Project Activities
Researchers will use Early Learning Network (ELN) data from approximately 3,200 children and over 2,300 classrooms in five states (Massachusetts, Ohio, North Carolina, Nebraska, and Virginia). ELN was designed as a longitudinal investigation that tracked children annually from PK into early elementary school. Researchers will examine the ways in which school and classroom factors are associated with ELN children's growth in academic achievement, executive function, and social skills. Information on school characteristics will be derived from three national datasets: the Stanford Education Data Archive (SEDA), the Civil Rights Data Collection (CRDC), and the Common Core of Data (CCD). Information on classroom processes and student outcomes are derived from ELN data. The data sources will be combined to analyze the associations between school characteristics, classroom processes, and student outcomes using multilevel structural equation modeling.
Structured Abstract
Setting
This proposed investigation includes data from children, classrooms, and schools across five diverse sites involved participating in the IES-funded ELN: Massachusetts, Ohio, North Carolina, Nebraska, and Virginia.
Sample
The sample includes 3,198 children who attended 2,309 classrooms in 341 schools between PK-1. Participants, drawn from each site's publicly-funded PK program, are predominately from low-income and racially/ethnically diverse families (27% White, 19% Black, 42% Latino/a, 5% more than one race/ethnicity, 2% other race/ethnicity, and 7% Asian).
Factors
School characteristics hypothesized to be associated with student outcomes are school academic performance (achievement level and achievement growth), accessed from SEDA data; school strain (novice teachers, teacher, and student absenteeism), accessed from CRDC data; and organization of resources (per-student expenditures, school expenditure choices, and teacher-student ratios); also accessed from CRDC data. Classroom processes hypothesized as mediators are quality of interactions (teachers' interactions with students to support, scaffold, and strengthen their academic and social learning), teacher-student relationships (teachers' perceptions of their closeness and conflict with students), and instructional rigor (teachers' provision of an instructional environment promoting learning), all of which will be accessed from the cross-site ELN data.
Research design and methods
ELN research teams collected classroom and student data in PK (2016–17) through grade 1 (2018–19). ELN data will be combined across sites and joined with school-level information from three publicly available national datasets: the Stanford Education Data Archive (SEDA), the Civil Rights Data Collection (CRDC), and the Common Core of Data (CCD). Researchers will use cross-site ELN data and school-level data to conduct analyses and examine proposed associations of interest.
Key measures
Measures include variables comprising school-level characteristics (academic achievement [level and growth], school strain [novice teachers, student, and teacher absenteeism], and organization of resources [per-student expenditures, expenditure choices, and teacher-student ratio]). Classroom-level measures include classroom interactional quality (as measured by the Classroom Assessment Scoring System), teacher- student relationships, (as measured by the Student-Teacher Relationship Scale) and instructional rigor (adapted from the Early Childhood Longitudinal Study Kindergarten Class of 2011). Child-level measures include assessments of children's: (1) language, literacy, and math, as measured with the Woodcock Johnson Battery, (2) executive function as measured by the Head-Toes-Knees Shoulders, NIH Toolbox, and the Behavior Rating Inventory of Executive Function, and (3) social skills as measured by the Teacher-Child Rating Scale and the Social Skills Improvement System.
Data analytic strategy
The research team will use a multilevel structural equation modeling framework to account for the nesting of children in schools and will adjust for a full set of child-, classroom-, and school-level covariates. Models will test proposed mediated and moderated relations. Sensitivity analyses using generalized propensity scores and Impact Thresholds for Confounding Variables will gauge the robustness of associations of interest. Missing data will be addressed using multiple imputation.
People and institutions involved
IES program contact(s)
Products and publications
Products: The research team will disseminate results through peer-reviewed publications and scientific presentations, social media, and reports for policymakers and practitioners.
Additional project information
Previous award details:
Previous award number:
R305A210277
Previous awardee:
Related projects
Supplemental information
Co-Principal Investigators: Ansari, Arya; Lin, Tzu-Jung; McCormick, Meghan; Purtell, Kelly; Witte, Amanda
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
