How's It Going? Adult and Student Interactions in Non-Classroom Settings within High Schools

Students are typically allowed greater independence in non-classroom settings within schools (e.g., begin to change classes, visit lockers) as they transition into adolescence; this increases opportunities for rule violations, as there is often less supervision during these times. It can also be an opportunity for students and school staff to contribute to a safe and supportive school climate, through informal yet positive interactions. In this study, we examined associations between observed adult and student behaviors in non-classroom settings. Observational data (N = 800) were collected from non-classroom locations (i.e. arrival, hallway/stairway, cafeteria, and dismissal) in 58 high schools participating in a study of school climate. Observers tallied identified teacher and student behaviors for 10 minutes in each location. Tallied teacher behaviors included actively connecting with students, proactive behavioral expectations, reactive behavior management, approvals, and disapprovals. Tallied student behaviors were non-compliance, horseplay, physical aggression and verbal aggression. Inter-observer agreement, reflecting the total number of agreements in tallies divided by the total number of agreements and disagreements multiplied by 100% (Barlow & Hersen, 1984), averaged 85% during initial training and 88% during recalibration. Following tallying of behaviors, observers completed a list of items on a global rating scale. Exploratory and confirmatory factor analysis (CFA) of the list of 51 global rating items was used to create 5 scales. Resulting scales include: adult-student relationships, adult monitoring, proactive behavior management, positive student behaviors, and negative student behaviors. The fit statistics for the adult scales were CFI = .986, TLI = .984, RMSEA = .047, and WRMR = 1.205 when accounting for clustering at the school level. All factor loadings were above .60. The fit statistics for the student scales were CFI = .985, TLI = .981, RMSEA = .040, and WRMR = .873 when accounting for clustering at the school level. All factor loadings were above .60. Detailed findings from the CFA will be presented as well as multilevel analysis of the associations between adult and student behaviors. Preliminary results show a positive significant association between tallied instances of teachers actively connecting with students and student positive behavior scales (r = .07, p < .05).