Association for Public Policy Analysis & Management

This paper will provide evidence about the effectiveness of all Early College High Schools (ECHS) in North Carolina, a state with the most ECHS in the nation. ECHS are small schools of choice limited to 400 students primarily located on campuses of two- or four-year colleges or universities. Students who attend ECHS have the opportunity to earn, at no financial cost to them, two years of transferable college credit or an associate's degree while simultaneously satisfying high school graduation requirements. The ECHS model includes a core set of design principles: college readiness, powerful teaching and learning, personalization, redefined professionalism, leadership, and purposeful design.

A prior study is examining impacts from 19 of the 70 ECHS sites in NC (Edmunds et al. 2010, 2012, 2013). The 19 sites used lotteries for admissions to ECHS. This study will use the experimental data and the data on students who attended non-lottery ECHS to build quasi-experimental models that can replicate the experimental impacts be applied, retrospectively, to the other 51 ECHS to estimate their effects. This will permit calculating a pooled estimate of all 70 ECHS and analysis of the generalizability of lottery impacts to ECHS not holding lotteries.

The study assembles statewide administrative data on all NC high school students (including ECHS), community colleges, universities, criminal involvement, and voting in North Carolina. Intermediate and ultimate outcomes include high school course taking and grades, NC standardized end of grade test scores, ACT scores, and high school graduation; postsecondary enrollment, course taking, GPA, and completion; incarceration; and voting. Pre-treatment covariates include levels and trends of middle school NC end of grade standardized test scores, absences and suspensions; parental education, poverty status, school enrollment changes, school-based educational classifications, and socio-demographic characteristics.

Using statewide administrative data and student-level information about lotteries conducted by ECHS, the best performing (lowest bias) propensity score models will be chosen from a within-study replication analysis that tests models using data from the 19 lottery sites as a benchmark. A variety of covariate sets and propensity score techniques will be tested separately for each outcome of interest. The comparison students are those attending a traditional public high school in the state of North Carolina.

This study will contribute to 1) the literature on the impact of this important intervention that has thus far had positive reported impacts from the RCT, 2) the literature on within-study replication and quasi-experimental design (Cook, Shadish, Wong, 2008; Pohl et al 2009; Bifulco 2012), and 3) the literature on the generalization of experimental impacts to schools implementing the same intervention but with no lotteries (Stuart et al 2012; Tipton 2013, 2014, Tipton et al 2014).