You are here
The gold standard in making causal inference on program effects is a randomized trial. Most randomization designs in education randomize classrooms or schools rather than individual students. Such "clustered randomization" designs have one principal drawback: They tend to have limited statistical power or precision. This study aims to provide empirical information needed to design adequately powered studies that randomize schools using data from Florida and North Carolina. The authors assess how different covariates contribute to improving the statistical power of a randomization design and examine differences between math and reading tests; differences between test types (curriculum-referenced tests versus norm-referenced tests); and differences between elementary school and secondary school, to see if the test subject, test type, or grade level makes a large difference in the crucial design parameters. Finally they assess bias in 2-level models that ignore the clustering of students in classrooms.
Citation: Zeyu Xu, Austin Nichols (2010). New Estimates of Design Parameters for Clustered Randomization Studies: Findings from North Carolina and Florida. CALDER Working Paper No. 43
This paper describes the school mobility rates for elementary and middle school students in North Carolina and attempts to estimate the effect of school mobility on the performance of different groups of students using student fixed effects models. School mobility is defined as changing schools at times that are non-promotional (e.g., moving from middle to high school). We used detailed administrative data on North Carolina students and schools from 1996 to 2005 and followed four cohorts of 3rd graders for six years each. School mobility rates were highest for minority and disadvantaged students. School mobility rates for Hispanic students declined across successive cohorts, but increased for Black students. Findings on effects were most pronounced in math. School mobility hurt the math performance of Black and Hispanic students, but not the math performance of white students. School mobility improved the reading performance of white and more advantaged students, but had no effect on the reading performance of minority students. "Strategic" school moves (cross-district) benefitted or had no effect on student performance, but "reactive" moves (within district) hurt all groups of students. White and Hispanic students were more likely to move to a higher quality school while Blacks were more likely to move to a lower quality school. The negative effects of school mobility increased with the number of school moves.
Citation: Zeyu Xu, Jane Hannaway, Stephanie D'Souza (2009). Student Transience in North Carolina:The Effect of School Mobility on Student Outcomes Using Longitudinal Data. CALDER Working Paper No. 22
This research brief estimates the overall extent of test measurement error and how this varies across students using New York City student- level longitudinal data across grades 3-8 from 1999- 2007. Results reinforce the importance of accounting for measurement error, as it meaningfully increases effect size estimates associated with teacher attributes. There are important differences in teacher effectiveness that are systematically related to observed teacher attributes. Such effects are important in the formulation and implementation of personnel policies. Also, effect sizes as traditionally measured have led analysts to understate the magnitudes of effects because the standard deviation of observed scores overstates the dispersion of true achievement in the student population.
Citation: Donald Boyd, Pamela Grossman, Hamilton Lankford, Susanna Loeb, James Wyckoff (2008). Overview of Measuring Effect Sizes: The Effect of Measurement Error. CALDER Working Paper No.
This research brief describes the legal and operational structure of the Texas longitudinal data system related to recent changes in the Family Educational Rights and Privacy Act of 1974 (FERPA)—which establishes the rights of parents to access their children's educational records and protects the confidentiality of student information—that more closely align law and practice. The U.S. Department of Education's FERPA Final Regulations Amendments took effect January 8, 2009.
Citation: Daniel M. O'Brien (2008). The Texas FERPA Story. CALDER Working Paper No.
This brief calculates graduation rates for the state of Florida using longitudinal data. We describe our measurement strategies and compare them with the state’s official measurement procedures. We calculate the diploma and GED attainment rates of six separate cohorts of Florida 9th graders who began high school between 1995/96 and 2000/01. We then present rates of both diploma receipt and GED receipt at four years and in later years. The results show an increasing trend in graduation rates in the state over the period studied and a substantial bump at five years, with growth flattening out after that time.
Using a unique longitudinal dataset covering all Florida public school students in grades 3–10 over a five-year period, we analyze the impact of classroom peers on individual student performance. Unlike many previous data sets used to study peer effects in education, our data allow us to identify each member of a given student's classroom peer group in elementary, middle and high school as well as the classroom teacher responsible for instruction. As a result, we can control for individual student fixed effects simultaneously with individual teacher fixed effects, thereby alleviating biases due to endogenous assignment of both peers and teachers, including some dynamic aspects of such assignments. We find some sizable, significant peer effects within nonlinear models, but not with linear specifications. We find peer effects depend on a student's own ability and on the ability of the peers under consideration. Peer effects tend to be smaller when teacher fixed effects are included, a result that suggests co-movement of peer and teacher quality within a student over time. We also find that peer effects tend to be stronger at the classroom level than the grade level.
This paper uses administrative data for the public K-12 schools of North Carolina to measure racial segregation in the public schools of North Carolina. Using data for the 2005/06 school year, the authors update previous calculations that measure segregation in terms of unevenness in racial enrollment patterns both between schools and within schools. They find that classroom segregation generally increased between 2000/01 and 2005/06, continuing, albeit at a slightly slower rate, the trend observed over the preceding six years. Segregation increased sharply in Charlotte-Mecklenburg, which introduced a new choice plan in 2002. Over the same period, racial and economic disparities in teacher quality widened in that district.
Citation: Charles Clotfelter, Helen Ladd, Jacob Vigdor (2008). School Segregation under Color-Blind Jurisprudence: The Case of North Carolina. CALDER Working Paper No. 16