College students were selected as participants because they are among the most frequent internet users and are highly exposed

to the risk of cyberbullying. After reviewing the study details, the participants completed a cyberbullying identification test

(Chao, 2021) and the Online Moral Disengagement Scale (Paciello et al., 2020). Subsequently, they were presented with two

hypothetical cyberbullying scenarios and asked to report their behavioral intention toward scenarios (these items were adapted

from Bastiaensens et al., 2014). For the analysis, we conducted a correlation analysis, simple linear regression analysis, and

moderation analysis with PROCESS for SPSS (Hayes, 2022) to test our four hypotheses.

Our analysis results revealed the following. First, cyberbullying awareness was not significantly correlated with bystanders’

defending behavior when the victim was cyberbullied without any apparent reason (r = .00, p = .99 > .05), which is inconsistent

with our first hypothesis. Second, cyberbullying awareness negatively predicted bystanders’ reinforcing behavior (adjusted R2 =

.03, F = 40.30, β = −.17, p < .001) and was not moderated by moral disengagement when the victim was cyberbullied without

any apparent reason (coefficient = −.02, p = .23 > .05), which is consistent with our second hypothesis. Third, cyberbullying

awareness negatively predicted bystanders’ defending behavior (adjusted R2 = .00, F = 4.75, β = −.06, p = .03 < .05) and was not

moderated by moral disengagement when the victim was cyberbullied because of his norm-violating behavior (coefficient = .02,

p = .29 > .05), which is inconsistent with our third hypothesis. Fourth, cyberbullying awareness negatively predicted bystander’s

reinforcing behavior (adjusted R2 = .02, F = 29.57, β = −.14, p < .001) and was moderated by moral disengagement when the

victim was cyberbullied because of his norm-violating behavior (coefficient = −.04, p = .03 < .05), which is consistent with our

fourth hypothesis.

Regardless of whether the victim was cyberbullied for no apparent reason or for violating social norms, the bystanders

exhibited only a moderate tendency to defend the victim, and this tendency was not influenced by the degree of cyberbullying

awareness. In other words, regardless of whether the bystanders perceived the ongoing event as a cyberbullying incident,

their inclination to assist remained the same. According to previous studies (Asanan et al., 2017; Elçi & Seçkin, 2019; Steer

et al., 2020), identifying cyberbullying can prevent bystanders from becoming bullies or victims. However, no direct evidence

currently exists regarding whether cyberbullying awareness increases the willingness of bystanders to assist. Therefore,

cyberbullying awareness is regarded as a prerequisite for initiating defending behavior, with willingness to take action

influenced by other factors.

In the two scenarios, cyberbullying awareness negatively predicted bystanders’ reinforcing behavior. In other words, when

the bystanders realized that the current event involved cyberbullying, they became less likely to engage in reinforcing behavior,

such as liking, sharing, reposting, or commenting on posts. These findings are consistent with those of Asanan et al. (2017),

who indicated that cyberbullying awareness has an inhibitory effect on reinforcing behavior. In this study, we further examined

the moderating effect of victim behavior and bystander moral beliefs on the relation between cyberbullying awareness and

bystander behavior. Our moderation analysis results indicated that when the victim violated social norms, moral disengagement

moderated the relationship between cyberbullying awareness and reinforcing behavior, whereas this was not the case in the

regular cyberbullying scenario. These findings support our second and fourth hypotheses, indicating that if bystanders lack

cyberbullying awareness and hold beliefs of moral disengagement, such as moral justification, dehumanization, or victim

blaming, they may become more inclined to rationalize their reinforcing behavior as an act of justice. Their moral sense may

also have a limited influence on their actions, resulting in increased reinforcing behavior. By contrast, if bystanders become

actively aware of their own behavior, even if they hold moral disengagement beliefs, they may become less likely to rationalize

their reinforcing behavior, which enables them to exercise restraint in their own behavior.

On the basis of our research findings, we offer three recommendations for preventing or intervening in cyberbullying

incidents. First, because cyberbullying awareness has an inhibitory effect on bystanders, we recommend that schools implement

antibullying programs that emphasize cyberbullying awareness among all members of the school community, including teachers,

students, and parents. Second, our study indicates that the reactions of bystanders are influenced by the victim’s behavior.

Bystanders are more likely to refrain from assisting or even engaging in reinforcing behavior if the victim has previously

violated social norms. Notably, cyberbullying differs from traditional bullying because online information tends to be one-sided,

which can challenge bystanders’ ability to fully grasp the situation and the victim, leading to premature judgments regarding the

victim’s behavior (Liu et al., 2019). Therefore, when bystanders witness a cyberbullying incident, they should attempt to first

understand the full context of the event, gather both positive and negative information, and refrain from premature judgments of the victim’s behavior. Lastly, in programs aimed at combating cyberbullying, we recommend that teachers incorporate

dilemma-based instruction and address the moral disengagement beliefs of their students. For students with high levels of moral

disengagement, teachers should explicitly communicate that cyberbullying is an unacceptable means of seeking justice.

    This study has some limitations. First, our sample comprised only college students, thus limiting the generalizability of our findings to other populations. Second, all data were collected using self-reporting instruments, potentially resulting in monomethod bias. Third, only one norm-violation scenario was examined. Future studies should include a wider range of scenarios to better understand how the perceived severity of cyberbullying incidents influences bystanders’ perceptions and actions.