BackHome
Refresh: OffView Personal Schedule
Individual Submission Summary
Share...

Direct link:

School Characteristics in Relation to PK-G1 Gains: Multisite Evidence from the Early Learning Network

Thu, March 23, 3:15 to 4:45pm, Salt Palace Convention Center, Floor: 2, Meeting Room 251 B

Abstract

Schools have widespread impacts on students’ educational attainment (Ansari & Pianta, 2018). However, little is known about how specific school characteristics affect young children’s development during the critical transition from prekindergarten (PK) to early elementary school. To address this need, the current study will examine relations between key school constructs (academic performance, strain to attract/retain teachers, and organization of fiscal resources) and children’s gains in literacy, language, and math from PK to first grade (G1).

This study uses multisite Early Learning Network (ELN) data from 3,118 children and 431 schools in five states (Massachusetts, Ohio, North Carolina, Nebraska, Virginia). ELN was designed as a longitudinal investigation following children from PK in 2016–2017 into early elementary school. Participants were predominately from families with low incomes and were racially/ethnically diverse.

Cross-site ELN child-level measures include individual assessments of children’s literacy, language, and math, as measured by Woodcock Johnson subtests (Letter-Word Identification, Picture Vocabulary, and Applied Problems) administered in the fall of PK and spring of each study year. ELN covariates include child gender, race/ethnicity, home language, and free/reduced lunch status. School characteristics derived from national datasets include academic achievement (Stanford Education Data Archive), organization of resources (per-student expenditures, teacher-student ratio; Civil Rights Data Collection [CRDC]), and school strain (novice teachers, student absenteeism, teacher absenteeism; CRDC). School covariates include proportion of students by disability and race/ethnicity, locale, and Title 1 status (Common Core of Data).

This study is currently being preregistered on Open Science Framework. To maintain the integrity of preregistration, we describe intended inferential analyses testing relations between school characteristics and child PK-G1 outcomes using multilevel structural equation models (MSEM). Missing data will be addressed using multiple imputation. Confirmatory factor analysis will represent three latent constructs of performance, strain, and organization of resources, with alternative factor structures considered using exploratory factor analyses if necessary. Two-level MSEM will estimate models with students at level 1 and schools at level 2. After establishing measurement models, we will test associations between school-level characteristics and children’s growth in school performance from PK-G1 and resulting G1 scores. Direct paths will be estimated from the three hypothesized latent constructs of school-level characteristics to intercepts and slopes of child outcomes. We anticipate that school performance and organization of resources will be positively associated with children’s outcomes, whereas school strain will be negatively associated.

Preliminary descriptive findings (Table 5) show cross-site variation in both child demographics and school characteristics, especially school strain (i.e., novice teachers, student/teacher absenteeism). In addition, children showed the largest PK-G1 growth in literacy and the smallest PK-G1 growth in English language skills (Figure 3). This multisite, multiyear study has the potential to shed light on how schools can target resources to promote student growth, and how policymakers can funnel resources to school improvement efforts. Moreover, the focus on PK into early elementary school is particularly important, as few studies to date have examined the influence of school-level indicators on children’s experience in classrooms upon school entry and their growth across key domains.

Authors