The Bill & Melinda Gates Foundation’s Early College High School Initiative is a reform movement designed to improve college enrollment and completion, particularly for students who are traditionally underrepresented in postsecondary education.  Early College Schools (ECSs) offer students the opportunity to simultaneously pursue a high school diploma and earn college credits (and even credentials) through partnerships with postsecondary institutions. This presentation will include findings from an evaluation of the impact of ECSs on students’ educational outcomes at a sample of ECSs that used lotteries to determine student admission. 

 The ECHSI Impact Study addresses two research questions:

1. Do ECS students have better outcomes than they would have had at other high schools?
2. Do the effects of ECSs differ for students from different backgrounds?

The study sample came from oversubscribed ECSs that conducted lotteries to determine enrollment. The study focuses on three student cohorts.  Cohort I students entered 9^th grade in 2005 and would have graduated in 2009 if they graduated on time; Cohort II entered 9^th grade in 2006 and graduated in 2010; and Cohort III entered 9^th grade in 2007 and graduated in 2011. The study includes 2,612 students who participated in admission lotteries in one of the three cohorts at 11 ECSs that used a lottery as part of their admission processes and met other study eligibility criteria.

Data came from both student educational records and a student survey. The student educational record data were collected from ECSs, districts, states, and the National Student Clearinghouse. The student survey examined students’ experiences in high school and post-high school experiences.  It was administered in fall 2011 to winter 2012 to a sample of students from cohorts II and III. In all, 1,416 students responded, a 90% response rate.

The main impact analyses provided intent-to-treat estimates, comparing all students assigned by lottery to ECSs to those not assigned, regardless of whether they attended.  We modeled the intercept as a random effect and the ECS effect as fixed at the lottery level. The primary outcomes for the study were consistent with the goals of the ECHSI: high school graduation, college enrollment, and college credential attainment (either while in high school or afterward).  We also examined differential impacts for ECSs based on gender, race/ethnicity, low-income status, parental education, and prior academic preparation.

We also examined the degree to which the educational environment experienced by treatment students differed from the environment experienced by comparison students.  The ECS model is designed to provide students with college exposure, small educational environments, and extensive supports.  However, with the national focus on college-readiness in high school, and in particular the boom in dual enrollment programs, many students who did not attend ECSs may have had access to experiences similar to those offered by ECSs. To examine the service contrast, we analyzed data from our student survey.