The Importance of Coding Time in Growth Models: Analyzing Graduation Rates in Title I Washington State High Schools 2019-2025

Abstract

Four-year graduation rates are a central accountability metric under the U.S. Every Student Succeeds Act (ESSA), yet limited research has examined how rates can evolve longitudinally across under-resourced schools, or how school-level characteristics relate to change over time. This study used multilevel growth modeling to predict graduation trajectories among N = 93 Title I high schools in Washington State across six academic years (2018–2019 through 2024–2025). In addition to investigating mean change over time, between-school heterogeneity, and associations with school enrollment, ESSA Section 1003 improvement funding, and Education Service District (ESD) membership. Data were drawn from publicly available administrative records from the Washington Office of Superintendent of Public Instruction. Results indicated that graduation rates increased slightly but significantly over time. The intraclass correlation coefficient (ICC ≈ .80) revealed that most variance in graduation outcomes was attributable to differences between schools rather than year-to-year fluctuations within schools. Enrollment size was a small but significant predictor of baseline graduation rates, while Section 1003 improvement funding was significantly associated with differences in growth trajectories—though not with baseline levels—showing that improvement funds relate more to longitudinal change than to initial standing. As well, how time points are coded can either smooth over trajectories or spotlight year-to-year differences. Taken together, findings have implications for how funding accountability systems can be designed, analyzed, and interpreted: single-time-point or smoothed-over comparisons alone risk obscuring meaningful heterogeneity whereas nuanced growth-oriented frameworks can better reflect the effects of targeted supports.