are associated with a range of mental and behavioral health symptoms and functional difficulties that diverge from those associated with combat exposure.
Appearance-based bullying represents a specific form of stigmatizing victimization that can undermine students' perception of school as a safe environment, particularly during adolescence when bodily characteristics become central to identity and peer dynamics. Despite its relevance, there is limited evidence on the prevalence of appearance-based bullying and how teachers' responses shape perceptions of school safety for students whose appearance deviates from dominant standards. This two-wave short-term longitudinal study-conducted with a sample of 968 students aged 10-18 from eight schools in southern Spain (Mage = 12.81, SD = 1.84; 45.8% girls; 68.8% secondary school)-pursued two aims: (1) to examine the prevalence of non-victimization, general bullying victimization, and appearance-based bullying victimization, and observe for differences by sex and educational level; and (2) to analyze whether sex, educational level, type of victimization, and students' perceptions of teachers' responses predict perceptions of school safety for students with diverse physical appearances. Appearance-based bullying victimization was reported by around 11% of students. There were no significant differences by sex or educational level in the distribution of victimization types. Latent profile analysis identified four distinct teacher response profiles: consistent intervention (67.3%), biased intervention (14.1%), ambivalent intervention (4.4%), and intermittent intervention (14.2%). Regression analyses showed that attending secondary school, experiencing appearance-based bullying, and perceiving inconsistent teacher responses were associated with perceptions of less school safety. The findings underscore the importance of coherent and sustained teachers' responses, as well as promoting school environments that explicitly affirm body diversity to foster inclusive and protective climates.
Nonphysical aggression and victimization are common in adolescence, yet their long-term associations remain poorly understood, particularly with regard to within-person processes. To disentangle stable between-person differences from dynamic within-person associations, this study employed a random-intercept cross-lagged panel model with four waves of data collected every 6 months over 2 years from 856 Taiwanese adolescents aged 13-16. At the between-person level, adolescents with higher average levels of nonphysical aggression tended to report higher average levels of nonphysical victimization. At the within-person level, positive deviations from an individual's mean level of nonphysical aggression were associated with subsequent negative deviations from their mean level of nonphysical victimization, whereas deviations in victimization did not predict later deviations in aggression. Multigroup analyses indicated no gender differences in within-person associations; however, boys had higher average levels of nonphysical aggression than girls at the between-person level. These findings suggest that adolescents' within-person changes in nonphysical aggression may reflect short-term functional processes that differ from the stable risk patterns observed at the between-person level. Intervention strategies are discussed.
While evidence indicates a connection between bullying involvement (perpetration and victimization) and peer relationships, the direction of this relationship remains debated. It is unclear whether peer relationships are a consequence or a precursor to bullying, and whether such association varies by gender. To address these gaps and capture fluctuations in daily school life, the present study employed a weekly diary method to examine the bidirectional relationships between bullying perpetration/victimization and peer relationships, as well as potential gender differences. A total of 452 adolescents (285 boys, Mage = 12.90, SDage = 0.48) completed the 7-week diary survey. Multilevel modelling revealed that weekly bullying/victimization significantly predicted subsequent decreases in peer relationships, while peer relationships did not predict later bullying or victimization, indicating a unidirectional pathway. Gender analyses showed that gender did not moderate these links but did differentiate bullying patterns: stratified analyses further revealed the unidirectional pathway only in boys, where verbal and relational perpetration/victimization predicted later declines in peer relationships. These findings not only theoretically expand our understanding of the dynamic interplay between bullying and peer relationships, unveiling unique cultural and social mechanisms of psychological bullying among boys, but also provide empirical evidence for developing targeted anti-bullying interventions for adolescent boys.
Impulsive acts of violence are perpetrated by a small proportion of society, which anecdotal evidence suggests often occurs in a rage state. Rage has generally been conceptualized as extreme anger, but it may be qualitatively different from extreme anger. Is rage best considered as part of the anger continuum or is it a different phenomenon altogether? The present research addressed this question in a large sample of university students (N = 545) who completed qualitative and quantitative measures of rage attacks, anger, emotional dysregulation, trait aggression and displaced aggression, angry rumination, adverse childhood events, head injuries, drug and alcohol use, and borderline personality disorder traits. Qualitative results showed that rage experiences are similar to anger experiences, but with some unique components. Path analysis found a similar pattern of correlates for rage and anger. Emotion dysregulation, trait physical aggression, trait displaced aggression, and borderline personality disorder traits predicted both rage and anger. Finally, in a latent profile analysis, we identified two subgroups of participants who are at risk for rage and anger. These findings largely support rage as a type of extreme anger with some additional features (e.g., loss of control, tunnel vision, aggression). Targeting the dysregulated aspects of people's personalities as early as possible may assist in reducing anger and rage.
Cyberbullying has emerged as a prominent form of online aggression, yet evidence on the role of online media remains fragmented because media-related variables are often treated as broad and undifferentiated constructs. This meta-analysis synthesized associations between six characteristics of online media use and cyberbullying perpetration. A systematic search of international databases identified 97 eligible studies published between 2004 and 2025, comprising 248,316 participants. Random-effects models were used to estimate pooled correlations, and moderator analyses examined cultural background and temporal patterns of content exposure. Five of the six focal variables were significantly associated with cyberbullying perpetration. Exposure to online violent content, exposure to online sexually explicit content, duration of online media use, and perceived online anonymity were positively associated with cyberbullying, whereas media literacy was negatively associated with it. Among these variables, perceived online anonymity showed the strongest association. By contrast, exposure to online prosocial content was not significantly associated with cyberbullying. Moderator analyses indicated that temporal pattern significantly moderated the association between exposure to online violent content and cyberbullying, with stronger coefficients for short-term than for long-term exposure. Most other subgroup differences were not significant. Sensitivity analyses showed that, despite substantial heterogeneity in most models, the overall pattern of findings remained stable. These findings suggest that online media should not be treated as a unitary correlate of cyberbullying. Distinguishing specific media characteristics may help refine theory and inform more targeted prevention efforts in online contexts.
Aggressive behavior has been linked to biases in social information processing, including heightened attention to threat and hostile interpretations of ambiguous social cues. The extent to which attention and interpretational biases uniquely or jointly contribute to aggression under provocation remains unclear. The present study investigated the effects of attentional bias toward angry faces and hostile interpretation bias on aggression using both self-report and behavioral measures in a non-clinical sample. Participants (n = 67) completed an emotional attentional blink task and a hostile interpretation bias measure, alongside self-report indices of aggression and antisocial behavior, and an experimental measure of provoked aggression. Hostile interpretation bias was positively associated with self-reported aggression and antisocial tendencies, whereas attentional bias toward angry faces was selectively associated with behavioral aggression. Critically, these cognitive biases were examined as potentially independent, overlapping, or interactive contributors to aggression under provocation, and attentional and interpretation biases interacted to predict aggressive behavior under conditions of high provocation, such that individuals high in both biases displayed the greatest levels of aggression. No interaction effects were observed for self-reported outcomes. These findings support social information-processing models proposing that multiple cognitive biases jointly contribute to aggression and highlight the importance of considering situational provocation and outcome modality when examining cognitive risk factors for aggressive behavior and preventive cognitive interventions in at risk populations.
Research has demonstrated the psychological and behavioral consequences of cyberbullying-for both perpetrators and victims alike. These effects highlight the importance of continued empirical attention towards validating efficacious cyberbullying interventions. However, the majority of cyberbullying interventions target youth and employ in-person and whole-school lessons that are derived from theory that does not distinguish traditional and cyberbullying. The current study aims to provide continued validity evidence of the "You're Not Anonymous" intervention, which (i) is derived from theory unique to cyberbullying, (ii) delivers the curriculum online using a short video, (iii) targets adults, and (iv) educates participants on how online behavior is not anonymous. US adult participants (N = 171) completed measures of cyberbullying perpetration, cyberbullying attitudes, and online disinhibition before being randomly assigned to view the YNA video or a control video. Afterwards the same online disinhibition and cyberbullying attitudes measures were completed. Finally, 1 month later, participants completed the cyberbullying perpetration measure again. Results showed that online disinhibition (specifically dissociative anonymity) decreased significantly for only YNA participants, and post-video online disinhibition mediated the relationship between the condition participants were assigned to and cyberbullying 1 month later. Taken together, these results suggest continued efficacy for the YNA intervention while highlighting the role that certain online disinhibition facets play in that efficacy.
Bullying has been shown to have adaptive functions, and some research has suggested this may be partly due to perpetrators making flexible use of both coercive and prosocial strategies. The current study probes this possibility further by examining whether bullying is differentially associated with two distinct affiliative strategies-genuinely cooperative affiliation and manipulative affiliation. Our study also examines whether bullying and the two affiliative strategies have differential relations with cooperative and competitive aspects of social power. A sample of 477 adolescents (45.5% girls; Mage = 15.50) completed peer nomination measures assessing direct and indirect bullying, cooperative affiliation, manipulative affiliation, and three facets of social power: social dominance (dominance and popularity), peer affection (likeability and number of friendships), and dating popularity. Multigroup path modeling by gender indicated that both direct and indirect forms of bullying were significantly associated with manipulative affiliation but not cooperative affiliation. Path analyses examining bullying and the two affiliative strategies in relation to social power found that bullying was positively associated with social dominance and dating popularity (direct bullying only), but these links depended on the level of manipulative affiliation. Manipulative affiliation had significant positive relations only with competitive aspects of social power (social dominance and dating popularity), whereas cooperative affiliation had significant positive associations with all three aspects of social power, and most strongly with peer affection. These findings underscore the importance of understanding how bullying may be combined with manipulative affiliation and highlight the potential of promoting genuine cooperation as an alternative to bullying within anti-bullying interventions.
Although social anxiety (SA) is typically conceptualized as an internalizing vulnerability, a subset of socially anxious individuals display heightened aggression. The present study examined whether vulnerable narcissism-particularly narcissistic self-centeredness-helps explain heterogeneity in the SA-aggression association. Chinese university students (N = 872) completed measures of SA (Social Interaction Anxiety Scale; Social Phobia Scale), vulnerable narcissism (Hypersensitive Narcissism Scale; self-centeredness and rejection sensitivity), and proactive/reactive aggression (Reactive-Proactive Aggression Questionnaire). Latent profile analysis using the two SA indicators and the two vulnerable-narcissism facets identified seven profiles. A high-SA/high-vulnerable-narcissism profile ("Vulnerable-Defensive"; 9.9% of the sample) reported aggression levels in the highest tier across all three aggression indices, together with a Moderate-SA/Moderate-VN profile (Class 5; 23.9%) and a High-SA/Low-VN profile (Class 6; 6.3%). Planned contrasts showed that, despite both profiles occupying the highest SA tier, the Vulnerable-Defensive profile-which additionally scored substantially higher on self-centeredness-reported higher reactive aggression (d = 0.32) and total aggression (d = 0.34) than a high-SA/low-vulnerable-narcissism profile. Complementary moderated regressions indicated that self-centeredness modestly strengthened the association between SA and proactive aggression (interaction B = 0.095, p = 0.032; f2 = 0.005) but not reactive aggression. These results suggest that SA is not uniformly associated with aggression; narcissistic self-centeredness helps differentiate high-SA individuals who are at heightened risk for aggressive responding.
Recent debates on violent video game (VVG) effects highlight the importance of analytical robustness and transparency in longitudinal research. Teng and Bushman reanalyzed our prior study, identifying several methodological concerns and arguing that our conclusions were unwarranted. In this response, we examine their reanalysis and assess whether their conclusions hold across alternative analytical specifications and operational decisions. Through comprehensive robustness checks available on OSF, we address concerns regarding model specification, missing data, measurement decisions, and time constraints. We further examine the unexpected empathy-aggression association and alternative VVG operationalizations. While some inconsistencies emerge across different specifications, the results do not consistently support General Aggression Model (GAM) interpretations in this dataset. We further identify inconsistencies and incomplete reporting in Teng and Bushman's presentation of results, data processing procedures, and interpretation of effect sizes. Finally, we argue that the current debate reveals a fundamental theoretical problem: without precise specifications of mechanisms, developmental patterns, and falsifiable predictions, GAM risks unfalsifiability when contradictory findings can all be accommodated as supporting evidence. We advocate for formal modeling approaches to enhance theoretical rigor and predictive specificity in violent media research.
With the growing prevalence of social media use among adolescents, its potential negative consequences warrant serious attention. This longitudinal study examined the associations between three common types of cyber social ostracism, namely cyber personal chat ostracism, cyber group chat ostracism, and cyber personal space ostracism, and adolescent suicidality and bullying behavior, as well as the mediating roles of thwarted belongingness and perceived burdensomeness. A total of 5478 adolescents were assessed across three waves spaced 6 months apart. The results showed that all three forms of online ostracism significantly predicted adolescents' suicidality through elevated thwarted belongingness, and that cyber personal chat ostracism and cyber personal space ostracism further increased suicidality via perceived burdensomeness. Multi-group analyzes demonstrated that the indirect pathway from cyber personal chat ostracism to suicidality via thwarted belongingness was significantly stronger for girls than for boys, whereas other pathways did not differ by gender. These findings highlight the distinct psychological mechanisms linking online ostracism to self-directed suicidality in adolescents, and underscore the importance of considering gender differences in prevention and intervention efforts.
Bullying during adolescence implies different forms of involvement that shape specific profiles, whose identification allows for understanding this phenomenon. However, it is essential to explore the mechanisms associated with the persistence or change of these profiles over time. Co-rumination may be crucial as a social-cognitive process in bullying development, reinforcing negative thought patterns among peers. The aim of this study was to examine involvement in bullying by identifying participation profiles and analyzing their longitudinal probability of stability or transition according to adolescents' levels of co-rumination. A total of 1537 adolescents aged 12-16 years (54% boys and 46% girls; MAgeT1 = 12.98, SD = 0.82) participated at two time points 1 year apart. The latent profile analysis conducted with Mplus 8.10 identified three bullying profiles: bully-victims, victims, and uninvolved, which showed stability across both time points. There was a tendency for bully-victims and victims to transition toward the uninvolved profile, although bully-victims also showed some probability of subsequently adopting the victim role. When considering co-rumination, adolescents with higher levels were more likely to belong to the victim profile across both assessments. Regarding transitions, the findings revealed that an increase in co-rumination was associated with a higher probability that victims maintain this profile, as well as with the transition of bully-victims toward victimization. These results provide a deeper understanding of the evolving dynamics of bullying profiles and offer valuable insights for designing interventions aimed at promoting effective coping strategies.
This study examined sources of heterogeneity in antisocial behaviors in a large-scale and nationally representative sample of n = 7655 youth from the Adolescent Brain and Cognitive Development Study. Separate latent profile analyses were conducted for males and females using indicators of physical aggression, relational aggression, callous-unemotional (CU) traits, and conduct symptom severity. Distinct profiles emerged within each sex. Among females, profiles were primarily differentiated by levels of CU traits, whereas among males, physical aggression played a more central role in distinguishing groups. In both sexes, a higher-risk profile characterized by elevated aggression and/or CU traits was identified alongside a lower-severity marked by minimal conduct problems and aggression. Profiles also differed significantly in overall distress, with the higher-risk male profile demonstrating the greatest distress and the lower-risk female profile the least. Findings highlight meaningful sex-specific configurations of aggression and CU traits and underscore the importance of person-centered, sex-informed approaches to developmental models, assessment, and intervention in youth with antisocial behavior.
Guided by social exchange theory, we conducted two experiments and a longitudinal field study to examine the role of trust as a mediator between victimization from two forms of online mistreatment (cyber-incivility and cyberbullying) and three relational outcomes: supervisor satisfaction, organizational citizenship behaviors (OCBs), and counterproductive work behaviors (CWBs). In Experiment 1, 90 employees were randomly assigned to read email exchanges that contained either uncivil, cyberbullying, or neutral messages between a co-worker and a team leader. Results indicated that participants perceived the leader as less trustworthy when cyber-incivility or cyberbullying messages were communicated. In Experiment 2, 136 employees were randomly assigned to read a series of email exchanges between a co-worker and a team leader that were either low or high in trust. Participants reported greater satisfaction and likelihood of engaging in OCBs and lower likelihood of engaging in CWBs with a highly (as compared to low) trustworthy leader. In the longitudinal field study, 313 employed adults completed four waves of an online survey. Findings indicated that trust mediated the relationships between cyber-incivility or cyberbullying victimization and supervisor satisfaction and OCBs (but not CWBs). Implications for future research and practice are discussed.
Although collective violence remains a pervasive issue affecting many societies today, the specific psychological mechanisms underlying individual differences in support for collective violence are relatively understudied. In four studies, using five samples from Lebanon, Syria, and Turkey (total N = 3758), we conceptualize and develop a new multidimensional scale for measuring individual differences in collective violence beliefs. Contrary to some prior theorizing and extant research on interpersonal violence, we found that people's justifications for collective violence are structured based on the target of the act rather than the intensity of the violent act. Consequently, we developed and validated a Two-Dimensional Collective Violence Beliefs Scale (CVBS: 2D) that distinguishes between violence targeted at outgroup members, referred to as diffuse collective violence, and violence targeted at outgroup leaders, which we term upward collective violence. We validate this novel scale in multiple contexts and discuss the implications of the two dimensions of collective violence revealed in this study.
Aggressive behavior when moving through traffic is recognized as a serious problem and safety hazard with respect to car drivers. By contrast, aggressive behavior by cyclists has received little attention, although it also has potential for harm to others and may lead to conflict between road users. The limited body of past research focused on hostile anger expression by cyclists as a way of expressing anger. By contrast, the current research provides evidence on cyclists' use of instrumental aggression in terms of making faster or smoother progress. In a cross-sectional study with a representative sample of cyclists in Germany ( N = 1240, 685 men, 555 women), participants completed measures of their aggressive cycling behavior, frequency of cycling, perceptions of others' aggressive cycling, and traffic-related stress. A subgroup of participants who also identified as car drivers ( n = 1135) additionally completed a measure of aggressive car driving. Self-reported aggressive cycling was higher among men than among women. Age and frequency of using the bike were negatively related to aggressive cycling, whereas traffic-related stress showed a positive association. Reflecting the “better-than-average” effect, participants believed that others showed significantly more aggressive cycling than themselves. Finally, we found a strong correlation between self-reports of aggressive cycling and aggressive car driving. The findings highlight the need for further research on aggressive cycling and the development of effective strategies for reducing aggressive behavior on the road in different modes of transport.
The association between cumulative risk and psychosocial adjustment typically follows either linear (additive model) or nonlinear patterns, with the latter comprising steeling and sensitizing models. This two-wave longitudinal study (baseline: T1; 6-month follow-up: T2) investigated how cumulative violence exposure relates to proactive versus reactive aggression and its dynamic in 1208 Chinese adolescents (Mage = 12.73 years; 51.4% boys). Hierarchical regression analyses showed significant nonlinear effects of cumulative violence exposure on both proactive and reactive aggression across waves. Restricted cubic spline analyses revealed that the relationship between cumulative violence exposure and T1 proactive aggression supported the sensitizing model, while the relationships with T2 proactive aggression and both T1/T2 reactive aggression aligned with the steeling model. Specifically, T1 proactive aggression exhibited escalating nonlinear growth with increasing exposure, whereas T2 proactive aggression and both T1/T2 reactive aggression demonstrated initial increases followed by declines as exposure intensified. These findings provide critical implications for interventions targeting violence-exposed adolescents.
Animal harm often results in lenient legal responses, with prosecutions primarily targeting offenses against companion animals, while cases involving farm animals remain significantly underreported. This study examined pet-favoritism bias in moral judgments, highlighting how animal category can shape perceptions of cruelty. Participants (N = 569) evaluated sadistic harm toward either pet or farm animals by rating proposed prison sentences and moral condemnation. They also reported pet ownership, dietary habits, political orientation, religiosity, and beliefs about human superiority. The results confirmed pet-favoritism, with harm to pets prompting greater moral condemnation and harsher punishment. Moderation analyses showed this bias was more pronounced among meat-eaters, non-pet owners, and those who believed in human superiority. We discuss the mechanisms underlying more lenient judgments of animal harm.
Aggression and emotional distress are all too common in at-risk adolescents in closed residential institutions, which often contain restrictive conditions such as overcrowding, lack of autonomy, and shifting staff-youth relationships. In war-torn areas like Israel, these restrictive conditions are amplified with additional stressors such as the threat of injury or even death. In such settings, feasibility studies are greatly needed to investigate how to reduce aggression and emotional distress in vulnerable adolescents. This feasibility randomized clinical trial evaluated the effectiveness of a virtual reality (VR)-based relaxation intervention in decreasing self-reported aggression tendencies and emotional distress and increasing positive psychological states among at-risk Israeli adolescents in residential care. Fifty-two participants from different residential facilities were randomly assigned to a treatment group (n = 27), who received twice-weekly VR relaxation sessions, or a waitlist control group (n = 25). Self-report measures of aggression and emotional distress were administered at baseline, midway through the intervention (2 weeks), and at the end of the intervention (4 weeks). Mixed-effects linear models suggest that the treatment group demonstrated clear reductions in both aggression and emotional distress across the intervention period relative to the control group (Hedges' gs: 0.85-1.42). In addition, immediately before and after each VR session, participants in the treatment group reported consistent increases in positive mood and flow experience (Hedges' gs: 0.45-0.73). These preliminary findings support the utility of brief VR relaxation practices in reducing negative emotional states and fostering well-being among high-risk youth in closed institutional settings. Trial Registration: This study was pre-registered on the Open Science Framework (OSF) prior to data collection. The pre-registration is available at: https://osf.io/pr3md/.
