Abstract
This mixed-methods study examined a group-based, social-emotional learning intervention for adolescents to identify group-level processes associated with positive relational outcomes and how these processes functioned in different settings. The Connection Project (TCP) was implemented with matched control groups in two settings: eight intervention groups from a small private school in Charlottesville, Virginia, and 40 intervention groups from public schools in St. Louis, Missouri. Pre- and post-intervention data were collected from all participants, including a sociometric measure of self-reported ‘relationship depth’ with classmates. A multi-level model predicting change in relationship depth controlling for baseline student demographics demonstrated significant group-level variation in both samples. Data were collapsed to the group-level and hierarchical regression analyses were used to predict change in relationship depth for TCP groups controlling for change in control groups, to produce a residual score for each intervention group. These were then rank ordered to identify the two ‘highest bonding’ and ‘lowest bonding’ groups in each setting, which were then qualitatively analyzed. Five themes emerged around (1) facilitator engagement, (2) trust, (3) discussions of marginalized identities, (4) the role of humor, and (5) rupture and repair processes. Findings suggested several markers of high bonding groups regardless of setting: ‘off-script’ group discussions, automatic trust in facilitators, deeper discussions about identity issues, effective rupture and repair processes, and student vulnerability met with student support. There were also key differences across settings, including the greater importance of discussions about racial and ethnic identities to group bonding among groups in which the majority of students held marginalized racial/ethnic identities. Also, the higher rates of negative humor among low bonding groups of students from a highly-resourced school where the cultural norm was that of high academic engagement. Together, findings suggest that the same program can be beneficial to youth in various settings, and should be tailored somewhat to students’ unique strengths and needs based on their ecological contexts.
