|Title:||Strengthening School Readiness through Pre-K for All: A University-District Partnership|
|Principal Investigator:||Morris, Pamela||Awardee:||New York University|
|Program:||Evaluation of State and Local Education Programs and Policies [Program Details]|
|Award Period:||5 years (9/1/17 – 8/31/22)||Award Amount:||$4,968,839|
Co-Principal Investigators: Josh Wallack (NYC Department of Education), Elise Capella (NYU)
Purpose: In 2014, New York City's Department of Education (NYC DOE) launched PreK for All, a program to make high-quality pubic prekindergarten available to all 4-year olds in the city. NYC DOE assigns each prekindergarten site to one of four tracks, each with its own focus: (1) PreK Explore, which has a math focus; (2) PreK Thrive, which focuses on child behavior and family engagement; (3) PreK Create, which has an arts focus; and (4) PreK Inspire, which allows sites to choose their own focus. NYC DOE provides professional development and some curriculum and materials specific to the first three tracks, while the fourth track, Inspire, receives more general professional development. The project will evaluate student education outcomes within each of the first three tracks and the costs of each.
Project Activities: The partnership team will examine the impacts of the first three PreK tracks listed above on classroom and site quality and on student education outcomes during prekindergarten and at the start of kindergarten. In addition, the team will examine the fidelity of implementation and the costs of each of the three tracks.
Products: The partnership team will provide evidence of the efficacy and cost of the three prekindergarten tracks being used in New York City's PreK for All program. The evidence will be directly provided to the NYC Department of Education through the participation of department personnel on the project and through ongoing joint meetings, briefings, and policy papers. The team will present results to the research community through presentations and peer-reviewed publications and to the practitioner and policymaker communities through a series of policy briefs, reports, seminars, and conferences.
Setting: The location is New York City.
Sample: The main analysis will include 2500 4-year olds chosen from PreK for All sites that are oversubscribed and use a lottery to choose their students. The sample will comprise two cohorts of prekindergarten students, half from the 2018–19 school year in 250 sites and half from the 2019–20 school year in 250 sites. Both English and Spanish speaking students will be included. Students whose first language is not English or Spanish and students in special education-inclusion classrooms will not be included. The study of longer-term impacts will draw its sample from 4-year olds who attended prekindergarten in 2016–17 and attended grade 3 in 2020–21.
Intervention: PreK for All began in 2014 and has expanded to include 68,500 students in 1,850 sties in 2015–16. Half the students are from low-income families. About 40% of students attend centers provided directly by NYC DOE in district schools and PreK centers and 60% of students attend sites that are under contract with NYC DOE, including community-based organizations and charter schools.
The NYC DOE's Division of Early Childhood Education assigns sites to one of four prekindergarten tracks and provides both general and track-specific professional development and materials. Lead and assistant teachers in all Pre-K for All sites, regardless of track assignment, participate in four full-day professional learning sessions each school year. All sites also receive on-site coaching throughout the year from a NYC DOE Instructional Coordinator and/or Social Worker. All sites must use a curriculum that aligns with the Prekindergarten Foundation for the Common Core and the Program Quality Standards. In addition, all sites may use the NYC Pre-K for All Interdisciplinary Units of Study. The tracks and their foci include:
Research Design and Methods: The research team will use both a lottery-based RCT and propensity score matching. The team will develop an algorithm and apply it after the first round of child pre-K applications is complete for each of the two cohorts to identify eligible (lottery) students and their sites. Comparisons of student education outcomes during prekindergarten and the start of kindergarten will be made for each track, with the treatment group representing students who won the lottery for their first-choice site. The counterfactual will comprise students who did not win their first-choice site and attended other sites. In addition, the researchers will examine comparisons of each track versus a counterfactual of students in the Inspire track, which represents a business-as-usual comparison. To examine the longer-term impacts, the research team will use only a propensity score match of students in prekindergarten in 2016–17 and grade 3 in 2020–21 to make the same type of comparisons as in the lottery study.
Control Condition: Within each track, students who did not gain entry to the site of their first choice through a lottery will serve as the comparison group for student who did gain entry through the lottery. In addition, the researchers will examine the use of a comparison group of students who attended sites within the Inspire track for each of the other three tracks.
Key Measures: The lottery study will measure student outcomes, site characteristics, and fidelity of implementation. Prekindergarten (fall and spring) and fall of kindergarten student outcomes will be measured using assessments of pre-language, vocabulary, word identification, math, and inhibitory control and there are two instruments measuring three constructs: behavior, self-regulation, and social interactions. Site characteristics will be measured using the Classroom Assessment Scoring System (CLASS) and the Early Childhood Environment Rating Scale (ECERS-R). Implementation will be measured using checklists for each track developed by the project team as well as coaching records and logs. For the longer-term study, student outcomes in grade 3 will include standardized test scores in math and English language arts, grades, attendance, disciplinary actions, and grade retention.
Data Analytic Strategy: The research team will analyze the lottery student outcome data using a regression for each track to estimate intent-to-treat results. In addition, the team will use a two-stage least square model with lottery as the instrument to estimate the local average treatment effect when the sample is adjusted for lottery winners who do not attend the site and lottery losers who do attend the site. A propensity score match will be used with the lottery data to estimate the average treatment effect on the treated for each track in order to adjust for potentially confounding site-level characteristics that may occur because sites are not randomly assigned to track. A similar propensity score analysis will be done using the grade 3 student outcomes for the earlier prekindergarten cohort.
Related IES Projects:
Academic Achievement Outcomes from a Pre-K Family and School Intervention (R305A100596)