Gratitude interventions reduce cyber-aggression in adolescents: gender and disposition effects

Abstract

The study explores the impact of gender and gratitude disposition on the effectiveness of a classroom-based gratitude intervention aimed at reducing cyber-aggression among Polish adolescents. Cyber-aggression, linked to maladjustment and mental health issues like depression and anxiety, is a growing concern. The intervention involved 548 students, divided into a control group (399) and an experimental group (149), and lasted for seven days. Participants completed the Cyber-aggression Types Questionnaire (CATQ) and a Gratitude Questionnaire. Results showed the intervention effectively reduced overall cyber-aggression, particularly aversive controlled types, but increased appetitive impulsive aggression. Girls responded better to the gratitude exercises, and those with lower gratitude levels saw the most significant reduction in cyber-aggression. However, students with medium levels of gratitude showed increased impulsive and controlled appetitive cyber-aggression. The study concludes that gender and gratitude disposition influence the success of gratitude interventions in reducing cyber-aggression.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-97214-w.

Introduction

Cyber-aggression is a complex multi-type maladaptive behavior that was recognized as one of the most concerning public health problems affecting adolescents^1,2. Despite the importance of studying distinct profiles of cyber-perpetrators related to dimensions of valence and control, research suggests the need for more attention to finding effective ways to allow individuals to avoid such aversive experiences due to their detrimental impact on adolescents’ development^3,4. Importantly, the meta-analysis showed that the prevalence of online perpetration among children and adolescents ranged from 5.3 to 66.2%⁵. To date, multiple studies have addressed the risk factors that may influence the rate of adolescent cyber-aggression, with the most commonly studied person-related characteristics. Evidence indicates mixed results for gender as a risk factor³. Moreover, given the prominent effects of gratitude intervention in promoting various benefits at every developmental stage, research on adolescents’ cyber-aggression and gratitude is scarce. Notably, some previous studies suggest that gratitude-based practices within natural social groups may effectively reduce negative states (aggression) and behaviors (cyber-bullying) among adolescents and adults with comparable positive effects to classical psycho-educational programs^6,7. Furthermore, previous studies have predominantly focused on a cross-sectional examination of the role of gratitude trait on the overall cyber-aggression, with limited evidence for the diversities caused by its various types. Hence, this project tends to provide valuable insight into the role of Classroom gratitude induction (CGI) with attention given to investigating the influence of personal factors, such as gender and low gratitude disposition.

Cyber-aggression among adolescent girls and boys

Cyber-aggression among adolescent girls and boys manifests in distinct patterns, reflecting differences in socialization, emotional expression, and online behavior. Research suggests that while both genders engage in cyber-aggression, the motivations and methods often differ^8,9. The heterogeneity of the cyber-offenders’ motivations contributing to developing several classifications of cyber-aggression¹⁰. Most commonly proposed types are based on classical typology of pro-active (instrumental aggressive actions focused on anticipated reward, planned and organized, with no emotional arousal defined as cold-blooded and high affective control), and reactive aggression (an impulsive response to perceived threat associated with adverse emotional states such as anger, tension or anxiety)^11,12. The second source of online aggression diversity is related to perpetrators’ motivation i.e. aversion induced aggression acts (aggression arising from aversive conditions such as violent conflicts) or appetitive motives (aggression initiated by the need for experiencing pleasure or fun that stems from attacking and fighting with others)^13,10. For example in respect to these typologies, Runions et al.² proposed four forms of cyber-attacks: aversive: impulsive and controlled; and appetitive: impulsive and controlled. More recently DeMarsico et al.¹⁴ have postulated eight forms of cyber-aggression: (1) Social Bonding - induced by a need to be affiliated by others; (2) Social Activism – initiated by the believes that the person have to advance or defend political/social issues; (3) Reactive Aggression – motivated by anger; (4) Interpersonal Distress – acts induced by relationship stress; (5) Impulsivity – conducted to mitigate impulsive urges; (6) Virtual Dissociation – motivated to take on a differentcyber-identity; (7)Thrill-Seeking – initiated to experience excitement; and (8) Vengeance – pursued to gain revenge. Noteworthy, adolescent girls are more likely to participate in relational forms of cyber-aggression, such as spreading rumors or social exclusion, which align with their tendency to value social connections and relationships¹⁵. On the other hand, boys are more inclined towards direct and overt forms of cyber-aggression, including threats and insults, which may be linked to traditional notions of masculinity and dominance¹⁶. These gendered differences highlight the importance of tailored interventions that address the specific ways in which girls and boys experience and perpetrate cyber-aggression. Understanding these distinctions is crucial for developing effective strategies to prevent and reduce cyber-aggression, as well as to support adolescents in navigating the complex dynamics of online interactions.

Cyber-aggression among adolescent girls and boys demonstrates notable differences in both prevalence and expression, which are influenced by gender-specific social dynamics and psychological factors. According to research by León-Moreno et al.¹⁷, gender plays a significant role in shaping how adolescents engage in cyber-aggression, with boys generally showing higher involvement in such behaviors compared to girls. However, the study also reveals that girls, particularly those who are rejected or neglected by their peers, may engage in cyber-aggression as a response to feelings of anger and helplessness caused by social exclusion. This behavior is often facilitated by the anonymity and indirect nature of online interactions, which can make it easier for these girls to express their frustrations without direct confrontation. Conversely, boys who are categorized as controversial—both liked and disliked by their peers—tend to exhibit the highest levels of cyber-aggression, suggesting that their involvement in both prosocial and antisocial behaviors in the physical world extends into their online interactions. Research conducted by Álvarez-García et al.¹⁸ reveals that boys tend to display higher levels of cyber-aggression than girls, with key factors such as impulsivity and antisocial behavior serving as significant predictors. Although girls are generally less involved in cyber-aggressive acts, they are not entirely exempt; their participation in such behaviors is often influenced by social factors, particularly peer relationships and experiences of victimization. According to Uddin and Rahman¹⁹, boys tend to exhibit higher levels of both cyber victimization and cyber aggression compared to girls. The study also found significant gender differences in emotion regulation strategies, with girls more likely to use cognitive reappraisal, which helps them manage their emotions effectively, thereby reducing the likelihood of engaging in cyber-aggression. On the other hand, boys were more inclined to use expressive suppression, which was positively correlated with both cyber victimization and aggression, suggesting that boys who suppress their emotions are more prone to retaliatory aggressive behaviors online.

Gratitude as a protective factor against cyber-aggression

The available scientific evidence advocate that positive youth development indicators that is character strengths such as grateful mood and predisposition toward thankfulness and appreciation are associated to a decreased likelihood of engaging in various risk behaviors (i.e. substance use, smoking, sexual risk behaviors)²⁰. Gratitude defined as a positive emotional state characterized by recognition for the kindness and benefits received from others, plays a crucial role in mitigating cyber-aggression, particularly among adolescents²¹. Consistently with Fredrickson’s broaden-and-build theory, the positive impact of gratitude on human behaviors stems from improving people’s monetary thought-action repertoires and the increase in accessibility of personal resources, what in a long-term positively impacts individuals’ mental health and social interactions²². As a protective factor, gratitude promotes emotional regulation, fosters empathy, and strengthens social connections²³. These traits not only enhance personal well-being but also create a buffer against negative emotions such as anger, envy, or frustration—key drivers of cyber-aggressive behaviors. Consequently, gratitude can contribute to a more respectful and empathetic online environment, reducing instances of harmful behaviors like cyber-aggression^7,22. The protective role of gratitude is especially significant in school environments, where adolescents are highly susceptible to peer influences and social dynamics. Research by Oliveira et al.²⁴ highlights the interplay between gratitude, forgiveness, and self-regulation in enhancing adolescents’ psychological well-being. This improved well-being fosters prosocial behaviors and reduces the likelihood of engaging in cyber-aggression. Adolescents who practice gratitude regularly show greater resilience to social challenges and are less likely to retaliate with aggression during online conflicts or provocations. Moreover, the relationship between gratitude and cyber-aggression is mediated by factors such as self-compassion and moral disengagement. Studies indicate that gratitude strengthens self-compassion, which in turn reduces moral disengagement—a process often associated with aggressive behaviors²⁵. This sequential mediation underscores the complex interplay of psychological traits that gratitude can influence to mitigate cyber-aggression. Individuals who consistently practice gratitude tend to experience positive emotions such as empathy and compassion, which are essential for fostering healthy interpersonal relationships online and offline. For example, DeWall et al.²⁶ found that gratitude encourages individuals to focus on the positive aspects of their social interactions, reducing the likelihood of engaging in harmful online behaviors. This is particularly relevant in digital spaces where anonymity and lack of physical presence often lead to dehumanization, making it easier to perpetrate cyber-aggression. By promoting a sense of connection and respect for others, gratitude discourages such behaviors. Gratitude’s influence extends beyond individual benefits to foster a supportive school culture. A school environment that promotes gratitude and other positive emotional practices can reduce the prevalence of cyber-aggression by creating a community where empathy and mutual respect thrive²⁵. Such environments not only discourage harmful behaviors but also prevent their escalation into more severe forms like cyberhate. Fulantelli et al.²⁷ argue that while cyberbullying and cyberhate differ, they share common predictors and consequences related to adolescents’ emotional well-being. By embedding gratitude into school programs, communities can strengthen their resistance to these harmful behaviors.

Gratitude interventions among children and adolescents

Numerous past studies have investigated the impact of increasing gratitude among adult or university students populations, with significant effects of such practices on decreasing adverse mental states i.e. depression, anxiety, and increasing the positive ones i.e. well-being, social functioning²⁸. Yet, the gratitude implementation among children and adolescents has received less attention. Among gratitude interventions dedicated to youth population most focused on private gratitude expressions (e.g., counting blessings, gratitude journaling, gratitude letters)²⁹. Scholars have induced gratitude in school through several ways i.e. Grateful Recounting (counting simple blessings received, list three good things), Grateful Reflection (reflecting on the kindness of others, participants are asked to nominate events, things or people they are grateful and which happened in a specified period), Grateful Expression (expressing gratitude and gratitude through letters or journals), Grateful Reappraisal (looking back at the positive side of negative events that have been experienced)³⁰. According to past gratitude literature the recommended duration of gratitude intervention carried out is at least two weeks, to give time for participants to process the information they have gained and provide an opportunity for applying what they have studied during the training³⁰. However, significant effects were detected also in short-term gratitude-based interventions³¹ confirmed that 1-day gratitude induction increased positive affect and psychological emotionality). According to meta-analysis performed by Kirca et al.³² intervention length and duration from baseline to final assessment did not significantly moderate effect sizes across studies. Scholars further assert that through the systematic cultivation of gratitude, various types of benefits have been achieved i.e. school connectedness, academic success, strong peer and family relationships, higher absorption in activities, better sleep quality³³, . The most cited studies on gratitude intervention among youths were conducted by Froh and colleagues³³. Eleven classrooms - youth aged 11–14 years (6-7th grades) took part in two weeks Counting blessings activities. After the intervention the experimental group presented greater optimism, life satisfaction and gratitude disposition, experienced less negative emotions compared to their counterparts without gratitude condition. Similar methodology was applied by Huebnar et al.³⁴, who induced gratitude towards school environment via 2-week intervention based on Counting good things at school. Students in gratitude condition reported higher school satisfaction, were more interested in school activities, were more positive towards attending school and engaging in learning. According to the authors the intervention mitigated adverse educational appraisals, and at the same time increase emotional and social competences. A 2-weeks positive Shamiri intervention implemented to 12–19 years old students, that included gratitude module has revealed a positive effect on reducing internalizing symptoms and improved academic outcomes³⁵. Similarly, Sahar et al.³⁶ found that the four-week improving gratitude program, which included counting blessings related to school life, friends, studies and teachers, writing gratitude letters and loving kindness meditation increases positive emotionality among adolescents aged 16 years. Chen et al.³⁷ showed a significant decline in stress and depressive symptoms over time after the 4-week writing of work-related gratitude diaries. Such an effect may have preventive impact on impulsively motivated aggression, mostly derived from frustration and distress². Nevertheless, empirical studies on gratitude intervention and aggression among youth are scarce. One of such gratitude- based interventions was a weekly group-based sharing gratitude and counting blessings exercises implemented by Deng et al.²¹ among prisoners. After the 5-weeks the authors found that participants in the gratitude sharing and blessing-counting subgroups had lower levels of aggression and higher of well-being than the controls. Prior research also confirmed that expressing gratitude leads to the reduction behavioral aggression in response to provocation, and to decrease the tendency of denigration of partners when confronting with threatened feedbacks^26,38. Interestingly, Cho et al.³⁸ argued that the ameliorating effect of gratitude expression on threatened power holders’ tendency to denigrate subordinates is mediated by increased perceptions of social worth. Similarly, Chamizo-Nieto et al.⁷ emphasized that programs designed to enhance emotional intelligence and gratitude in 12–18 aged adolescents yield promising results in mitigating cyber-aggression. These programs focus on fostering a deeper appreciation for positive emotions and social bonds, equipping individuals to navigate online interactions constructively. By integrating gratitude practices into educational and community initiatives, it is possible to develop a more resilient and empathetic digital culture that curtails the spread of cyber-aggression.

Differences in gratitude experiences across demographic characteristics

A review of past gratitude literature revealed that the feeling of gratitude and the ability to express this complex emotion depends on a wide variety of contextual factors, including chronological gender (females experience gratitude more frequently, with a greater intense, and are more sensitive to gratitude stimuli) Skalski, Pochwatko³⁹, age (older people are more willing to experience grateful mood and present their appreciation towards other⁴⁰, however the findings regarding school-age children and youth are mixed. Empirical work by O’Brien et al. (2018) suggested that older school-aged children express gratitude more often compared to the their younger counterparts. Chan⁴¹ observed a positive association between expressing gratitude and age during adolescence period. In contrast, no significant age-effect was detected by other scholars in samples aged 10–19 years old⁴². These results may be related to different ways of grateful reactions. Baumgarten-Tramer⁴³ proposed four types of gratitude expressions: (a) verbal reaction, which is saying kindly thank you” to the benefactor, without necessarily denoting any emotions; (b) concrete (repaying benefactor with something valuable for child) (c) connective(defined as the creation of a spiritual relationship with the benefactor) (d) finalistic (child is repaying a favor through positive actions that assists in the objects’ attainment, or that promotes own personal development. All those types were related to age. More specifically, youth aged 11–14 more frequently expressed verbal and connective gratitude, whereas those aged 7–10 years old its concrete and connective types. Interestingly, the finalistic type was recognized as typical for older adolescents⁴⁴ (de Lucca Freitas et al., 2011). Similarly Morgan et al.⁴⁵ found developmental and cultural differences in relation to more sophisticated reasoning around gratitude, such as recognition of ulterior motives. More specifically, Australian adolescents were less impacted by the ulterior motive than were their UK counterparts, and the feature of obligation/indebtedness was named significantly more in the UK sample than in the Australian cohort. Other results suggest that subjects from collectivistic cultures tend to report higher levels of gratitude than those from individualistic ones⁴⁵. Moreover, gratitude was associated with more negative features in the UK than in the US⁴⁶. Similarly, Parker et al.⁴⁷ observed that higher self-focused and more autonomous interpersonal style people experience less gratitude.

Present study

The current research used a two-wave longitudinal design to explore the effectiveness of classroom gratitude intervention (CGI) in reducing cyber-aggression among early adolescents from Poland. The primary objective of this study was to examine the pre- and post-intervention levels of cyber-aggression types to identify the distinct impact of CGI on these maladaptive behaviors. Based on findings from past research^6,7,34, the following hypotheses were formulated:

1. Adolescents will demonstrate a decrease in cyber-aggression after the 7-day CGI intervention.

2. The effectiveness of CGI in minimizing the frequency of youth cyber-aggression will vary based on gender and gratitude disposition.

3. Early adolescent girls with a high level of gratitude will exhibit the greatest decrease in overt cyber-aggression.

4. Adolescents who identify as male and have low levels of gratitude will not exhibit a significant decrease in cyber-aggression following the CGI intervention.

Method

Participants and procedure

Approximately 548 adolescents from 7-8th grade primary school participated in the paper-pencil survey conducted during 2023. The randomly chosen primary schools located in urban and rural areas of Poland were invited to participate in the quasi-experimental survey. There were no specific criteria, except attending 7-8th grade. We randomly chose classes from those schools that agreed to participate in the project. After receiving consent classes were randomly assigned to intervention vs. control conditions (there was no variation in assignment – a similar number of schools from urban and rural areas took part in the survey). All subjects provided assent, with counselors and parental consent, to participate in a study. The psychological assessments in the control group were introduced by teachers, and completed in the classroom during classes. Similarly, teachers provided information on a one-week intervention, with a measurement one week before the study and a one-week follow-up in the experimental group. The control sample consisted of 399 participants (177 girls, 222 boys) aged 12–15 years (Mage = 13.7, SD = 0.67). The experimental sample included 149 adolescents (80girls,69 boys) aged 12–15 years (Mage = 13.18, SD = 0.55) at Time1, and 142 youth at Time2 (77 girls, 65 boys)aged 12–15 (Mage = 13.66, SD = 0.55). Seven participants from the experimental group did not complete all parts of the survey (4.7% of the sample), due to the absence at Time 2 at school. This paper refers a part of larger project on school burnout, personality traits and cyber-aggression among early adolescents in Poland. The controlled group was examined only at Time 1 with measurements of Cyber-aggression (CATQ) scale, and demographic items, whereas the experimental sample was tested twice time, that is at Time 1 we used CATQ scale and demographic items, at Time 2 (3 weeks later) they fulfilled CATQ scale and gratitude (GQ) scale. The gratitude intervention was delivered to students by trained teachers, who participated in a set of online meetings with the researchers before the project started. The teachers were learned how to conduct the intervention, they were delivered gratitude materials to improve their knowledge of the subjects, and they were also informed how to use an online tool i.e. an online gratitude classroom book. One week after the first examination students took part in a 7-day Classroom Gratitude Intervention (CGI), that includes enhancing knowledge on the importance of expressing gratitude by watching a film and discussion during lectures with teachers and classmates, gratitude exercises such as counting blessing, gratitude journaling completed jointly with other students, gratitude challenge completed privately. Each gratitude activity was implemented by trained teachers during classes (the sessions lasted 45 min per day during school week), who instructed each gratitude exercise by presenting “A Gratitude Week” researchers’ PowerPoint presentations (a gratitude quiz, a link to gratitude film, a hand of gratitude, a gratitude self-reported level, and an example of classroom gratitude book). After these sessions, students were divided into small sub-groups and completed collectively gratitude journal in the classroom setting (the task for each group was to find a gratitude sentence for each day and discuss it publically). Students were then encouraged to fulfill the private gratitude challenge during weekend (the implementation of this exercise was not controlled). Prior the survey we obtained an approval from the University of National Educational Commission Research Ethics Committee (Nb. DNa.0046.13.2023).All studies were conducted in accordance with relevant guidelines and regulations. Informed consent was obtained from all participants and their parents based on the information provided to them by teachers (students were informed during classes and parents during parent-teacher evenings). The study was conducted in accordance with the Declaration of Helsinki.

Measurements

Cyber-aggression Types Questionnaire (CATQ) developed by Runions et al.² measures four types of online aggression aversive impulsive and controlled, appetitive impulsive and controlled. The instruments includes29 items rated on a 4-point Likert scale (1-Very unlike me; 4-Very Like me) (e.g. Sometimes I can be mean to people online to get what I want).The polish version of the CATQ was prepared by Tomaszek and Muchacka-Cymerman⁴⁸.The validity of the CATQ in Poland was confirmed by estimating correlations with offline aggression (r = .42, p < .0001)²⁸.Reliability in this study was very high (Control sample (n = 399): α /ɷ over 0.97 for CATQ Total score, and α/ɷ equal or over 0.90 for its subtype; Experimental sample (n = 149): α/ɷ over 0.90 for CATQ Total score, and α/ɷ equal or over 0.86 for its subtypes).

Gratitude Questionnaire (GQ) developed by McCullough et al.⁴⁹, and adapted by Kossakowska and Kwiatek⁵⁰ assess individuals’ gratitude disposition levels by using 7-point Likert scale (1-strongly disagree, 7 – strongly agree). The subject respond on the 6 items (i.e. “I have so much in life to be thankful for”).The significant positive relationships between gratitude and other psychological positive measurement i.e. life satisfaction (r = .29,p < .05), religiosity (r = .35, p < .05), ethical openness (r = .46, p < .05), and harmony (r = .35, p < .05) confirmed the validity of the GQ scale in Poland.The GQ reliability in this study was high, i.e. α/ɷ equal to or higher than 0.88.

Demographic items– all participants were asked to report their biological gender (Male vs. Female), and age.

Data analysis

To test the effectiveness of the CGI intervention on levels of cyber-agression types, we conducted paired-parametric and non-parametric statistics (tStudent, rankWilcoxon, U Mann-Whitney). Required minimum sample size equal to 45 (tStudent test for matched pairs), and 47 (Wilcoxon signed-rank test – matched pairs) (G*Power 3.1.9.7.calculator conditions: effect size = 0.50, α = 0.05, 1-ß = 0.95). To compare samples unequal in the number of participants (experimental vs. control; girls vs. boys) we calculated the tStudent_Welch, and the U Mann-Whitney statistics. Required sample 105 (tStudent independent sample), and 110 per group (G*Power3.1.9.7. calculator conditions: two tails, d = 0.50, α = 0.05, 1-ß = 0.95). To examine the differences between students varied in the levels of gratitude disposition we divided the experimental sample into three sub-groups: Low gratitude n = 32; medium gratitude n = 79, high gratitude n = 17 (χ2(2) = 95.39, p < .001), and used Kruskal-Wallis statistic to examine the sub-group differences. Required minimum sample size equal to 66 (G*Power 3.1.9.7. calculator conditions: effect size = 0.50, α = 0.05, 1-ß = 0.95, number of groups 3). The pre-and post intervention results within the three gratitude conditions were examined by calculating paired Wilcoxon rank test. Required sample size reported above. Finally, we used Repeated Measures ANOVA (RM-ANOVA) to estimate interaction effect between gender and gratitude. This procedure was used because regardingthe violations of normalityit is robust (i.e. Type I error and power of the F-statistic are not altered) with skewness ≥ 2.31, and kurtosis ≥ 8⁵¹. Required minimum sample size equal to between factors: 51 subjects, and within factors: 12 subjects (G*Power 3.1.9.7. calculator conditions: f = 0.50, α = 0.05, 1-ß = 0.95). The conduct statistical analysis were performed by using the IBM SPSS-22 (IBM Corp., NY) and the Jamovi 2.3.28 free software.

Data availability

The datasets generated and analysed during the current study are not publicly available due to ethical restrictions related to participant privacy and the confidentiality agreements required by the institutional review board. However, the data are available from the corresponding author on reasonable request.

Results

Descriptive statistics

The CGI intervention significantly decreased overall level of cyber-aggression (small effect size), and its aversive controlled type (F2) (large effect size), however, slightly increased appetitive impulsive cyber-aggression (F3) (small effect size) (see Table 1).

Table 1.

Descriptive statistics and comparison between pre-and post gratitude intervention measures of CATQ scores in experimental group (n = 149).

Variables			M(SD)	Skewness	Kurtosis	Reliability		tStudent/ Wilcoxon rank	p	Effect size
Cyber-aggression						α	ɷ
Aversive	Impulsive (F1)	Time 1	18.2(7.31)	1.40	1.68	0.91	0.92	−1.02	0.309	0.09
		Time 2	19.1(9.51)	1.41	1.49	0.96	0.96
	Controlled (F2)	Time 1	15.2(3.16)	−0.16	−0.41	0.90	0.91	12.9	< 0.001	6.24
		Time 2	9.71(4.77)	1.42	1.48	0.91	0.91
Appetitive	Impulsive (F3)	Time 1	8.30(3.49)	1.95	3.77	0.86	0.87	1299	0.020	0.29
		Time 2	9.39(4.85)	1.44	1.36	0.92	0.93
	Controlled (F4)	Time 1	7.11(3.36)	2.04	1.48	0.88	0.90	2105	0.328	0.11
		Time 2	7.88(4.26)	3.91	1.20	0.93	0.94
CATQ	Total score	Time 1	49.9(13.9)	1.48	2.55	0.96	0.96	6326	0.007	0.26
		Time 2	46.0(22.7)	1.45	1.55	0.98	0.98
Gratitude		Time 2	27.2(9.44)	0.20	−1.00	0.88	0.89	–	–	–

Paired Samples tStudent test and Wilcoxon rank test were used for comparisons with hypothesis Time ≠ Time 2.

Control vs. experimental sample

The controlled sample scored significantly lower in CATQ_Time1 (z(546) =−7.56, p < .001, effect size = 0.42 - average, Mean difference = 6.50), and F2 _Time1 (z(546) =−14.44, p < .001, effect size = 0.77 - average; Mean difference = 6.67) before CGI. After the intervention we detected one significant difference in aversive controlled aggression (F2) (z(539) =−3.18,p = .001). Particularly, CGI intervention caused the decrease in overall CATQ score among adolescents from experimental sample, hence, the controlled and experimental group were equal in CATQ level (Mean difference = 3.6, effect size = 0.05). In addition, students from experimental group still scored higher in aversive controlled aggression (F2) (Mean difference = 1.18; Effect size = 0.17 - small), however the difference was much smaller than at Time 1. An in-depth analysis reflecting participants gender revealed that before the intervention boys from control and experimental sample scored higher in aversive controlled aggression (F2) (z(289) =−10.18, p < .001, Mean difference = 6.81), and CATQ (z(289) =−5.29, p < .001, Mean difference = 7.57). However, unexpectedly after the CGI boys from experimental sample scored significantly higher in some CATQ indicators (CATQ z(285) = −2.45, p = .014, Mean difference = 10.61; F1 z(285) = −2.12, p = .034, Mean difference = 3.62; F3 z(285) =−3.25, p = .001, Mean difference = 2.62). Importantly, differences in F2 between both samples at Time 2 were smaller than at Time 1 (F2: z(285) =−3.53, p < .001, Mean difference = 6.81), indicating a decrease in aversive controlled cyber-aggression type. In girls sample the CATQ (z(255) =−5.47, p < .001, Mean difference = 5.42), and aversive controlled aggression (F2) (z(255) =−10.10, p < .001, Mean difference = 6.51) were higher in experimental sample. After the intervention we detected significantly lower scores in Aversive Impulsive type (F1) (z(252) =−2.28, p = − .023) among girls from the experimental sample. Importantly, the scores in Aversive Controlled type (F2) (z(252) =−0.99, p = .324), and CATQ (z(252) =−1.22, p = .224) decreased in experimental sample, so both groups were similar in F2 (Mean difference = − 0.10) and CATQ (Mean difference = −2.31).

Gender differences

Before the gratitude intervention the insignificant differences between girls and boys were observed in almost all cyber-aggression factors, except appetitive controlled type (F4) – boys scored slightly higher in this type of cyber-aggression compared to girls (t_Welch (147) = −2.34, p = .021, d = .39-average effect size, 95%CI[−2.45;−.20]). After the intervention girls scored significantly lower in all examined characteristics(F1: t_Welch(140) = −3.33, p = .001, d = .57, 95%CI[−8.41;−2.14]; F2: t_Welch(140) = −3.06, p = .003, d = .52, 95%CI[−4.05;−.87]; F3: t_Welch(140) = −3.54, p < .001, d = .60, 95%CI[−4.44;−1.25]; F4: t_Welch(140) = −2.73, p = .007, d = .46, 95%CI[−3.37;−.54]; CATQ Total score: t_Welch(140) = −3.31, p = .001, d = .56, 95%CI[−20.00;−5.02]).The pre-post intervention comparison in girls sample’ revealed a significant decrease in CATQ total score (average effect size), and its two aversive dimensions: impulsive (F1) (small effect size), and controlled (F2) (large effect size), and insignificance of changes in both appetitive dimensions (F3, F4). In boys’ sample we detected an increase in aversive impulsive cyber-aggression (F1) (average effect size), and a decrease in aversive controlled cyber-aggression (F2) (large effect size), and appetitive impulsive cyber-aggression (F3) (average effect size). The levels of appetitive controlled cyber-aggression (F4), and CATQ total score were similar to pre-intervention ones (see Fig. 1). To sum up, study hypothesis 1 was mostly confirmed.

Fig. 1.

The effectiveness of CGI in reducing cyber-aggression among early adolescents’ girls and boys.

Gratitude disposition

The findings suggest significant differences after the gratitude intervention in all three gratitude subgroups were in cyber-aggression total score (χ² (2, N = 142) = 7.36, p = .025, ε²=0.06)and its three types, that is: aversive impulsive (F1:χ² (2, N = 142) = 6.18, p = .046, ε²=0.05), aversive controlled (F2: χ² (2, N = 142) = 7.09, p = .029, ε²=0.06); appetitive impulsive (F3: χ²(2, N = 142) = 8.69, p = .013, ε²=0.07). We also observed a difference on the level of statistical tendency in appetitive controlled cyber-aggression type (F4: χ² (2, N = 142) = 5.71, p = .058, ε²=0.04).The Dwass-Steel-Critchlow-Fligner pairwise comparisons showed that participants from medium gratitude sample scored significantly higher in cyber-aggression, and its two types, that is F2, and F3 measured at Time 2 compared to their low gratitude counterparts (W = 3.54, p = .033, W = 3.71, p = .024; and W = 3.43, p = .041, respectively). A non-parametric Wilcoxon rank statistic found a significant decrease in CATQ total score, and aversive impulsive (F1) and controlled (F2) cyber-aggression among subjects with low gratitude sample. The aversive controlled cyber-aggression (F2) significantly decreased, whereas the appetitive impulsive and controlled cyber-aggression significantly increased after the intervention among individuals with medium gratitude disposition. Finally, among adolescents with high gratitude disposition we detected a significant decrease only in aversive impulsive cyber-aggression (see Table 2; Fig. 2). Considering our study hypothesis 2 we did not find support for the notion that a high level of gratitude disposition will have the most positive impact on the decrease in cyber-aggression.

Table 2.

Means and standard deviation among low gratitude (n = 32), mediumgratitude (n = 79, and high gratitude (n = 17) samples before and after gratitude intervention.

Variables		Gratitude level	Before intervention	Afterintervention	z	p	Hedges g
			M(SD)	M(SD)
Aversive	Impulsive (F1)	Low	18.63(7.11)	16.09(7.32)	-2.13	0.034	0.35
		Medium	17.86(6.62)	21.18(10.46)	-1.69	0.090	0.38
		High	16.41(5.01)	16.59(7.33)	-0.05	0.964	0.03
	Controlled (F2)	Low	14.19(3.14)	8.03(3.03)	-4.48	< 0.001	2.00
		Medium	15.30(3.32)	10.68(5.32)	-5.75	< 0.001	1.04
		High	16.00(3.37)	9.29(4.82)	-3.36	0.001	1.61
Appetitive	Impulsive (F3)	Low	8.50(3.19)	8.06(3.72)	-0.94	0.348	0.13
		Medium	7.81(3.06)	10.68(5.31)	-3.79	< 0.001	0.66
		High	8.24(3.63)	8.11(3.90)	-0.07	0.943	0.03
	Controlled (F4)	Low	7.06(2.14)	6.72(3.39)	-1.26	0.209	0.12
		Medium	6.89(3.20)	8.81(4.82)	-2.46	0.014	0.61
		High	6.94(3.53)	6.88(3.10)	0.00	1.00	0.02
CATQ	Total score	Low	48.38(12.24)	39.91(16.92)	-3.05	0.002	0.57
		Medium	47.86(12.62)	51.15(25.26)	-1.00	0.921	0.16
		High	47.59(13.12)	40.88(18.63)	-1.33	0.185	0.42

Non-parametric Wilcoxon rank statistic was used for group comparison.

Fig. 2.

The effectiveness of CGI in reducing Cyberaggression among individuals with different level of gratitude disposition.

The interaction effect of gender and gratitude disposition

The within subjects analysis: The RM-ANOVA showed significant interaction effects for time and gender for cyber-aggression total score, and its three types, i.e. aversive impulsive (F1), aversive controlled (F2); appetitive impulsive (F3). Time and gratitude interaction was significant in cyber-aggression total score, and its type appetitive impulsive cyber-aggression (F3). This interaction was on the level of statistical tendency (i.e. p = .057) in aversive impulsive cyber-aggression (F1) (see Table 3).

Table 3.

Within subjects effects for cyber-aggression.

Variables	Aversive impulsive		Aversive controlled		Appetitive impulsive		Appetitive controlled		CATQ total score
	F	η²p	F	η²p	F	η²p	F	η²p	F	η²p
Time	0.31	0.003	119.72***	0.50	2.08	0.02	1.13	0.01	2.04	0.02
Time x Gender	7.28**	0.06	4.82*	0.04	4.36*	0.03	1.02	0.01	5.57*	0.04
Time x Gratitude	2.94	0.05	1.82	0.03	4.56*	0.07	2.56	0.04	3.39*	0.05
Time x Gender x Gratitude	1.09	0.02	1.58	0.03	0.29	0.01	0.83	0.01	1.06	0.02

Type 3 Sums of Squares; F statistic and Partial η² are reported;

The between subjects analysis: The findings indicate a significant differences in cyber-aggression between participants with low, medium, and high gratitude disposition in both aversive types of cyber-aggression (i.e. impulsive and controlled). The results for gender, as well as the interaction between gender and gratitude were insignificant (see Table 4).

Table 4.

Between subjects effects for cyberaggression.

Variables	Aversive impulsive		Aversive controlled		Appetitive impulsive		Appetitive controlled		CATQ total score
	F	η²p	F	η²p	F	η²p	F	η²p	F	η²p
Gender	0.57	0.01	0.84	0.01	2.39	0.02	2.90	0.02	1.52	0.01
Gratitude	3.05*	0.05	4.12*	0.06	1.45	0.02	1.91	0.03	2.86	0.05
Gender x Gratitude	1.07	0.02	1.15	0.02	1.19	0.02	0.76	0.01	1.15	0.02

Type 3 Sums of Squares; F statistic and Partial η² are reported.

Discussion

We investigated the effect of gender and gratitude disposition on the effectiveness of a 7-day classroom gratitude intervention in reducing cyber-aggression among early adolescents from Poland.

The first study hypothesis referred to the decrease in cyber-aggression after the 7-day CGI intervention. We observed a significant reduction in cyber-aggression (small effect size), and its aversive controlled (F2) type (large effect size), however, slightly increased in appetitive impulsive cyber-aggression (F3) (small effect size). Compared to adolescents in a control group, those who completed the CGI intervention reported significant decrease in cyber-aggression (g = 0.42 – average effect size). Our finding aligns with prior studies that highlighted the positive effect of gratitude interventions on youth behavioral functioning^20,26. For example, Ma et al.²⁰ found that greater life-orientation gratitude was associated with abstinence from sexual intimacy, sexual intercourse, likelihood of engaging in sex during primary school, and abstinence from drug/alcohol use. According to the authors, higher appreciation leads to personally and socially productive behaviors such as avoidance of high-risk behaviors, but also serves as a resources reservoir that can be used when young people confront with life demands. Our findings, partially confirm the Wood’s coping hypothesis, according to which increasing gratitude utilizes positive coping mechanisms, and reducing avoiding behaviors (i.e. risky activities) when confronting with socially challenges situations i.e. conflicts. The data are also in line with the notion that gratitude activates moral self-schemas that are positively related to self-control, and manifests as a greater emotional and behavioral regulation (i.e. reduces proactive and reactive aggression)⁵².

In our research, the classroom gratitude induction shared with natural counterparts was found to have a significant positive effect mostly its aversive controlled online acts. This type of cyber-aggression is an effect of perceived provocation that is related to uncomfortable emotional states (e.g., shame, embarrassment). At the same time, individuals maintain high capacities for self-control, hence aggressive acts are more likely to take the form of calculated payback, or revenge². The observed decline in aversive controlled cyber-aggression may be explained through the typical gratitude-based benefits i.e. an increase in positive emotions and gratitude disposition^33,53 what causes more positive attitudes towards others, and makes people less reactive to provocations. For instance, studies conducted on a large sample of adolescents aged 12–17 years old by García-Vázquez et al.⁵² showed that forgiveness and gratitude had an indirect relationship by decreasing both proactive and reactive aggression through their positive effects on self-control. Corresponding to the abovementioned explanation prior research has found that practices based on cultivating gratitude activate greater generosity and sensitivity exhibited towards others, as well as more empathetic and compassionate reactions, hence attenuating students’ aggression^21,26. According to the broaden-and-build framework, even a single positive experience (i.e. expressing thanks for small gifts) can extend and intensify due to positive creating upward spirals of emotions and actions²². Such a positive circle may initiate the deep transformation of possessed interpersonal relationships, hence the social environment is perceived and actually becomes more friendly and safe. The prosocial effect of cultivating gratitude was confirmed by many scholars i.e. McCullough et al.⁴⁹ posited that gratitude prompts people to change their behaviors into more social. According to these authors the prosociality of gratitude stems from its three core functions: (1) a moral barometer function (being the beneficiary of someone’s moral actions), (2) a moral motive function (motivation to behave morally toward the benefactor), and (3) a moral reinforce function (being encouraged to behave morally in the future). These gratitude functions may also directly reduce the tendencies to behave immorally i.e. taking revenge through an online aggression. It is worth adding, that moral disengagement mediated the relationship between gratitude and cyberpetration²⁵. Gratitude decreased the tendency to engage in unethical behavior without guilt or self-sanctions, because activated the need for maintaining positive moral self-image. Noteworthy, the laboratory induced gratitude also increased the distribution of self-resources to another⁵⁴, which in turn may decrese the provocative acts from others. Another possible mechanism underlying the effectiveness of gratitude-based interventions is related to an increase in self-worth³⁸. Prior studies have shown that expressing gratitude activates positive self-schemas and shifts individuals’ thoughts on self-strengths, hence people are less prone to perceive conflicts as threatened, and are more willing to look for adaptive and creative ways of solving such situations via reconciliation⁵⁵. For example, teams worked under the gratitude condition experienced an increase in information elaboration and generated highly creative ideas more than those in the neutral condition⁵⁶. Dizon⁵⁵ argued that participation in a gratitude journaling by intimate partners increases adoption of more positive conflict resolution styles, and decreased adherence of negative ones.

On the contrary, the CGI was found to be ineffective in appetitive impulsive cyber-aggression, named recreation. This type of aggression refers to the self-generated violent acts by offenders, with a positive perception of violence perpetration, recognized as fascinating and appealing^2,57. According to Runions et al.² it reflects spontaneous and immediate aggressive reactions, conducted without heed of long term consequences. This is an instrumental and goal-oriented type of aggression, defined by the primary intrinsic enjoyment of aggressive activity⁵⁸. The origins of appetitive aggression are related to the adverse childhood experiences, that manifest through PTSD symptoms later, produced by an ongoing climate of violence in families⁵⁷. According to some scholars it can be considered a stable and long-term adaptation to cope with an insecure and negative environment⁵⁹. The instances of impulsive-induced appetitive aggression are in the service of individuals’ protective mechanism (attaining social status) and striving for significance. In this context, increasing grateful moods through short-term interventions may ineffective, as this group of students may found hard to produce positive appraisals of benefits. It is possible, that this group of young people found it difficult to notice help from benefactor i.e. interpret al.l benefits as less valuable, costless or non-altruistically intended. Prior research has found that adolescents, who experienced family abuse and school violence presented lower level of gratitude disposition⁶⁰. Noteworthy, although gratitude decreases PTSD symptoms among victims, individuals with traumatic experiences are characterized by diminished level of gratitude disposition⁶¹. According to Kim et al.⁶¹ such a regularity is related to negative self-attributions i.e. an increased amount of self-stigma (negative perspective and opinion one carries about themselves) and social stigma. It is also worth to add, that the relationship between gratitude and cyberperpetration was found to be mediated by other psychological characteristics i.e. self-compassion and moral disengagement²⁵. In accordance, the I³ metatheoretical model of aggression by Finkel and Hall⁶² posits three orthogonal processes that causes violent reactions. The first one is instigation, which includes immediate environmental stimuli (e.g., provocation), the second one is impellance, situational or dispositional qualities that decide on the intensity of aggressive response, and the third one is inhibition characterized by situational or dispositional qualities responsible for overridden the proclivity. According to Zeng et al.⁶³ gratitude and all character strengths may be recognized as inhibitors of aggression, because grateful people tend to be more compassionate and less critical towards themselves, which in turn inhibits violent acts in space and contributes a higher tendency to develop healthy attitude towards others (i.e. all human beings deserve kindness). Finally, as cyber-aggression may be an element of classroom in-group rivalry, the ineffectiveness of CGI may be an effect of class social hierarchy. According to social distance theory of power higher-power individuals experience higher social distance towards others compared to low-powered ones⁶⁴. In addition, those high in power are more prone to perceive relationships in the context of social exchange. Hence, the level of engaging in group-gratitude activities may be lower and the gratitude indebtedness may be higher among high-power people. Indeed, studies conducted by Anicich et al.⁶⁵ revealed that individuals with relatively low social power tend to be characterized by increased feelings and expressions of gratitude after benefiting from a favor. Drawing on the above-presented findings, it is recommended to precede the gratitude-based intervention with a shifting perspective from rivalry to equality, and to include activities that increase classroom positive climate.

The next three study hypothesis referred to gender and gratitude disposition as a significant factors for the effectiveness of CGI in minimizing the frequency of youth cyber-aggression. Our findings confirmed that the effect of CGI on cyber-aggression depended on participants’ gender (girls benefit more from this intervention compared to boys), and gratitude disposition (students with low gratitude levels experienced the highest decrease in cyber-aggression, whereas the medium gratitude appeared as a risk factor for the increase in aversive impulsive and controlled as well as appetitive controlled cyber-aggression).

Gender differences: Gender was found to be a significant factor that interacted with time and explained changes in adolescent cyber-aggression, (i.e. aversive impulsive and controlled types, and appetitive impulsive type). Additionally, we detected that boys exhibit a higher rate of cyber-aggression (before and after the intervention) compared to girls. The results remain a consensus with the past findings suggesting that being an adolescent boy is a risk factor for engaging in online perpetration^66–68. Moreover, the CGI intervention appeared ineffective among boys. Particularly, the changes in CATQ total score were either insignificant or indicated an increase in some indicators - aversive and appetitive impulsive types of cyber-aggression. In contrast, we observed a small decrease in the aversive controlled cyber-aggression type among boys’ experimental sample. The CGI program seems to be more beneficial when implemented among adolescent girls. We observed a decrease in the overall level of cyber-aggression at pre-to-post intervention, and its all types. The comparison focused on cyber-aggression subtypes revealed a significant decrease in two aversive dimensions: impulsive and controlled, but no changes in its appetitive types. The above results suggest that gratitude intervention may not be a fruitful practice among all adolescents as its effectiveness is limited to some cyber-aggression types and varies regarding recipients’ gender. The findings coincide with prior notions that practitioners should differentiate cyber-aggression preventive strategies by sex⁶⁹. Particularly, the main explanation for this suggestion stems from diversity addressed for boys and girls in: (1) gender socialization, (i.e. a greater approval of direct – physical aggression for boys, and an indirect – relational aggression for girls)⁷⁰, (2) social norms (i.e. the experiences of being perpetrator or victim are grounded in hegemonic masculinity - masculine dominance and gender inequalities)⁶⁹, and (3) the relatives’ normative expectations regarding resolving conflict situations (i.e. the belief that boys will attack during the confrontation, and girls will look for compromise) and building social bonds (i.e. gender-normative girls have a greater vested interest than gender-normative boys in maintaining friendships and resolving conflicts)⁷¹. To understand more about sex diversity in cyber-perpetration it is worth adding that gender is a significant factor in motives, types, and consequences of this maladaptive behavior. More interestingly, individuals with more feminine traits were found to engage in more cyber-relational aggression via social networking sites and mobile phones, whereas those who identify themselves as more masculine carried more often cyber-verbal aggression through online gaming⁷². Moreover, boys not only report more tolerance toward cyberbullying but also present more frequently impulsive cyber-aggression geared towards fulfilling their immediate needs or heightening their social status and power (i.e. intrasexual competition between males)^72,73. In contrast, girls are mostly motivated to conduct spontaneous cyber-attacks in order to gain peer’s attention⁷⁴. Therefore, the observed in our study higher decrease in cyber-aggression among girls may be related to satisfying their need to attract attention and approval from classmates during CGI. However, the CGI practices could not support the participants’ need for domination or power, which may explain the low effectiveness of gratitude induction among males. Additionally, according to Kashdan⁷⁴, the promptness to express positive emotions (i.e. gratitude) is also mediated by gender. Particularly, men are less willing to feel and present their appreciation to a benefactor, but also are more critical and suspicious of gratitude induction, thus deriving less goodness from such practices. Moreover, gender significantly differentiates the preferred ways of expressing gratitude and the urge need to reciprocate⁷⁵. Consistently to these findings, the CGI might be related to a higher feeling of debt activated by gratitude induction among males and the preference of showing it privately to the benefactor. In other words, improving gratitude among boys’ adolescents may be more effective when includes individual face-to-face rather than classroom exercises. Finally, it is also worth mentioning, that past research pointed to the adolescents’ identification with the role or motive they played in online aggressive behaviors. Particularly, engaging in aggressive online attacks was more associated with the form of aggression rather than role (i.e. sending aggressive posts more frequently was motivated by reactive reasons, whereas creating hostile websites with proactive ones⁷⁶. The above diversity in actions and motives underlying online aggression may also be related to the sex differences in CGI effectiveness (i.e. higher decrease in impulsive compared to controlled aggression, and an increase in appetitive controlled aggression). Gratitude interventions are focused on exercises, that enhance the willingness to express this feeling towards the benefactor and re-pay the goodness by prosocial behaviors. Therefore, such practices mostly strengthen the positive meaning of individuals’ lives (i.e. activates the perspective of being supported and loved), and the capacity to control negative impulses (i.e. activates cognition by initiating positive thoughts about others), however, for those who are motivated to commit aggression by instrumental purposes (increase the power or domination over others), that is emotionally callousness proactive aggression, typical strategy to increase prosocial emotions in order to reduce aggressive acts may be ineffective because must be preceded by an inner change of personal goals and fundamental values.

Gratitude disposition: Our study indicates that students with a low level of gratitude experienced a significant decrease in cyber-aggression, and in its two aversive types, whereas participants from the medium gratitude sample scored significantly higher in cyber-aggression, and in its two appetitive types. A decrease in aversive impulsive cyber-aggression type was observed among medium and high gratitude adolescents. Significant interaction effects were detected between the gratitude disposition and the time when explaining cyber-aggression total score and its appetitive Impulsive type. The above findings are partly in line with McCullough’s et al.⁷⁷ resistance hypothesis, reflecting that a high grateful mood inhibits the benefits of discrete positive emotional episodes because of a lack of reasons for looking new benefits to be appreciated for. In fact, youth high in gratitude disposition may believe that a grateful attitude is something obvious, and there is no need to present privately experienced emotions jointly with the classmates as the benefactor has already received the thanks. Additionally, low gratitude individuals may experience the “halo effect” related to a new meaningful perspective (i.e. ‘Being thankful for all the advantages in your life makes your life better) activated by the CGI, and thus they might start to see more positives around them, which stopped them from acting out aversive emotions through the digital devices. It is worth highlighting that these findings are not in line with the notion that low gratitude adolescents typically magnify the daily adversities’ distress by blindly adopting accidental and inappropriate coping strategies, and thus they are constantly frustrated which results in inner disturbance and adverse states⁷⁸. Therefore, momentarily activated appreciation by the CGI might have initiated the process of looking for highly valued benefit, which is shared by the entire class group. In this context, the relative size of goodness proposed by the classmates might allow for changing the perspective regardless of the low gratitude trait. In other words, youth may be guided by trained group leaders to be grateful for all things received from others facilitating a positive perception of a person’s life and the owned strength and resources. Similarly to other studies⁷⁹ gratitude alleviated stress related to school events and reduced the frequency of externalized behaviors, such as cyber-aggression.

The fact that medium gratitude students experienced an increase in cyber-aggression after the CGI is more difficult to explain, as somehow is inconsistent with past research⁵². Notably, our results suggest that gratitude intervention is more effective in minimizing reactive cyber-aggression, that is an impulsive action initiated by a provocation or an actual or conceived threat⁸⁰. Moreover, it is hostile aggression that “is a primordial means for establishing one’s sense of significance and mattering, thus addressing a frustration of a fundamental social-psychological need”⁸¹. Past studies confirmed that by enhancing gratitude individuals activate cognitive planning, emotional self-control, and emotional stability, but also increase the satisfaction of fulfilling extrinsic and intrinsic needs⁸². Thus individuals are less motivated to react aggressively in response to frustration. The detected ineffectiveness of the CGI in reducing appetitive controlled online aggression may be also related to the following characteristics: (1) most cyber-attacks are done by one or a few perpetrators, (2) cyber-aggressors are usually from the same class group (3) cyber-attack often lasts about a week and is committed outside of school⁸³. The CGI intervention was provided only for one week, inside the school, with no control for the internee’s engagement and expectations in the proposed grouped gratitude exercises. The complex long-term adverse relation between perpetrator and victim, and the self-identification of aggressor with the played role may be so strong that a one-week lasting induction of gratitude perspective may not be enough to change these grounded class-positions. Therefore, although momentary cyber-aggression attacks inside school may be less frequent, this effect may not last after schooling. Additionally, even though gratitude motivates individuals to be more sensitive or more focused on others’ needs, and promotes prosociality, and by doing so mitigates intentional aggression directed to harm others, the key determinant of these benefits may lay in the level of empathy²⁶. More specifically, higher empathy mediates the positive relational behaviors conveyed by gratitude expressions, since is a social anchor, reducing antipathic behaviors during face-to-face contact⁸⁴. Enhancing empathetic skills is also thought to be an effective strategy in reducing youths’ involvement in cyber-perpetration^3,50. Nevertheless, some evidence demonstrates that affective empathy and anti-empathy (the tendency to feel a contradictory response to others’ emotions) significantly accounted for aggression, but its cognitive component did not⁸⁵. Knowing the appetitive controlled cyber-aggression is weakly related to empathy¹², we may conclude that CGI, which is built on emotional induction, may not be effective as a single strategy to mitigate aggressive acts on the Internet committed for violence lust. Importantly, gratitude is defined as an interpersonal process, directly related to social relationships, and depends on expected social costs. For example, Oishi et al.⁸⁶ found that expressing gratitude towards someone important vs. for something good is more costly, because elicits a higher rate of indebtedness, and adverse emotions (guilt, shame). Furthermore, feelings of elevation and indebtedness caused by writing gratitude letters were found to be unique to socially relevant expressions of gratitude (i.e., gratitude to a specific benefactor)⁸⁷. Considering the above limitations of gratitude interventions, if the CGI activated among perpetrators negative emotions such as guild: ‘I have received so much and I have not repaid anyone’; or shame: ‘I have received so much and I myself do not give good things’ the recipients may try to discard them by minimizing the meaningfulness of gratitude exercises (i.e. making jokes from those who express gratitude), or not engaging in those gratitude activities that were supposed to be done with counterparts. As a consequence, the effectiveness of gratitude practices may be disturbed. Another barriers, that limit the beneficial effects of gratitude interventions, are related to the suppression of emotional expression of the experienced gratitude due to social expectations. More specifically, Kumar⁸⁸ pointed out that people underestimate the goodness or overestimate the costs of showing gratitude, which creates a misplaced barrier to interpersonal interaction. Consequently, it undermines the pro-social values in daily life and creates social distance⁸⁹. Noteworthy, past results suggest that the positive effect of gratitude intervention on persons’ mental health is rather short-lived and small^90,91. Complementing the above explanations, it is also worth adding that the ability to respond to gratitude induction may depend on the development of social-cognitive and emotional processes (i.e. empathetic attitude towards classmates), that were found as significant mediators between social normative beliefs about aggression and aggressive behaviors⁹². Additionally, according to the bi-factor structural theory of gratitude, it consists of either cognitive or affective dimension, however when the emotional aspect of gratitude (i.e. the feeling of being grateful to a benefactor) positively predicts mental health outcomes (in cross-sectional and longitudinal studies), the cognitive gratitude (i.e. one’s appreciation for the benefaction – reflecting it as goodness) did it negatively. Considering that most GQ items refer to cognitive gratitude, one week after the CGI intervention data collection could not capture the accurate level of gratitude feelings, which are rather connected to day-to-day interactions. Furthermore, the meta-analysis revealed that the vast number of cyber-aggression interventions did not show positive effects in the long term, especially as they were focused only on social skills and implemented only through the classroom without delivering a wider program focused on the entire school society⁹².

To our knowledge, no previous studies have analyzed the impact of gratitude-based intervention on different types of cyber-aggression. According to our findings, the short-term activation of grateful attitudes among youth effectively reduces aversive, but not appetitive cyber-aggression. Next, adolescent girls benefit more from gratitude intervention than adolescent boys (among the girls’ sample we observed a decline in all types of cyber-aggression, whereas among boys only in the aversive controlled type). Finally, the last valuable result is related to the level of gratitude disposition, the highest decrease was observed among the low-gratitude subgroup in aversive cyber-aggression. In summary, we must conclude that the effectiveness of Classroom Gratitude Intervention aimed to reduce adolescent cyber-aggression is limited regarding some early adolescent groups, therefore practitioners should adopt a broader path of preventive strategies that is not focused only on strengthening positive emotions, but provides students with knowledge about the phenomenon of cyber-aggression itself and its adverse long-term consequences.

Study limitations

The study identified several important limitations that need to be considered. First, the effectiveness of the Gratitude Classroom Intervention (CGI) was not consistent across all participant groups. In particular, boys and students with medium levels of gratitude did not experience the same positive effects as others, and in some cases, their levels of certain types of cyber-aggression actually increased. This raises concerns about the generalizability of the intervention and suggests that it may not be equally effective for all groups. Next, 4.7% dropouts from Time 1 to Time 2 in the experimental group were observed. Failing to measure the analyzed variables among these students might impact the findings. We thus recommend using mixed methods to improve gratitude, to avoid possible participants attrition or boredom. Additionally, the short duration of the intervention—lasting only one week—may have limited its potential to produce lasting behavioral changes. Interventions of longer duration could provide more substantial effects and allow for a better understanding of how gratitude practices influence behavior over time. Another significant limitation is the reliance on self-reported data through the CATQ and GQ questionnaires. While these tools are validated, self-reported data may introduce bias, such as social desirability or inaccurate self-assessment. Future studies could mitigate this limitation by incorporating additional objective measures, such as behavioral observations or peer reports, to provide a more comprehensive evaluation of outcomes. The study also measured outcomes only within the school setting, without considering external influences or behaviors that occur outside of the classroom. This limitation makes it challenging to fully assess the overall impact of the gratitude intervention, as many relevant interactions and experiences may have taken place in other environments. Lastly, the study’s sample was limited to a specific group of Polish adolescents, which restricts the ability to generalize the findings to other populations, age groups, or cultural contexts. Broader studies that involve more diverse samples would be necessary to draw conclusions that are applicable to a wider range of settings.

Future directions

Building on the limitations identified in the current study, future research should aim to address several key areas. One of the main recommendations is to extend the duration of gratitude interventions. The one-week timeframe used in this study may not have been sufficient to foster lasting change in participants’ behaviors and attitudes. Future research should consider implementing longer interventions that include follow-up activities or booster sessions to reinforce the concepts learned and help maintain the benefits over time.

Furthermore, the mixed methods of gratitude induction, including verbal, concrete, connective, and finalistic ways of its expression seems reasonable to increase the effectiveness of such practices among early adolescents. Another critical direction for future research is the need to tailor interventions to account for gender differences. The study’s findings suggest that gender significantly influences how participants respond to gratitude interventions, with girls benefiting more than boys. Future programs should therefore consider offering more personalized activities that cater to the distinct needs, motivations, and social dynamics of different genders. Additionally, expanding the scope of research to include more varied and larger samples is crucial. Future studies should aim to include participants from different cultural, socioeconomic, and geographic backgrounds, as this would provide more robust and generalizable findings regarding the effectiveness of gratitude-based interventions. Finally, it is essential to investigate the long-term outcomes of such interventions. Conducting longitudinal studies that track participants’ behaviors over extended periods would help determine whether the observed effects of gratitude practices are enduring or whether they diminish over time. This type of research would provide valuable insights into the sustained impact of gratitude interventions and inform the design of more effective programs.

Conclusions

The findings of this study indicate that gratitude-based interventions hold promise as a tool for reducing cyber-aggression among early adolescents, but their effectiveness is not universal and is influenced by several factors, including gender and participants’ initial levels of gratitude. While the intervention was successful in reducing aversive controlled types of cyber-aggression, it paradoxically led to an increase in impulsive appetitive aggression in some cases. These results underscore the complexity of using gratitude as a prevention strategy and highlight the importance of tailoring interventions to the specific needs and characteristics of the target audience. The study suggests that while gratitude can be a valuable component in efforts to curb cyber-aggression, it should be integrated into more comprehensive approaches that combine emotional regulation strategies with educational programs focused on the broader issue of cyber-aggression and its consequences. Overall, the study emphasizes the need for more nuanced and flexible approaches to addressing cyber-aggression, ones that take into account the diverse factors that influence adolescent behavior.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

K.T. and A.M-C. wrote the main manuscript text and all prepared figures and tables. All authors reviewed the manuscript.

Data availability

The datasets generated and analysed during the current study are not publicly available due to ethical restrictions related to participant privacy and the confidentiality agreements required by the institutional review board. However, the data are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.