Search
On-Site Program Calendar
Browse By Day
Browse By Time
Browse By Person
Browse By Room
Browse By Unit
Browse By Session Type
Search Tips
Annual Meeting Housing and Travel
Sign In
PreK can have strong effects on Kindergarten-readiness, but these academic advantages typically fade out by Grade 3 (Deming, 2009; Lipsey et al., 2015; Puma et al., 2012). However, little is known about the pattern of fade out before third grade because standardized tests do not begin until Grade 3. Our district partnership allowed us to leverage unique data on student outcomes from Kindergarten through Grade 2 to shed light on these patterns. Here we examine the impact of two preschool programs (preK and TK) on the skills of early elementary school students at several time points between Kindergarten and Grade 2. Because preK programs have larger effects for enrollees from low-income households (Cascio & Schanzenbach, 2013; Feller et al., 2016; Gormley & Gayer, 2005; Kline & Walters, 2015), we also examined how patterns of literacy and math skill convergence vary within a targeted, means-tested preK program as a function of household socioeconomic status.
This study uses administrative data from 72 schools in our partner district, which offers both a targeted, means-tested preK program and an age-based universal TK program. We focus on students in general education classes in Kindergarten through Grade 2, from five cohorts who could have attended preK or TK from 2015-2016 through 2019-2020 (N = 32,550). To assess literacy outcomes, we leverage the Fountas and Pinnell pre-literacy (in Kindergarten) and literacy assessments (in each of Kindergarten through Grade 2). To assess mathematics outcomes, we leverage a district-wide assessment of mathematics milestones tasks.
We use entropy balancing to compare the skills of preK enrollees and nonenrollees across several grades. Similar to other matching methods such as propensity score or Mahalanobis distance matching, entropy balancing involves preprocessing the data to synthesize comparison groups that are balanced in their covariates so the estimate of the effect of a treatment is independent of observable traits or demographics (Ho et al. 2007; Sekhon 2009; Abadie & Imbens 2007). We match on students’ ages, races/ethnicities, English language proficiencies, special education statuses, and genders, as well as percent free and reduced price lunch eligibility at the school-level. We match separately for each analysis by group (i.e. preK, subsidized preK and tuition-paying preK, and TK), assessment (i.e. pre-literacy assessments, reading levels, and mathematics) and grade-level (i.e.. each of Kindergarten, Grade 1, and Grade 2).
While literacy and math skills of preK or TK enrollees are higher on average at Kindergarten-entry than nonenrollees’, these skills largely converge by the end of Kindergarten. This corroborates and extends recent analyses of convergence in literacy and socioemotional skills before Grade 3 (Ansari et al., 2020; McCormick et al., 2021). The targeted preK program appears to mitigate economic gaps, but only at Kindergarten-entry. This suggest that to sustain the benefits of preK and TK, school practitioners may do well to focus on instructional practices both in the preK programs and in the Kindergarten through Grade 2 environments, including a renewed focus on the teaching of unconstrained skills in preK and curricular alignment from preK through Grade 2.