To address this gap, the High School Persistence Study is testing the impact of Check & Connect on school engagement, attendance, and completion outcomes among general education students at high risk of dropping out, using a block-randomized controlled trial. The study is being conducted in a large urban school district. The study sample includes the students in ten high schools who were most at risk of dropping out based on 8^th/9th grade risk factors in the year prior to the start of the study (2010-11). To identify at-risk students, we first identified the risk factors from grades 8 and 9 that were most predictive of on-time graduation, using historical data from the district. Students with the lowest predicted probabilities of on-time graduation were then randomly assigned within each school to receive Check & Connect mentoring or to a business-as-usual control group. Study implementation began at the end of students’ 9^th grade year, and is continuing through the students’ expected graduation date in spring of 2014 (2011-12, 2012-13, 2013-14).

This paper will present full study findings about program implementation and impacts of Check & Connect on student outcomes, including engagement, progress toward school completion, and graduation. In impact analyses to date, we have found no significant impact of Check & Connect on interim measures of engagement and student progress toward graduation, including credit accrual, standardized test scores, behavior measures, and enrollment status. These results are not consistent with previous studies. However, ongoing analyses are examining potential impacts on engagement in credit recovery opportunities; these analyses, along with those through the end of the 2013-14 school year, will be completed by or before September 2014, well in advance of the APPAM conference. In addition to presenting full impact analyses, this presentation will provide a summary of our rich implementation data, which capture the frequency, duration, and focus of mentors’ work with their individual students. We will consider the findings in terms of cost-benefit of intensive (e.g. one-on-one) interventions for students, and in so doing, provide valuable insight to policymakers and practitioners struggling to find effective ways to improve outcomes for at-risk students across the nation.