Project Activities
The research team will investigate the efficacy of City Connects through three quasi-experimental studies that build on established partnerships between Boston College and the participating school districts and surrounding community agencies. The first study will analyze intervention effects for students randomly assigned to City Connects elementary schools through an oversubscribed enrollment lottery. The second study will consider whether the impact of City Connects is replicated in elementary schools that have been identified as persistently underperforming. The third study will analyze outcomes for students who participate in City Connects during middle school.
Structured Abstract
Setting
The studies will take place using student and school data from high poverty public elementary/K-8 schools in urban school districts in Massachusetts and Connecticut.
Sample
Across all three studies, data from more than 20,000 grade K-8 students in 41 schools (34 in Massachusetts, 7 in Connecticut) from diverse socioeconomic, racial, ethnic, and language backgrounds will be included.
Intervention
Harnessing school district and local community resources, City Connects addresses the unique strengths and needs of every student in a school across a variety of areas (e.g., academic, behavioral, family). A full-time City Connects Coordinator (Masters-level licensed school counselor or social worker) in the school collaborates with teachers and school staff to identify each student's strengths and needs. Based on this assessment, the Coordinator connects referred students to a tailored set of services and enrichments provided by the school district or the local community (e.g., arts and music programs, health and wellness classes, tutoring, mental health or family counseling). In a proprietary web-based system, the Coordinator documents and tracks the service plan for each student. A documented, standardized set of practices, oversight mechanisms, and fidelity tools guide implementation across sites and services.
Research design and methods
For the first study, the research team will take advantage of a natural experiment by analyzing seven years of data (2006-2013) from Boston Public Schools' random lottery assignment system for over-enrolled schools, comparing outcomes for those who applied to City Connects schools and were randomly assigned to City Connects or not. For the second study, the research team will conduct a replication study of City Connects in 18 persistently underperforming schools that are new to the intervention. In this study, the research team will use propensity-score weighted multi-level analyses and a longitudinal Difference-in-Differences (DID) analytic approach to compare student performance before the introduction of City Connects to student performance after the program is implemented. For the third study, the research team will analyze outcomes for students who participate in City Connects during middle school (grades 6-8) using multi-level models that control for student and school characteristics and include propensity score weights.
Control condition
The control condition includes all other schools in the same district that have not participated in City Connects. In the lottery study, City Connects students will be compared with students who requested, but were not assigned to, a City Connects school. Student support structures and processes in a sample of non-City Connects schools will be documented for comparison purposes.
Key measures
Academic outcomes include student report card and state standardized assessment scores, retention in grade, and dropout status. Social behavioral outcome measures include teacher-rated academic effort, behavior, and work habits; school absenteeism; and suspensions. The team will also document the number and type of services and the match between identified student needs/strengths and services.
Data analytic strategy
In Study 1, the research team will estimate treatment effects using multi-level models with consideration given to covariate balance and control for multiple comparisons. In Study 2, researchers will use propensity-score weighted multi-level analyses to examine student- and school-level outcomes and to estimate treatment effects using a Difference-in-Differences (DID) analytic approach. In Study 3, the research team will implement propensity score weighted multi-level models and a multiple baseline interrupted time series (ITS) design with a comparison group.
People and institutions involved
IES program contact(s)
Products and publications
Products: The research team will produce evidence of the efficacy of the City Connects program to improve the social-emotional development and academic achievement of elementary school students in high poverty urban school districts and produce peer reviewed publications.
Supplemental information
Co-Principal Investigator: Laura O'Dwyer
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
