These schools do not receive the PRAISE intervention. These schools can have their standard aggression and bullying practices such as community meeting, restorative circles, Second Step program, and the Olweus program. We will measure what classrooms in the comparison condition are doing, but we will not intervene.

These schools will receive the PRAISE program which includes training and coaching from the research team.

PRAISE is a universal, classroom-based aggression and bullying prevention program that addresses all forms (physical, verbal, relational, cyber), with a strong focus on the relational and reactive behaviors. All emotional and social skills taught focus on applying and practicing the skills within peer relationships. This program has gone through iterative development and adaptations using a community-based participatory approach whereby the elementary students and staff were active partners with researchers in refining the program. Specific attention was paid to ensuring that the program is feasible for schools to implement and sustain while also being relevant and relatable for students from a variety of backgrounds and school communities (e.g., under-resourced, higher levels of community violence). This updated PRAISE includes ~14 sessions (amounts can vary by grade) and training and coaching for school-based facilitators of the program.

Measured via peer-reports. Students rate the frequency that each classmate engages in relational (1 item), verbal (1 item) and physical (1 item) aggression on a scale from 1- Not at All to 5- A Whole Lot. Each individual child gets an average score based the classroom peer reports with higher scores reflecting more of that outcome.

Student self-report of social information processing knowledge will be gathered using the Knowledge of Anger Problem Solving measure (9 items) where items are scored a 1 if correct and 0 if incorrect and summed (Range: 0-9), higher scores reflect better social information processing

Student self-report of hostile attributions will be assessed using the Cartoon-Based Hostile Attributional Bias measure, with 2 relational vignettes and 2 physical vignettes, with one question asking whether a behavior was intentional or unintentional. Each vignette has a range of 0 (no bias) to 4 (strong bias). Scores on the 2 items per vignette type (relational, physical) are summed where higher scores reflect a worse outcome.

Student self-report of their use of non-aggressive or violent responses when in a conflict with a peer will be assessed using 9 items drawn from the measure developed for the Multisite Violence Prevention Project with a range of 1-Not at to 4-Very confident. Items are averaged with higher scores reflecting more conference in their ability to not use violent or aggressive responses.

Student self-report of cognitive empathy using the perspective taking scale of the Interpersonal Reactivity Index (6 items; Range: 1-Never to 5-All the Time). Items are averaged where higher scores indicated more empathy.