Project Activities
The study sample consists of 12,740 children who applied to Boston Pre-K in 2007–2011. The research team previously followed these students through the end of third grade. In the proposed study, researchers will access educational records and estimate the impacts of Boston Pre-K in 4th -12th grade and through age 20 on student's school progress, school engagement, academic achievement, educational trajectory, and educational attainment. They will also examine whether impacts vary by important student characteristics (race/ethnicity, cohort, free-reduced lunch status, home language, and gender). Finally, researchers will describe students' experiences during the COVID-19 crisis and explore whether the timing of crisis held different implications for students in later high school grades versus middle school/early high school.
Structured Abstract
Setting
This study will take place in Boston, Massachusetts. BPS offered Pre-K in ~70 out of 81 district elementary schools in the study years. The program was open to all four- year-olds in Boston, regardless of income or other factors.
Sample
The study sample consists of all children whose families applied for a BPS Pre-K slot in 2007-2010. About 25% of these children applied to over-subscribed schools and were ultimately randomized in or out of the program through Boston's school choice system. Children were diverse in their demographic characteristics (i.e., 39% Latino, 21% Black, 28% White, 7% Asian; 51% eligible for free-reduced lunch; 56% English home language in the lottery sample).
Boston Pre-K students were enrolled in a full-day school day for 180 days/year in a Boston public school; their teachers are paid at parity with K-12 teachers, with a BA minimum and MA within 5 years; and they experienced evidence-based curricula (Opening the World of Learning for language, literacy, and social-emotional skills; Building Blocks for math).
Research design and methods
Researchers will estimate the impact of winning a first-choice Pre-K lottery (intent to treat; ITT) and of enrolling in the program (complier average causal effect; CACE). Within the full applicant sample, they will examine the generalizability of lottery-based impacts using propensity-score weighting (PSW). They will use these same approaches to estimate subgroup effects (race/ethnicity, cohort, free-reduced lunch status, home language, and gender).
Control condition
Overall, 88% of lottery control group children who complied with their random assignment and did not enroll in Boston Pre-K were enrolled in another center-based program. For all appliers (the PSW sample), 76% of those who did not enroll in Boston Pre-K attended another center-based preschool program.
Key measures
Measures for the follow-up study are drawn from statewide administrative data. Key measures include school progress and engagement (special education placement; grade retention; attendance; and discipline); educational trajectories/experiences (BPS enrollment and persistence; advanced work course acceptance and enrollment; exam school application, admission, and enrollment); and academic achievement and educational attainment (standardized reading and math tests; taking algebra in 8th grade or earlier and earning a passing grade; on track for graduation in 9th grade; credits towards graduation; on-time graduation; graduation within 5 years; took SAT; SAT score; took Advanced Placement (AP) course; number of APs taken; postsecondary enrollment; enrolled in a 2-year institution). BPS will share with the research team measures of children's experiences during COVID-19 (i.e., school enrollment, credits earned, remote learning log-in records, attendance, and learning format (in-person, hybrid, or remote)).
Data analytic strategy
Researchers will leverage students' first-choice lotteries to estimate the impact of winning a lottery on children's outcomes and the impact of actually enrolling. With the full sample of applicants, they will also examine the generalizability of the lottery-based impacts using propensity-score weighting (PSW). In this approach, the research team will replicate lottery sample findings using PSW before fitting PSW models in the full sample. They will also examine effects for important subgroups (race/ethnicity, cohort, free-reduced lunch status, home language, and gender). Finally, the research team will examine trends in children's experiences during COVID-19 using basic descriptive statistics to try to understand broader findings.
Cost analysis strategy
The research team identified and analyzed the costs of the ingredients of the program in in a previous IES grant.
People and institutions involved
IES program contact(s)
Project contributors
Products and publications
The research will take place within the context of an established research-practice partnership. This partnership will help to make research and research products more relevant to stakeholders and aid in communication. The research team will communicate study findings to key stakeholders, including the BPS Department of Early Childhood, the Massachusetts Department of Elementary and Secondary Education, the Massachusetts Department of Early Care and Education, practitioners, the broader early education field, and families. Products of the research will include peer-reviewed journal articles; policy briefs; presentations at academic and practitioner-focused conferences; blog posts and op-eds; and "fact sheets" for parents that will be shared via multiple channels. The findings may inform public preschool decisions and investments in Massachusetts and other states.
Publications:
Wu, T., & Weiland, C. (2024). Leveraging Modern Machine Learning to Improve Early Warning Systems and Reduce Chronic Absenteeism in Early Childhood. EdWorkingPaper No. 24-1081. Annenberg Institute for School Reform at Brown University.
Related projects
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.