Longitudinal Comparison of the Adoption and Implementation of Community Schools

Background and Purpose:

Community Schools foster collaborations that leverage family and community assets, coordinate a variety of school-based and school-linked services, and ultimately meet the diverse needs of students and families (Blank, Melville, & Shal, 2003). By crafting interventions to fit within local contexts, Community Schools respond to the variety of ecological systems impacting students and their communities. Studies show improved school- and student-level outcomes are associated with the adoption of Community Schools, including attendance and math achievement (Anderson-Butcher et al., in press; Biag & Castrechini, 2016; Houser, 2016;). Research on the adoption of Community Schools, however, rarely examines achievement across time and in comparison to other schools who do not adopt the Community Schools model.  

Method:

The current study sought to examine how the adoption of Community Schools in four Title I elementary schools (K-5) influenced student achievement in one school district in Utah. The study used longitudinal achievement data (e.g., percent proficient in English/language arts) gathered over a 7 year time period. Achievement in the Title I schools was compared not only across the district in comparison to the other 25 elementary schools in the district, but also across pre- (2010-2012) and post-intervention time periods (2013-2016). Researchers used a piecewise hierarchical linear modeling (HLM) technique to evaluate whether or not achievement over time nested within schools improved across pre- and post-intervention time periods. Two unconditional models containing a single fixed effects and a random effect at Level 1 and Level 2 was tested. The percent of students living in poverty was included as a covariate in the model.

 

Results:

Two unconditional models were estimated. We first estimated a model that predicted the initial intercept of ELA proficiency across the pre- and post-intervention time periods; testing the effect of the adoption of Community Schools (e.g., intervention) on achievement. Results indicated no significant differences in the achievement of the schools adopting the Community Schools model and other schools in the district (B = 0.68, p = 0.63). The percent of students living in poverty was a significant predictor of achievement in the model (B = -4.06, p = 0.00). In the second model, we tested the interaction between poverty and the intervention. Results showed the percent of students living in poverty (B = -4.00, p = 0.00), intervention (B = 2.54, p = 0.02), and the interaction between poverty and the intervention (B = -3.22, p = 0.00) were significant predictors of achievement.

 

Conclusions and Implications:

Results suggest the intervention interacted with poverty such that Title I schools with a smaller percentage of students living in poverty had increased achievement, while schools with a higher percentage of students living in poverty did not see the same gains in achievement. Researchers, district leaders, and school officials can use these results and advanced methodological techniques to examine how schools respond to interventions that seek to coordinate community resources and services to improve student achievement.