Bullying events are mostly known by bystanders and the rest are victims and perpetrators (Evans & Smokowski, 2015). However, most bystanders just look and do not try to stop it (Padgett & Notar, 2013). One of the factors causing low pro-social behavior in bystanders is the low of self-efficacy, thus making the bystanders prefer to remain silent and avoid the occurrence of bullying events. Fear will be the next target of the perpetrator and do not know what to do, make bullying happen continuously. So, the role of bystanders to intervene in bullying is very important to prevent and stop the perpetrators. The respondents in this study are 24 students that indicated as bystanders of bullying through peer-rating. Respondents consisted of 23 female students and 1 male student.

There are two studies in this research. The first study aims to examine the effectiveness of self-efficacy training "Aku Peduli" to improve self-efficacy in the respondents. The second study aims to determine whether there is a relationship between self-efficacy and pro-social behavior on bystanders bullying. The first study used experimental methods with quasi-experimental pretest-posttest and follow-up design. The effectiveness of the "Aku Peduli" efficacy training is measured using self-efficacy scales. The second study used a quantitative method of correlation. The relationship between the two variables in study 2 was measured using the pro-social behavior scale and self-efficacy scale.

Analysis of research data in study 1 using statistical test Wilcoxon Signed-Rank and showed the results p = 0.001 (p <0.05). So, it can be concluded that self-efficacy training "Aku Peduli" effective to improve self-efficacy in the respondents. Furthermore, data analysis in study 2 used Spearman Rho correlation test and showed p = 0,008 (p <0,05). So, it can be concluded that self-efficacy has a positive relationship with pro-social behavior on bystanders bullying. The higher the self-efficacy of bystanders bullying, the higher the pro-social behavior.