KiVa, an Antibullying Program developed at the University of Turku, Finland, specifically focuses on and targets bystander behavior. KiVa asserts that its cornerstones are prevention, intervention, and monitoring (What is Kiva?).¹ The other major school-based prevention programs, Second Step and the Olweus program, provide only general instructions to pupils about not supporting the bully, reporting bullying incidents to an adult, etc. Bystander behavior is featured prominently in a much briefer intervention for secondary-school students called Say Something (Booker et al., 2023). This intervention consists only of a single 150-min workshop. Despite the small scale of the intervention, the intervention participants reported intervening to support victims of bullying more frequently than members of the control group. Further study and, especially, follow-up data are needed.

KiVa equips teachers with a variety of resources including training modules, materials, and lesson plan manuals. It also includes resources for parents such as a guide that includes general information about bullying, bullying detection, and bullying prevention. According to its website, KiVa is the most rigorously studied antibullying program. The creators of KiVa hypothesized that the less positive reinforcement a bully receives from his or her classmates, the less rewarding bullying will be, which will decrease the motivation to bully to begin with (Kärnä et al., 2011b). Indeed, KiVa assumes that bullying is a social, group phenomenon, (Salmivalli et al., 1996) meaning that bystander behavior modulates bullying (Kärnä et al., 2011a). For this reason, KiVa puts a particular emphasis on shaping the attitudes and behaviors of students who are bystanders.

The bystander material in KiVa does address some components of the situational model of bystander behavior that has emerged from broader research on the bystander effect, introduced earlier. In KiVa, potential bystanders learn very clearly that bullying is a serious problem. They also learn what they can and should do if they witness bullying at school. The identification of bystanders who have fulfilled their responsibilities reduced the anonymity that has been found to lead to standing by idly. Besides being the antibullying program that focuses most specifically on bystander behavior, it is also the one bolstered by the greatest among evaluation data.

The program consists of three separate versions, each tailored to a specific age group (Grades 1–3 (ages 6–9), Grades 4–6 (ages 10–12), and Grades 7–9 (ages 13–16)). In all three versions of the program, KiVa is divided into two types of actions: universal actions and indicated actions. Universal actions are targeted to the entire school and are meant to be implemented, regardless of the circumstances. These actions include 20 h of classroom instruction time, consisting of 10 lessons that cover a variety of topics including peer pressure, respect, and the negative outcomes of bullying. The objective of these lessons is to increase students’ empathy toward bullying victims, spread awareness about the issue of bullying as well as the role the “group” plays in addressing and preventing it, and equip kids with adequate skills to intervene in situations of bullying, thereby increasing their self-efficacy. Students can strengthen their knowledge of these subjects and improve their defending skills by participating in a computer game as part of KiVa (Poskiparta et al., 2012). The game is broken into three parts: “I KNOW,” “I CAN,” and “I DO.” The “I CAN” and “I DO.” Parts of the game specifically target bystander behavior by teaching students how to do the right thing in situations of bullying and motivating them to defend victims (“I CAN”) and then giving them real-world scenarios where they can put these new defending skills into practice (“I KNOW”). Indicated actions are only utilized in instances when bullying is reported to the school. A team consisting of three teachers (or other school faculty) and the bullied student’s classroom teacher is created to address the specific case of bullying. The classroom teacher holds a meeting with between two and four students who exhibit prosocial behaviors and who are considered to be of high social status. Currently, there is no standardized method of picking these students, so it may be beneficial to systematize thus in future evaluation research and/or future revisions revision of the KiVa program. Individual and small group discussions with victims and bullies are implemented, complemented by follow-up meetings.

The studies by Salmivalli and her Finnish colleagues that document the effectiveness of KiVa are many, including several featuring longitudinal follow-up and control conditions. The very first RCT of KiVa took place during the 2007–2008 academic year when KiVa was implemented in schools throughout Finland in grades 4 through 6 whilst being studied for effectiveness (Kärnä et al., 2011a). 78 schools participated in the study, 39 of which were in the KiVa condition (4,207 students) and the other half of which were in the control condition (4,030 students). The study established KiVa as an effective program to target bullying and demonstrated that after only 9 months of implementation, KiVa was effective at reducing both victimization and bullying. In addition to demonstrating the overall effectiveness of KiVa, Kärnä et al. (2011a) highlighted KiVa’s potential to increase prosocial bystander attitudes and empathy, as well as improve bystander behavior. At Wave 2, students in the intervention condition showed increased anti-bullying attitudes and increased empathy. Unfortunately, these effects were rendered null by Wave 3. A similar trend was seen when bystander behavior was analyzed. At Wave 2, students in the intervention condition exhibited defending behaviors more frequently than their counterparts in the control condition. Yet, once again the effect of the intervention on defending behavior was not significant by Wave 3. Despite this, there was a decrease in assisting and reinforcing behaviors at Wave 3. These results are indicative of KiVa’s effectiveness at improving bystander responses to bullying, and the areas where KiVa was not effective (i.e., significant effects at Wave 2 that were nonsignificant at Wave 3) guide future research. Future studies should investigate why might this change be occurring, and how KiVa can be improved to create a long-lasting impact on bystander behavior.

The goal of another study by Saarento et al. (2015) was to uncover which mechanisms in KiVa lead to this reduction of reported victimization amongst students in grades 4 through 6 who have undergone the program. The results confirmed that students in KiVa schools perceived a significant increase in the number of classmates as defenders of the victim and an increase in their classmates’ defending behavior. In contrast to their hypothesis, however, these perceptions were not predictive of future bullying. Participants in the KiVa program showed an increase in affective empathy as well as in antibullying attitudes (Saarento et al., 2015). Results of the study revealed that antibullying attitudes acted as a mediator for self-reported bullying measures, though this was not the case for affective empathy. As well, participants also perceived an increase in their teacher’s overall disapproval of bullying, which led to a statistically significant reduction in self-reported bullying. Results showed that the teacher’s attitudes toward bullying also mediated KiVa’s effects.

Garandeau et al. (2021) similarly considered KiVa’s impact on empathy, querying whether the KiVa program increases cognitive and/or affective empathy in its participants. Further, they questioned if such effects were consistent across all populations (i.e., people of different genders, races, etc.). They discovered that KiVa had a positive impact on affective empathy, and this effect was seen consistently across students of all demographic backgrounds. At the individual level, affective empathy was significantly higher for participants in the KiVa program, though the effect size was quite small. At the classroom level, a significant positive effect of KiVa on affective empathy was also seen and had a slightly larger, but nonetheless small effect size. The same positive effects were not seen on cognitive empathy. While Garandeau et al. (2021) did not measure KiVa’s direct influence on bystander behavior, the increase in affective empathy after participation in KiVa might act as a mediator variable and lead to an increase in defending behaviors, considering previous research that demonstrates a positive correlation between bystander defending and affective empathy (Deng et al., 2021).

Garandeau et al. (2023) sought to determine the degree of impact KiVa has on defending behaviors, as well as the individual-level factors that influenced such defending behaviors as measured by the Participant Role Questionnaire (Salmivalli and Voeten, 2004). The results showed that at 5 months and 9 months, KiVa had a positive effect on defending behavior. While the result was not significant at 9 months when controlling for baseline and 5 months at the same time, there was still a significant indirect effect of KiVa on defending behaviors at 9 months via defending at 5 months (Garandeau et al., 2023). The only significant mediator found to have a consistently positive effect on defending behaviors was responsibility to intervene, when measured at 5 months.

Despite these positive results, one study raises possible concerns regarding KiVa’s effectiveness on bystander behavior. The results of Kärnä et al. (2013), an RCT that studied KiVa’s effectiveness on a population of students in Grades 1–3 and Grades 7–9 (populations that, at the time of the study, had yet to be studied for KiVa’s effectiveness) showed that those in the intervention condition exhibited a decrease in defending behaviors after undergoing KiVa. All students were also asked to nominate an unlimited number of peers as defenders. Students were able to simply report “no one,” if they could not identify a peer whom they saw as a defender. While the intervention effect of KiVa was statistically significant, it was not in the hypothesized or desired direction: defending actually decreased in the intervention condition. This was consistent across all age groups in the study, though the effect increased with age. Future research should seek to address these differing results between studies.

KiVa has also proven to be effective outside of its country of origin. In a Dutch trial of KiVa, a support-group approach was added to the indicated actions protocol, and its subsequent impact on victimization and defending was measured (Van Der Ploeg et al., 2016). To measure any change in defending bystander behavior, data was collected from students in October and again in May. Students who indicated that they were victims of bullying were asked to nominate defenders. Change was indicated by measuring the difference between the total number of defenders each victimized student nominated in October and May. To analyze the differences between students who participated in the support group approach and those who did not, students in both conditions were statistically matched to one another. Victims with a support group had more defenders by May than those who did not have a support group. Descriptive statistics also convey these results, with 52.6% of victims in the support group condition reporting an increase in defending as opposed to only 34.7% of victims in the control condition reporting an increase in defending. The results of Van Der Ploeg et al. (2016) show that a support-group approach to increasing bystander defending is effective.

In further support of KiVa’s effectiveness at improving bystander behavior in countries other than just Finland, data from a Chilean trial of KiVa revealed improvement in the psychosocial adjustment of youths enrolled in schools that offered the KiVa program (Valenzuela et al., 2022). Psychosocial adjustment was measured by a self-report which included questionnaire items related to prosocial behavior, which probably includes helpful bystander behavior although bystander helping was not specifically isolated. Further analysis of this data could allow the researchers to pinpoint if the prosocial behavior was a main component impacting the improvement in psychosocial adjustment, or if it may have been another factor.

The following table summarizes the eight major trials of KiVa that have yielded results and taken place outside of its country of origin, Finland, during the period covered by our literature review. The inclusion criteria for our literature review were that the study had to include, first of all, quantitative data on the effects of Kiva. It had to either be a published study indexed in PSYCINFO before DATE or a study mentioned in the articles already retrieved; and the trial of KiVa had to be conducted outside of Finland (Table 1). Eight studies emerged, including Nocentini and Menesini (2016), Van der Ploeg et al. (2016), Swift et al. (2017), Green et al. (2020), Axford et al. (2020), Huitsing et al. (2020), López-Catalán et al. (2022), and Valenzuela et al. (2022).

The table shows that, in seven out of the eight studies, KiVa had a statistically significant positive effect on reducing both victimization and bullying on self-report measures (all studies) and peer-report measures (2 studies). The main findings of these studies, as outlined in the table, and they highlight that KiVa is effective and can be implemented outside of Finland. This bolsters the support for KiVa as an anti-bullying intervention, as it confirms that the effectiveness of KiVa is not bound to one setting or culture, but rather that it is an intervention that can be universally implemented to address school bullying. Moreover, when determining the effectiveness of any psychosocial intervention, it is important for similar or replication studies to be conducted by scholars who were not creators of the intervention. Of the XX studies conducted outside of Finland, YY did not involve researchers involved in the development of KiVa. Despite these positive results, it is also important to consider that the results of two of these studies, Huitsing et al. (2020) and Van der Ploeg et al. (2016), demonstrate that KiVa’s effectiveness appears to dwindle over time, consistent with the results of some of the original Finnish studies. The reason for this is unclear and future research must be done to determine why this occurs and how maintain KiVa’s effectiveness over time. We theorize that adding “booster” sessions of KiVa could be beneficial.

The possible benefits of moving well beyond preventive programming to more intensive peer-mediated school interventions are just beginning to be explored. These interventions are inspired by generations of research on peer-mediated interventions for children with special learning needs, very often conditions on the autism spectrum. The interventions are administered by classmates who show leadership potential under the guidance of teachers and/or mental health professionals (Strain and Bovey, 2015).

The very limited results to date come from two studies conducted in Spain, with mixed results. Pina et al. (2021) developed the “Count on Me” program. Pupils elected the student leaders whom they wanted to serve as peer mediators among the students who volunteered. The teachers helped the volunteer leaders learn how to recognize, point out, and praise positive behaviors by both teachers and pupils. Based on specific rules and guidelines, the mediators learned how to resolve conflicts. The peer mediation intervention continued for several months. The experimental group that participated in the peer mediation program experienced reduced violence and increased prosocial behavior in comparison with the control group. The effects of the Count on Me program on bystander behavior have yet to be studied. It is possible to conceive of this type of program as guided practice in becoming a prosocial bystander. However, another study in Spain on a peer-mediation program failed to yield a significant improvement in pro-victim attitudes, bystander roles, or school climate (Villanueva Badenes et al., 2022).

Pending further study, there is no strong reason to believe that these more intrusive peer-mediated interventions improve bystander behavior more than the better-established KiVa program. However, these two small studies on peer mediation, with conflicting results, cannot be considered conclusive. Future studies should seek to focus on which specific components and mechanisms of KiVa lead to positive change in situations of school bullying. As well, the null results produced by some studies of KiVa should be investigated. It is important to note that the studies available to review establish the effectiveness of KiVa on bullying and victimization. These studies do not necessarily indicate that it is the bystander focus that accounts for the outcomes reported or that KiVa is the only possible intervention that could be effective; these issues would have to be explored in new studies in which these issues are explored systematically as different intervention conditions.