Introduction

The safeguarding of children, as the foundation of a nation’s future and the aspiration of its populace, is of paramount importance. Worldwide, a significant problem is improving the governance and prevention of adolescent delinquency. China has achieved notable advancements in this domain, owing to collaborative endeavours across several societal sectors, augmented investment in judicial resources, and the extensive dissemination of legal education (Yao 2009). Nevertheless, concerning trends endure. According to the White Paper on Juvenile Prosecution Work (2024) published by the Supreme People’s Procuratorate (2025), juvenile crime currently exhibits intertwined characteristics of younger age and increasing complexity. Procuratorates nationwide handled 9317 cases involving 14 to 16-year-olds for review and prosecution, among which 34 cases of severe violent crimes committed by 12 to 14-year-olds were approved for prosecution by the Supreme People’s Procuratorate. It is particularly alarming because a substantial proportion of offenders are students from vocational schools. A recent incident in Jiangsu Province featured a 21-year-old vocational school student who, frustrated by his failure to graduate and inadequate internship remuneration, perpetrated intentional homicide, leading to 8 fatalities and 17 injuries. These events highlight the pressing necessity to investigate the root causes of violent crimes in this group, prompting essential inquiries that require careful consideration and response.

This is also a significant issue in the United States. Researchers in the USA have increasingly concentrated on the spatiotemporal distribution patterns of juvenile violent crime. International researchers have investigated the correlation between educational attainment and crime rates, emphasizing education’s possible function in mitigating crime (Lochner and Moretti 2004; Lochner 2010). Nonetheless, the concentration effect of school education may intensify social interactions among adolescents, potentially resulting in increased conflicts and group-based crimes among students, which could inadvertently elevate crime rates, especially concerning violent offenses (Jacob and Lefgren 2003). American researchers have examined preventive interventions for adolescent violent crime from 3 primary perspectives: family, schools, and social institutions or community groups (Greenwood 2008; Webster-Stratton and Hammond 1997).

Many Chinese experts have acknowledged the problem of adolescent delinquency and have performed thorough research on it. These works predominantly encompass fields including criminology, law, sociology, psychology, and education, illustrating the fundamentally multidisciplinary character of this topic. Criminological study predominantly focuses on the analysis of the causes and contributing factors of juvenile delinquency. Such conversations frequently encompass social, psychological, and educational aspects. Significant research in China has investigated the correlations between left-behind experiences, deviant peer affiliations, and their association with violent offenses perpetrated by juveniles (Sun et al. 2023). Legal scrutiny, especially in criminal law and procedure, has focused on substantive matters, such as the precise application of age-related provisions for juvenile criminal responsibility (Qiu 2023), and procedural concerns, including the protection of the legal rights of young offenders (Yu 2024). These fields jointly investigate diverse methodologies and tactics for tackling juvenile misbehaviour. Furthermore, the Frustration-Aggression Theory (FAT), a significant paradigm with a history of 80 years, is extensively cited in China. A substantial body of research utilizes FAT to elucidate aggressive behaviour in teenagers, further broadening its explanatory framework by integrating psychological resilience and self-control as mediating factors (Sun et al. 2023; Liu et al. 2022). Research concerning the vocational school population (VSP) primarily highlights their job development opportunities and educational structures, while neglecting their mental health and experiences of bullying inside the school environment. Nevertheless, research about the participation of this cohort in violent crimes is still very limited. Moreover, current research has yet to establish a definitive connection between FAT and the root causes of violent crimes among these individuals, indicating significant potential for further investigation.

This study concentrates on a certain population, specifically the VSP. The main aim of this research is to examine the fundamental reasons of violent crimes perpetrated by persons in this group via the perspective of FAT. We aim to examine the mechanisms connecting frustration events to aggressive actions, focusing on how these episodes affect the perpetrator’s selection of methods and targets in violent crimes. To do this, we executed an extensive data collection process, employing both questionnaire surveys and in-depth interviews with members of VSP. The survey collected generic information, whereas the interviews offered profound insights into the individuals’ experiences with dissatisfaction and how these events influenced their engagement in violent behaviour. Our research indicates that frustration events substantially influence this demographic, leading to an increased propensity for violent criminal behaviour. These individuals frequently prey on vulnerable victims and utilize objects like belts as their main means of aggression, a behaviour directly shaped by their experiences of frustration. These findings provide essential insights that can inform the formulation of more efficient and targeted strategies to regulate and reduce violent crimes within VSP.

Vocational School Population (VSP)

Definition of VSP

The term “Vocational School Population” (VSP) in this study denotes individuals who either leave the educational system after finishing junior high school, a crucial element of China’s compulsory education, or engage in vocational education at multiple tiers, encompassing secondary vocational education at the high school level and higher vocational education at post-secondary institutions. This category comprises persons who, having exited formal education prematurely or opted for vocational training, participate in skill development programs tailored for certain occupations or crafts.

Vocational education is a fundamental element of the global educational framework, underscored by the founding of UNESCO-UNEVOC in September 2000 in Bonn as UNESCO’s centre for technical and vocational education and training (TVET). UNESCO underscores the importance of vocational education in meeting societal and labour market requirements, ensuring that education stays pertinent and adaptable to the changing economic landscape.

The United States has a longstanding legacy of vocational education, referred to as Career and Technical Education (CTE). CTE encompasses several institutions and programs aimed at providing learners with skills for specific trades, applied sciences, contemporary technology, and overall career readiness. These programs integrate academic and vocational coursework, offering a comprehensive education to equip students for the labour market. Consequently, vocational education in the USA has evolved to be more comprehensive, with the objective of cultivating versatile professionals for the changing technological environment.

The United Kingdom, as a forerunner in Vocational Education and Training (VET), possesses a lengthy and distinguished history, having cultivated a robust and internationally acknowledged system. A significant accomplishment is the establishment of the National Vocational Qualifications (NVQ) system, which amalgamates general and vocational qualifications and has developed into the present Regulatory Qualifications Framework (RQF). These systems have been crucial in improving educational quality and the societal acknowledgment of vocational graduates.

The advancement of VSP

The middle school test in China, occurring at the conclusion of compulsory education, is a pivotal juncture in students’ academic trajectories, as not all are able to progress to regular high schools. Students who do not secure entrance may seek overseas education, enrol in private high schools, or undertake vocational education, with many ultimately advancing to higher vocational institutes or joining the workforce. The 2023 National Educational Development Statistical Bulletin reports 7085 secondary vocational schools with an enrolment of 4.54 million students, alongside 5.55 million students in higher vocational institutions, highlighting the significance of this considerable and varied demographic in the vocational education system Ministry of Education of the People’s Republic of China (2024).

In the United States, vocational education is pivotal, with one million sub baccalaureate certificates and 1billion associate degrees awarded in 2021–22, as reported by the NCES “Report on the Condition of Education 2024” (Irwin 2024). Unlike in China, where vocational tracks frequently stem from academic deficiencies, in the USA, the decision to engage in CTE is primarily influenced by employment opportunities. CTE programs are more concise and cost-effective than four-year degrees, facilitating expedited entry into the workforce. Furthermore, numerous CTE programs present ample internship chances, furnishing practical experience that aids students in acclimating to professional requirements.

The frustration encountered by VSP

VSP encounters numerous problems that can be classified into 3 primary dimensions: individual-level academic and social interaction challenges, familial-level issues and societal-level factors. These frustrations can range from acute stressors like a specific academic failure to chronic conditions stemming from structural inequalities such as persistent economic hardship or systemic discrimination against vocational education. From an individual standpoint, VSP generally demonstrate inferior academic performance relative to students enrolled in conventional high schools in China. Studies indicate that the enrolment rate in secondary vocational schools is persistently low, and academic achievement in these institutions is typically subpar (Liu and Xie 2021). Furthermore, VSP endure considerable test anxiety, with female students indicating elevated anxiety levels compared to their male peers (Zhou 2012). Academic performance significantly impacts parents’ preferences concerning their children’s subsequent education, as superior academic outcomes diminish the probability of parents endorsing vocational education options (Shen and Wan 2024). VSP encounter difficulties in interpersonal communication, with research indicating that social skills are the fourth most critical psychological need for these pupils (Chen 2015). Moreover, interpersonal challenges are a prevalent psychological concern, with 64.75% of vocational school students indicating such issues (Li et al. 2012).

In familial matters, VSP encounter considerable problems, including the experiences of being left-behind children and exposure to domestic violence. The “White Paper on the Development of Vocational Education in China” indicates that more than 70% of vocational school students originate from rural regions, with a significant number also categorizing themselves as left-behind children and adolescents. Students who are left behind frequently possess restricted developmental resources, which is associated with a significant discrepancy in comparison to their non-left-behind counterparts (Zhang et al. 2022). Research suggests that 57.44% of vocational students from left-behind backgrounds encounter mental health concerns, a rate markedly greater than that of their non-left-behind peers. Extended separation from parents throughout crucial phases of personality development can result in enduring adverse impacts on mental health (Zhou et al. 2021). Moreover, research indicates that vocational students from homes affected by domestic violence likely to report greater self-experience disharmony and demonstrate less self-harmony relative to their counterparts from non-violent households (Na and Xu 2017).

Beyond individual and familial challenges, VSP often face frustrations rooted in structural violence. For example, China’s high school entrance examination system effectively streams many students into vocational tracks, which are often perceived as less prestigious and may offer limited upward mobility. This educational stratification can contribute to feelings of inadequacy and blocked opportunities. Furthermore, VSP may encounter employer discrimination in the labour market, reinforcing their marginalization and economic precarity. These systemic factors contribute to a pervasive sense of frustration and disadvantage. Research reveals various detrimental sociocultural elements affecting VSP’s illegal and illicit behaviours in both media and activity contexts. VSP are affected by exposure to hazardous content in literature, audiovisual media, and digital platforms, while entertainment establishments such as karaoke bars, dance halls, and internet cafes also have adverse consequences. Moreover, societal stigmatization of vocational schooling frequently causes VSP to perceive their dignity as unrecognized, engendering a deep sense of inferiority. Neglected social changes, often undervalued, also impact this demographic. Furthermore, virtual interactions in online environments may be linked to a deficiency in accountability for VSP, resulting in role displacement and fostering a persistent sense of emptiness and disillusionment (Na and Xu 2017; Yin 2022).

In the United States, researchers have investigated the difficulties encountered by VSP. Data from the American Association of Community Colleges (AACC) indicates that community college graduation rates have persistently remained low. The National Student Clearinghouse (NSC) monitors student advancement over a 6-year period across many institutions. The NSC’s completion report indicates that the 6-year graduation rate for students who commenced their studies in fall 2010 at public 2-year schools was 27%. Furthermore, 3.3% completed their programs at an alternative 2-year university, whilst 9% graduated from a four-year institution. In all, 39.3% of community college students attained program completion within 6 years, whether at their initial institution or elsewhere (Juszkiewicz 2017).

The investigation of violent offenses perpetrated by VSP

China has made notable progress in tackling juvenile delinquency with a criminal strategy that balances leniency and strictness, resulting in beneficial outcomes. Yet, the incidence of violent crimes among VSP continues to be a problem. The research on violent crimes committed by VSP has concentrated on crime trends, causative factors, and management techniques. Research has delineated primary crime categories, offender profiles, and crime hotspots, shown by the Supreme People’s Court’s 2016 report, which revealed that school violence is predominantly associated with specific types, with offenders from vocational high schools constituting 26.42%. The determinants of violent crime have been examined at the individual, familial, and social levels. At the individual level, warped values, economic adversity, detrimental social relationships, and academic underachievement are significant causes. Family-related issues comprise parenting methods and parental marital status, whereas societal factors involve inadequate school supervision and community oversight. Management techniques advocate for synchronized efforts across various levels to cultivate a supportive environment for VSP’s robust development (Wang 2019; Zhou et al. 2012). The findings indicate a comprehensive strategy for addressing youth violent crime, incorporating individual, familial, and social interventions.

Scholars worldwide have conducted extensive research and discussions on the causes of violent crimes committed by VSP. Theories such as Social Learning Theory, Labelling Theory and Strain Theory have provided perspectives for explanations. Social Learning Theory emphasizes that violent behaviour is not instinctive but learned through observing and imitating violent role models in social environments (Burgess and Akers 1966). For VSP, frequent exposure to violent models in families and schools may lead to acquisition of violent behavioural patterns. Labelling Theory posits that an individual’s identity as a criminal does not stem from the behaviour itself but from the internalization of negative social labels (Becker 1963). For VSP, when they are repeatedly labelled as troubled youth or dangerous group, they may gradually identify with this identity. Strain Theory argues that individuals experience strain when they cannot achieve socially approved goals such es economic success through legitimate means, then violent behaviour may serve as a way to release strain (Merton and Robert 1938). For VSP, structural barriers may cause long-term stress and increase the risk of violent crime.

Frustration-Aggression Theory (FAT)

The substance and evolution of FAT

The Frustration-Aggression Theory (FAT), an early psychological model aimed at elucidating violent behaviour, possesses a substantial and enduring international scientific legacy. In the pursuit of their objectives, individuals inevitably face both internal and external challenges. During such instances, a prevalent response is to eradicate these impediments, with violence serving as one of the most expedient means, characterized by aggressive conduct directed at eliminating the obstruction. In 1939, Dollard et al. introduced the Frustration-Violence Theory, which officially postulated the connection between frustration and violence. They contended that an individual’s frustration precipitates a sort of aggressive behaviour, which is a subsequent expression of that frustration. The frustration that incites aggressive behaviour is not a direct action but a stimulus produced by the individual following frustration, which results in aggression, and this inducible stimulus either instigates or hastens the manifestation of aggressive behaviour. They posited an intrinsic causal link between frustration and crime: frustration engenders aggression, and aggression is invariably dependent on the past occurrence of frustration (Dollard et al. 1939). Berkowitz, via his research advancements, found that frustration does not invariably result in extreme behaviours, leading to a revision of FAT (Berkowitz 1989). Berkowitz posited that violent conduct arises when frustration induces emotional arousal or a state of readiness in an individual, which will only culminate in violence if the environment is conducive to such behaviour. In other words, if there is merely arousal or preparedness in the absence of a threatening environment, it is challenging for the individual to exhibit violent conduct. This elucidates why certain individuals do not display aggressive behaviour despite encountering frustration. He also posited that frustration creates a readiness for aggression, but aggressive behaviour typically requires the presence of releasing environmental cues or triggers. This refinement moves beyond a simple linear model to acknowledge the critical role of the situation. Anderson and Bushman expanded upon prior research on FAT to introduce the General Aggression Model (GAM), which asserts that personal and environmental factors shape an individual’s cognitive processes, consequently impacting aggressive behaviour, with aggressive attitude cues being a significant factor. Behavioural triggers, including frustration, provocation, pain, and discomfort, can precipitate violent conduct. The hypothesis posits that frustration is a significant contextual determinant of violence. GAM provides a more comprehensive framework, suggesting that frustration, as a situational variable, interacts with personal variables, such as aggressive personality to influence internal states, which in turn affect appraisal and decision-making processes leading to aggressive or non-aggressive outcomes (Anderson & Bushman, 2002).

The mechanistic pathway of FAT

Rosenzweig posited that irritation elicits many reaction patterns, which he categorized into 3 primary types: extropunitive, intropunitive, and impunitive (Rosenzweig 1934). Extropunitive reactions arise when individuals ascribe their frustration to external circumstances, prompting them to channel their anger outward, frequently resulting in verbal or physical aggressiveness. This response is closely associated with aggressive behaviour and is seen as a major factor in violent crime. Conversely, intropunitive reactions entail the internalization of displeasure, which may contribute to emotions of guilt and self-directed ire. These emotional responses may result in melancholy or, in severe instances, suicidal inclinations. Nevertheless, in contrast to extropunitive reactions, intropunitive responses are less prone to exhibit outward violence and hence present a diminished risk for criminal conduct. Impunitive reactions entail downplaying or neglecting the displeasure, eschewing both external attribution of blame and internalization of guilt. This strategy may mitigate aggressive outbursts, but it fails to address the underlying anger. When frustrated experiences aggregate or escalate beyond an individual’s coping capacity, impunitive responses may transform into tropunitive ones, resulting in heightened violence and destructive behaviours. This shift can intensify the adverse effects relative to a straight extropunitive response, resulting in increased hostility and more severe consequences.

The Implementation of FAT

FAT has developed from a theory of violent crime to a comprehensive framework with several applications, including the analysis of aggressive driving behaviour. In 1998, Shinar empirically implemented this concept in the context of driving, positing that road-related frustrations, such as traffic congestion, incite drivers’ violence (Shinar 1998). The idea has been utilized in China to elucidate workplace aggressiveness (Long and Zhou 2007) and has also offered insights into the origins of medical-related violence (Zhao 2023). A significant portion of current research remains concentrated on utilizing FAT to examine the mechanisms of violent crime (Li 2014), teenage aggression, and school bullying (Qu et al. 2012).

VSP is attracting considerable attention in both scholarly and practical domains. This has resulted in increased attention to the difficulties encountered by VSP. FAT, originally designed to elucidate aggressive conduct and violent crime, has progressed to investigate the mechanisms of frustration in adolescent aggressiveness and school violence. Current research has enhanced comprehension of the challenges encountered by VSP. Nonetheless, there is a deficiency of studies that particularly apply FAT to this demographic. Considering the increasing incidence of violent crimes associated with VSP, it is essential to utilize FAT to investigate how frustration episodes prior to violent crimes may incite later aggressive behaviours, thereby establishing a theoretical basis for more effective prevention methods.

The Vulnerability to frustration and violent behaviour of VSP

In China, the susceptibility to frustration and tendency toward violent behaviour among VSP are not isolated phenomena but are deeply rooted in specific social and legal contexts. China’s criminal liability age system adopts a hierarchical approach to juvenile accountability. Adolescents aged 14 to 16 are only liable for 8 categories of serious crimes. This legal framework aligns closely with the age distribution of vocational school students who are predominantly 15–18 years old, potentially weakening behavioural constraints on this group. Moreover, the 2021 Criminal Law Amendment (XI) expanded liability to include specific violent crimes committed by 12 to 14-year-olds, directly addressing the trend of juvenile delinquency reflected in “little devil” cases, aiming to deter adolescent violence through legal means. For minors not subject to criminal liability, China primarily relies on parental supervision and specialized schools for rehabilitation. These institutions use quasi-military management, psychological intervention, and vocational training to guide troubled youth toward normalization. The high school entrance exam tracking system channels many middle school graduates into vocational schools, structurally labeling vocational students as “academic failures.” Besides, unlike in foreign countries, vocational schools and their students in China occupy a disadvantaged social position. Meanwhile, left-behind children constitute large part of VSP. The absence of familial support remains a potential trigger for frustration and violence. These overlapping factors collectively forming the unique causes of high frustration sensitivity and violence among VSP, creating a frustration accumulation–violence release cycle.

Research Questions

This research aims to explore the types of frustration events that affect the perpetration of violent crimes by VSP, including frustration events at the individual, family, and societal levels, and how these events influence their choice of methods and targets for committing violent crimes. The aforementioned frustration events may also play a significant role in influencing their choices of tools and targets in the commission of violent crimes.

What types of frustration events may influence VSP to commit violent crimes, and to what extent do these frustration events affect their choice of crime tools and targets?

Methods

Participants

This study sought to investigate the negative experiences faced by VSP before their engagement in violent criminal activities, utilizing field research done at a juvenile detention centre in a designated area. It comprised 30 male participants, each of whom had been convicted of violent offenses and was serving their sentences during the study period. All individuals were minors when they committed the violent offenses, but had attained adulthood by the time the research was done. To maintain ethical compliance, no personally identifiable information, including participants’ precise ages, was gathered during the study. This group’s violent crimes included robbery, rape, and the deliberate infliction of serious physical damage. It should be noted that during the research process, authorization to interview female VSP individuals and a non-offender group was not secured. Consequently, baseline characteristics for these groups and comparative analyses with the study participants are not included. The juvenile detention centre investigated in this study, one of the designated area’s largest youth correctional institutions, houses young offenders originating from various districts and counties within the designated area, as well as from other provinces and cities across China.

Instruments

Questionnaire surveys and interviews were used in this study to perform a thorough analysis of both quantitative and qualitative research methodologies. Due to time limits, the questionnaire survey is restricted to evaluating participants’ levels of irritation and aggression, hence complicating the capturing of the personal, familial, and social hardships encountered by participants before engaging in violent crimes. Therefore, it is imperative to amalgamate the results from the questionnaire survey with focused interviews of representative persons to acquire a more nuanced comprehension of their particular situations. This method enables a more profound examination of the impact of frustration-inducing events on the perpetration of violent acts by VSP.

This study employs the Sense of Frustration Scale (SFS), experimental objects intended to elicit frustration, the Reactive-Proactive Aggression Scale (RPAS), and a Self-Constructed Scale (SCS) created by the authors.

Firstly, the SFS, created by Keller and Dauenheimer in 2003, is intended to assess an individual’s perceived degree of irritation. The assessment consists of 4 items, each evaluated on a 7-point Likert scale, with the cumulative score reflecting the level of frustration. Elevated scores indicate increased levels of frustration. The Cronbach’s alpha coefficient for the annoyance scale assessed prior to the writing activity was 0.961. The Cronbach’s alpha coefficient for the frustration scale assessed post-writing assignment was 0.881.

Secondly, the experimental materials for creating frustration scenarios are derived from the recall priming paradigm established by Duclos et al. in 2013. These materials intend to provoke frustration by encouraging individuals to recollect and contemplate their adverse experiences, especially those from the recent past or preceding their confinement. Participants are directed to concentrate on negative situations from their lives that may have resulted in considerable emotional turmoil. Participants are prompted to remember instances where their endeavours faced rejection, their anticipations were unmet, or commitments made by others were not realized. The instructions encompass examples such as experiencing rejection in a novel social setting despite exhibiting enthusiasm, failing to attain an academic objective despite committed effort, or occurrences of unfulfilled promises made by parents.

Thirdly, the RPAS utilized in this study is derived from the version created by Raine in 2006 and then modified by Fu Junjie in 2009. This instrument comprises 23 items, categorized into 2 dimensions: reactive aggression (11 items) and proactive aggression (12 items). The items assessing reactive aggression consist of statements numbered 1, 3, 5, 7, 8, 11, 13, 14, 16, 19, and 22, whereas the other items evaluate proactive aggression. The questionnaire has high reliability, with an overall Cronbach’s Alpha coefficient of 0.924, 0.866 for the reactive aggression dimension, and 0.857 for the proactive aggression component.

Finally, taking into account the distinct attributes of the target demographic in this study, characterized as persons who were either students or employed previous to their criminal crimes, the authors developed the SCS. This scale was developed to capture specific school-related aggressive behaviours relevant to the VSP context prior to their incarceration, as existing instruments did not fully cover these nuances. Based on a review of literature on the research of violent behaviours of VSP, this bespoke tool comprises 8 items evaluated on a 3-point Likert scale, focusing on dimensions such as interpersonal connections in educational environment, scholastic difficulties, and instances of violent or aggressive conduct. This customized method guarantees that the distinct experiences and situations of the participants are thoroughly documented for analysis. The Cronbach’s alpha coefficient for this scale is 0.790. While it demonstrated acceptable internal consistency, we acknowledge it has not been validated against external objective records, such as school disciplinary reports, which was not feasible for this incarcerated sample retrospectively.

This study aimed to thoroughly investigate the impact of setbacks in personal, familial, and social contexts on the perpetration of violent crimes by VSP. It focused on identifying the specific types of setbacks that significantly influence their violent criminal behaviour and how these setbacks affect their selection of targets and methods of attack. To achieve this, findings from a questionnaire survey were employed to select 6 representative individuals for detailed, one-on-one interviews using a purposeful sampling strategy. The selection was based on 2 criteria. Firstly, scores on the frustration scale. We selected individuals with both high and medium scores to capture a range of experiences. Secondly, the type of violent crime committed, including representatives from both robbery and rape cases to ensure diversity of perspectives. They were requested to elucidate the circumstances behind their crimes and the challenges they faced on personal, familial, and social levels prior to their offenses. The study examined the psychological states of the interviewees after committing violent acts, emphasizing how their defeats affected their choices of targets and attack techniques. All interview data were meticulously recorded in textual format.

Execution procedures

The study process was thoroughly clarified to all participants prior to their engagement in the survey. The approach employed in this study comprised the following steps. Initially, a situational frustration induction experiment was executed. Participants were first instructed to complete the SFS while experiencing a tranquil emotional state. Subsequently, they received explicit instructions pertaining to the situational frustration induction experiment. Participants were thereafter instructed to spend 15–20 min recollecting past incidents that had induced frustration and to record these feelings in writing. Upon concluding the recall and recording process, participants were directed to complete the SFS once more. They successively completed the RPAS and the SCS. Third, 6 participants were chosen for comprehensive one-on-one interviews based on the selecting criteria. The interviews were performed individually, and the results were methodically recorded for subsequent study.

Analysis

The quantitative research segment of this study employed SPSS 27.0 software as the principal analytical instrument. Before analysis, the data gathered from the questionnaire survey were input into an Excel spreadsheet and then imported into SPSS 27.0 for processing. The analytical approach commenced with an evaluation of the reliability of each questionnaire element. An assessment of common technique bias was subsequently performed utilizing Harman’s single-factor test. The quantitative study included pre- and post-assessments for the SFS, the RPAS, and the SCS. A paired-sample T-test was used to compare the pre- and post-measurement values of irritation to validate the frustrating condition. Subsequently, Pearson’s product-moment correlation analysis was employed to investigate the correlations among the 5 variables: post-test frustration, reactive aggression, proactive aggression, total aggressive behaviour score, and aggressive behaviour in school. Ultimately, following the outcomes of Pearson’s product-moment correlation analysis, where post-test frustration served as the independent variable and reactive aggression, proactive aggression, total aggressive behaviour score, and aggressive behaviour at school constituted the 4 dependent variables, a linear regression analysis was performed.

This study utilizes Nvivo 12 Plus software as the analytical instrument for the qualitative research component. The qualitative data from the 6 in-depth interviews were transcribed verbatim. Analysis was conducted using the software and followed Braun and Clarke’s (2006) thematic analysis approach, integrated with a multi-tier coding process. Initial codes were generated by systematically working through each transcript, identifying segments of text relevant to the research questions, particularly focusing on experiences of frustration and their perceived link to violent acts, tool and victim selection. These codes were often descriptive and close to the participants’ own words. This generated 73 initial concepts. Similar initial codes were then grouped into potential themes. For example, codes related to academic difficulties, teacher conflicts, and dislike of studies were grouped under individual-level academic and social interaction challenges. We looked for patterns and relationships between codes. This process involved moving between the coded extracts and the developing set of themes. Themes were then reviewed against the coded extracts and the entire dataset to ensure they accurately represented the data and answered the research questions. This led to the refinement and synthesis into third-level codes.

Results

The results of quantitative research

Harman’s single-factor test

This study employs anonymous questionnaires in the quantitative research component to mitigate any errors in the measuring process and regulate the administration procedure. The utilization of self-report scales in the questionnaire survey component of this study may result in common method bias. A Harman’s single-factor test was performed for analysis, and exploratory factor analysis was utilized to assess the existence of significant common method bias. All items from the 4 scales included in the quantitative research underwent exploratory factor analysis. The findings reveal that there are 9 components with eigenvalues exceeding 1. The cumulative variation accounted for by the first factor is 35.89%, which falls short of the essential threshold of 40%. This signifies the absence of a substantial common method bias concern in this study.

Analysis of setback situation validity

A paired sample T-test was performed on the scores of SFS obtained from participants prior to and following the writing activity to evaluate the efficacy of the frustration scenario arrangement. This study obtained 30 valid data sets, with the statistical results presented in Table 1. Table 1 indicates that the average frustration score of participants was significantly elevated following the writing activity compared to before it (p < 0.05). This suggests that employing the memory induction paradigm to prompt individuals to recount their past negative experiences can elicit feelings of annoyance, hence demonstrating the efficacy of this frustration-inducing scenario.

Table 1 The results of paired sample T-test.
Full size table

Correlation analysis

A correlation analysis was performed utilizing Pearson’s product-moment correlation method to examine the relationship between participants’ irritation scores after writing assignment and their reactive aggression, proactive aggression, total aggressive behaviour, and aggressive behaviour in school. The findings are presented in Table 2.

Table 2 The results of Pearson’s product-moment correlation analysis.
Full size table

Table 2 indicates a substantial association among post-test frustration, reactive aggression, proactive aggression, overall aggressive behaviour score, and aggressive conduct in school. Post-test frustration exhibits a substantial positive link with reactive aggression (r = 0.409, p < 0.05), total aggressive behaviour score (r = 0.387, p < 0.05), and aggressive behaviour in school (r = 0.466, p < 0.01), while showing no significant correlation with proactive aggression. Reactive aggression exhibits a substantial positive link with proactive aggression (r = 0.731, p < 0.01), overall aggressive behaviour score (r = 0.929, p < 0.01), and hostile behaviour in academic settings (r = 0.744, p < 0.01). Proactive aggressiveness exhibits a substantial positive association with the overall aggressive behaviour score (r = 0.931, p < 0.01) and aggressive behaviour in educational settings (r = 0.884, p < 0.01). The overall score of aggressive behaviour exhibits a substantial positive link with aggressive behaviour in school (r = 0.876, p < 0.01).

Linear regression analysis

Owing to the absence of a substantial correlation between post-test frustration and proactive aggressiveness in the correlation analysis, proactive aggression was excluded as a dependent variable in the linear regression analysis. A linear regression analysis was performed using post-test frustration as the predictor variable, with reactive aggression, total aggressive behaviour, and aggressive behaviour in school as the dependent variables, respectively. The statistical findings are presented in Table 3.

Table 3 The results of linear regression analysis.
Full size table

Table 3 illustrates that post-test irritation significantly predicts reactive aggression and aggressive behaviour in educational settings. Post-test frustration accounts for 16.7% of the variance in reactive aggression (R = 0.409, R2 = 0.167, adjusted R2 = 0.137). Post-test frustration accounts for 15.0% of the variance in total aggressive behaviour (R = 0.387, R2 = 0.150, adjusted R2 = 0.120). Post-test irritation accounts for 18.9% of the variance in aggressive conduct in school (R = 0.466, R2 = 0.217, adjusted R2 = 0.189).

The results of qualitative research

A total of 73 initial concepts were generated using open coding analysis. The preliminary concepts were further classified, leading to the development of 9 primary codes (Table 4). Subsequent axial coding was performed, resulting in the refinement of 7 second-level codes (Table 5). Consequently, 4 third-level codes were synthesized and retrieved (Table 6). For example, a participant’s statement, “My teacher always criticized me because of my poor grades,” was coded as the initial concept relationships with classmates and teachers. Similar initial concepts were then grouped. Codes like relationships with classmates and teachers was grouped into the second-level code social interaction and crime, which was then synthesized into the core code Individual-Level Frustration (Table 7).

Table 4 Results of Open Coding Analysis.
Full size table
Table 5 Results of Axial Coding Analysis.
Full size table
Table 6 Results of Core Coding.
Full size table
Table 7 Nodes and Reference Points of Frustration -Attack for VSP.
Full size table

Discussion

The validity test results for the frustration induction approach utilized in this study indicate that the recollection induction paradigm successfully provoked frustration among participants. Participants were directed to undertake a writing activity intended to elicit frustration, and a notable disparity in their frustration levels was recorded before and after the task’s completion. The results validate the efficacy of the frustration manipulation, along with Kim’s (2008) study, which indicates that recollecting previous frustrating experiences consistently elicits negative feelings. This outcome establishes a solid foundation for the next questionnaire-based inquiry undertaken in this research.

The descriptive statistical analysis indicates that, under the effective induction of a frustration scenario, participants had elevated levels of aggressive behaviour, with an overall aggression score of 36.767 ± 7.968. Reactive aggression (19.933 ± 4.259) was significantly greater than proactive aggression (16.833 ± 4.308). The findings suggest that, even in a juvenile detention centre where participants face punitive consequences, VSP exhibits a quantifiable level of aggressive behaviour. This group comprises teenagers during a developmental phase marked by swift physical and psychological maturation, however their social adaptation is yet insufficiently developed. Consequently, they are especially susceptible to maladaptive behaviours and are prone to difficulties in domains such as education, interpersonal interactions, and daily living. These challenges are frequently associated with an increased tendency for aggressive behaviour. These findings are congruent with FAT and correspond with other empirical studies on teenage violence. Accordingly, it is essential for pertinent authorities to meticulously observe aggressiveness levels in this population during the duration of criminal sentences. Delivering focused assistance and educational strategies may significantly improve their adaptability, alleviate the effects of additional setbacks, and avert the intensification of aggressive conduct.

This study’s correlation and regression analyses revealed a relatively strong positive association between the frustration experienced by VSP and both the total score of aggressive behaviour and reactive aggression. Frustration accounted for 15% of the variance in violent behaviour and 16.7% of the variance in reactive aggression. These findings correspond with the preponderance of previous studies investigating the relationship between adolescent dissatisfaction and aggressive behaviour, and they are in accordance with the theoretical framework of FAT (Azemi et al. 2020). The findings highlight the imperative of instituting effective measures to alleviate the frustration experienced by VSP currently incarcerated, as this could significantly contribute to diminishing and preventing aggressive actions. Criminological research has clearly established that violent crimes among minors result from the interaction of multiple factors. For instance, individual trait factors such as intelligence and personal characteristics as well as social environment factors like family parenting are strong correlated with aggressive behaviour (Farrington and West 1971).

This study’s finding of no significant link between frustration and proactive aggression may be attributed to 2 primary factors. Firstly, this discrepancy may arise because the RPAS assesses proactive aggression as a general, instrumental, and goal-oriented trait, which might be suppressed or less salient in the current detention environment, or not strongly triggered by the specific frustration recall task among this sample. Besides, the differences in theoretical applicability may play a role in this insignificance. Proactive aggression is motivated by instrumental goals such as resource acquisition or control, while reactive aggression is triggered by immediate frustration or threat. Conversely, the qualitative data reveal specific instances of premeditated criminal acts, such as robbery for economic gain, which were likely driven by chronic, deep-seated frustrations including economic deprivation as well as lack of opportunity, rather than solely by the acute frustration state induced in the experiment. Thus, the qualitative data suggest context-specific premeditation linked to specific life frustrations, while the quantitative measure might be capturing a more general, and in this context, perhaps less relevant or less activated, form of proactive aggression. This highlights the complexity of proactive aggression and the value of mixed methods in uncovering different facets of a phenomenon. Secondly, all participants in this study are VSP individuals incarcerated in juvenile detention centres, potentially resulting in an absence of substantial link between frustration and proactive aggression. The participants differ from the subjects in the majority of prior frustration-aggression research undertaken in China. Current studies on violent behaviour predominantly concentrate on minors or college students, neglecting the VSP group. VSP in this study has already spent time in a juvenile detention centre and received legal education from pertinent authorities, which may instill a fear of retribution for exhibiting proactive aggressive behaviour. This anxiety may result in a reduction in proactive aggressiveness, therefore influencing the relationship between frustration and proactive aggression. This study’s lack of pre-task measurement of aggressiveness may impose constraints on the aforementioned hypothesis, necessitating future research to validate its underlying mechanisms. Meanwhile, unmeasured factors may also moderate the relationship. For instance, VSP with strong social support might alleviate frustration through external resources, weakening its impact on proactive aggression.

Considering that most participants in this study were vocational school students at the time of their crimes, we utilized a self-developed measure to evaluate their violent behaviour in school. The results of the correlation and regression analyses reveal a substantial positive link between frustration and aggressive behaviour in school, with frustration explaining 18.9% of the variance in this behaviour. Given that the participants encounter the problem of societal reintegration post-sentencing, which potentially entails a return to educational environments, it is crucial to examine their levels of violent conduct in school and its correlation with frustration. During detention, measures should be implemented to alleviate frustration through interventions like legal education and psychological counselling, hence reducing the probability of hostile behaviour in educational environment. Furthermore, these findings have significant consequences for the physical and emotional health of VSP. Campus violence, encompassing campus bullying, which may not necessarily fulfill the threshold for criminal prosecution yet nevertheless represents aggressive conduct, continues to be a substantial concern worldwide. Educators must endeavour to reduce frustration among VSP to avert violent behaviours stemming from setbacks, so promoting a supportive environment for their holistic development.

While the Keller and Dauenheimer (2003) frustration scale provides a standardized measure of perceived frustration, our qualitative findings reveal a more complex, multidimensional picture of deprivation, encompassing experiences like the rural-urban divide and left-behind phenomena, which may not be fully encapsulated by the quantitative instrument. The qualitative data thus serves to contextualize and elaborate upon the types of frustrations contributing to the scores on the SFS.

The qualitative research results demonstrate that for VSP, grievances at the interpersonal, familial, and societal levels influence their involvement in violent criminal conduct. VSP has generally encountered frustrations with academics and social interactions at the individual level. Academic frustration constitutes 47.62%, and social interaction frustration comprises 52.38%. Academic discontent may also impact their social relationships at school, especially with teachers. One respondent stated, “During my schooling, I maintained a strained relationship with my teachers due to my reluctance to study, resulting in frequent reprimands.” At the familial level, many respondents reported parental divorce or parents employed away from home, with some being reared by their grandparents or experiencing a complete absence of parental oversight (44.44%). Such familial backgrounds significantly impacted their economic resources (55.56%), as one respondent articulated, “My parents work in another city. I am a left-behind child from a rural area, raised by my grandfather. We lacked economic resources.” On a social level, the detrimental influence of unhealthy entertainment habits and venues facilitated impulsive actions and engagement in violent criminal behaviour.

Subsequent analysis of the interview content indicates that the frustrations encountered by VSP may affect both their selection of victims and the instruments employed in perpetrating violent crimes. Particularly, owing to financial hardships, including inadequate revenue or insufficient resources, they often resort to utilizing no tools or readily accessible materials such as soft iron sheets, belts, and similar objects when choosing implements for their criminal activities. One respondent stated, “I did not bring a knife. I utilized a belt to strike,” and “I employed a broom and soft iron sheets to hit.” Furthermore, many respondents indicated experiencing bullying from peers in the same or higher grades during their educational years, which influences their selection of victims. One respondent remarked, “I was bullied by my grade’s supervisor,” and “I selected this victim because I perceived him as more manageable and less likely to resist.” The qualitative findings suggest that frustration influences the choice of tools and victims. The selection of manageable victims or those perceived as less likely to resist, especially after experiences of being bullied themselves, can be interpreted through the lens of displaced aggression, where aggression is directed towards a less powerful or more available target when the original source of frustration is inaccessible or too threatening. Similarly, the use of readily accessible materials such as soft iron sheets, belts points towards the weapon availability effect or opportunistic armament, where frustration-induced arousal may lead to aggression utilizing whatever objects are at hand, rather than pre-planned weaponry in all cases. These choices reflect not only the intensity of frustration but also cognitive appraisals of risk, opportunity, and perceived vulnerability of targets, which are important mediating processes between frustration and the specific form of aggressive acts.

The regression analyses showed that post-test frustration accounted for approximately 15-19% of the variance in different aggression measures. While statistically significant, this indicates that frustration, as measured quantitatively, is one of several factors contributing to aggression. The qualitative findings, which suggest a decisive role for frustration, should be understood as highlighting the perceived significance and richness of frustration experiences in the narratives of the participants. The qualitative data help to explain the types of frustrations that contribute to the overall frustration levels and provide context for how these frustrations might translate into aggression, even if other unmeasured factors also play substantial roles. For example, the qualitative data reveal deep-seated frustrations stemming from structural disadvantages and prolonged negative experiences, the cumulative impact of which might be more profound than what is captured by a single-time frustration induction and measurement.

The findings should be interpreted in light of FAT’s emphasis on situational interactions. While the quantitative data show a direct link between recalled frustration and aggression, the qualitative interviews provide glimpses into environmental triggers. For instance, experiences of bullying within school, which is an environmental factor, or the specific context of the detention centre itself, though it is not a focus for pre-crime frustration, it is relevant to current aggression levels, could act as moderators or eliciting cues and also be consistent with Berkowitz’s model.

Conclusion

This study utilizes a mixed-methods approach, integrating quantitative data from surveys with qualitative data from interviews for an exhaustive analysis. The findings validate that FAT effectively elucidates the incidence of violent crime within VSP. Nonetheless, not all instances of frustration provoke aggressive conduct. Frustration from social interactions and economic challenges are more likely to incite violent crimes by VSP. Their experiences with frustration events markedly affect their selection of victims and methods, typically favouring the targeting of vulnerable populations and employing readily available objects as weapons.

This study possesses some shortcomings that require attention. Firstly, there are significant shortcomings regarding the research participants. The limited number of questionnaire respondents and interviewees may impact the generalizability of the findings. We specifically recruited only 30 participants from the VSP, a sample size that is diminutive in comparison to analogous studies. Furthermore, the study was performed in Chongqing, a direct-controlled municipality in China, and the unique attributes of this locale may have affected the results, thereby constraining the generalizability of the findings to other settings. Secondly, a major limitation of this study is the lack of examination of potential mediating or moderating variables in the relationship between frustration and aggression. Factors such as self-control, social support, coping mechanisms, sense of relative deprivation, and psychological resilience are well-documented in the literature as influencing this relationship (Sun et al. 2023; Liu et al. 2022). By not measuring and analysing these variables, our model presents an oversimplified direct pathway from frustration to aggression. For example, high social support or effective coping strategies might buffer the impact of frustration on aggressive behaviour, while low self-control or high relative deprivation might exacerbate it. The modest variance explained by frustration in our regression models strongly suggests that such unmeasured variables may play a substantial role. Ultimately, regarding research methodology, this study exclusively utilized questionnaires and interviews, excluding experimental procedures. A key limitation is the cross-sectional design of the study, which prevents the establishment of temporal precedence between frustration and aggression. It is possible that aggressive traits could amplify perceived frustration, or that unmeasured confounding variables influence both frustration and aggression. Therefore, while the findings are consistent with FAT, causal inferences cannot be definitively made. Besides, due to the lack of authorization to interview female VSP individuals and the non-offender VSP populations was not obtained, baseline characteristics for these groups and comparative analyses with the study participants are not included.

This research’s principal contribution is its novel perspective, particularly on VSP, which has been significantly neglected in both contemporary and historical studies. This methodology considerably expands the utilization of FAT in comprehending the determinants of violent crimes, especially among persons in this distinct cohort. The study further investigates the characteristics of frustration events, analysing their effect on the selection of aggressive targets and strategies, along with the magnitude of their effects. The results of this research offer significant insights for developing therapies targeted at those profoundly immersed in violent behaviours. The study highlights the necessity of addressing the frustration events experienced by individuals within the FAT framework when creating systems aimed at preventing recidivism and promoting their resocialization. By alleviating the adverse impacts of previous frustration occurrences, society can more efficiently assist this group in their reintegration endeavours. This research also highlights the imperative to deepen our comprehension of the root causes of violent crimes and to execute focused governance solutions specific to VSP.

It is essential to examine the factors contributing to violent crimes among VSP within the context of FAT. Future research may enhance this field by diversifying the participant pool through increased sample sizes, varied geographical representation, and critically, the inclusion of female VSP individuals and non-offender comparison groups. When possible, longitudinal studies may be conducted to track individuals with a history of violent crime, allowing researchers to assess recidivism rates and investigate the correlations between frustrative experiences and ensuing criminal behaviours. Longitudinal studies are crucial to establish the temporal order of frustration and aggressive behaviours among VSP and to explore potential reciprocal relationships or the influence of confounding variables over time. Additionally, comparison studies may be undertaken to assess the disparities in managing frustration and its ensuing aggressive stimuli between VSP and their counterparts in conventional educational environments. Future study should also contemplate the inclusion of specific characteristics, such as experiences of domestic abuse or the status of left-behind children, as potential mediators. Furthermore, the implementation of more stringent scientific experimental techniques could greatly facilitate the reconstruction of these persons’ frustrating experiences, resulting in a more accurate and empirically based comprehension of the pathways linking frustration and aggressiveness. Future research should more directly measure and analyse such environmental triggers, including peer violence culture within vocational schools, and potentially the conditions within detention environments if studying post-incarceration aggression. In the future, targeted frustration—aggression interruption programs for VSP should be designed. It is suggested to provide a non-violent parenting guidance to families around the country. The practical and theoretical efficacy of these measurements must be routinely assessed to guarantee they are both actionable and impactful.