Bystanders tend to defend victims in a supportive classroom climate: A cluster randomized control trial and an observational study
通过两项研究（一项整群随机对照试验和一项相关研究）发现，教师采用自主支持性教学能营造支持性课堂氛围，从而增加学生保护受害者的行为，减少被动旁观和强化欺凌的行为。
When bystanders reinforce bullies, bullying tends to escalate; when bystanders defend victims, bullying tends to de-escalate. Recognizing this pattern, we adopted a social-ecological perspective within a self-determination theory framework to conduct two studies. Study 1 was a pre-registered cluster randomized control trial in which 38 Korean secondary teachers (9 females, 29 males; 19 experimental, 19 control) participated in an intervention to learn how to teach in highly autonomy-supportive and not-at-all controlling ways. We hypothesized that this approach to teaching would create a supportive peer-to-peer classroom climate, which in turn would increase defending- and decrease passive- and reinforcing-bystanding. In three waves over an academic year, 1084 adolescent students (490 females, 594 males) reported their classroom climate and bystanding behaviors. According to a doubly latent multilevel SEM analysis, experimental group teachers created a more supportive climate (Β = 0.55, p < .001) and this classroom-level effect increased defending-bystanding (Β = 0.55, p = .001), decreased passive-bystanding (Β = -0.52, p < .001), and decreased reinforcing-bystanding (Β = -0.40, p = .006). Study 2 was a correlational study in which 629 adolescent students (398 females, 231 males) reported on their teacher's autonomy-supportive and controlling motivating styles and the 11 teachers (four females, seven males) rated each student on the three bystanding behaviors and the extent to which they contributed to two dimensions of classroom climate (i.e., supportive and conflictual). A SEM analysis showed that students' perceived autonomy-supportive teaching predicted teacher-rated supportive climate (Β = 0.23, p = .036) and students' perceived controlling teaching predicted teacher-rated conflictual climate (Β = 0.11, p = .041). According to mediation analyses, supportive climate ratings then predicted teacher-rated high defending (Β = 0.28, p = .006) and low passive (Β = -0.29, p < .001) bystander behavior, whereas conflictual climate ratings predicted teacher-rated low defending (Β = -0.22, p = .008) and high passive (Β = 0.26, p = .001) and high reinforcing (Β = 0.37, p < .001) bystander behavior. We conclude that teachers can learn how to create a supportive climate that orients students toward defending and away from passive and reinforcing bystanding.