Beyond Classroom Averages: Reframing and Measuring Quality at the Individual Child Level in Early Childhood Education

Abstract

Early childhood education (ECE) quality is theoretically expected to promote child development through warm, responsive interactions and cognitively stimulating learning opportunities. Yet empirical associations between classroom-level quality measures and child outcomes remain modest and inconsistent, creating a persistent “quality-outcome puzzle.” This dissertation proposes that this puzzle reflects a mismatch between developmental theory, which emphasizes individualized proximal processes, and dominant measurement and research design practices, which assess quality at the classroom- or program-level. Using video-based observations of 175 prekindergarten children across 26 classrooms in Seattle Public Schools, three interconnected studies examine whether individual-level quality assessment offers a more theoretically aligned and empirically powerful approach.Paper 1 introduces the Quality Exposure Index (QEI), an individual-level observation tool developed as part of this dissertation, and establishes initial psychometric evidence within Kane’s argument-based validation framework. Individual-level quality assessment is psychometrically feasible, though feasibility is subscale-specific: some dimensions achieve acceptable reliability under current observation conditions while others require expanded occasions, additional raters, or purposeful context selection. Paper 2 tests the field’s implicit assumption that contextual variation averages out across observations. This assumption does not hold: children’s quality experiences shift meaningfully across activity formats and learning content in subscale-specific, theoretically interpretable ways. Contextual variation largely reflects genuine differences in children’s opportunity structures, constituting construct-relevant variance under QEI’s experience-based quality framework, rather than measurement noise. Paper 3 directly tests whether the quality-outcome puzzle reflects a measurement mismatch rather than a failure of quality to matter for development. Substantial within-classroom heterogeneity demonstrates that classroom averages obscure the quality experiences individual children actually receive. When quality is measured at the child level, it predicts vocabulary, task orientation, and other outcomes where CLASS does not. Asymmetric attenuation, whereby CLASS effects diminish when QEI is controlled but QEI effects remain stable, identifies individual-level quality as the active predictive mechanism. Together, these studies advance a reconceptualization of early childhood quality as a distribution of experiences each child actually receives, with implications for research design, professional development, and accountability systems that define what it means for early childhood education to be high quality.