Testing Latent Profiles of Cohesion in the Neighborhood, Family, and School As Predictors of Emotional and Behavioral Outcomes in a Sample of Youth

Background: Cohesion refers to feelings of trust, mutual support, belonging, and quality relationships. A broad body of research suggests that cohesion is associated with positive youth outcomes. Though these findings are informative, a key limitation in previous research is that researchers typically focus on cohesion in one community context such as neighborhood, family, or school.

Findings presented in a separate SSWR abstract identified three latent profiles of cohesion and community context. Profile one (n = 206, 5.9%) is labeled “neighborhood cohesion,” and is comprised of individuals who report higher levels of neighborhood social cohesion and lower levels of school and family cohesion. Profile two (n = 2,468, 71.6%) is labeled “high cohesion” because respondents reported high levels of cohesion in all community contexts. Profile three (n =770, 22.4%) is “neighborhood and school” because respondents reported lower levels of family cohesion compared to the “high cohesion” group.

The present study tests these latent profiles as predictors of three youth outcomes: wellbeing, externalizing symptoms, and internalizing symptoms.

Method: Data for this study are drawn from the sixth wave of the Future of Families and Child Wellbeing Study. Missing data was accounted for using multiple imputation, resulting in a sample of 3,444 individuals. Data for this study were drawn from the Future of Families and Child Wellbeing study (n = 3,444). Roughly half the youth were male (51.3%), Black (49.1%) and were 16-19 years old (45.7%).

Child wellbeing was measured using the EPOCH Measure of Adolescent Wellbeing (α = 0.794). Externalizing (α = 0.906) and internalizing (α = 0.828) symptoms were measured using the Child Behavior Checklist. Ordinary least squares regression was used to test predictors of wellbeing and Poisson regression was used to test for predictors of externalizing and internalizing symptoms because the variables were highly skewed. Youth in the “high cohesion” profile will serve as the reference category for analyses.

Results: Youth in the “neighborhood cohesion” profile reported lower levels of wellbeing (β = -0.268, p < .05) and higher levels of externalizing (OR = 1.826, p < .05) and internalizing symptoms (OR = 1.709, p < .05). A similar pattern was observed for youth in the “neighborhood and school” profile (wellbeing: β = -0.269, p < .05; externalizing: OR = 1.331, p < .05; internalizing: OR = 1.390, p < .05).

Conclusions: Our findings support for the positive impact on cohesion on youth outcomes. Youth in the “high cohesion” profile reported higher wellbeing compared to youth in the “neighborhood cohesion,” and “neighborhood and school” profiles. This finding suggests that building social cohesion across community contexts can have a more positive impact on youth than focusing on any one context alone.

Our findings also underscore the importance of family cohesion. The main difference between youth in the “high cohesion,” and “neighborhood and school” cohesion profiles was perceived family cohesion. However, the “neighborhood and school” profile was significantly associated with poor outcome. This finding suggests that family cohesion can have a positive impact on youth that they carry into other community contexts.