Educational Opportunity in North Carolina: School District Achievement and Growth Between Third and Eighth Grade

• In Event: Poster Session 09
  In Poster Session: PS 09 Topic - Education, Schooling

Abstract

Reardon (2018) defined and measured two dimensions of educational opportunity using data from all public-school districts in the U.S.: (1) the average level of third grade achievement in a school district was used to measure the cumulative set of educational opportunities available to students up to age 9 (termed “early childhood educational opportunity”); (2) the average rate of growth in achievement from grades 3 to 8 was used to measure educational opportunities available to students in a school district from age 9 to 14 (termed “middle childhood educational opportunity”). Reardon (2018) showed that average achievement was highly correlated with community socioeconomic disadvantage, while achievement growth was not—suggesting that achievement growth is a better measure of school-district effectiveness compared to average achievement. Today, many states use measures of average achievement and achievement growth for high-stakes accountability and quality monitoring.

We extend Reardon’s (2018) work by examining the extent to which average achievement and achievement growth were related to the level of public funding expended per pupil in the school district (per pupil expenditure; PPEs)—a fundamental dimension of educational input. We examined this issue among public-school districts in North Carolina (NC)—a state where average achievement and achievement growth are used for accountability and monitoring. We measured average achievement and achievement growth for 14 cohorts of students who progressed from third to eighth grade in each public-school district in NC. We used test scores of individual students on end-of-grade (EOG) reading and mathematics assessments (transformed to normal curve equivalent scores, with M=50 and SD=21.063 each grade/year). Cohorts were defined as all students who took the third¬–eighth grade EOG assessments during a six-year period (e.g., the first cohort was defined by all students who took the third grade EOG in 1997, the eighth grade EOG in 2002, and/or any EOG assessment in between).

We estimated multi-level growth curve models for each cohort with repeated EOG assessments (Level-1) nested within students (Level-2), and students nested within school districts (Level-3), with separate models for math and reading assessments (28 models total). For each school-district cohort, these models produced (a) an intercept value to index average third-grade achievement and (b) a linear slope value to index achievement growth between third and eighth grade (see Table 1). We then examined bivariate correlations between (a) intercept and slope values and (b) the average level of PPEs in the school district during the six-year period experienced by the cohort (see Table 2).

We found that PPEs showed moderate-to-large magnitude correlations with intercept values (r = |0.18–0.53|), and small-to-moderate correlations with slope values (r = |0.02–0.15|). Additionally, state/federal PPEs were negatively correlated with intercept/slope values, while local PPEs were positively correlated—suggesting that greater levels of state and federal investments were applied toward lower-achieving and lower-growth schools. Given that generally small correlations were found with slope, but not intercept values, these findings suggest that achievement growth is largely unrelated to school spending—providing support for achievement growth as a measure school-district effectiveness in NC.

Authors

• Robert Carr, Duke Univeristy
  Presenting Author

• Tyler Watts, Teachers College, Columbia University
  Non-Presenting Author

• Jade V Marcus Jenkins, University of California - Irvine
  Non-Presenting Author

• Kenneth A Dodge, Duke University
  Non-Presenting Author